Thorax trauma-induced experimental lung injury

Thorax trauma-induced experimental

lung injury

DISSERTATION

zur Erlangung des akademischen Grades eines

Doktors der Naturwissenschaften (Dr. rer. nat.) des Fachbereiches für Biologie

an der Universität Konstanz

vorgelegt

von

Katja Eichert

Konstanz, im Juli 2003

Dissertation der Universität Konstanz

Datum der mündlichen Prüfung: 16.10.2003

1. Referent: Prof. Dr. Albrecht Wendel

2. Referent: Prof. Dr. Klaus Schäfer

Table of Contents

I

Table of Contents 1. Introduction....................................................................................................... 1

1.1 The lung.................................................................................................................. 1 1.1.1 Lung morphology and aerodynamics.................................................................... 1 1.1.2 Physiological parameters of the lung .................................................................... 2

1.2 Pulmonary oedema ................................................................................................ 3 1.2.1 Alveolar liquid clearance....................................................................................... 4

1.3 Mechanisms of host defence ................................................................................ 5 1.4 Multiple trauma..................................................................................................... 11

1.4.1 Impact of trauma on patients and society ........................................................... 11 1.4.2 Blast injury.......................................................................................................... 11 1.4.3 Chest trauma...................................................................................................... 12 1.4.4 Molecular injury .................................................................................................. 15 1.4.5 Management of pulmonary blast injury ............................................................... 16 1.4.6 Immunomodulation after trauma......................................................................... 16

2. Aims of the Study ........................................................................................... 18

3. Materials and Methods ................................................................................... 19 3.1 Animals................................................................................................................. 19 3.2 Chemicals ............................................................................................................. 19

3.2.1 Chemicals used in the isolated perfused rat lung experiments ........................... 19 3.2.2 Chemicals used in the in vitro experiments......................................................... 20 3.2.3 Other chemicals ................................................................................................. 20 3.2.4 Solutes ............................................................................................................... 21 3.2.5 Anaesthesia and analgesic................................................................................. 21 3.2.6 Cell culture material............................................................................................ 21

3.3 Laboratory equipment and technical devices .................................................... 21 3.4 The in vivo blast wave thorax trauma in rats ..................................................... 22

3.4.1 The blast wave generator apparatus .................................................................. 22 3.4.2 Pressure wave monitoring .................................................................................. 23 3.4.3 Protocol of anaesthesia and analgesic ............................................................... 24 3.4.4 Experimental protocol......................................................................................... 25 3.4.5 Determination of an injury score and validation .................................................. 26

3.5 The isolated ex vivo perfused and ventilated rat lung....................................... 27 3.5.1 Isolated perfused rat lung preparation ................................................................ 27 3.5.2 Experimental Setup ............................................................................................ 27 3.5.3 Experimental design of the perfused lung studies............................................... 28

3.6 In vitro stimulation of primary rat alveolar macrophages ................................. 29

Table of Contents

II

3.6.1 Preparation and culturing of rat alveolar macrophages....................................... 29 3.6.2 Stimulation with dead, fluorescent bacteria......................................................... 30 3.6.3 Stimulation with endotoxin.................................................................................. 30 3.6.4 Fluorescence labelled E. coli/ S. aureus phagocytosis assay ............................. 30 3.6.5 Assay for superoxide production ........................................................................ 31 3.6.6 Preparation of washed and haemolysed erythrocytes and haemolysat............... 32 3.6.7 Lysis of erythrocytes........................................................................................... 32 3.6.8 Experimental design of the in vitro experiments.................................................. 33

3.7 Lung lavage .......................................................................................................... 33 3.8 Determination of the alveolar haemorrhage....................................................... 34 3.9 Total and differential cell counts and cell viability ............................................ 34 3.10 Determination of total protein content................................................................ 34 3.11 Cytokines determinations.................................................................................... 34 3.12 Measurement of eicosanoids .............................................................................. 35 3.13 Determination of NO production......................................................................... 35 3.14 Measurement of lactate dehydrogenase ............................................................ 36 3.15 Measurement of CINC-3....................................................................................... 36 3.16 Gelatin zymograpy ............................................................................................... 36 3.17 Lung Wet/ Dry-Ratio............................................................................................. 37 3.18 Histological examinations ................................................................................... 37 3.19 Statistics............................................................................................................... 37

4. Results............................................................................................................. 38 4.1 System characteristics ........................................................................................ 38

4.1.1 Variability and influence of physical parameters ................................................. 38 4.2 Functional changes of alveolar macrophages and interaction with blood in vitro after trauma ............................................................................................................. 43

4.2.1 Trauma-induced impairment in host defence mechanisms ................................. 43 4.2.2 Role of the trauma-induced alveolar haemorrhage in impaired macrophage

functions......................................................................................................................... 48 4.2.3 Influence of different blood components on the macrophage function ................ 50 4.2.4 Summary of the in vitro findings ......................................................................... 53

4.3 Functional consequences and endogenous modulation in vivo ...................... 55 4.3.1 Blast injury-induced disruption of the alveolar capillary barrier............................ 55 4.3.2 Blast injury-related mediator release .................................................................. 60 4.3.3 Infiltration of inflammatory cells .......................................................................... 64

4.4 Ex vivo lung perfusion: Trauma-induced changes in lung function................. 69 4.4.1 Interrelationships between physical and physiological injury .............................. 69

Table of Contents

III

4.4.2 Time-dependent recovery of the lung function after trauma................................ 73 4.4.3 Pharmacological intervention.............................................................................. 76

5. Discussion....................................................................................................... 83 5.1 A laboratory model for studying primary blast injury ....................................... 84

5.1.1 Rationale for the development of the blast thorax trauma rat lung model ........... 84 5.1.2 Standardization and monitoring of the blast pressure wave................................ 84 5.1.3 Standardization of the blast thorax trauma in rats............................................... 86

5.2 Combination of the in vivo thorax trauma model with ex vivo lung perfusion 86 5.2.1 Match of physical and biological impact.............................................................. 87 5.2.2 Time-dependence of the lung dysfunction .......................................................... 88 5.2.3 Pharmacological modulation of the injury ........................................................... 89

5.3 Thorax trauma-related pathophysiological changes in vivo over time ............ 91 5.3.1 Structural integrity of the lung tissue after blast injury......................................... 91 5.3.2 Mediator release in response to primary blast injury........................................... 93 5.3.3 Thorax trauma-related chemokine release and inflammatory cell infiltration ....... 97 5.3.4 Oedema generation and resorption after thorax trauma: Comparison of the ex

vivo and in vivo results.................................................................................................... 98 5.4 Thorax trauma-related impairment of alveolar macrophage function in vitro . 98

5.4.1 Alveolar macrophage population after trauma .................................................... 98 5.4.2 Phagocytic capacity, microbial killing and TNF release....................................... 99 5.4.3 Involvement of trauma-related alveolar haemorrhage in impaired macrophage

functions....................................................................................................................... 103

6. Summary ....................................................................................................... 105

7. Deutsche Zusammenfassung ...................................................................... 107

8. References .................................................................................................... 110

9. Abbreviatons................................................................................................. 120

Introduction

1

1. Introduction

1.1 The lung

The lung is the central organ for gas exchange of land-living animals, including mammals. The

respiratory system is one of the primary interfaces between the organism and the

environment. As a result of air circulating through the respiratory system, the latter represents

a target for toxic gases, airborne particulates, and infectious agents. Because the entire output

of the right heart also passes through the vascular bed of the lung, this organ is also a target

for blood-borne toxicants and pathogens. Therefore, the lung also has many non-respiratory

functions, such as metabolism, mediator synthesis, host defence and the clearance of

substances from the pulmonary circulation. The respiratory system, especially the conducting

airways and gas exchange area of the lungs, is organized in a highly polarized fashion. As a

consequence, most of the pathological responses within the lung tend to be highly focal and

generally target one or more small subpopulations of the over 40 different cell types found in

this organ. Thus, the cellular composition and architectural organization of airway and gas

exchange tissue have major impacts on the pathological responses within the system. This

chapter will only focus on physiological aspects and metabolic functions pertinent to the

present study.

1.1.1 Lung morphology and aerodynamics

Most of the thoracic cavity is occupied by the right and the left lungs, which are divided into

three and two lobes, respectively. The anatomical structure of the lung combines the

respiratory system and the blood circulation. During breathing, the inhaled air enters the lung

via the trachea and flows towards the respiratory zone driven by pressures, generated by

constriction of the inspiratory muscles. The airways divide about 25 times, depending on the

species. The trachea branches out into the two main bronchi, which then further divide in the

lobular bronchioles, the respiratory bronchioles and the alveolar ducts, which end blindly in the

alveoli. The upper airways are stabilized by cartilage; towards the lower airways less cartilage

but more contractile muscle fibers and elastic fibers are found. The total number of airways

grows exponentially and with each division the total diameter increases, while the diameter of

the single airways decreases. The total area of the 300 million alveoli is estimated to vary

between 80 and 140 m2 (depending on the publication) in the adult lung. The alveoli are 0.2 –

0.3 mm in diameter and are surrounded by a net of capillaries with 5 µm in diameter. Here, at

the very thin alveolar-capillary membrane (0.2 – 0.6 µm), oxygen and carbon dioxide

Introduction

2

exchange occurs. This membrane consists of capillary epithelial cells, a very narrow

interstitium and endothelial cells. Two different systems ensure the blood supply of the lung:

i) the pulmonary circulation (low pressure system), which starts at the right ventricle,

leaving the heart via the pulmonary artery which carries the oxygen-poor blood to

the lung. After saturation with oxygen, the blood leaves the lung by the pulmonary

vein, which ends in the left atrium of the heart.

ii) The bronchial circulation (high pressure system) supplies the central airways and

vessels as well as lung nymph nodes with blood.

The alveolar epithelium is composed mainly of cuboidal type II pneumocytes and large, flat

type I pneumocytes. Alveolar type II pneumocytes synthesize, store and secrete the highly

surface active pulmonary surfactant, a heterogeneous mixture of lipids and proteins, that

reduces surface tension along the air-liquid interface and prevents alveolar collapse upon

expiration. The main function of the type I pneumocytes is the formation of the air-blood

barrier and therefore they facilitate the gas exchange. Alveolar and interstitial macrophages,

eosinophils and neutrophils are also located in the lung, in order to maintain its immunological

functions.

1.1.2 Physiological parameters of the lung

The functional properties of the lung, i.e. the airway and vascular mechanics, are

characterized by different lung function parameters, including the tidal volume, the pulmonary

compliance, the airway resistance, the vascular resistance, and the extent of extravascular

lung water (oedema). The determination of these parameters is not only interesting for the

diagnosis used by the medical specialist, but also for the scientist. Various experimental

approaches in whole animals or in isolated organs, such as the isolated perfused rat lung, 1

have been done to investigate these lung parameters.

1.1.2.1 Airway mechanics

The tidal volume is the amount of air which enters the lung during a single breathing cycle of

inspiration and expiration. It can be calculated by integrating the airflow velocity during

respiration. The driving force for the airflow in the airways is the transpulmonary pressure, e.g.

the difference between the alveolar and the pleural pressure. During inspiration, the thorax

expands, thus the pleural pressure decreases, and the transpulmonary pressure increases,

causing inspiratory airflow from the outside into the alveoli. During expiration, the thorax

relaxes forcing expiratory airflow back into the atmosphere. At a given transpulmonary

pressure, tidal volume depends mainly on two parameters: airway resistance and lung

compliance. The airway resistance is an index for the resistive forces against the airflow in the

Introduction

3

airways and depends on the diameter and the length of the airways. Because the total

diameter is lowest in the upper airways, this region accounts for the major part of the airway

resistance. The airway resistance can be calculated from the relation between transpulmonary

pressure and airflow velocity. The airway resistance increases, as a consequence of

narrowing the airways due to bronchoconstriction or obstructive processes, e.g. bronchial

oedema or enhanced mucus deposition.

The pulmonary compliance is a marker for the functional stiffness of the lung and thus

depends on properties of the peripheral lung tissue, such as elastic fibers. The compliance

can be calculated from the relation between tidal volume and transpulmonary pressure and

thus is also affected by the air volumes at which it is calculated and the lung volume history.

The pulmonary compliance decreases during restrictive pathological changes e.g. atelectasis,

fibrosis, loss of elastic fibers, pulmonary oedema or disturbed surfactant secretion.

1.1.2.2 Vascular mechanics

The pulmonary blood flow is regulated by contractile elements of the pulmonary vasculature

and depends mainly on the smooth muscle tone of the small arteries. The vascular resistance

is calculated from the ratio between the blood pressure difference (Part - Pven) and blood flow

rate. This parameter is increased as a consequence of vasoconstriction, but also of air emboli

and vascular obliteration, due to thrombus formation or blood cell aggregation.

1.2 Pulmonary oedema

The term oedema denotes the presence of abnormally high amounts of water in the

interstitium (interstitial oedema) or in the alveoli (alveolar oedema). Under physiological

conditions, the positive hydrostatic blood pressure, interstitial pressure and oncotic forces are

in balance and small amounts of excess water are drained of by the lymphatic system. Only

when the lymph`s capacity is exceeded e.g. due to massive interstitial fluid accumulation or

loss of the alveolar-capillary barrier integrity, water gets access into the alveoli. The

consequences are impaired gas exchange and a loss of pulmonary compliance. Depending

on the cause, two different kinds of oedema formation can be determined: Cardiogenic or

hydrostatic oedema and non-cardiogenic or high permeability oedema. The former is due to

enhanced capillary pressure, leading to subsequent water influx into the interstitial space,

whereas the latter is independent of the hydrostatic pressure and is related to an impaired

permeability for water and/ or proteins. Non-cardiogenic oedema and hypotension represent

the hallmarks of ARDS (adult respiratory distress syndrome).

Introduction

4

1.2.1 Alveolar liquid clearance

Sodium and fluid transport across the intact respiratory epithelium

Although, for many years, Starling forces (hydrostatic and protein osmotic pressures) were

thought to play a major role in maintaining the alveolar space free of fluid 2, there is now

strong evidence that active ion transport across the alveolar epithelium creates an osmotic

gradient that leads to water reabsorption both during the perinatal period 3 and in adulthood 4

(reviewed by Matthay and coworkers, 5; Sartori and Matthay, 6). Despite significant species

differences in the basal rates of sodium and fluid transport 7-12, the basic mechanisms seem to

be comparable. On the apical membrane of alveolar type II cells there is a sodium uptake by

amiloride-sensitive cation channels, such as the amiloride-sensitive epithelial sodium channel

(ENaC) as well as by several other non-selective cation channels 3,4,13. On the basolateral

surface of the cell, sodium is discharged into the interstitium by the ouabain-sensitive sodium/

potassium adenosine triphosphatase (Na+/ K+-ATPase) 14,15. In addition to the sodium

transport, a pathway for chloride transport, involving CFTR (cystic fibrosis transmembrane

regulator chloride channel) is suggested to play an important role in fluid resorption 16. CFTR

activation was recently suggested to regulate ENaC 17. Due to the osmotic gradient, water

follows passively, probably through water channels (the aquaporins, AQPs) 18,19, although the

presence of these water channels is not required for maximal alveolar epithelial transport in

the lung 20. This process can be upregulated by several catecholamine-dependent and

independent mechanisms.

Regulation of alveolar transepithelial fluid transport

Beta2-adrenergic agonists can upregulate alveolar fluid clearance in isolated perfused rat

lungs 21,22, ventilated rats 8, sheep 7 and dogs 23 as well as in mice 24, partly due to cyclic

adenosine monophosphate (cAMP)-dependent mechanisms 25. Furthermore β1-adrenergic

stimulation is reported to be also effective in upregulating fluid resorption. Both β1- and

β2-receptors are detectable on both the apical and basolateral surface of the alveolar

epithelium 26.

Hormones like epidermal growth factor (EGF) 27, keratinocyte growth factor (KGF) 28 or TGFα 29 are capable of inducing a catecholamine-independent increase of the alveolar fluid

clearance, primarily by stimulating alveolar type II cell proliferation.

Introduction

5

Sodium and fluid transport across the respiratory epithelium in acute lung injury

A profound injury to the alveolar epithelium can disrupt the integrity of the alveolar barrier and/

or downregulate ion transport pathways, thus reducing net alveolar fluid reabsorption and

enhancing the extent of alveolar oedema. Endogenous catecholamines in experimental rat

models of hyperoxia 30, haemorrhagic shock 31, septic shock 32 and in a canine model of

neurogenic pulmonary oedema 33 were reported to stimulate the alveolar fluid clearance.

Alternatively, catecholamine-independent stimulation of the fluid transport may occur due to a

mechanism involving the lectin-like domain of TNF 34,35.

Pharmacological stimulation of the alveolar fluid clearance

Beta2-agonists are attractive therapeutic agents because of their minimal side effects even in

critically ill patients 36. They accelerate not only the resolution of experimental induced alveolar

oedema, e.g. aerosolized salmeterol in sheep 37, moderate injury due to hyperoxia in rats 38,

but also increase the secretion of surfactant and perhaps exert an anti-inflammatory effect 39,

thus helping to restore the vascular permeability of the lung 40. With regard to accelerate re-

epithelialization of the alveolar barrier in the case of acute lung injury 41, benefits with KGF

pre-treatment 42 have been achieved. The combination of KGF and β2-agonist treatment

results in an additive upregulation of fluid clearance 43. This suggests that there may be

mechanisms for providing both short-term (β2-agonist) and long-term (growth factors)

upregulation of fluid transport, that might hasten the resolution of clinical pulmonary oedema.

1.3 Mechanisms of host defence

The respiratory tract is situated at the interface between the environmental air and the internal

tissues of the organism. During normal ventilation or as a result of aspiration, noxious

materials, including infectious agents, are deposited on mucosal surfaces of the airways and

penetrate into the depths of the lung parenchyma. Foreign material encounters a highly

integrated system of natural and acquired defence mechanisms, that prevent injury, infections,

and invasion of host tissue 44-46. This system of host defence includes mechanical (filtration,

cough, mucociliary clearance), molecular (airway secretions), phagocytic, and antigen-specific

immune mechanisms (B- and T-cell-mediated reactions) of resistance (reviewed in 47).

Introduction

6

1.3.1.1 The alveolar macrophage

The alveolar macrophage (AM) is the primary resident respiratory cell responsible for

maintaining the sterility of the alveolar space and thus represents the first line of defence. As a

mobile cell and representative of the mononuclear phagocyte system, the AM is believed to be

the central modulator, as both regulator and effector of the degree of inflammation and anti-

inflammation within the alveolar space 48. This metabolically highly active cell scavenges

particulate matter, removes macromolecular debris, kills microorganisms, functions as an

accessory cell in immune response, maintains and repairs the lung parenchyma and provides

surveillance against neoplasms 49. Since the pulmonary alveolar macrophages are enveloped

by surfactant in vivo, and since their lysosomes contain hydrolytic enzymes necessary for

surfactant degradation 50, it is believed that alveolar macrophages are also involved in the

clearance of surfactant from the alveoli. When confronted with a particularly large inoculum of

microorganisms, the AM supplements its direct antimicrobial capabilities by recruiting and

activating polymorphonuclear leukocytes (PMN) from the bloodstream, which may represent

the second line of defence. Approximately 40% of the body’s PMN’s are marginated within the

microvasculature of the lungs, facilitating the recruitment to the alveolus, or to other sides of

the body 51.

1.3.1.2 Location and origin

Pulmonary macrophages constitute the most abundant non-parenchymal cell type in the lung 52. Macrophages are present in the alveoli (alveolar macrophages, AM), interstitial spaces

(interstitial macrophages), intravascular spaces (intravascular macrophages), conducting

airways (airway macrophages), pleura (pleural macrophages), and lymph nodes (lymph note

macrophages) 53,54.

Alveolar macrophages initially encounter materials, that reach gas exchange units and

subsequently perform phagocytosis and microbial killing. Interstitial macrophages serve as

precursors for physiologic renewal of alveolar macrophages and for their expansion during

pulmonary inflammation 53. Adhering to the pulmonary endothelial cells, intravascular

macrophages remove circulating particles, but may also contribute to pathogenesis during

sepsis 48,55. Airway macrophages are responsible for the reactivity of the conductive airways in

response to inhaled stimuli in patients with asthma 47.

The ultimate source of pulmonary macrophages are monocyte progenitor cells in the bone

marrow 56, which enter the lung from the vascular bed, and adapt to the local environment by

maturation into tissue macrophages. Maturation results in an increase in cell size, in the

number of cytoplasmatic organelles, lysosomal enzyme activity, phagocytosis capacity, and

expression of Fc, complement and cytokine receptors on surface membranes 54. Alveolar

Introduction

7

macrophages vary considerable in size and morphologic features. They represent both

phenotypically and functionally a heterogenous population of cells 57. Pulmonary macrophages

live for weeks to months and possess some limited replicative potential 58. It is believed that

the effector arm of the cell-mediated immune response depends on both the activation of

resident macrophages as well as on the recruitment of circulating monocytes into the alveoli 58. However, the proportion of monocyte influx and local replication during the pulmonary

inflammation has to be defined.

Chemotactic factors like complement component C5a, transforming growth factor beta (TGF-

β), formylated peptides and monocyte-chemotactic peptide (MCP-1), stimulate the influx of

monocytes into the lung (reviewed in 50). To exert their effector functions, alveolar

macrophages have to be either activated non-specifically by adherence 59 or usually through

ligand-to-receptor coupling. Among a large number of receptors expressed on alveolar

macrophages 57, immunoglobulin (Ig) and complement (C) receptors (CR1, CR3) participate in

receptor-mediated phagocytosis. The former bind to the Fc portion of Ig molecules (IgG, IgA,

IgE), whereas the latter are important natural opsonins for microbial organisms and

particulates 60. Macrophages further express C5a, a potent chemotactic, cell-activating and

pro-inflammatory molecule, as well as various cytokine receptors for interleukin-1 (IL-1), tumor

necrosis factor (TNF), interferon γ (INF-γ), growth factor receptors and cell activation factors

e.g. lectins, lipoproteins and glucocorticoids (reviewed in 50,57).

CD14, a 55-kDa phosphatidylinositiol-anchored membrane glycoprotein expressed on the

surface of monocytes, macrophages, and, to a lesser extent on neutrophils 61 has been

identified as an important receptor for LPS, in association with LPS-binding protein (LBP). The

stimulation of CD14 triggers a concentration-dependent macrophage activation 62. This

activation results in the release of TNF, IL-6 and IL-8 63,64 and has been described as both

LBP-dependent and -independent 63. Haugen and coworkers reported a higher expression of

CD14 on monocytes (MO) than on AM, with AM of different size and maturity having different

CD14 expression levels. The expression of CD14 can be modulated in response to LPS, but

the LPS binding capacity of AM and MO does not correlate with their CD14 levels 65. Indeed,

several other macrophage receptors that function as LPS receptors have been described 66.

1.3.1.3 Secretory function

Macrophages are pluripotent cells. The range of products secreted by alveolar macrophages

is broad, with over 100 molecular species having been identified, including reactive oxygen

species, proteases and anti-proteases, bioactive lipids, cytokines as well as growth factors

(reviewed in 50).

Introduction

8

Reactive oxygen radicals are generated via the “respiratory burst”. Together with several

proteases and other enzymes that function intracellularly, reactive oxygen species are

involved in microbial killing. Secreted proteinases (metalloproteinases, serine proteinases)

affect fibrinolysis and extracellular matrix remodelling 50. Stimulation of alveolar macrophages

with IgG or IgE immune-complexes leads to the production of both cyclooxygenase-derived

(thromboxane A2 and prostaglandin E2 and D2) and lipoxygenase-derived (leukotriene B4)

compounds 67. In the IgG immune complex animal model of acute lung injury 68,

bronchoalveolar lavage fluids contain large amounts of biologically active TNF and IL-1. These

“early response cytokines” function in a pro-inflammatory manner, by playing a key role in the

upregulation or induction of adhesion molecules in the lung vasculature, such as ICAM-1 and

E-selectin, which are critically necessary for neutrophil recruitment. In this regard, the α-

chemokine family (IL-8 or its family relatives such as MIP-2, CINC) are described to have

chemotactic activity primarily for neutrophils and to a lesser extent for T-cell subsets, while the

β-chemokines (e.g. MIP-1a, b; MCP-1, 2 and 3) are thought to be predominantly chemotactic

for monocytes and lymphocytes. In contrast, there are various anti-inflammatory cytokines

such as IL-10, IL-4, IL-6 or IL-13 produced by alveolar macrophages to suppress the cytokine

production and therefore regulate the intensity of the inflammatory response (reviewed in 69).

In addition to cytokines, also lipid mediators such as leukotrienes (LT) upregulate directly the

antimicrobial potential of macrophages as well as PMN’s from different species against both

bacteria and fungi (reviewed in 70). Leukotriene levels have been shown to be elevated in

bronchoalveolar lavage fluid from patients with bacterial pneumonia and also in lung tissue in

animal models of pneumonia 70.

1.3.1.4 Phagocytic function

Macrophages remove materials present in the local environment through both pinocytosis and

phagocytosis. In contrast to pinocytosis, phagocytosis is an energy-dependent process,

characterized by adherence through receptors and engulfment, leading further to

internalisation of the particulates and finally to digestion in phagolysosomal vacuoles. Efficient

phagocytosis depends on opsonization. Opsonins defined in the lung include IgG1, IgG3, C3b,

surfactant protein A/ D (SP-A/ D) 71,72, mannose and lipopolysaccharide-binding proteins 60.

Furthermore, SP-A has been reported to enhance the phagocytosis of C1q-coated particles by

alveolar macrophages 73. However, binding of inhaled environmental particles must be

accomplished without the benefit of opsonization by specific antibodies. The identities of

receptors on AM’s that mediate unopsonized particle binding are not fully known, whereas the

role of some scavenger receptors (SR) have been recently reported 66,74,75.

Introduction

9

1.3.1.5 Microbicidal function

The production of reactive oxygen species (ROS) upon receptor-mediated stimulation of

phagocytosis was first recognized in phagocytes such as macrophages, a process

determining their microbicidal activity and referred to as respiratory burst. This process results

from the assembly and activation of the nicotinamide dinucleotide phosphate (NADPH)

oxidase, a multicomponent enzyme that catalyses the one-electron reduction of molecular

oxygen to superoxide 76, which is subsequently converted to other oxygen-derived species, by

either spontaneous or enzyme-mediated reduction. ROS generated in an inflammatory milieu

act in an autocrine and paracrine manner, in order to rapidly amplify the effector potential of

Fcγ R on quiescent phagocytes, by means of altering signal transduction (reviewed by Pricop

and Salmon, 77). Controlled upregulation of oxygen generation is important, since

dysregulation or overproduction results in acute tissue injury 78,79. With regard to this, it has

been reported that during the respiratory burst of phagocytic cells in vitro, the superoxide

anion production per cell shows an inverse correlation with the cell density, a phenomenon

described as autoregulation 80. Furthermore, it has been shown that the decrease in individual

cell response is due to a significant increase in the amount of basal responses of the

macrophages, thus, concomitantly, the number of reactive cells remains unchanged,

irrespective of the cell density of the population 78.

In vitro, the respiratory burst can be triggered e.g. by certain soluble (PMA) or particulate

opsonized zymosan (OZ) agents. Both agents produce an intense oxidative burst, although

mediated by different signal transduction pathways: PMA is soluble and associated to cell

membrane perturbations that is dependent on protein kinase C (PKC) activation, whereas

particulate OZ reflects processes related to PKC-independent phagocytic mechanisms 81.

Moreover both stimuli utilize more than one activation pathway to stimulate NADPH-oxidase 82, revealing that different stimuli produce the same reaction in terms of respiratory burst

autoregulation at the single cell level. Together with the LPS-induced oxidative burst (CD14

receptor-dependent) these pathways depend on ERK 1/ 2 activation. Activated ERK kinases

also control the production of TNF in both LPS and PMA membrane activation of human

alveolar macrophages. Whereas LPS activates NFκB nuclear translocation, PMA does not,

but rather activates alveolar macrophages through the ERK 1/ 2 MAP kinase pathway. It has

been shown that IgA, a predominant Ig isotype of respiratory secretions, regulates LPS-and

PMA-induced oxidative burst and TNF through both dependent and independent modulation

of ERK pathways, revealing a role of IgA in both lung protection and inflammation 83. The LPS-

related oxidative burst, TNF and IL-10 release of human monocytes can also be modulated by

IL-9, through an upregulation of TGF-β 84. In contrast, ADP stimulates the respiratory burst,

without activation of ERK in rat alveolar macrophages 85.

Introduction

10

The physiological generation of ROS production has now been clearly implicated in the

activation of signalling pathways, resulting in a broad array of physiological responses ranging

from cell proliferation to gene expression and apoptosis 79,86. In this regard, and as reviewed

by Forman and Torres, 79, it has been suggested that: 1. hydrogen peroxide and superoxide

act as second messengers; 2. anti-oxidant enzymes are implicated in the “turn-off” phase of

signal transduction; 3. the primary physiological role of the respiratory burst in macrophages

may be redox signalling, rather than microbicidal activity.

In contrast to neutrophils, oxygen-dependent killing by macrophages is myeloperoxidase–

independent because they lose this enzyme during maturation 47.

In macrophages from rodents, a second oxidant-generating pathway features inducible nitric

oxide synthase (iNOS), that is inducible by TNF, IL-1, and interferon-γ 69.

1.3.1.6 Immune function

Macrophages contribute importantly to the induction, expression and regulation of immune

responses 47. Alveolar macrophages possess the capacity to serve as antigen presenting cells

(APC’s) and secrete co-stimulatory cytokines, both required for T-cell activation. In contrast to

resting macrophages, alveolar macrophages activated by chronic pulmonary inflammation or

by T-lymphocyte-derived cytokines are more effective in antigen presentation 48. Furthermore,

pulmonary macrophages are important effector cells for T-cell mediated immunity and regulate

immune responses. Thus IFN-γ, secreted by activated T-lymphocytes potently activates

resting macrophages and upregulates these host defence functions 87. The balance between

macrophage-derived enhancing and -suppressive signals determines the degree with which

antigen presentation and induction of immune responses occur. Therefore, not only T-cells,

but also dendritic cells (DCs), another antigen-presenting cell, are maintained in a down-

modulated state in the airways, tightly controlled by high numbers of nearby macrophages 47.

Immuno-modulatory factors involved in this process include among others: AM-derived nitric

oxide (NO) 88, prostaglandin E2 (PGE2) 89,90, TGFβ 91 and IL-10 92.

Introduction

11

1.4 Multiple trauma

1.4.1 Impact of trauma on patients and society

Trauma is generally the third leading cause of death in the United States and specifically the

leading cause of death in men under 40 years of age 93, due to accidents and military causes.

Speed limits combined with improvements in seatbelts, airbags, and vehicle constructions

lead to a decline in the number of accidental deaths, but also result in an increase in the

number of accidental victims suffering from blunt chest trauma. Trauma further represents the

second leading cause of life-threatening acute lung injury, i.e. in its severe form the Adult

Respiratory Distress Syndrome (ARDS) 41,94. Prevention and intervention are thus both a

socio-ethical and economic necessity. After initial, non-fatal injury thoracic trauma is

considered to be an important clue to single or multi-organ failure, itself making the patient

more susceptible to infections 95-97. With exception of changing the ventilation strategy 98, none

of the numerous interventions made it into clinical practice. This emphasizes the need for

further basic understanding of both pathogenesis and consequences of intervention.

1.4.2 Blast injury

Blast injuries are a special form of blunt trauma, with serious internal injuries, often without

evidence of external lesions. Blast injury occurs following a sudden change in environmental

pressure, originating from an explosion. The process of the explosion is equivalent to the

energy release which is then transmitted as a radial pressure wave, called shock-wave, from

the source into the surroundings 99,100. After a short travelling distance, a new shock front with

lower pressure peak and initial velocity is formed in the air, called the blast wave. It travels

supersonically and exerts its effects at comparably long distances from the explosion. The

simple blast wave, described as the Friedlander wave form, rises very rapidly, then decays

slowly (the vacuum phase) and may even drop below the previous ambient pressure. If the

blast is confined by reflective surfaces, then a complex wave-form is formed. Complex blast

waves contain multiple overpressure peaks 101. These wave-forms are usually expressed by

their impulse (integral of pressure changes over time, P/ dt) rather than by their maximum

peak overpressure value alone. Several factors can alter the response to the blast, such as

peak overpressure, duration of the overpressure, impulse (in case of complex wave forms),

the surrounding medium (e.g. water is a dense medium), and the distance from the explosion,

because the effect is inversely proportional to the distance and the body orientation towards

the blast wave front (reviewed by Elsayed, 102). Blast injuries are divided into 4 groups 103,104:

1. primary blast injury due to the blast wave, the effect of which is greatest at gas-liquid

Introduction

12

interfaces, 2. secondary blast injury resulting from flying debris, 3. tertiary blast injury due to

victims being thrown against an object, and 4. miscellaneous blast injuries including exposure

to dust, thermal burns or fire by the blast.

The pressure wave of the primary blast injury, encountering media of different densities in the

body, is reflected, causing turbulence and cavitation. The phenomenon termed “spalling”

describes the process where fluid is thrown from a more dense to less dense medium.

Another effect is the “secondary” explosion or implosion, as a consequence of compression

and re-expansion of gas pockets in the organs by the traversing blast wave. As the blast wave

reaches the surface of the victim’s body, a pressure difference develops between the outer

surface of the body and the internal organs. In air-containing organs like the lung, the air in the

alveoli is easily compressible and thus rupture of the alveolar capillary membrane occurs, with

blood driven from the capillaries into the alveolar space, due to the pressure differences

created. Finally, inertia is the shear created when the pressure wave moves tissues of

different densities at different speeds 105,106.

1.4.3 Chest trauma

The organs most affected by blast injuries are the hollow, gas-filled organs such as the ears,

the lungs, the gastro-intestinal tract 103,104,107-110, and to a lesser extent the cardiovascular and

central nervous system 111. The lungs are almost always affected by blast injuries 103,107-109.

Postulated explanations are: 1. The positive pressure is transmitted through the airways into

the lung, causing rupture of the alveoli and the capillaries, 2. the longer lasting negative phase

produces the alveolar capillary disruption, and 3. direct compression of the lung.

1.4.3.1 Pacemaker function of thorax trauma

Thoracic trauma seems to be an important trigger of morbidity and mortality in polytraumatic

patients and may lead itself to severe lung injury 93. Indeed, the lethality of polytrauma patients

without thorax involvement is 4%, as compared to 23% with simultaneous thorax trauma 112.

Therefore, thorax injury is a major negative determinant for the long-term consequences after

trauma 104,113. One of the major problems for clinical evaluation of the early phase after trauma

is the missing of clinical signs of pulmonary dysfunction 114,115. This means that the severity of

the consecutive injury is unpredictable at this stage.

Introduction

13

1.4.3.2 Pathology of pulmonary blast injury

Pulmonary contusion and haemorrhage

Experimental lung injury from air blast 116,117 showed that the most consistent lesion was

bilateral traumatic haemorrhage, variable in extent and distribution. Furthermore, oedema is

usually prominent 116-118. Post-mortem data of fatal blast injury in humans is rare.

Ultrastructural examination of lungs 24 hours after blast exposure, reveals haemorrhagic

areas in the lung contralateral to the side of blast exposure which initially microscopically

appeared only slightly congested 118. Loss of structure of the alveolar epithelium and changes

to the type II pneumocytes 118 and thus loss of surfactant production may be important in the

development of respiratory insufficiency.

Haemopneumothorax and air embolism

Tension pneumothorax has been observed in rats exposed to air blast under laboratory

conditions 116. Air embolism, pulmonary contusion and haemorrhage account for the majority

of immediate and early death 104,110,119,120. Whether air emboli are important in survivors is

unclear, but early and persisting neurological deficits might be the result of this phenomenon 103. Whether air-embolism is caused by mechanical ventilation is still a matter of debate 121.

However, recently it has been shown that air embolism, present in blast victims, is not a mere

ventilation-induced artefact 119.

1.4.3.3 Acute physiologic responses to primary blast

The reported physiologic responses in blast-injured patients vary considerably, largely as a

result of varying observation times and the influence of other injuries. The most consistent

physiological data is derived from animal experiments from which common cardiorespiratory

and haemodynamic changes can be determined.

Pulmonary response

Moderate to severe blast exposure in animals is usually followed by an immediate, variable

period of apnoea, lasting from a few seconds to up to more than one minute 122. This apneic

period is followed by a fast and shallow breathing for a variable time before returning to pre-

blast values 122. Blast-induced effects on gas-exchange and ventilatory function are variable

throughout different animal experiments. The reduction of arterial blood oxygen tension

(PaO2) that returns gradually to control levels, causes transient hypoxemia of varying degree

Introduction

14

of severity. Carbon dioxide tension (PaCO2) however is not affected and has been related to a

ventilation-perfusion mismatch, due to a massive pulmonary haemorrhage 116.

Cardiovascular response

The most consistent finding of experimental blast injury on the heart rate appears to be an

immediate bradycardia, which is more severe with higher intensity blast waves 116. Bomb blast

survivors are occasionally found in profound shock and hypoxia without external signs of injury 115. An immediate fall in the mean arterial pressure (MAP) to less than 50% has been shown in

various animal species following blast exposure 116. This response to blast wave injury is

sometimes bimodal 116,117 and graded with higher intensity blasts, resulting in lower MAP and

slower recovery. Recovery time to pre-blast values ranges from two hours 122 to up to several

days 123. Concomitant with the reduction in MAP, a fall in the cardiac index (CI) can be

observed. The CI response to injury, like MAP, is lower with exposure to higher intensity blast

pressure waves. Despite hypotension and low CI, no compensatory peripheral

vasoconstriction is detectable 116. This triad of apnoea, bradycardia, and hypotension 116,122 is

only seen in animals in which just the thorax is exposed to the blast, but not in animals

undergoing abdominal blast exposure 122. These changes are attributed to air embolism and

direct cardiac injury 120 and probably mediated by vagal reflexes 124,125. Experimental animal

studies 126 and clinical observations 127 following blunt chest trauma reveal a variety of EKG

disturbances, from ventricular extracystoles to ventricular fibrillation, that are usually

temporary but might account for some fatalities following blast injury and may exacerbate the

triad of apnoea, bradycardia, and hypotension.

Progressive pulmonary insufficiency (PPI)

Several different mechanisms, in addition to the pressure wave-induced disruption of the lung

architecture, are considered to be responsible for the various presentations of blast-induced

pulmonary insufficiency (“blast lung”) and delayed respiratory insufficiency. Respiratory failure

occurring 24 - 48 hours after blast exposure is unlikely to be caused solely by the primary

blast 103,128. The pulmonary results of injury, namely the combination of blast effects, inhalation

injury, tissue injury, and fluid resuscitation are frequently diagnosed as the Adult Respiratory

Distress Syndrome (ARDS) 41,94. The late mortality after trauma is related to multi-organ failure

(MOF), as a consequence of shock or sepsis 97. The pathophysiology of injury-induced organ

MOF is poorly characterized, but has been linked to systemic inflammation, as a result of

infection (either obvious or occult) or massive tissue injury (systemic inflammatory response

syndrome, SIRS). Post-traumatic immunosuppression, called the counterregulatory anti-

Introduction

15

inflammatory response syndrome (CARS), also contribute to the development of MOF

(reviewed by Cobb and coworkers, 129), 95,130.

1.4.4 Molecular injury

Recent information indicates that there is a complex cellular and molecular generic response

to injury, leading to multi-organ failure.

First of all, there is the direct tissue injury leading to interstitial or alveolar haemorrhage and

oedema formation. The intensity of pulmonary oedema was positively correlated with the

length of the survival time 119. As a result, hypoxemia and pulmonary air embolism may occur 120. Such events can also cause oxidative stress in the lung, which is characterized by an

antioxidant depletion (water soluble and lipid soluble), an increase in lipid peroxidation, an

increased methaemoglobin (metHb) content 131, and an inhibition of ATP-dependent Ca2+

transport 132. Supplementation with vitamin E and lipoic acid, but not vitamin C, increases the

Hb oxygenation 133, however the pro-oxidant action of water soluble antioxidants via redox

cycling of oxyHb and metHb may promote oxidative stress rather then prevent it 102,131.

Due to tissue trauma itself, as well as due to blood components and cell debris, an activation

of host defence mechanisms occurs, including soluble plasma factors (complement and

clotting cascades) and immunocompetent cellular components (neutrophils, monocytes,

macrophages, and endothelial cells). Such activated cells in turn produce potentially toxic host

mediators, such as tumor necrosis factor (TNF) and interleukin-1 (IL-1), as well as chemical

species such as kinins, eicosanoids, platelet-activating factor (PAF), and nitric oxide. This

sequential increase of the inflammatory response may variably lead to shock, multiple organ

failure and death. Eicosanoids have been shown to play an important role in the pathogenesis

and modulation of pulmonary oedema associated with blast injury. Significantly elevated

plasma levels of the prostanoids, thromboxane A2 (TXA2), prostacycline (PGF1a), PGF2a and

PGE2 are reported for polytrauma patients with thorax involvement 134,135 and after surgery

induced lung tissue injury 136, implying a relative prognostic value of these early mediators with

regard to the intensity and the pattern of the injury. Concomitantly elevated plasma levels of

type II a phospholipase A2 (PLA2) are suggested to bear some prognostic value 137,138.

Unchanged levels of prostaglandin M (PGM) indicate an unaltered pulmonary metabolic

capacity 136. Pharmacological inhibition of 5-lipoxygenase, immediately prior to blast exposure,

reduces oedema formation, accumulation of neutrophils and generation of lipid peroxidation

products in injured rabbit lungs 139 and permits respiratory compensation of metabolic acidosis

in the general circulation, in spite of increased hypotension and acidosis in the venous

circulation 140. Neither experimental haemorrhage alone, nor haemorrhage in combination with

blunt trauma increases circulating TNF in pigs, implying no involvement in the vasomotor

Introduction

16

collapse observed in this setting 141. Unlike TNF and IL-1, increased IL-6 levels are commonly

detected during disease 135,142, thus IL-6 serum concentrations received more attention as a

prognostic indicator for both severity and outcome.

1.4.5 Management of pulmonary blast injury

Success or failure of the treatment may depend on the judicious use of resuscitative fluids and

respiratory support 128. In trauma victims, increased interstitial and intra-alveolar oedema

impairs gas exchange. The initial ventilator settings in the trauma victim are somewhat

different from those used in other patients, because the decreased functional residual

capacity, due to oedema formation requires a lower tidal volume. Thus, modern ventilator

strategies include a low tidal volume with a positive end-expiratory pressure (PEEP) and high

inspired oxygen concentrations, in order to recruit alveoli and thus avoid atelectasis. However,

the beneficial effects of supplementary oxygen and mechanical ventilation remain

controversial and depend on the extent of the respiratory insufficiency 143. Concomitant

respiratory acidosis is thereby compensated by therapeutic bicarbonate application. Together

with pharmacological manipulations that increase the cardiac output, the function of both the

heart and the lung, as well as other organs, can be augmented. The outcome of

pharmacological interventions manipulating the reflex mechanisms and therefore altering

haemodynamic functions remains unclear 128.

Various attempts have been done to interrupt the interaction of the neutrophils, endothelial

cells, cytokines, and free radicals 133,144. Despite limited clinical and experimental success, this

approach may prove to be worthwhile in the future for treating the severely injured patient 145.

1.4.6 Immunomodulation after trauma

Besides clinical characterization, recent research data indicate that the prognosis of trauma

patients is strongly associated with a post-traumatic imbalance of the immunological system 146. For the outcome of patients with multiple injuries, it seems very important to gain an

inflammatory mediator homeostasis as fast as possible. Overwhelming immune activation,

however, can result in a systemic inflammatory response syndrome (SIRS) and septic shock.

To control the potentially harmful pro-inflammatory response, the immune system, as a

mechanism of counterregulation, releases anti-inflammatory mediators inducing the so-called

anti-inflammatory response syndrome (CARS). In addition to the autoregulatory pathways of

the immune cells, the systemic immune response is controlled by two neuroimmune

pathways: 1. the hypothalamic-pituitary-adrenal (HPA) axis and 2. the sympathic nerve system

leading to glucocorticoid and catecholamine release, respectively, that affect primarily

Introduction

17

monocytes and macrophages 130. However, if the anti-inflammatory mediators predominate,

immunosuppression (“immunoparalysis”) with ineffective eradication of microorganisms and

septic complications may follow 147. In this context, the early presence of soluble TNF

receptors (sTNFRs) 142 and interleukin 1 receptor antagonist (IL-1ra) in the circulation of

trauma patients has been reported, even with a strong correlation to mortality 148,149. Traumatic

injuries cause the liberation of various mediators such as IL-6, IL-10, and PMN elastase, with

more or less an association between anatomic injury and the pattern of mediator release 135,142. For IL-6, an anti-inflammatory role in controlling the levels of other pro-inflammatory

cytokines 150 and by inducing the release/ formation of sTNFRs has been suggested 151. In

addition, PMN elastase is thought to increase the sTNFR release by proteolytic mechanisms 152. Since PGE2 is significantly elevated in the plasma of injured patients and since the

immune cells (primarily the macrophages) from traumatized and burn patients hypersecrete

PGE2 upon stimulation 153, this potent immunosuppressive agent is regarded to be a pivotal

mediator in the post-trauma immune dyshomeostasis 154. Not only increased synthesis of

PGE2, but also decreased removal and degradation have been reported in burn injury and

trauma (reviewed in 89). Furthermore, it has been shown that trauma results in delayed

macrophage hypersecretion of inflammatory mediators, such as TNF, IL-6 and end products

of the respiratory burst, such as H2O2, that is associated with persistent functional

macrophage defects in antigen presentation 155. The paradoxical combination of suppressed

macrophage function and hypersecretion of inflammatory mediators, simultaneously renders

the host susceptible to both infectious complications and the immune-mediated sequelae of

SIRS and MOF. However, despite immunosuppression immediately after injury, a recovery

from an early impairment occurring 3 to 6 days after trauma is also reported 156.

Aims of the Study

18

2. Aims of the Study

The blast trauma model is based on a rodent model originally developed by Irwin et al. 1998, 117.

The objective of this study was to use and to improve this model, in order to reproduce the

clinical spectrum of injuries seen in blast victims, for studying the pathophysiology and

potential treatment approaches of thorax trauma in rats.

The first aim was to standardize the blast pressure wave and its application, with regard to

assessing characteristical physical blast properties, in order to study its biological correlate.

The second aim was to asses and to characterize the physiological and functional changes

after in vivo trauma by means of:

I. Assessing lung physiological parameters over time in order to characterize trauma-

induced changes, by combining our primary blast injury model with our ex vivo

isolated perfused rat lung model. These studies include the investigation of

pharmacological interventions, as to their therapeutic potential.

II. Assessing pro-and anti-inflammatory mediator release, oedema generation and

resorption, as well as pulmonary infiltrate formation in vivo over time.

The third aim represented the study aspects of local and general immunodeficiency after

thoracic trauma, that might clinically result in a higher number of infections. Due to their key

role as modulator of pulmonary inflammation, macrophage functions will be assessed in vitro,

as to their phagocytic capacity, respiratory burst and mediator release.

Materials and Methods

19

3. Materials and Methods

3.1 Animals

Female Wistar rats (230 ± 20 g) obtained from Harlan Winkelman GmbH (Borchen, Germany)

and kept under controlled conditions (24°C, 55% humidity, 12 hours day/ night rhythm) on a

standard laboratory chow and water ad libitum were used as lung donors for experiments with

isolated perfused rat lungs, for the isolation of primary alveolar macrophages and for the in

vivo experiments.

3.2 Chemicals

3.2.1 Chemicals used in the isolated perfused rat lung experiments

Substance Producer Solute in Concentration Storage

Amiloride Sigma-Aldrich, Deisenhofen, Germany

DMSO *, NaCl 10-4 M freshly prepared

Bovine serum albumin fraction V, receptor-grade; Lot No. 12353

Serva, Heidelberg, Germany

buffer 2% freshly prepared

Formoterol AstraZeneca, Zug, Switzerland

saline, 0.05% Na2S2O5

1 nM freshly prepared

Propranolol Sigma-Aldrich, Deisenhofen, Germany

saline 10-4 M freshly prepared

Terbutaline Sigma-Aldrich, Deisenhofen, Germany

NaCl 10-4 M stock at -20°C

Table 3.1 Substances used in the isolated perfused rat lung experiments. * 0.03% DMSO.


20

3.2.2 Chemicals used in the in vitro experiments

Substance Producer Solute in Concentration Storage

Escherichia coli (K-12 strain) tetramethylrhodamine conjugate

Molecular Probes, Eugene, OR, USA

PBS 250 µg/ ml 20 mg/ ml stock at 4°C

Liquemin Hoffmann-La Roche, Grenzach-Wyhlen, Germany

NaCl (1:10) 8 Units/ ml Heparin

4°C

LPS from Salmonella abortus equi

Metalon, Wusterhausen, Germany

PBS 100 µg/ ml – 10 ng/ ml

1 mg/ ml stock 4°C

LTA from Staphylococcus aureus

LS Wendel, University of Konstanz, Germany

PBS 100 µg/ ml – 2 µg/ ml

5 mg/ ml stock -70°C

SOD (superoxide dismutase)

Sigma-Aldrich, Deisenhofen, Germany

HBSS * 1000 U/ ml freshly prepared

Staphylococcus aureus (wood strain without protein A), tetramethylrhodamine conjugate

Molecular Probes, Eugene, OR, USA

PBS 250 µg/ ml 10 mg/ ml stock at 4°C

WST-1 (C19H11IN5NaO8S2, tetrazolium salt)

Probior, München, Germany

HBSS * 500 µM freshly prepared

Table 3.2 Substances used in vitro. * Hanks balanced salt solution.

3.2.3 Other chemicals

Brij, Tween 20, 0.4%Trypan blue, 3,3´,5,5´-tetramethyl-benzidine liquid substrate solution,

LPS from Salmonella abortus equi, sodium dodecylsulfate (Sigma-Aldrich Chemie GmbH, Deisenhofen, Germany); HEPES (ICN Biomedicals, Ohio, USA); OptEIATM, biotin labelled

polyclonal anti mouse/ rat TNFα (Pharmingen, Hamburg, Germany); Hoechst 33342, Sytox

green (Molecular Probes Europe BV, Leiden, Netherlands); rabbit anti-TNFα- serum, K533

(LS Wendel, Biochemical Pharmacology, University of Konstanz, Germany); Streptavidin-

peroxidase (Jackson Immuno Research, West Grove, PA, USA); MMP-2, matrix

metalloproteinase 2, human rheumatoid synovial fibroblast, MMP-9, matrix metalloproteinase

9, human neutrophil granulocyte, Eosine-methylene blue, Azur-eosine-methylene blue (Merck

KGaA, Darmstadt, Germany); Coomassi brilliant blue R250 (Serva, Heidelberg, Germany);


21

Temed (Biorad, München, Germany); APS (Amersham, Little Chalfont, UK); Pierce BCA

protein assay reagent (Bender & Hobein GmbH, Heidelberg, Germany); ELISA/ EIA: IL-6, IL-

10 (Biosource Europe SA, Nivelles, Belgium); TXB2, 6-keto-PGF1α (BioTrend GmbH, Köln,

Germany); CINC-3, mouse sTNFR1, sTNFR2 (R&D Systems, Wiesbaden-Nordenstadt,

Germany). All standard chemicals were purchased from established companies, primarily

Sigma-Aldrich Chemie GmbH (Deisenhofen, Germany).

3.2.4 Solutes

Distilled water; Dimethylsulfoxid, DMSO (Riedel de-Haen, Seelze, Germany); Ethanol; NaCl

0.9% (Delta-Pharma Boehringer-Ingelheim, Germany); Phosphate-buffered saline solution,

PBS (PAA, Germany); Triton X-100 (Sigma-Aldrich Chemie GmbH, Deisenhofen, Germany);

EDTA, ethylendiamine tetraacetic acid (Sigma-Aldrich Chemie GmbH, Deisenhofen,

Germany).

3.2.5 Anaesthesia and analgesic

• Halothane, 2-bromo-2-chloro-1, 1, 1-trifluoroethane (Sigma-Aldrich Chemie GmbH,

Deisenhofen, Germany)

• Pentobarbital sodium (Narcoren), (Wirtschaftsgenossenschaft deutscher Tierärzte,

Hannover, Germany)

• Buprenorphin (Temgesic), (Tierforschungsanlage, Universität Konstanz, Germany)

3.2.6 Cell culture material

RPMI 1640: with L-Glutamine, with/ without phenol red (PAA, Pasching, Austria); Hanks

balanced salt solution (PAA, Pasching, Austria); Penicillin-Streptomycin (Gibco BRL Life

Technologies, Eggenstein, Germany); Fetal calf serum (FCS) (Boehringer Mannheim GmbH,

Mannheim, Germany); Cell culture plates (Greiner, Frickenhausen, Germany)

3.3 Laboratory equipment and technical devices

Centrifuges: Eppendorf 5417R (Netheler & Hinz, Hamburg, Germany); Minifuge RF (Heraeus

Instruments, Hanau, Germany); ELISA reader: SLT Spektra Rainbow Photometer (SLT

Labinstruments, Grailsheim, Germany); Microplate Fluorescence Reader FL 600 (Deelux

Labortechnik, Gödenstorf, Germany); Isolated perfused rat lung (Harvard Apparatus, March-

Hugstetten, Germany); Osmometer 3MO (Advanced Instruments, Needham Heights, MA,


22

USA); Eppendorf ACP 5040 (Netheler & Hinz, Hamburg, Germany); Microscope: Zeiss

Axiovert 35 (Zeiss, Oberkochen, Germany)

3.4 The in vivo blast wave thorax trauma in rats

3.4.1 The blast wave generator apparatus

The blast wave generator was originally designed and constructed by Irwin and colleagues in

1998 117 and optimized by Liener and coworkers to discharge a global blast pressure wave in

a safe laboratory environment. We have modified this setting to our specific needs in order to

improve the pressure wave monitoring and the reproducibility of pressure wave exposure. The

blast wave generator (Wissenschaftliche Werkstätten, Universität Konstanz) consists of three

parts: a pressure reservoir, a blast nozzle, and a 190 A µm Mylar polyester film (Strohmeier,

Rheda-Wiedenbrück, Germany) that separates the two compartments. Compressed air from

the storage tank was slowly delivered by a pressure reducer to fill up a 0.75 l compressed air

bottle, that is located close to the pressure reservoir of the generator. When a working

pressure of 18 bar was reached the air supply was stopped. The valve (opening time: ≈ 25

ms) (HEE-D-MINI-24, Festo AG & CO, Esslingen, Germany) between the compressed air

bottle and the pressure reservoir was triggered by the computer. Opening of the valve resulted

in a fast pressure build-up in the reservoir. As soon as the burst strength of the Mylar

diaphragm was reached, a blast pressure wave was discharged towards the rat by the nozzle.

Figure 3.1 shows the blast wave generator.

Figure 3.1: Schematic diagram of the blast wave generator (front view). Dimension is in millimetre (mm).


23

3.4.2 Pressure wave monitoring

Different sensors at varying positions offered an exact monitoring of the exposed blast

pressure wave (Erne B., Abteilung Elektronik, Wissenschaftliche Werkstätten, Universität

Konstanz):

1. A pressure transducer (PR-6ST-80400.XX-20, sensitivity: 39.7 mV/ bar, excitation: 4 mA,

natural frequency: < 30 KHz; Keller AG, Winterthur, Switzerland) in the pressure reservoir

determined the cracking pressure meaning the rupture of the polyester film. The measured

time course of the pressure wave curve indicated the velocity of the pressure increase

followed by the decay. The fast pressure build-up and pressure decline is a requirement for

explosively discharging of the pressure wave (Fig. 3.2).

2. Two pressure transducers (2 MI PAA 110-050-020, sensitivity: 8.03/ 7.99 mV/ bar,

excitation: 1 mA, natural frequency: > 400 KHz; Keller AG, Winterthur, Switzerland) on both

sides of the rat measure the pressure peaks on rat thorax level (Fig. 3.2).

3. Another transducer (Halleffekt-IC 634-SS2, sensitivity: 7.5 – 10.6 mV/ mT; RS Components

GmbH, Mörfelden-Walldorf, Germany) records the breathing frequency of the animal.

Therefore a string connected with the breathing sensor was placed above the thoracic cage.

Via this transducer the valve was triggered automatically depending on the breathing situation

(e.g. inspiration, expiration or resting expiratory position) (Fig. 3.2). In addition the breathing

frequency of the anaesthetized animals, before and after the pressure wave exposure was

recorded.

All the measured pressure wave signals by the sensors were transmitted through a high

accuracy measuring instrumentation amplifier Burr-Brown (Texas Instruments GmbH,

Freising, Germany) either direct to the computer (cracking pressure transducer) via an A/ D-

converter (MAX 186, Maxim Corporate Headquarters, Sunnyvale, CA, USA) or first (pressure

peak transducer) to the HAMEG oscilloscope HM 407 (HAMEG instruments, distributor

VELMA, Großkrotzenburg, Germany). Then pressure wave data were analysed by a special

developed software (Heine G., Abteilung Elektronik, Wissenschaftliche Werkstätten,

Universität Konstanz) for cracking pressure of the polyester film (Rp), wave form, pressure

peaks (Pp(r), Pp(l)), impulse or area under the curve (AUC(r), AUC(l)), duration (t(r), t(l)) and

pre-and post breathing frequency (Fpre, Fpost). For calculation of these parameters the

following equations were used:


24

pressure peaks (Pp(r), Pp(l)): 12

2

1

tt

UP

mi

mii

−=

∑=

= (U = voltage; t = time), whereas the voltage U is

proportional to the pressure P (cracking pressure transducer: 0.5 bar = 1 bar, pressure peak

transducer: 0.1 bar = 1 bar);

area under the curve (AUC(r), AUC(l)): ∑=

=

=2

1

ti

tiiUAUC ,

and pre-and post breathing frequency (Fpre, Fpost): breath/ min.

As an index for variations between different experiment the variation coefficient CV was

calculated using the equation: [ ] 100% ⋅=medianSDCV .

Figure 3.2: The blast wave generator including pressure transducers at different positions. A cracking pressure transducer in the pressure reservoir, two overpressure sensors on the left and right side of the rat and a transducer for recording the breathing frequency.

3.4.3 Protocol of anaesthesia and analgesic

The rats were anaesthetized in a bell jar flooded with a Halothane®/ O2 gas mixture through a

Halothane-evaporator (flow: 2 L/ min oxygen, 4% Halothane). By the time the breathing

frequency was slow enough the animals were weighed and placed on a special immobilized


25

pad (Wissenschaftliche Werkstätten, Universität Konstanz). The rats were positioned in a

supine position and the paws were fixed with tapes on the pad. Up to the pressure wave

exposure the animals were inhalation-anaesthetized through a anaesthesia mask (flow: 1.5 L/

min oxygen, 4% Halothane). In addition the animals received the analgesic Buprenorphin

(0.03 mg/ kg BW) (Temgesic, Tierforschungsanlage, Universität Konstanz) subcutan. After

the blast exposure the animals were provided with oxygen (flow: 4 - 5 L/ min) to get them out

from the anaesthesia. The rats were terminally anaesthetised by an intraperitoneal injection of

sodium pentobarbital (Narcoren®, 160 mg/ kg BW).

3.4.4 Experimental protocol

After positioning and fixing the rat in the supine position and starting the inhalation-

anaesthesia through the anaesthesia mask, the string connected with the breathing

transducer was placed above the thoracic cage detecting the breathing frequency. Triggered

by the breathing sensor the rats were subjected to the pressure wave at the top of expiration.

The blast pressure wave was centered on the xiphoid with the tip of the blast nozzle located

either 3.5, 5, 7 cm away from the sternum. The blast animals were provided with oxygen (flow:

4 - 5 L/ min) to get out from the anaesthesia afterwards and the reset of the breathing was

recorded for one minute via the breathing sensor. For studies immediately after blast exposure

the animals were directly, terminally anaesthetised by an intraperitoneal injection of sodium

pentobarbital (Narcoren®, 160 mg/ kg BW). In case of long-time studies the animals were

sacrificed at 1.5, 3, 6, 12, 24 and 96 hours following the blast respectively. If necessary these

animals got a second injection of the analgesic Buprenorphin (0.03 mg/ kg BW) (Temgesic,

Tierforschungsanlage, Universität Konstanz) subcutan. Control animals were subjected to the

same experimental protocol but did not receive a blast.

Subsequently either the bronchoalveolar lavage (BAL) was performed (3.7) to determine total

and differential BAL cell counts (cytospin) (3.9), total protein content (3.10), erythrocyte counts

(3.8) and the release of different mediators and enzymes (3.11 - 3.16) or the wet/ dry ratio

(3.17) of the lung was assessed.

Figure 3.3 shows the timescale of the in vivo experiments.


26

Figure 3.3: Timescale of the in vivo experiments.

3.4.5 Determination of an injury score and validation

3.4.5.1 Haemorrhage score

The different lung lobes were each examined macroscopically for areas of pleural and

subpleural bleeding using an injury score:

0 = no haemorrhage,

1 = < 30% haemorrhage,

2 = 30 – 60% haemorrhage,

3 = > 60% haemorrhage

the total score resulted from the summation of the scores of each lobe whereas the score of

the upper left lobe (≈ double size) was duplicated. The inter-observer agreement was 93%.

3.4.5.2 Oedema score

Analogous to the haemorrhage score an oedema score was defined depending on the amount

of macroscopically visible oedematous areas. The same score scale (3.4.5.1) was used. The

inter-observer agreement was also > 90%.

in vivo

in vivo• thorax trauma (3.5 cm distance)• sham controls

~ + 10 min ~ + 6 h~ + 3 h

lung wet/ dry ratio

~ + 1.5 h ~ + 12 h ~ + 24 h ~ + 96 h

• BAL: - mediator measurement- cytospin- protein content- erythrocyte counts - LDH- MMP-2, -9

histological examination

in vivo

in vivo• thorax trauma (3.5 cm distance)• sham controls

~ + 10 min ~ + 6 h~ + 3 h

lung wet/ dry ratio

~ + 1.5 h ~ + 12 h ~ + 24 h ~ + 96 h

• BAL: - mediator measurement- cytospin- protein content- erythrocyte counts - LDH- MMP-2, -9

histological examination


27

3.5 The isolated ex vivo perfused and ventilated rat lung

3.5.1 Isolated perfused rat lung preparation

The lungs were perfused and prepared essentially as described previously 1. Briefly, female

Wistar rats were anaesthetized by intraperitoneal injection of 160 mg/ kg pentobarbital sodium

(Narcoren®). The positive pressure ventilation was performed through a tracheal cannula.

After exsanguination, two catheters were placed in the pulmonary artery and the left atrium

and then the non-recirculating perfusion was started to remove the blood from the pulmonary

circulation. The heart-lung block was excised and put in the artificial thorax chamber (38°C). In

order to obtain a correct continuous assessment of the lung weight via the weight transducer

integrated into the chamber lid 157, it was checked that the lung is suspended without contact

between catheters and water-jacked glass chamber wall.

3.5.2 Experimental Setup

The apparatus and technical equipment was purchased from Hugo Sachs Electronics-

Harvard Apparatus GmbH, Germany. The Setup is almost the same as described previously 1.

Excised lungs are perfused at constant hydrostatic pressure (14 cm H2O) in a recirculating

fashion, with a total volume of about 100 ml. For perfusion a Krebs-Henseleit buffer (305 - 315

mosm, 37°C) containing 2% bovine serum albumin (BSA), 0.1% glucose and 0.3% HEPES

was used. The pH value of the perfusate before entering the lung was kept at 7.35 by

automatic bubbling of the buffer with carbon dioxide as soon as the pH exceeded this value.

The perfusate flow (NarcomaticTM RT 500, Narco Biosystems, Texas, USA), the arterial

pressure PA, the venous pressure PV (both Isotec™ transducer, Quest Medical, Dallas, USA)

and pH (WTW pMX 3000) were continuously monitored.

The suspended lungs were ventilated by the trachea with a rotary vane compressor pump

(ventilation control module VCM) with negative pressure and ambient air. Breathing frequency

was 80 breaths/ min. The pulmonary end-expiratory (PEEP), –inspiratory (PIP) pressures

were set to 2 and 7 cm H2O reaching a tidal volume between 1.8 – 2.2 ml in untreated control

lungs. To avoid the generation of atelectasis as well as to recruit collapsed lung areas an

hyperinflation of -16 cm H2O was triggered automatically every 5 minutes (timer counter

module, TCM). Pulmonary ventilation pressure was measured with a differential pressure

transducer (Validyne DP 45-24, Northridge, Ca, USA) and air flow velocity was measured with

a pneumotachograph connected to a differential pressure transducer (Validyne DP 45 - 14,

Northridge, Ca, USA).


28

All data were transmitted to a computer and analysed by a special software (Pulmodyn

software): pulmonary pressure (P), tidal volume (Tv), airway resistance (RA), dynamic lung

compliance (CL), perfusate flow (Q), pulmonary arterial pressure (PA), pulmonary vein

pressure (PV), vascular resistance (RV), perfusate temperature, perfusate pH and lung weight

(G). For calculation of lung mechanics the following equation was used:

LL

RTvC

P +⋅

= 1

⋅dtdTv ; t = time.

Vascular resistance RV was calculated from the equation: ( )QPP vA −

.

In order to obtain a minimal vital ventilation of 1ml in injured lungs, it was necessary to

increase the standard ventilation pressure, depending on the extent (3.4.4) of blast injury.

The Setup of the isolated perfused rat lung is shown in Figure 3.4.

Figure 3.4: The Setup of the isolated perfused rat lung. The isolated rat lung is perfused in a recirculating fashion, blood-free under constant pressure conditions and with a negative pressure ventilation.

3.5.3 Experimental design of the perfused lung studies

After the lungs were fixed in the artificial thorax chamber, the ventilation and perfusion

parameters were controlled and adjusted. If procurable the injured lungs were perfused over

150 minutes like the untreated control lungs. The experimental design of the ex vivo studies is

shown in Figure 3.5.


29

Figure 3.5: Experimental design of the ex vivo studies.

3.6 In vitro stimulation of primary rat alveolar macrophages

3.6.1 Preparation and culturing of rat alveolar macrophages

Anaesthetised rats were exposed to the blast pressure wave as already described previously

(3.4). The animals were sacrificed 10 minutes (min), 24 and 96 hours following this

experimental induced blunt chest trauma, non-traumatized animals served as sham controls.

The alveolar macrophages were harvested by bronchoalveolar lavage (BAL). Therefore lungs

were tracheotomized and five times rinsed with 10 ml cold NaCl: The latter three times while

softly massaging the thoracic cage to increase the amount of the isolated BAL cells. Recovery

of the NaCl was always > 75%. The lavage fluids were pooled and centrifuged (4°C) for 10

min at 270 x g, and the pellet was resuspended in 20 - 50 ml cold RPMI 1640 and centrifuged

for another 5 min (190 x g, 4°C). The pellet was again resuspended in RPMI 1640 with phenol

red, 10% heat-inactivated FCS and 1% penicillin/ streptomycin. BAL cells were cultured at an

initial density of 5 x 105/ ml (except 3.6.3/ 3.11) on 96-well plates in 200 µl/ well RPMI 1640

(phenol red, 10% FCS, 1% penicillin/ streptomycin) for two hours at 37°C. The non-adherent

in vivo ex vivo

end of perfusion

in vivo• thorax trauma (3.5, 5, 7 cm distance)• sham controls

rat lung preparation

~ - 15 min~ - 6 h ~ - 3 h

start of perfusion

0 min 4 min

infusion of:• terbutaline (10-4 M),• formoterol (1 nM),• amiloride (10-4 M),• propranolol (10-4 M)into the perfusate.

150 min

• lung wet/ dry ratio• BAL

~ - 10 min

in vivo ex vivo

end of perfusion

in vivo• thorax trauma (3.5, 5, 7 cm distance)• sham controls

rat lung preparation

~ - 15 min~ - 6 h ~ - 3 h

start of perfusion

0 min 4 min

infusion of:• terbutaline (10-4 M),• formoterol (1 nM),• amiloride (10-4 M),• propranolol (10-4 M)into the perfusate.

150 min

• lung wet/ dry ratio• BAL

~ - 10 min


30

BAL cells were removed by washing the cells three times with fresh medium. The remaining

adherent alveolar macrophages were cultured in 200 µl/ well RPMI 1640 (without phenol red,

10% FCS, 1% penicillin/ streptomycin) either non-stimulated, with endotoxin or whole, dead

bacteria (3.6.2, 3.6.3). Depending on the isolated BAL-cell counts all experiments were run at

least in duplicate.

3.6.2 Stimulation with dead, fluorescent bacteria

The experiments were started by incubating the alveolar macrophages with either

fluorescence labelled Staphylococcus aureus (S. aureus) or Escherichia coli (E. coli)

(Molecular Probes) in a final concentration of 250 µg/ ml. Both reagents were stored as

concentrated stocks at 4°C and diluted with medium after sonicating (2 min) immediately

before use. After three hours incubation at 37°C the cells and supernatant were either

harvested and analysed (3.11) or subjected to different assays as described below (3.6.4,

3.6.5).

3.6.3 Stimulation with endotoxin

The cells were cultured at an initial density of 106 cells/ ml (100 µl/ well) (3.11) or 5 x 105/ ml

(200 µl/ well) (3.6.5) in the presence of various concentrations of lipopolyssacharide (LPS)

from salmonella abortus equi (Metalon) (0, 10, 100 ng/ ml, 1, 10, 100 µg/ ml) or lipoteichoic

acid (LTA) from staphylococcus aureus (Konstanz) (0, 2, 10, 50, 100 µg/ ml). Both reagents

were stored as concentrated stocks at 4°C and –20°C respectively and diluted with medium

immediately before use. Restored endotoxin was sonified for 2 minutes before use. After three

(3.6.5) or four (3.11) hours incubation at 37°C the supernatant was either harvested and

analysed (3.11) or subjected to another assay as described below (3.6.5).

3.6.4 Fluorescence labelled E. coli/ S. aureus phagocytosis assay

This assay measures the ability of phagocytes to engulf fluorescence labelled E. coli/ S.

aureus particles. Prepared BAL cells were cultured in 96-well plates (5 x 105 cells/ ml, 200 µl/

well), allowed to adhere for 2 hours and stimulated with dead, labelled E. coli/ S. aureus (final

concentration: 250 µg/ ml) as described in 3.6.5. After 3 hours incubation the phagocytosis

was stopped by washing the cells three times with PBS (room temperature) to remove non-

phagocytosed bacteria. The remaining PBS was removed by quickly turning the plate and

gently pushing it on a lint-free paper towel. Then cells were lysed by addition of 100 µl/ well

PBS + 0.1% Triton X-100. Fluorescence was determined at 530 nm excitation and 590 nm


31

emission wavelengths using a fluorescence microplate reader. Cells without bacteria were

used to determine the background fluorescence. From every experiment a digital photograph

was taken (Nikon Coolpix 995 through a special ocular with a Zeiss Axiovert 50).

Viability assays like Trypan blue or Sytox/ Hoechst staining at different steps during the assay

procedure were used as an index for physiological culture conditions.

The used particle concentration of 250 µg/ ml as well as the incubation time of 3 hours is

based on previous experiments, were the concentration signal relation and kinetic of particle

phagocytosis was measured. Furthermore complete inhibition of ingestion at 0°C was a

criteria for a quantitative phagocytic assay 158.

3.6.5 Assay for superoxide production

The production of superoxide after stimulation with either endotoxin or whole, dead bacteria

was performed by assaying the superoxide sensitive reduction of the water soluble and cell

impermeable tetrazolium salt WST-1. The advantage of this system is the higher sensitivity

and lower background in comparison to the method using the reduction of ferrycytochrome C 159. Isolated BAL cells were seeded in 96-well plates as described in 3.6.1. After 2 hours

adhesion the non-adhered BAL cells were removed by washing. Then 200 µl of 500 µM WST-

1 in Hanks balanced saline solution (HBSS) were added to each well.

In the experiments with whole, dead bacteria (see 3.6.2) to half of the wells either 250 µg/ ml

E. coli or S. aureus were added respectively. If endotoxin was used for stimulation to half of

the wells 0.1 µg/ ml LPS or 100 µg/ ml LTA was added. Therefore, three different groups (non-

stimulated, stimulated with endotoxin or bacteria) can be defined. To one well of each of these

groups 1000 U/ ml superoxide dismutase (SOD) was added. Then the plates were incubated

at 37°C for 3 hours.

After 3 hours, the extinction of the single wells was read by a plate reader at a wavelength of

λ = 450 nm. The concentration of superoxide and the rate of its production was calculated

considering the following facts:

The number of moles n is defined by the product of volume V of the sample ( 610100 ⋅=V l)

and its molar concentration C: CVn ⋅=

According to Lambert/ Beer, the molecular concentration C is calculated by dividing E∆

(change in WST-1 extinction caused by its reduction by superoxide) through the product of

WST-1`s molar extinction coefficient ε (37000 M-1 cm-1) and the diameter d of the sample (d =

0.3cm): dEVn⋅

∆⋅=ε


32

The difference in extinction E∆ was calculated for each group by forming the difference

between extinction En of each single well of a group and the average extinction SODE for the

appropriate wells containing SOD: SODn EEE −=∆

Since superoxide – in contrast to cytochrome C – is reduced by two electrons, 2 moles of

superoxide are needed to reduce 1 mol of WST-1, the formula has to be multiplied by the

factor 2, resulting in the final equation: dEVn⋅

∆⋅⋅=ε

2

The rate of production of superoxide R is the quotient of the number of molecules produced

during a certain time t, which was 180 minutes in these experiments:

⋅⋅⋅⋅∆==td

VEtnR

ε2

After inserting the known constants, the average rate for one group can be expressed as:

∑ ⋅∆⋅= 2.2001 En

R [pmol superoxide/ minute].

3.6.6 Preparation of washed and haemolysed erythrocytes and haemolysat

Peripheral venous blood was drawn from rats. To prevent coagulation Liquemin in NaCl (1 :

10) was added. The blood was diluted 10-fold in ice-cold NaCl and centrifuged for 5 minutes

at 270 x g and 4°C. The supernatant was carefully removed and the washing procedure was

repeated twice. Then a 10-fold volume of RPMI 1640 (without phenol red) was added and the

erythrocyte cell count was determined using the Neubauer counting chamber. After another 5

minutes centrifugation step at 270 x g and 4°C, cell pellet was diluted in RPMI 1640 (without

phenol red) to a final concentration of 8 x 109 erythrocytes/ ml corresponding approximately to

whole blood erythrocyte counts. The experiments were performed with 1.25 % washed

erythrocytes. To determine the impureness with leukocytes the leukocyte counts were

assessed (3.9). To get haemolysed erythrocytes the washed erythrocytes were frozen in liquid

nitrogen and then thawed at RT. This procedure was repeated three times. The haemolysat

was obtained after centrifugation for 10 minutes at 10600 x g and 4°C.

3.6.7 Lysis of erythrocytes

BAL fluid was centrifuged for 10 minutes at 270 x g and 4°C and the pellet was resuspended

in medium (1:20) containing 0.17 M ammonium chloride to lyse the erythrocytes and

incubated for 5 minutes at RT. To stop the lysis a five-fold excess of fresh medium was added,

pellet was resuspended and centrifuged again for 5 minutes at 190 x g and 4°C.


33

3.6.8 Experimental design of the in vitro experiments

Anaesthetised rats were exposed to blast wave thorax trauma (3.5 cm distance from the

nozzle). Animals were sacrificed 10 min, 24 and 96 hours after trauma, non-traumatized

animals served as sham controls (3.4.4). The alveolar macrophages were harvested by BAL

(3.6.1) and cultured either non-stimulated, with endotoxin (LPS or LTA) (3.6.3) or whole, dead

bacteria (E. coli or S. aureus) (3.6.2) for four and three hours, respectively.

Supernatants were either harvested for TNF measurement (3.11) or subjected to a

phagocytosis (3.6.4) and superoxide (3.6.5) production assay respectively.

In some experiments either whole blood, plasma, washed erythrocytes, haemolysed

erythrocytes or haemolysat (3.6.6) was added together with the bacterial stimulus.

In addition total and differential BAL cell counts (3.9) were determined.

Figure 3.6 summarizes schematic the experimental design of the in vitro experiments.

Figure 3.3: Timescale of the in vitro experiments.

3.7 Lung lavage

Lungs were tracheotomized and gently rinsed with 6 ml ice-cold NaCl. Recovery of the buffer

was always > 50%. The BAL fluids were centrifuged at 340 x g for 8 min and the supernatants

were stored at -70°C.

• phagocytosis assay• superoxide production• TNF release

in vitroin vivo

in vivothorax trauma/ sham controls

isolation/ culturingof thealveolar macrophages, 4°C

~ + 10 min, ~ + 24 h, ~ + 96 h

2 h adhesion, 37°C

in vitro stimulation:• dead, fluorescent E. coli/ S. aureus• LPS/ LTA

3 h or 4 h stimulation, 37°C

3 x washing

in case of oxidative burst measurement:• WST-1(500 µM) • SOD (1000 U/ ml)

cytospin

viability control

viability control

0 min

• phagocytosis assay• superoxide production• TNF release

in vitroin vivo

in vivothorax trauma/ sham controls

isolation/ culturingof thealveolar macrophages, 4°C

~ + 10 min, ~ + 24 h, ~ + 96 h

2 h adhesion, 37°C

in vitro stimulation:• dead, fluorescent E. coli/ S. aureus• LPS/ LTA

3 h or 4 h stimulation, 37°C

3 x washing

in case of oxidative burst measurement:• WST-1(500 µM) • SOD (1000 U/ ml)

cytospin

viability control

viability control

0 min


34

3.8 Determination of the alveolar haemorrhage

Pulmonary haemorrhage was quantified using the total erythrocyte count in the BAL fluid. The

erythrocyte number was assayed in a Neubauer counting chamber.

3.9 Total and differential cell counts and cell viability

To determine the total BAL cell count, the BAL fluid was mixed with Trypan blue solution (1:2)

and incubated for 1 minute at RT. Cell number was assayed using a Neubauer counting

chamber. Cells that excluded the Trypan dye were regarded as viable, blue cells as being

dead.

Cytospin preparations were used to determine the BAL cell differential. Depending on the total

cell counts the BAL samples were first diluted (2 – 4 folds) in isolation medium before pipetting

200 µl in the cytospin inserts. The samples were centrifuged at 90 x g (Minifuge RF (Heraeus

Instruments, Hanau, Germany) for 7 min at RT. The cells were fixed and stained with May-

Grünwald/ Giemsa solution (Merck KGaA, Darmstadt, Germany). From every experiment two

cytospins were made and a digital photograph was taken (Nikon Coolpix 995 through a

special ocular with a Zeiss Axiovert 50). The numbers of alveolar macrophages, lymphocytes

and infiltrated PMN were assessed by using standard morphological criteria for differentiation.

Total blood leucocytes were determined in the Neubauer counting chamber mixing heparin

blood (Liquemin in NaCl, 1:10) with Türk solution (1:10). For the differential blood count 10 -

15 µl whole blood were dispensed on a microscope slide, dried and stained with May-

Grünwald/ Giemsa solution.

3.10 Determination of total protein content

The BAL supernatants were stored at -70°C. The standard 2 mg/ ml BSA was diluted (2000,

1000, 500, 250, 125, 62.5, 31.25, 0 µg/ ml). 10 µl sample and standard were added on a

microtiter plate respectively. 190 µl/ well Pierce BCA protein assay reagent was added and

incubated for 30 min at 37°C. The extinction was measured at 550 nm in an ELISA reader.

Protein content was calculated using the BSA standard curve.

3.11 Cytokines determinations

For TNF measurements, BAL samples or cell culture supernatants were stored at –70°C and

quantified by sandwich ELISA. Antibody pairs (polyclonal rabbit anti-rat/ mouse Ab) and

recombinant TNF used as standard were purchased from Pharmingen (San Diego, CA, USA).


35

ELISA plates were coated over night at 4°C with 50 µl/ well with coat Ab (0.5% rabbit anti-TNF

serum) diluted in coating buffer (1:200). After blocking with 200 µl/ well PBS/ 3% BSA, pH 7.0

for 2 hours at room temperature (RT), the plates were washed twice with PBS/ 0.05% Tween

20. Samples and standard (50 µl/ well) were added and incubated for 3 hours at RT. After four

wash cycles, 100 µl/ well 0.5 mg/ ml tracer antibody in PBS/ 3% BSA were added and

incubated for 45 minutes. After another six wash cycles, plates were incubated for 30 minutes

with 100 µl/ well 50 ng/ ml streptavidin-peroxidase (Jackson Immuno Research, West Grove,

PA, USA) in PBS/ 3% BSA. Following eight wash cycles, 100 µl/ well TMB liquid substrate

solution (Sigma, Deisenhofen, Germany) was added and incubated at RT for 5 – 15 minutes.

After addition of 50 µl/ well stop solution (1 M H2SO4), absorption was measured at 450 nm

using a reference wavelength of 690 nm in an ELISA reader. The detection limit of the assay

was 10 pg/ ml.

IL-10 and IL-6 were determined using a commercially available ELISA kit (Biosource Europe

SA, Nivelles, Belgium). The detection limits of the assays were 15.6 pg/ ml for IL-10 and 31

pg/ ml for IL-6. Assay and analysis were performed according to the instructions of the

supplier.

3.12 Measurement of eicosanoids

BAL samples were stored at –70°C. Thromboxane A2 (t1/2 = 30 sec) was assessed as the

stable by-product thromboxane B2 (TXB2). Both TXB2 and and prostacycline (6-keto-PGF1α)

were determined by EIA (BioTrend GmbH, Köln, Germany) according to the manufacturer`s

instructions. The detection limits of the assays were 13.7 pg/ ml TXB2 and 3.2 pg/ ml for 6-

keto-PGF1α. The cross reactivity of the detecting antibody was TXB2 100%, 2,3-dinor TXB2

7.1%, prostaglandins < 0.01% and 6-keto-PGF1α 100%, 2,3-dinor 6-keto-PGF1α 3.17 %,

PGF2α 1.67%, prostaglandins 0.2-0.6% respectively.

3.13 Determination of NO production

The NO release into the alveolar space was determined indirectly by measuring the nitrate

level in the BAL fluid. Therefore 350 µl BAL fluid was filtrated by centrifugation in a microcon

concentrater (cut off 10 KD) at 470 x g and 4°C until a filtrate volume of approximately 300 µl

was reached. The nitrate reduction was performed in Eppendorf cups by mixing 300 µl BAL

filtrate with Nitrate reductase (0.1 U/ ml), FAD (5 µM), and NADPH (30 µM), incubating at

37°C for 15 minutes. For the NADPH oxidation, LDH (0.1 KU/ ml) and Na-pyruvat (0.3 mM)

were added and incubated at 37°C for 10 minutes. For di-azzo cuppling reaction, one part

sulfanilamide (1 mM) was mixed with one part HCL (6 N) and incubated for 15 minutes at 4°C.


36

A standard curve with different concentrations of NaNO3 (0 to 10 µM) was performed to

calculate the concentration of nitrate in the BAL fluid. After adding of N-naphtylethyldiamine

(NEDA) (0.8 mM) the absorption was measured at 560 nm to the reference at 690 nm.

3.14 Measurement of lactate dehydrogenase

Lactate dehydrogenase (LDH) was measured in BAL fluids according to Bergmeyer 160.

Measurement of LDH activity was performed in an Eppendorf ACP 5040 Analyser (Netheler &

Hinz, Hamburg, Germany). The reduction of nicotinamid adenine dinucleotide (NAD) to NADH

was measured at 340 nm.

3.15 Measurement of CINC-3

Samples taken from bronchoalveolar lavages were stored at -20°C. Rat CINC-3 was

assessed with a Duo Set ELISA Kit (Cat. No. DY525) and performed according to the

manufacturers instructions (R&D Systems Europe).

3.16 Gelatin zymograpy

To detect the gelatinolytic activity of matrix metalloproteinase-2 and –9 in BAL fluids, the

samples were evaluated by gelatin zymography 161. 300 µl BAL fluid were concentrated by

centrifuge in a microcon concentrater (cut off 10 KD) at 470 x g and 4°C until a volume of

approximately 50 µl was reached. The supernatant were treated under non-reducing

conditions and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-

PAGE) in 10% acrylamide gels containing 1.2% gelatin (300 Bloom, pig) at a constant current

of 120 V at RT. After electrophoresis, the gels were washed three times for 15 minutes with

TNCA buffer (40 mM Tris-HCL (pH 7.6), 10 mM CaCl2, 2 µM ZnCl2, and 0.1% Brij) including

3% Triton X-100 and then once for 5 minutes with TNCA buffer without Triton X-100 to remove

all trace of SDS. The gels were then incubated over night at 37°C in the TCNA buffer (without

Triton X-100). Following incubation, the gels were fixed in a solution of 10% acetic acid and 40

% methanol and stained with 0.2% Coomassie brilliant blue R250 for 15 minutes. The gels

were decolorized in a solution of 5% acetic acid, 25% methanol and 70% water. Gelatin

digestion was identified as a clear lytic zone against a blue background. The gels were

scanned and converted to digital images by scanner.

As markers we used the purified pro-matrix metalloproteinase-2 (1 ng MMP-2, isolated from

human rheumatoid synovial fibroblasts) and pro-matrix metalloproteinase-9 (0.1 ng MMP-9,

isolated from human neutrophil granulocytes) (Merck KGaA, Darmstadt, Germany).


37

3.17 Lung Wet/ Dry-Ratio

After the animal was sacrificed and exsanguinated the lungs were excised “en bloc”, cleaned

of extraneous tissue and blood. The different lung lobes were carefully separated from the

bronchi. The lung wet weight (LWW [g]) and the lung wet weight to body weight ratio (LWR)

was determined before drying the lung tissue in an oven at 50°C. The mean LWR for the

animals exposed to blast was expressed as a multiple of the LWR of non-traumatized

controls; this ratio was designed the injury quotient (Qi). Thus Qi = LWR (injured animals)

divided by LWR (controls). The extent of injury was determined as minor or uninjured for Qi <

1.2, moderate for Qi = 1.2 to 1.5, severe for Qi = 1.5 to 1.9, and very severe for Qi > 1.9 162.

After three days a stable dry weight of the lungs was achieved and the wet-to-dry weight ratio

was calculated.

3.18 Histological examinations

After exsanguination and opening of the thoracic cage the cardiac apex was cut and a

catheters was placed in the pulmonary artery. Then a non-recirculating perfusion (30 ml/ min)

with ice-cold PBS was started to remove the blood from the pulmonary circulation. The heart-

lung block was excised, cleaned and infused intratracheally with 4% buffered formalin (15 cm

H20). The paraffin embedded lung tissues were cut in five-micrometer sections and stained

with haematoxylin and eosin (H & E) and examined by light microscopy. Representative

sections are shown (in the chapter “Results”).

3.19 Statistics

Data in the figures are given as means ± SEM, data in the tables as mean ± SD. The entire

curves data from two experiments were compared by a Two-way ANOVA design. This test,

determines how a response is affected by different factors e.g. drug treatment. Values of p <

0.05 were considered statistically significant. Data from end- or other time-points were

analysed by One-way ANOVA. In case of differences among the groups, post-tests were

performed as indicated in the legends (Dunnet’s Multiple Comparison Test, Tukey’s Multiple

Comparison Test or Bonferroni's Multiple Comparison Test); Statistics were performed with

the Graph Pad Prism 3.0 software (Graph Pad Software, San Diego, CA, USA).

Results

38

4. Results

4.1 System characteristics

Since it turned out that the blast wave generator originally designed and constructed by Irwin

and coworkers in 1998, 117 and modified by our colleagues in Ulm (U.C. Liener, F. Gebhard, L.

Kinzl) was lacking reproducibility concerning the variability of the measured lung function

parameters, we tried to modify this setting to our specific needs in order to optimise the

reproducibility of pressure wave exposure and the monitoring of the pressure wave.

4.1.1 Variability and influence of physical parameters

A well-defined pressure wave exposure requires a standardized rupture of the Mylar

polyester film. There is a good correlation between the speed of the air supply of the pressure

reservoir, the rupture of the diaphragm and the degree of standardization of the expansion of

the pressure wave. In order to obtain a fast air supply from the storage tank into the pressure

reservoir of the blast wave generator, we installed a 0.75 l compressed air bottle in between.

In addition, the bottle was located as close as possible to the pressure reservoir. The working

pressure of 18 bar in the air bottle was determined higher than the burst strength of the Mylar

diaphragm (≈ 8.5 bar). Thus a complete rupture of the polyester film was reached, as

compared to the incomplete rupture in the case of a slow air supply via a compressed-air hose

directly to the valve and the pressure reservoir. By pressing the diaphragm along a

predetermined breaking point between pressure reservoir and nozzle instead of only fixing it

with screws, the rupture of the film was further optimised (Fig. 4.1).

Figure 4.1: The nozzle of the blast wave generator in A) the top view and in B) the predetermined breaking point in detail.

Results

4.1.1.1 Monitoring and standardization of the blast pressure wave

For an exact monitoring of the blast pressure wave as the basis of comparability between

different experiments, we assessed the pressure wave data by different sensors at different

positions. As exemplarily presented in Figure 4.2 the pressure measurement by the

transducer in the reservoir revealed a fast pressure increase of up to 8.7 bar within 19.2 ms,

depending on and limited by the opening time of the valve, an immediate decay to zero

indicating a fast rupture of the diaphragm and therefore a fast exposure of the pressure wave

by the nozzle, and a negative phase or suction.

P

l

(

a

w

ressure curves over time measured at the thorax level by two transducers on the right and

eft side of the rat revealed a nearly instantaneous 100 percent rise time of 85 and 295 µs

Pp(r), Pp(l)) to the pressure peaks 3.4 and 3 bar (Pp(r), Pp(l)) and a duration (t(r), t(l)) of 654

nd 625 µs respectively (Fig. 4.3). The calculated area under the curves (AUC(r), AUC(l))

ere 101.35 and 100.37 V*s, respectively.

Figure 4.2: Time course of pressure increase in the pressure reservoir until rupture of the diaphragm. The working pressure in the compressed air bottle was 18 bar. Under control of a valve the pressure charged the pressure reservoir. The time course of the rupture pressure (Rp) was measured by the transducer in the reservoir. 0 5 10 15 20 25

-10123456789

10

t [ms]

Rp

[bar

]

4.0

39

Figure 4.3: Pressure wave monitoring at rat thorax level. The pressure wave data were measured over time by two sensors on the right Pp(r) and left side Pp(l) of the rat. Pp(r): , n = 1; Pp(l): ---, n = 1. 0 250 500 750 1000 1250 1500

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

t [µs]

Pp [b

ar]

Results

40

Table 4.1 summarizes the pressure wave data and the calculated area under the curves at a

3.5 cm distance from the nozzle measured in the present study.

Table 4.1: Pressure wave data measured by different transducers at various positions.

Rp

[bar]

Pp(r)

[bar]

AUC(r)

[V*s]

t(r)

[µs]

Pp(l)

[bar]

AUC(l)

[V*s]

t(l)

[µs]

mean±SD 8.56 ± 0.21 2.69 ± 0.37 95.04 ± 5.81 612.6 ± 17.42 3.63 ± 0.94 101.73 ± 10.4 644.71 ± 55.01

n 218 219 220 219 220 218 217

CV [%] 2 14 6 3 26 10 9

Pressure wave data were measured at a distance of 3.5 cm from the nozzle by three different transducers at different positions: cracking pressure in the pressure reservoir (Rp), pressure peaks at thorax level with two transducers on the left and the right side (Pp(r), Pp(l)). The corresponding area under the curves (AUC(r), AUC(l)) and the duration of the pressure wave t(r), t(l) were calculated. Data are expressed as mean ± SD, number of individual experiments (n), variation coefficient (CV).

The mean values of the pressure wave data measured and calculated from the right and left

transducer were assumed to be the pressures obtained above the rat thorax (Tab. 4.2).

Table 4.2 Assumed pressure conditions above the rat thorax.

3.5 cm

distance

Pp(r+l)

[bar]

AUC(r+l)

[V*s]

t(r+l)

[µs]

mean ± SD 3.16 ± 0.43 98.35 ± 6.18 628.78 ± 29.31

n 219 220 216

CV [%] 14 6 5

Pressure wave data were measured at a distance of 3.5 cm from the nozzle by three different transducers at different positions: mean of the pressure peaks at thorax level (Pp(r+l)), mean of the calculated area under the curve: (AUC(r+l)) and mean of the duration of the pressure wave (t(r+l)). Data are expressed as mean ± SD, number of individual experiments (n), variation coefficient (SD/ median) (CV).

To further improve the reproducibility of the setting, the animals were always blasted in the

same breathing situation during spontaneous breathing at halothane anaesthesia, namely at

top of expiration. This was achieved by a string above thorax that allows respiratory

monitoring and which triggers automatically the time point of valve opening and thus of the

blast. Figure 4.4 shows exemplarily the breathing frequency at halothane anaesthesia that

was uniformly 36 breaths per minute compared to physiological 100 - 130 breaths/ minute

indicating a deep anaesthesia (Fig. 4.4 A). The animals had a low breathing frequency

Results

immediately after blast exposure (Fig. 4.4 B), which was sometimes only transient in nature

depending on anaesthesia.

FBA

T

±

4

T

e

0

s

m

c

a

A) B)

41

igure 4.4: Respiratory monitoring before and after pressure wave exposure. reathing frequency was monitored with a string above the rat thorax connected with a sensor: ) before and B) after pressure wave exposure, exemplary for n = 1.

he mean respiratory rate before blast at 3.5 cm distance throughout the experiments was 38

12 breath/ minute (n = 224).

.1.1.2 Distance dependency of the pressure wave parameters

he mean pressure peak (Pp(r+l)) measured at rat thorax level decayed rapidly as it

xpanded from the blast wave generator. This pressure decrease correlated significantly (r = -

.58) with the distance (Fig. 4.5 A), allowing rats` exposure to different blast intensities by

imply varying the distance between the blast wave generator and the rat. Moreover, the

ean duration of the pressure wave (t(r+l)) and therefore the impact on the rat and the

alculated area under the curve (AUC (r+l)) correlated significantly with the distance (r = 0.24

nd r = 0.51) respectively (Fig. 4.5 B+C).

0.0 0.5 1.0 1.5 2.0 2.5 3.0

insp

iratio

nex

pira

tion

t [s]0 10 20 30 40 50

insp

iratio

nex

pira

tion

t [s]

Results

A)

2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5

1

2

3

4

5

distance below nozzle [cm]

Pp (r

+l) [

bar]

B)

2.5 4.5 6.5 8.5 10.5

0

250

500

750

1000


time

(r+l

) [µs

]

C)

42

Figure 4.5: Correlation between pressure wave parameters and distance below the nozzle. The pressure wave data were measured at different distances below the nozzle: A) mean pressure peak (Pp(r+l); B) mean duration of the pressure wave (t(r+l)) and C) mean calculated area under the curve (AUC(r+l)). The number of experiments (n) was n = 220. The statistical analysis was done by correlating A) - C) with the distance and linear regression revealing a Pearson r of r = 0.58 (p < 0.0001), r = 0.24 (p = 0.0003) and r = 0.51 (p < 0.0001), respectively.

2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5

0

25

50

75

100

125

150


AUC

(r+l

) [Vs

]

Results

4.1.1.3 Pressure wave data of survivors versus non-survivors

The mortality was about 14.46% in the group exposed to the blast at a distance of 3.5 cm,

mostly due to vessel rupture. Comparing the pressure wave data of the survivors versus the

non-survivors, the latter showed a higher rupture pressure (Rp, p = 0.0004) resulting in a

pressure wave curve with no significant higher pressure peak at rat thorax level (Pp(r+l), p =

0.89) but a longer duration (t(r+l), p = 0.025) and therefore an increased area under the curve

(AUC(r+l), p = 0.025) (Tab. 4.3). Since the Mylar polyester films show differences in their

physical characteristics (supplier information) and since different lots of Mylar polyester film

were used in the present study, the differences in the necessary rupture pressures may be

explained. A relationship between mortality and pre-blast breathing frequency with regard to a

prolonged recovery time from anaesthesia, could not be observed (survivors: 38 ± 12 breath/

minute (n = 224); non-survivors: 44 ± 16 breath/ min (n = 36)).

Table 4.3: Comparison of the pressure wave data from survivors versus non-survivors.

The presrighttwo the cData

4.2

4.2.

Alth

com

impa

func

imm156,16

sur

non

sur

distance

[cm] n

Rp

[bar]

Pp(r+l)

[bar]

AUC(r+l)

[V*s]

t(r+l)

[µs]

vivors 235

8.56 ± 0.21

(218)

3.16 ± 0.43

(219)

98.35 ± 6.18

(220)

628.78 ± 29.31

(216)

-3.5

34 8.72 ± 0.39 3.17 ± 0.33 101.05 ± 7.34 642.31 ± 45.18

43

animals were exposed to the pressure wave at a distance of 3.5 cm below the nozzle. The sure wave data were recorded by a transducer in the pressure reservoir (Rp), two sensors at the and left side at rat thorax level (Pp(r), Pp(l)). The data represent the recorded mean values of the transducers: mean pressure peak (Pp(r+l)), mean duration (t(r+l)) and mean calculated area under urve (AUC(r+l)). Data represent mean ± SD, number of individual experiments (n) in parentheses. were analysed by an unpaired t-test (* p < 0.05, *** p < 0.0001).

Functional changes of alveolar macrophages and interaction with blood in vitro after trauma

1 Trauma-induced impairment in host defence mechanisms

ough it is reported that severe trauma contributes to an increase in infectious

plications through an impairment of host defence mechanisms 163, the cause of such

irment remains unknown 164. Most studies have focused on the suppression of immune

tion determined during the first few hours after trauma, implying that this

unosuppression may be responsible for an impairment of host defence mechanisms 5. Since the alveolar macrophage, mobile representative of the mononuclear phagocyte

vivors (34) *** (33) (32) * (32) *

Results

system, is believed to be the central modulator as both regulator and effector of inflammation/

anti-inflammation within the alveolar space 48, macrophage function following experimental

blunt chest trauma with regards to phagocytotic capacity, respiratory burst and mediator

release was assessed.

4.2.1.1 Differences in the amount of isolated macrophages depending on the time after trauma

The macrophage is by far the most abundant non-parenchymal lung cell: about 50 to 100

macrophages are found per alveolus, and they dominate lymphocytes by a ratio of

approximately 5 to 10 49. The number of macrophages recruited from the alveolar space of

rats after thoracic trauma, ranging from 3.11 ± 1.08 x 106 cells/ ml (10 minutes after trauma) to

9.86 ± 0.34 x 106 cells/ ml (96 hours after trauma) of bronchoalveolar lavage exudates

depending on the time after blast exposure (Fig. 4.6), was significantly different from that of

non-traumatized controls (6.32 ± 0.92 x 106 cells/ ml). The observed increase after 96 hours

may be due to an excess of macrophages recruited into the alveolar space.

4

I

d

c

S

0

***

44

.2.1.2 Trauma-related modulation of phagocytosis of dead, fluorescence labelled E. coli and S. aureus by alveolar macrophages

nitially after the blast pressure wave exposure, the alveolar macrophage phagocytosis rate of

ead, fluorescence labelled Escherichia coli (E. coli) was reduced by 82% (Fig. 4.7) and

ontrol levels were reached 96 hours after the blast. In contrast, the phagocytic capacity for

taphylococcus aureus (S.aureus) particles was not significantly impaired after trauma (p =

.075).

Figure 4.6: Thoracic trauma-induced changes in the amount of resident alveolar macrophages. Animals were sacrificed 10 min (n = 37), 24 (n = 23) and 96 (n = 13) hours after trauma, non-traumatized animals served as sham controls (n = 57). The alveolar macrophages were harvested by BAL. Data represent means ± SD, number of individual experiments (n) in parentheses. Data were statistically analysed by One-way ANOVA and Tukey's Multiple Comparison Test: ***p < 0.001 vs. sham control, 96 hours.

sham

contro

l

+10m

in+2

4h+9

6h

2.5×1006

5.0×1006

7.5×1006

1.0×1007

time after blast exposure

*** ***

BA

L ce

ll co

unt [

cells

/ml]

Results

4

N

r

t

“

s

S

S

p

o

b

t

T

s

400

FcTcncaD

45

.2.1.3 Oxidative burst capacity after trauma induced by Gram positive and Gram negative stimuli

ext, we investigated whether the oxidative burst capacity of the alveolar macrophages in

esponse to either whole, dead bacteria or endotoxin was impaired, following experimental

horax trauma, using the superoxide sensitive dye WST-1 (as described in the chapter

Materials and Methods”). Thoracic trauma yielded a marked decrease in the production of

uperoxide within the first 24 hours after trauma, both with or without the addition of E. coli or

. aureus, with recovery only after 96 hours (Fig. 4.8 A).

imilarly, the trauma-related initial decrease of superoxide generation was independent of the

resence of lipopolysaccharide (LPS) or lipoteichoic acid (LTA) (Fig. 4.8 B). Superoxide levels

f non-traumatized controls were reached already at 24 hours. Furthermore, 96 hours after the

last the superoxide production was significantly increased in all groups, as compared to non-

raumatized controls.

aken together, these results indicate that thoracic trauma decreases the production of

uperoxide by primary rat alveolar macrophages.

E. coli S. aureus0

100

200

300

sham controltrauma: +10mintrauma: +24htrauma: +96h

**

rela

tive

units

igure 4.7: Thorax-trauma related modulation of phagocytosis of dead, fluorescence labelled E. oli/ S. aureus by primary alveolar macrophages. he alveolar macrophages (AM) were cultured either non-stimulated or stimulated with fluorescent E. oli and S. aureus particles (250 µg/ ml) for three hours respectively. Data represent means ± SEM, umber of individual experiments (n) in parentheses: AM harvested either from non-traumatized sham ontrols (n = 10); from recently (10 min) traumatized rats (n = 7) and from rats traumatized 24 h (n = 8) nd 96 h (n = 6) in advance. Data were analysed by One-way ANOVA (on log-transformed data) and unnett's Multiple Comparison Test: * p < 0.05, ** p < 0.01 vs. control.

Results

Fmd(t(stO

A)

non-stim

ulated

E. coli

S. aureu

s0

20

40

60

***

****

***


supe

roxi

de p

rodu

ctio

n [p

mol

/min

]

*

B)

46

igure 4.8: Thoracic trauma decreases the production of superoxide by primary rat alveolar acrophages. The alveolar macrophages were seeded either non-stimulated or stimulated with whole,

ead bacteria or endotoxin for three hours. Data are means ± SEM, number of individual experiments n) in parentheses: A) E. coli or S. aureus stimulation (250 µg/ ml) of alveolar macrophages from non-raumatized sham controls (n = 8); from recently traumatized rats (10 min) n = 9) or 24 h (n = 7) and 96 h after trauma (n = 5), and B) LPS (0.1µg/ ml) or LTA (100 µg/ ml) timulation of alveolar macrophages either from non-traumatized sham controls (n = 5); from recently raumatized rats (n = 4) or from rats 24 h (n = 4) and 96 h (n = 3) after trauma. Data were analysed by ne-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05, ** p < 0.01 vs. control.

non-stim

ulated LTA

LPS0

50

100

150

200

250


**

**

**

**

**

**

supe

roxi

de p

rodu

ctio

n [p

mol

/min

]

Results

4.2.1.4 Trauma-related modulation of TNF release after Gram positive and Gram

negative stimulation

Due to the key role of alveolar macrophages (AM) as modulators of pulmonary inflammation,

we further investigated macrophage-induced tumor necrosis factor α (TNF α) release after

experimental thorax trauma alone and in response to Gram-positive and Gram-negative

stimuli. The TNF release of non-stimulated AM was more than 3-fold increased 96 hours after

blast, indicating that the AM are primed to release TNF following thoracic trauma. Additional

stimulation of the AM with whole, dead E. coli or S. aureus induced a marked TNF release

already 24 hours after trauma. In contrast, initially after trauma the stimulation with bacteria

did not increase the TNF release as compared to sham controls. In the case of E. coli

stimulation, the TNF release was even decreased at this time point (Fig. 4.9 A). In these

experiments E. coli stimulation was more potent in inducing TNF release from AM than S.

aureus.

I

T

t

a

t

i

t

A)

47

n line with the results described above, there was a more than 1.5-fold and 3-fold increase in

NF release in response to LPS and LTA stimulation (100 µg/ ml), respectively, 24 hours after

rauma. In contrast, no endotoxin-induced increase in TNF release could be detected initially

fter trauma, as compared to non-traumatized sham controls (Fig. 4.9 B+C). With regard to

he TNF release, the AM are more susceptible towards LPS than LTA stimulation. These data

ndicate that 24 hours after blast injury the AM showed an increased TNF release in response

o both whole bacteria or endotoxin.

non-stim

ulated

E. coli

S. aureu

s0

1000

2000

3000

4000

5000

6000


**

**

**

**

TNF α ααα

[pg/

ml]

Results

48

4.2.2 Role of the trauma-induced alveolar haemorrhage in impaired macrophage functions

The most frequent and obvious injury, detected in our rat blast model, was bilateral pulmonary

haemorrhage, due to a disruption of the alveolar parenchyma and capillaries, with exudation

of blood into the interstitial and alveolar spaces. Therefore, we tested whether the observed

macrophage dysregulation after trauma is due to the accompanying alveolar haemorrhage.

Thus, we lysed the red blood cells (RBCs) in the harvested BAL fluids (as described in the

chapter “Materials and Methods”) before seeding, stimulation and assessing the macrophage

function. This manipulation did not lead to any changes in macrophage functions (data not

shown).

4.2.2.1 Role of alveolar haemorrhage in reduced phagocytosis

Based on the total reduction of 70 ± 26% of the phagocytic capacity for labelled E. coli

particles of alveolar macrophages from traumatized animals, as compared to non-traumatized

sham controls, the removal of red blood cells (RBCs) resulted in a more than 2-fold increase

in phagocytic capacity of AM from traumatized rat lungs (p = 0.0078). These experiments

Figure 4.9: Thoracic trauma leads to alterations in TNF release. The alveolar macrophages were seeded either non-stimulated or stimulated with E. coli and S. aureus particles (250 µg/ ml) or with LPS (100, 10, 1, 0.1 and 0.01 µg/ ml) and LTA (100, 50, 10, and 2 µg/ ml) for three and four hours, respectively: A) E. coli and S. aureus stimulation of alveolar macrophages harvested either from non-traumatized sham controls (n = 8); from recently traumatized rats (n = 7) and from rats 24 h (n = 7) and 96 h (n = 5) after trauma, B) LPS and C) LTA stimulation of alveolar macrophages either from non-traumatized sham controls (n = 6); from recently traumatized rats (n = 7) and from rats 24 h (n = 5) after trauma. Data were analysed by One-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05, ** p < 0.01 vs. control. Data represent means ± SEM, number of individual experiments (n) in parentheses.

100 10 1

0.1

0.01

non-

stim

ulat

ed

0

2500

5000

7500

10000 sham controltrauma: +10mintrauma: +24h

TNF α ααα

[pg/

ml]

LPS [µg/ml]

**

*

**

*

B)

100 50 10 2

non-

stim

ulat

ed

0

500

1000

1500

2000

2500 sham controltrauma: +10mintrauma: +24h

LTA [µg/ml]

TNF α ααα

[pg/

ml]

**

***

C)

Results

Fptntr

suggested that the massive alveolar haemorrhage is at least in part responsible for the

impaired phagocytic capacity after trauma in this model (Fig. 4.10).

4

A

o

(

p

T

i

100

Figure 4.10: Impairment of phagocytosis upon thorax trauma is partly due to the accompanying alveolar haemorrhage. E. coli (250 µg/ ml) stimulation of alveolar macrophages from non-traumatized sham controls (n = 6) and recently traumatized rats (10 min) (n = 6) with or without lysis of the red blood cells (RBCs) respectively. Data were analysed by an unpaired t-test: ** p < 0.01 trauma (lysed RBCs) vs. trauma. Data represent means ± SEM, number of individual experiments (n) in parentheses.

.2.2.2 Role of alveolar haemorrhage in reduced oxidative burst

s already mentioned in 4.2.1.3 thoracic trauma yielded a marked decrease in the production

f superoxide, as compared to non-traumatized sham controls. Lysis of the red blood cells

RBCs) showed an up to 6-fold (p = 0.0248) and 8-fold (p < 0.0001) increase in superoxide

roduction of AM from traumatized rat lungs with or without E. coli stimulation, respectively.

hese data provided further evidence that alveolar haemorrhage plays a role in the trauma-

nduced decrease in superoxide production by alveolar macrophages (Fig. 4.11).

25

50

75

trauma: +10mintrauma: +10min(lysed RBCs)

sham controlsham control(lysed RBCs)

sham control trauma: +10min

**

rela

tive

units

100 sham control]

49

igure 4.11: Alveolar haemorrhage plays a role in thorax trauma-related decrease in superoxide roduction. Alveolar macrophages from non-traumatized sham controls (n = 4) and recently

raumatized rats (10 min) (n = 3) with or without lysis of the red blood cells (RBCs) and cultured either on-stimulated or stimulated with E. coli (250 µg/ ml) respectively. Data were analysed by an unpaired -test: * p < 0.05 vs. trauma (E. coli stimulation), *** p < 0.0001 vs. trauma (non-stimulated). Data epresent means ± SEM, number of individual experiments (n) in parentheses.

0

25

50

75

non-stimulatedE. coli

no RB

Cs

trauma: +10min

no RB

Cs

*** *

supe

roxi

de p

rodu

ctio

n [p

mol

/min

Results

50

Figure 4.12: Alveolar haemorrhage plays a minor role in TNF release after trauma. Alveolar macrophages were harvested from non-traumatized sham controls (n = 6) and recently traumatized rats (10 min) (n = 6) with or without lysis of the red blood cells (RBCs) and cultured either non-stimulated or stimulated with E. coli (250 µg/ ml) respectively. Cells without bacteria were used to determine the background and were subtracted. Data were analysed by an unpaired t-test. Data represent means ± SEM, number of individual experiments (n) in parentheses.

4.2.2.3 Role of alveolar haemorrhage in diminished TNF release

Next, we investigated whether the trauma-associated alveolar haemorrhage affects the TNF

release from alveolar macrophages in response to E. coli particles. Figure 4.12 shows that in

the presence of RBCs the TNF release was reduced by 48 ± 21% in consequence of pressure

wave exposure. However, in the absence of RBCs, the TNF release from AM from

traumatized animals was not significantly increased (p = 0.0887). In contrast to the

phagocytosis and oxidative burst capacity, these data exclude a major role of the alveolar

haemorrhage in trauma-related diminished TNF release.

1000

2000

3000

4000

sham control(lysed RBCs)

trauma: +10mintrauma: +10min(lysed RBCs)

sham control

TNF α ααα

[pg/

ml]

4.2.3 Influence of different blood components on the macrophage function

Since at least a part of the trauma-induced macrophage dysregulation was shown to be due to

the associated alveolar haemorrhage, we tested different blood components, which could be

responsible for the observed effects. For this purpose, we prepared washed and haemolysed

red blood cells (RBCs), haemolysate of RBCs, and plasma from peripheral venous blood (as

described in the chapter “Materials and Methods”) and tested whether these different blood

components impaired the macrophage functions. To preserve the in vivo conditions after

thorax trauma, whole blood and the prepared components were used in a final concentration

of 1.25%, corresponding to the red blood cell counts found in bronchoalveolar lavage fluids

from traumatized rats (data not shown).

4.2.3.1 Influence on phagocytosis

The phagocytic activity was significantly diminished in the presence of whole blood and

washed red blood cells (RBCs), whereas the addition of haemolytic blood, plasma, haemolytic

RBCs and RBC haemolysate had no effect (Fig. 4.13). These results indicate that only intact

blood cells, but neither cellular debris nor plasma components decreased phagocytosis.

Results

250

FmsnnaE

51

E. coli

+ whole

blood

+ hae

molytic

blood

+ plas

ma

+ was

hed RBCs

+ hae

molytic

RBCs

+ RBC hae

molysate

0

50

100

150

200

* *

rela

tive

units

igure 4.13: Blood cells attenuate phagocytosis of fluorescent E. coli particles by alveolar acrophages. Alveolar macrophages harvested from non-traumatized sham controls (n = 8) were

timulated with fluorescent E. coli (250 µg/ ml) particles together with either whole blood (1.25%, = 4), haemolytic blood (1.25%, n = 5), plasma (1.25%, n = 3), washed red blood cells (RBCs) (1.25%, = 5), haemolytic RBCs (1.25%, n = 5) or RBC haemolysate (1.25%, n = 5). Data were statistically nalysed by One-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05 vs. . coli control. Data represent means ± SEM, number of individual experiments (n) in parentheses.

Results

4.2.3.2 Influence on oxidative burst

In contrast to the reduced phagocytosis by whole, intact blood cells, the oxidative burst

capacity was significantly reduced by haemolytic blood, both with or without the addition of E.

coli (Fig. 4.14). Furthermore, the addition of plasma yielded a decrease of superoxide

generation, although to a lesser extent than with E. coli and afflicted by variations between the

individual experiments. The plasma-related effect may be explained by a plasma

contamination with intracellular components due to centrifugation.

Fmcb(r*S

70

52

igure 4.14: Intracellular blood cell components decrease the oxidative burst by alveolar acrophages. Alveolar macrophages harvested from non-traumatized sham controls (n = 8) and

ultured either non-stimulated or stimulated with whole, dead E. coli (250 µg/ ml), together with whole lood (1.25%, n = 4), haemolytic blood (1.25%, n = 5), plasma (1.25%, n = 3), washed red blood cells RBCs) (1.25%, n = 5), haemolytic RBCs (1.25%, n = 5) and RBC haemolysate (1.25%, n = 5) espectively. Data were analysed by One-way ANOVA and Dunnett's Multiple Comparison Test: p < 0.05, ** p < 0.01 vs. non-stimulated and E. coli-stimulated control, respectively. Data are means ± EM, number of individual experiments (n) in parentheses.

+ whole

blood

+ hae

molytic

blood

+ plas

ma

+ was

hed RBCs

+ hae

molytic

RBCs

+ RBC hae

molysate

0

10

20

30

40

50

60

non-stimulatedE. coli

** ****

supe

roxi

de p

rodu

ctio

n [p

mol

/min

]

Results

4.2.3.3 Influence on TNF release

In line with the reduced oxidative burst capacity by haemolytic blood, the E. coli-induced TNF

release was also significantly reduced, whereas the addition of the other blood components

was without effect (Fig. 4.15).

T

c

s

d

t

s

4

T

5000

Fmsnndan

53

he experiments described above were also assessed with whole blood and prepared blood

omponents in a final concentration of 5%. Compared to 1.25% there were no statistically

ignificant differences in the phagocytotic- and oxidative burst activity or TNF release

etectable (data not shown). Whole intact blood cells as well as cellular debris seeded

ogether with harvested macrophages interfered with the macrophage adhesion (data not

hown).

.2.4 Summary of the in vitro findings

he results of the different in vitro experiments are summarized in the Table 4.4.

E. coli

+ whole

blood

+ hae

molytic

blood

+ plas

ma

+ was

hed RBCs

+ hae

molytic

RBCs

+ RBC hae

molysate

1000

2000

3000

4000

**

TNFα ααα

[pg/

ml]

igure 4.15: Haemolytic blood decreases E coli-induced TNF release by macrophages. Alveolar acrophages from non-traumatized sham controls (n = 8) were cultured either non-stimulated or

timulated with E. coli (250 µg/ ml) together with whole blood (1.25%, n = 4), haemolytic blood (1.25%, = 5), plasma (1.25%, n = 3), washed red blood cells (RBCs) (1.25%, n = 5), haemolytic RBCs (1.25%, = 5) and RBC haemolysate (1.25%, n = 5) respectively. Cells without bacteria were used to etermine the background TNF levels and were subtracted. Data were analysed by One-way ANOVA nd Dunnett's Multiple Comparison Test: ** p < 0.01 vs. E.coli control. Data represent means ± SEM, umber of individual experiments (n) in parentheses.

Results

54

A) Thorax trauma-related changes in macrophage functions.


stimulus read out sham control 10min 24h 96h

phagocytosis

oxidative burst non-stimulated

TNF release

phagocytosis / / / /

oxidative burst / / / / E. coli/ S. aureus

TNF release / / / /

oxidative burst / / / / LPS/ LTA

TNF release / / / -

B) Influence of different blood components on macrophage functions.

Table 4.4: no change, upregulation, downregulation (compared to sham control A) and non-stimulated and E. coli stimulation B), respectively), and - not measured.

non-traumatized controls

non-stimulated E. coli

addition of: phagocytosis oxidative

burst

TNF

release phagocytosis

oxidative

burst

TNF

release

cell culture

medium

whole blood

haemolytic

blood

plasma

washed

RBCs

haemolytic

RBCs

RBC

haemolysate

Results

55

4.3 Functional consequences and endogenous modulation in vivo

After an initial non-fatal injury, thoracic trauma is considered to play a pivotal role in the

development of single-or multi-organ failure 97, probably making the patient more susceptible

to infections 95. The initial physiologic responses to thoracic trauma have not been adequately

investigated and none of the numerous interventions studied until today made it into clinical

practice. Therefore, we investigated the pathophysiological consequences occurring in rat

lungs in response to primary blast exposure in vivo over time.

4.3.1 Blast injury-induced disruption of the alveolar capillary barrier

As already mentioned in 4.2.2, histological examinations of the blasted rat lungs showed

characteristic signs of pressure wave-induced injury (Fig. 4.16 A-E). The non-traumatized

sham controls showed a normal lung structure (Fig. 4.16 D+E). In contrast, at 10 minutes, i.e.

immediately after blast pressure wave exposure, a massive alveolar haemorrhage and

exudation of fluid into the interstitial and alveolar spaces was detectable (Fig. 4.16 A).

Disruption of the alveolar septa was also observed (not shown in the present visual field). At 3

hours after blast exposure, there was a marked focal alveolar haemorrhage, but there were

virtually no PMN infiltrates yet observed (Fig. 4.16 B). On the contrary, erythrocyte-laden huge

alveolar macrophages were observed 24 hours after the blast accompanying by a few

detectable PMN in the alveolar space (Fig. 4.16 C).

Figure 4.16: Histopathology of rat blast injury. Rats were exposed to the pressure wave at 3.5 cm distance from the nozzle. Animals were sacrificed either 10 min A), 3 hours B) or 24 hours C) after blast exposure, non-traumatized animals served as sham controls E) + D). Visualized at an original magnification of 40-fold A) – C) + E) and 10-fold in D).

Results

10

4.3.1.1 Thorax trauma-related BAL protein accumulation

In order to determine pressure-induced cell destruction, we assessed the total protein content

in the bronchoalveolar lavage fluids. Figure 4.17 illustrates a significant protein accumulation

in the alveolar space of traumatized lungs, as compared to non-traumatized controls (p <

0.0001). From 12 hours on, the protein levels decreased and returned to control levels 96

hours after trauma, indicating that some kind of repair and/ or active protein transport took

place 166.

4

I

t

f

d

a

FmedCe

56

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

2

4

6

8


****

**

prot

ein

cont

ent [

mg/

ml]

.3.1.2 Thorax trauma-induced pulmonary haemorrhage

ntra-alveolar haemorrhage quantified by red blood cell (RBC) counts in BAL fluids of

raumatized rat lungs (Fig. 4.18) showed a massive increase within the first 10 minutes,

ollowed by a decrease from 3 hours on. At 96 hours after blast exposure no RBCs were

etectable. These results provided evidence for an enormous clearance over time, which is

lso reported for clinical situations of alveolar haemorrhage 167.

igure 4.17: Blast injury-related protein accumulation in BAL fluids. Animals were sacrificed 10 in (n = 5), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast

xposure, non-traumatized animals served as sham controls (n = 9). Protein concentration was etermined in the BAL fluid. Data were analysed by One-way ANOVA and Dunnett's Multiple omparison Test: ** p < 0.01 vs. control. Data represent means ± SEM, number of individual xperiments (n) in parentheses.

Results

10.0

4

T

m

b

0

T

e

a

h

i

F(efa

57

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0.0

2.5

5.0

7.5


*

** **R

BC

s [1

07 /ml]

.3.1.3 Trauma-induced cell destruction

he release of the cytosolic enzyme lactate dehydrogenase (LDH) into the BAL was used as a

arker for cellular integrity after blast. In accordance with the protein accumulation and red

lood cell counts, LDH activity was significantly increased upon pressure wave exposure (p =

.0047) (Fig. 4.19). This also confirmed the mechanical-induced cell damage in this setting.

he levels were still elevated 24 hours after the blast, probably due to the long half-life of the

nzyme and/ or possible delayed cell death. The activity returned to control levels 96 hours

fter blast. Thus, the BAL LDH activity, in contrast to the different kinetics of the alveolar

aemorrhage and total protein content, may be a suitable marker for the extent of the initially

nduced injury.

igure 4.18: Blast injury-related pulmonary haemorrhage. The animals were sacrificed 10 min n = 5), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 3), 24 h (n = 3) or 96 hours (n = 3) after blast xposure, non-traumatized animals served as sham controls (n = 8). RBCs were determined in BAL luids. Data were analysed by One-way ANOVA and Dunnett's Multiple Comparison Test: *p < 0.05 nd ** p < 0.01 vs. control. Data are means ± SEM, number of experiments (n) in parentheses.

Results

200

4

P

s

o

e

(

w

q

6

d

v

58

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

50

100

150


**

**

*LD

H ac

tivity

[uni

ts/l]

.3.1.4 Thorax trauma related pulmonary oedema formation

ulmonary injury, quantified by the lung wet to dry weight ratio (wet WT/ dry WT ratio), was

ignificantly increased within the first 90 minutes after blast (p < 0.001), indicating pulmonary

edema formation (Fig. 4.20). The reduced ratios assessed from three hours after blast

xposure on may suggest that oedema generation stopped and/ or fluid resorption increased

1.5 h vs. 6 h: p < 0.001). However, although not significant different, the calculated lung wet

eight (LWW) to body weight (BW) ratio (LWR) (p = 0.27) and the corresponding injury

uotient (Qi) (described in detail in the chapter Material and Methods, 162) examined between

and 24 hours after blast exposure may indicate a secondary deterioration (Tab. 4.5). This

iscrepancy may be explained by pulmonary infiltrates in these lungs, which were not obvious

isible in the lung wet to dry weight ratios.

Figure 4.19: LDH release after thoracic trauma. The animals were sacrificed 10 min (n = 5), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 8). LDH activity was determined in BAL fluids. Data were analysed by One-way ANOVA and Dunnett's Multiple Comparison Test: *p < 0.05 and ** p < 0.01 vs. control. Data are means ± SEM, number of experiments (n).

Results

T

TM

FaDv

** **

59

able 4.5: Time course of wet WT/ dry WT ratio and LWR

time after blast n BW [g] wet Wt/ dry WT

ratio LWR Qi

sham control 3 250 ± 13 5.32 ± 0.21 0.0045 ± 0.0006

10 min 4 234 ± 11 6.58 ± 0.99 0.0082 ± 0.0034 1.84

1.5 h 6 246 ± 6 6.77 ± 0.49 ** 0.0083 ± 0.0022 1.85

3 h 4 234 ± 7 5.64 ± 0.72 0.0064 ± 0.0017 1.44

6 h 4 233 ± 10 5.05 ± 0.34 0.0065 ± 0.0011 1.46

12 h 3 243 ± 4 5.15 ± 0.44 0.0081 ± 0.0033 1.82

24 h 3 237 ± 2 4.66 ± 0.13 0.0073 ± 0.0014 1.63

able 4.5: Data represent mean ± SD. Data were analysed by One-way ANOVA and Bonferroni's ultiple Comparison Test: **p < 0.01 1.5 h vs. sham control and 6 h after blast, respectively.

igure 4.20: Thorax trauma-induced weight gain. Lung wet weight and dry weight was determined t different time points after pressure wave exposure. The wet to dry weight ratios were calculated. ata were analysed by One-way ANOVA and Bonferroni's Multiple Comparison Test: **p < 0.01 1.5 h s. sham control and 6 h after blast, respectively.

sham

contro

l

10min

1.5h 3h 6h 12

h24

h

4

5

6

7

8


wet

WT/

dry

WT

ratio

Results

Fs(aTS

4.3.2 Blast injury-related mediator release


patients is strongly associated with a posttraumatic imbalance of the immune system 146. Post-

traumatic immunosuppression (“immunoparalysis”) 95,130 is a highly complex interplay of cell

interactions, cytokines, central nervous and humoral system, cell demise, and further factors.

Thus we focused on a restricted list of mediators that may explain the observed mitigation of

immune defence (described in 4.2).

4.3.2.1 Eicosanoid release

Since it is reported that there is a pattern-related release of prostanoids in patients suffering

from thoracic trauma 134, we examined the release of thromboxane and prostacycline into the

alveolar space over 96 hours, following blast pressure wave exposure. As expected and in

line with clinical data obtained from polytrauma patients with thorax involvement 138, there was

a marked immediate release of both thromboxane and prostacycline within the first 10 minutes

after blast followed by a significant decrease by 90 minutes. Within 96 hours after blast

exposure, we could not detect a secondary increase at the time points investigated (Fig. 4.21

A+B). Together with the lack of COX-2 induction, determined by RT-PCR analysis (data not

shown) within the first 6 hours after blast exposure, an inflammatory reaction, involving these

mediators during this time appeared to be rather unlikely.

A) B)

60

igure 4.21: Prostanoid release in the early phase after thoracic trauma. The animals were acrificed 10 min (n = 6), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 9). A) TxB2 release nd B) 6-keto PGF1α were determined in BAL fluids. Data were analysed by One-way ANOVA and ukey's Multiple Comparison Test. *** p < 0.001: 10min vs. all other groups. Data represent means ± EM, number of experiments (n) in parentheses.

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

200

400

600

800

1000

1200

1400

1600

1800


***

TxB

2[p

g/m

l]

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

500

1000

1500

2000

2500

3000


***

6-ke

to P

GF1

α ααα [p

g/m

l]

Results

61

4.3.2.2 Nitric oxide release

Since there was no COX-2 induction detectable, we were further interested whether there is

an induction of the NO-synthases (NOS) after thoracic trauma. Figure 4.22 may suggest a

two-phase release of nitric oxide related nitrite and nitrate with a delayed maximum 12 hours

after trauma. However, RT-PCR analysis revealed no induction of the inducible NOS within 12

hours (data not shown). These results indicate rather an initial trauma-induced NO release

due to the constitutive endothelial NOS, than an involvement of the iNOS during the first four

days following blast exposure in this setting.

4.3.2.3 Cytokine release

High synthesis of pro-inflammatory cytokines after trauma has been correlated with poor

outcome 168,169. Therefore, we examined the release of TNF-α and interleukin (IL)-6 after

thoracic trauma over time. Increasing BAL TNF-α levels were detectable from 3 hours after

trauma on, reaching a maximum 12 hours after blast exposure, followed by a continuous

decrease over 96 hours (Fig. 4.23 A). Plasma TNF-α was not detectable within 96 hours after

trauma, neither by ELISA nor by WEHI-assay (data not shown). In addition to TNF-α, IL-6

levels seemed to increase within 3 hours after trauma and reached significant peak levels

after six hours (Fig. 4.23 B). This early increase is in accordance with plasma IL-6 levels

obtained from polytrauma patients 142.

Figure 4.22: Thoracic trauma-induced nitric oxide release. The animals were sacrificed 10 min (n = 5), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 8). Nitric oxide release was determined as nitrate in BAL fluids. Data were statistically analysed by One-way ANOVA (p = 0.0054) and Dunnett's Multiple Comparison Test: * p < 0.05 vs. sham control. Data are means ± SEM, number of experiments (n) in parentheses.

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

1

2

3

4

5 *


nitr

ate

[µM

]

Results

F

c

l

p

Fs(Bam

A) B)

urthermore, we measured interleukin (IL)-10, as a possible indicator for an anti-inflammatory

ounterregulation after trauma. In comparison to non-traumatized sham controls, the IL-10

evels in BAL fluids of traumatized animals were significantly decreased at any measured time

oint within 24 hours and returned to reference ratios 96 hours after blast (Fig. 4.24).

igure 4.23: Time course of TNF and IL-6 release after thoracic trauma. The animals were acrificed 10 min (n = 6), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 9). A) TNF-α and ) IL-6 release were determined in BAL fluids. Data were statistically analysed by One-way ANOVA nd Dunnett's Multiple Comparison Test: * p < 0.05 and ** p < 0.01 vs. sham control. Data represent eans ± SEM, number of experiments (n) in parentheses.

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

15

30

45

60

75


**

****

*TNFα ααα

[pg/

ml]

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

250

500

750

1000 **


IL- 6

[pg/

ml]

75

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

15

30

45

60


***

IL-1

0 [p

g/m

l]

Figure 4.24: Decrease in BAL IL-10 after thoracic trauma. The animals were sacrificed 10 min (n = 6), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 9). IL-10 release was assessed in BAL fluids. Data were statistically analysed by One-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05 and ** p < 0.01 vs. sham control. Data are means ± SEM, number of experiments (n) in parentheses.

62

Results

63

4.3.2.4 Measurement of soluble tumor necrosis factor receptors

Some publications have provided evidence for the presence of 55- and 75-k Da soluble tumor

necrosis factor receptors (sTNFRs) in injured patients, with a strong correlation with mortality 148,149. Because of the lack of detectable plasma TNF, together with the relatively moderate

levels of secreted TNF into the BAL after blast exposure, we examined sTNF R1 - and

sTNFR2-BAL levels in response to blast injury, in order to test whether the TNF is bound to

the soluble TNF receptors and therefore not active and possibly not detectable. Patterns of

BAL 55 and 75 kDa sTNFR at serial intervals after injury are illustrated in Figure 4.25 A. Both

soluble receptors were significantly elevated 90 minutes after blast. Then levels declined,

followed by a second marked increase within the first 6 to 12 hours, and reaching near

baseline levels within 24 hours. Within the first 12 hours after blast there were no remarkable

differences between 55 and 75 kDa sTNFR release. In contrast, 24 and 96 hours after trauma

the sTNFR p75 levels were significantly higher than the sTNFR p55 levels (p = 0.0074 and p =

0.04, respectively). Comparison of the 55 and 75 kDa sTNFR levels of non-traumatized

control lungs showed higher levels of sTNFR p75 than of p55 (p = 0.0113). The soluble tumor

necrosis factor receptor p55/ p75 ratio was significantly elevated within 6 hours of trauma and

returned to reference ratios after 12 hours (Fig. 4.25 B). A high correlation between both

sTNFR subtypes could be shown over 96 hours (r = 0.96; p < 0.0001). Correlation between

different detected plasma mediators revealed a correlation between both soluble TNF

receptors and IL-6 levels (4.3.2.3) over 96 hours after blast exposure (r = 0.595; p = 0.0002

and r = 0.557; p = 0.0006 for p55 and p75 to IL-6, respectively).

0 3 6 9 12

0

25

50

75

100

125

0

50

100

150

200

250

300

24 48 72 96

trauma: sTNFR p55trauma: sTNFR p75sham control: sTNFR p55sham control: sTNFR p75

**** **

**

**

**

A)

time after blast exposure [h]

sTNF

R p5

5 [p

g/m

l] sTNFR p75 [pg/ml]

Results

B)

64

4.3.3 Infiltration of inflammatory cells

To further elucidate a possible inflammatory response to blast injury, we investigated whether

an infiltration of inflammatory cells into the BAL took place.

4.3.3.1 PMN infiltration

The analysis of the amount of different cell types in lavage fluids from traumatized rats,

isolated at different time points, may suggest an permanent increase in neutrophil (PMN)

infiltration into the alveoli, starting from 90 minutes after trauma on. However, significant

amounts of neutrophils were detected not until 6 and 12 hours after blast exposure followed

by a continuous decay within 96 hours (Fig. 4.26). The number of lymphocytes in the BAL fluid

did not vary significantly over time (data not shown). Figure 4.27 shows the corresponding

cytospins of the different BAL fluids. The cytospins of the non-traumatized sham controls

showed a quite homogeneous population of alveolar macrophages. Only very few

lymphocytes can be found, and no PMN (Fig. 4.27 A). At 24 hours post-trauma the PMN were

as frequent as the alveolar macrophages. The alveolar macrophages were carrying

Figure 4.25: Time course of sTNFR p55 and p75 BAL concentrations in A) as well as the p55/ p75 ratio in B). The animals were sacrificed 10 min (n = 6), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 9). Levels of sTNFR p55 and p75 were assessed in BAL fluids. Data were analysed by One-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05 and ** p < 0.01 vs. sham control. Unpaired t-test was performed to compare differences in p55 and p75 levels at different time points: **p = 0.0074 for p75 vs. p55 and * p = 0.04 for p75 versus p55 at 24 hours and 96 hours after blast, respectively. Data are means ± SEM, number of experiments (n) in parentheses. Linear regression between p55 and p75 release revealing a Pearson r of r = 0.96; p < 0.0001.

0 3 6 9 12

0.0

0.1

0.2

0.3

0.4

0.5

24 48 72 96

sham

con

trol

****

** *

time after blast exposure [h]

ratio

sTN

FR p

55/ p

75

Results

intracellularly high amounts of erythrocytes due to the important alveolar haemorrhage (Fig.

4.27 B).

Fe

W

c

e

4

T

w

C

i

I

i

100

65

igure 4.27: Cytospin pictures after blast exposure. A) sham control, B) 24 hours after blast xposure. AM: alveolar macrophages; PMN: neutrophils; Lym: lymphocytes; RBC: red blood cells.

ith respect to a possible post-traumatic immunosuppression, we quantified the white blood

ell counts, revealing no statistically significant differences within 96 hours following blast

xposure (data not shown).

.3.3.2 CINC-3 measurement in the BAL fluids

o elucidate the role of chemokines in the thoracic trauma-induced infiltration of neutrophils,

e measured the α-chemokine CINC-3 (chemokine-induced neutrophil chemotactic factor,

INC). CINC-3 is one of three functional rat homologues to human IL-8 and plays an

mportant role in an animal model of acute lung injury induced by intrapulmonary deposition of

gG immune complexes 170. The CINC-3 levels in the BAL fluids of traumatized rat lungs were

ncreased 6 hours after trauma and returned to sham control levels after 96 hours. The time

Figure 4.26: Time-course of PMN infiltration into the alveoli after thoracic trauma. The animals were sacrificed 10 min (n = 6), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast exposure, non-traumatized animals served as sham controls (n = 9). Statistical analysis was performed by One-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05 vs. sham control. Data are expressed as mean ± SEM, number of experiments (n).

A)

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

20

40

60

80


**

% B

AL n

eutro

phils

Results

course of the CINC-3 release correlated with the infiltration of neutrophils into the alveolar

space (r = 0.63, p = 0.0007) (Fig. 4.28).

4

W

i

o

k

l

r

b

G

l

b

a

N

f

T

t

e

Fp(wv

500 **

66

.3.3.3 Matrix metalloproteinase activity

ith regard to the observed neutrophil infiltration and extracellular matrix remodelling, we

nvestigated the release and activation of the matrix metalloproteinases (MMP)-2 and MMP-9

ver time after thoracic trauma. Gelatinolytic bands of approximately 134, 92, 85, 72, and 66

Da were detected in BAL fluids (Fig. 4.29). These bands correspond to the pro-MMP-9

ipocalin complex, pro-MMP-9, active form of MMP-9, pro-MMP-2, and active form of MMP-2,

espectively (according to Moritaka and coworkers, 161). Within the first 24 hours following

last exposure, intense lytic bands corresponding to pro – and active MMP-2 were detected.

elatinolytic bands corresponding to MMP-9 and coexpressed bands of the pro-MMP-9

ipocalin complex were detected after six hours and weak after 12 and 24 hours, whereas lytic

ands equivalent to active MMP-9 were only barely detected at these time points. In addition,

weak gelatinolytic band corresponding to active MMP-9 appeared initially after trauma.

either pro- nor active MMP-2 or MMP-9 were detectable after 96 hours and in BAL fluids

rom non-traumatized sham controls, respectively.

he MMP-9 activity after six hours correlated with the increased neutrophil infiltration at this

ime point (4.3.3.1). MMP-2 is reported to be preferentially secreted by fibroblasts and

pithelial cells 171 and may therefore be released after trauma-induced cellular damage.

igure 4.28: CINC-3 measurement in the BAL fluid of traumatized rat lungs at different time oints after blast exposure: 10 min (n = 6), 1.5 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n=3), 24 h n = 3) and 96h (n = 3). Non-traumatized animals served as sham controls (n = 9). Statistical analysis as performed by One-way ANOVA (p = 0.0014) and Dunnett's Multiple Comparison Test: ** p < 0.01 s. sham control. Data are expressed as mean ± SEM, number of experiments (n).

sham

contro

l

10min

1.5h 3h 6h 12

h24

h96

h0

100

200

300

400


CIN

C-3

[pg/

ml]

Results

67

Furthermore, increased neutrophil related myeloperoxidase (MPO) activity was determined in

BAL fluid investigated 12 after blast exposure (p < 0.01), as compared to non-traumatized

controls (12 h: 0.43 ± 0.24 relative units, n = 4 and sham controls: 0.07 ± 0,004 relative units,

n = 9). This elevated MPO activity may related with the infiltration of neutrophils into the

alveolar space (4.3.3.1).

Figure 4.29: Gelatin zymograms of BAL fluids from traumatized rats and controls. Gelatin zymogram of BAL fluids from animals sacrificed 10 min (n = 5), 1.5 h (n = 3), 3 h (n = 3), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3) or 96 hours (n = 3) after blast exposure, and from non-traumatized sham controls (n = 8). Gelatinolytic bands of approximately 134, 92, 85, 72, and 66 kDa correspond to the pro MMP-9 lipocalin complex, pro-MMP-9, active form of MMP-9, pro-MMP-2, and active form of MMP-2, respectively. Purified pro-MMP-9 and pro-MMP-2 were used in the concentrations of 0.1 ng and 1 ng, respectively were used simultaneously as positive controls for gelatin zymographic analysis of BAL fluid samples to standardize the measurement results.

pro-MMP-2 (72kDa)active MMP-2 (66kDa)

pro-MMP-9 (92kDa)/ active MMP-9 (85kDa)

lipocalin complex (134kDa)

pro-

MM

P-9

pro-

MM

P-2

6h3h1.5

h

10m

in

sham

con

trol

12h

24h

96h Mw

(kDa)


pro-MMP-2 (72kDa)active MMP-2 (66kDa)

pro-MMP-9 (92kDa)/ active MMP-9 (85kDa)

lipocalin complex (134kDa)

pro-

MM

P-9

pro-

MM

P-2

6h3h1.5

h

10m

in

sham

con

trol

12h

24h

96h Mw

(kDa)


Results

68

4.3.3.4 Summary of time-dependent BAL-findings

time course in vivo after blast injury

read out peak/

minimum

maximal

recovery conclusions

total protein content 1.5 h 96 h

alveolar

haemorrhage 1.5 h 96 h

LDH activity 10 min 96 h

disrupted alveolar-

capillary barrier/

cell death

wet/ dry ratio 1.5 h 6 h pulmonary infiltrates

TxB2 10 min 3 h

6-keto PGF1α 10 min 3 h early prostanoid release

nitrite/ nitrate 12 h > 96 h constitutive NO release

TNF α 12 h 96 h

IL-6 6 h 96 h

pro-inflammatory

cytokine release

IL-10 10 min 96 h early decrease

CINC-3 6 h 96 h pro-inflammatory

chemokine release

sTNFR p55/ p75

ratio 10 min 24 h

anti-inflammatory

counterregulation

neutrophil infiltration 6 - 12 h > 96 h

10 min 96 h MMP-2

MMP-9 6 h 96 h

MPO 12 h 96 h

inflammation

time-

depe

nden

t rec

over

y

Table 4.6: upregulation, downregulation (compared to non-traumatized sham controls).

Results

69

4.4 Ex vivo lung perfusion: Trauma-induced changes in lung function

To further assess and to characterize the physiological and functional changes after in vivo

thorax trauma, we combined our blast trauma model with our ex vivo isolated perfused rat

lung model, allowing the assessment of physiological lung function parameters without

interference of other organs.

4.4.1 Interrelationships between physical and physiological injury

The pressure peak decayed rapidly as it expanded from the blast wave generator (described

in the chapter 4.1.1.2) and therefore theoretically allowing rats exposure to different blast

intensities by varying the distance between the blast wave generator and the rat. Therefore,

we investigated whether physical and biological impact match in this setting.

Pathological evaluation:

The animals were subjected to a blast pressure wave with the tip of the blast nozzle located

either 3.5, 5, or 7 cm above the sternum. The alveolar haemorrhage showed a marked

increase in rats positioned at 3.5 cm distance below the nozzle, as compared to non-

traumatized control lung (Fig. 4.30 A, Tab. 4.8). The red blood cell counts correlated not only

with the distance between the animal and the nozzle (r = - 0.49, p = 0.033), but also with the

particular defined injury score (p = 0.011, r = 0.565) for each group (described in 3.4.5). The

latter correlation underscores the validity of this macroscopically examined injury score (Fig.

4.30 B, Tab. 4.8).

Results

sham

contro

l7 c

m5 c

m3.5

cm

2.0×10 06

4.0×10 06

6.0×10 06

8.0×10 06

1.0×10 07

**

distance belowthe nozzle

A)R

BC

s/ m

l BA

L flu

id

2 4 6 8 10 12 14

5.0×10 06

1.0×10 07

1.5×10 07

2.0×10 07

B)

haemorrhage score

RB

Cs/

ml

Figure 4.30: Distance-dependent alveolar haemorrhage. A) alveolar haemorrhage expressed as red blood cells (RBCs) per ml BAL fluid and assessed after 150 minutes perfusion time and B) RBCs/ haemorrhage score correlation. The animals were located either 3.5 (n = 12), 5 (n = 4), or 7 cm (n = 3) distance from the nozzle, non-traumatized animals served as sham controls (n = 5). Statistical analysis was performed on A) by One-way ANOVA and Dunnett's Multiple Comparison Test: ** p < 0.01 vs. sham control. Linear regression revealed a Pearson r of r = - 0.49, p = 0.033. Data are expressed as mean ± SD, number of experiments (n) and B) by correlating RBCs with the score and linear regression revealing a Spearman r of r = 0.565.

F

(

f

s

c

t

s

T

Ov

70

urthermore pulmonary injury, quantified by lung wet weight (LWW)/ body weight (BW) ratio

LWR), was significantly increased at 3.5 cm distance below the nozzle, indicating oedema

ormation (Tab. 4.7) in these lungs. In addition, the individually defined oedema score showed

ignificant higher values in the 3.5 and 5 cm group compared to non-traumatized sham

ontrols (Tab. 4.8). Comparison of the calculated injury quotients (Qi) revealed a Qi of 1.94 for

he 3.5 cm group, 1.31 for the 5 cm group and 1.20 for the 7 cm group, indicating “very

evere“, “moderate” and “minor” lung injury, respectively (Tab. 4.7) 162.

able 4.7: Lung wet weight to body weight ratio at varying distances from the nozzle.

distance n BW [g] LWW [g] LWR Qi

sham control 4 269 ± 30 1.83 ± 0.13 0.0069 ± 0.0013

3.5 cm 3 243 ± 6 3.21 ± 1.25 * 0.0130 ± 0.0053 * 1.94

5 cm 5 241 ± 6 2.14 ± 0.23 0.0089 ± 0.0009 1.31

7 cm 3 238 ± 12 1.94 ± 0.06 # 0.0082 ± 0.0006 # 1.20

ne-way ANOVA and Tukey's Multiple Comparison Test: # p < 0.05 vs. 3.5 cm and * p < 0.05 s. sham control. Data are expressed as mean ± SD, number of experiments (n).

Results

Table 4.8: Selected results from exposure to graded blast waves.

distance RBCs x 106/ ml haemorrhage score oedema score

sham control 0.12 ± 0.15 (5) 0 ± 0 (5) 0 ± 0 (4)

3.5 cm 7.13 ± 4.72 (12) ** 9.6 ± 2.2 (12) *** 10.0 ± 3.6 (3) ***

5 cm 4.68 ± 3.11(4) 8.3 ± 2.1 (4) *** 7.2 ± 1.9 (5) **

7 cm 1.33 ± 0.44 (3) 5.7 ± 3.1 (3) ** # 4.0 ± 1.4 (3) #

Lung function parameters:

The initial tidal volume (Tv) and therefore the compliance (CL) (data not shown) of the lungs

immediately after blast, obtained with standard ventilation pressures (PEEP/ PIP: -2/ -7 cm

H2O) was significantly decreased in the 3.5 and 5 cm distance group. (Tv: 3.5 cm: 0.98 ± 0.35

ml, 5 cm: 1.33 ± 0.52 ml, and 7 cm: 1.72 ± 0.5 ml), as compared to the non-traumatized sham

controls (1.95 ± 0.33 ml) (Fig. 4.31). The decrease in initial Tv correlated with the distance

below the nozzle (r = 0.58, p = 0.0005), indicating that thoracic trauma deteriorates the lung

function depending on the pressure wave intensity.

As already mentioned above (described in d

used a negative pressure ventilation (PEEP/

With these ventilation pressures tidal volu

sham

contro

l7 c

m5 c

m3.5

cm0.0

0.5

1.0

1.5

2.0

2.5

distance belowthe nozzle

**

*

Tv [m

l]

One-way ANOVA and Tukey's Multiple Comparison Test: ** p < 0.01, *** p < 0.001 vs. sham control; # p < 0.05 vs. 3.5 cm. Data are expressed as mean ± SD, number of experiments (n) in parentheses.

Figure 4.31: Pressure wave intensity-related decrease in initial tidal volume after thoracic trauma. Rats were exposed to the pressure wave at different distances from the nozzle (3.5 cm, n = 9; 5 cm, n = 16; 7 cm, n = 7). The rat lungs were perfusedimmediately after blast. Non-traumatized animals served as sham controls (n = 7). The initial tidal volume (Tv) was assessed with standard ventilation pressures (PEEP/ PIP: -2/ -7 cm H2O). Data are expressed as mean ± SD, number of experiments (n). Statistical analysis was performed by One-way ANOVA and Dunnett's Multiple Comparison Test: * p < 0.05 and ** p < 0.01 vs. sham control. Correlation between distance and Tv was performed by linear regression revealing a Pearson r of r = 0.58, p = 0.0005.

71

etail in the chapter “Materials and Methods”) we

PIP: -2/ -7 cm H2O) in our lung perfusion model.

mes of 1.95 ± 0.33 ml are obtained in non-

Results

72

traumatized control lungs (Fig. 4.31). The ventilation pressures in traumatized lungs were

adjusted in order to obtain a minimal vital ventilation of 1 ml tidal volume. Rats exposed to

blast pressure wave at a distance of 3.5 cm needed a significantly higher inspiratory pressure

PIP) to achieve 1 ml tidal volume than non-traumatized controls (Tab. 4.9).

Lung function parameters of traumatized rat lungs measured at the end of 150 minutes

perfusion time were significantly deteriorated in the 3.5 cm distance group, as compared to

non-traumatized controls with respect to tidal volume (TV) (Tab. 4.9), airway resistance (RA)

(Fig. 4.32, Tab. 4.9), and also lung weight gain (ΔW) (Fig. 4.32, Tab. 4.9). The vascular

parameters showed no significant differences (data not shown). The comparison of the airway

resistance and the lung weight gain of the different groups revealed a distance-dependency

(Tab. 4.9). This was not the case for the tidal volume; despite the correlation between distance

and initial tidal volume deterioration, at the end of perfusion no distance-dependency could be

measured. The following experiments were all done at a distance of 3.5 cm below the nozzle.

Figure 4.32: Bronchoconstriction and pulmonary oedema ex vivo as a function of distance between rat thorax and blaster nozzle. Rats were exposed to the pressure wave at different distances from the nozzle (3.5 cm, 5 cm, and 7 cm). Immediately after blast the lung perfusion was started. Non-traumatized animals served as sham controls. Data are expressed as mean ± SEM. A) RA/ RA0min: airway resistance and B) ΔW: weight gain were all normalized to the value at t = 0 min.

0 20 40 60 80 100 120 140

1.0

1.2

1.4

1.6

1.8

trauma: 5 cm distancetrauma: 7 cm distance

trauma: 3.5 cm distance

sham control

A)

time [min]

RA/R

A0m

in

0 20 40 60 80 100 120 140

0

250

500

750

1000

1250

trauma: 5 cm distancetrauma: 7 cm distance

trauma: 3.5 cm distance

sham control

B)

time [min]

∆ ∆∆∆W

[mg]

Results

73

Table 4.9: Thoracic trauma-induced deterioration of lung function parameters.

in vivo trauma ex vivo perfusion

~ -15 min ~ 3 min 150 min 150 min 150 min

distance PIP

(cm H2O)

Tv

(ml)

RA

(cm H2O*s/ ml)

ΔW

(mg)

sham

control

7 ± 0

(7)

1.60 ± 0.15

(7)

0.33 ± 0.05

(7)

579 ± 184

(6)

3.5 cm 8.36 ± 1.36 *

(16)

0.66 ± 0.21 ***

(13)

0.47 ± 0.15 *

(15)

1099 ± 441 **

(15)

5 cm 7.31 ± 0.64 #

(11)

0.68 ± 0.23 ***

(11)

0.34 ± 0.05 #

(10)

723 ± 158 #

(9)

7 cm 7.07 ± 0.07 #

(7)

0.90 ± 0.31 ***

(7)

0.29 ± 0.02 ##

(7)

789 ± 114

(6)

4.4.2 Time-dependent recovery of the lung function after trauma

With regard to the reduced wet to dry weight ratios assessed in vivo from three hours after

blast exposure on, we further investigated whether the pulmonary oedema formation and the

deteriorated lung function observed ex vivo initially after blast is also a function of time.

Therefore, the animals were subjected to a blast wave injury at a distance of 3.5 cm, and

sacrificed 10 minutes, three hours and six hours after trauma. Subsequently lungs were

perfused over 150 minutes.

Pathological evaluation:

As expected and in accordance with the results obtained in vivo (4.3.1.2), the red blood cell

counts examined after 150 minutes perfusion time did not significantly differ within the three

Rats were exposed to the pressure wave at different distances from the nozzle (3.5 cm, 5 cm,

and 7 cm). Immediately after blast, lung perfusion was started and the lung function parameters

were recorded. Non-traumatized animals served as sham controls. Inspiratory pressure (PIP),

tidal volume (Tv), airway resistance (RA) or weight gain (ΔW). Data are expressed as mean ± SD,

number of experiments (n) in parentheses. One-way ANOVA and Tukey's Multiple Comparison

Test was performed: * p < 0.05, ** p < 0.01, *** p< 0.001 vs. sham control, and # p < 0.05, ## p <

0.01 vs. 3.5 cm, respectively

Results

10 m

in 3 h 6 h0.0

0.5

1.0

1.5

2.0

time after blastexposure

**

**

Tv [m

l]

groups (10 minutes: 7.13 ± 4.72 x 106/ ml (n = 12), 3 h: 6.74 ± 2.71 x 106/ ml (n = 4), 6 h: 10.9

± 7.65 x 106/ ml (n = 3)).

Lung function parameters:

Linear correlation showed that the initial tidal volume (Tv) obtained with standard ventilation

pressures (PEEP/ PIP: -2/ -7 cm H2O) was improved depending on the time after blast

exposure (r = 0.74, p < 0.0001). Thus, the initial tidal volume was significantly improved 3 and

6 hours after blast exposure, as compared to immediately perfused traumatized lungs (10min:

0.98 ± 0.35 ml, 3 h: 1.60 ± 0.29 ml, 6 h: 1.85 ± 0.35 ml) (Fig. 4.33). Within six hours after the

blast the tidal volume recovered and reached sham control levels (1.95 ± 0.33 ml), indicating

that some kind of repair takes place in vivo over time. These data are in line with the improved

compliance (CL) over time (data not shown).

Due to this time-dependent amelioratio

necessary inspiratory pressure (PIP) to

and 6 hours after blast as compared to

significant improved initial tidal volume

compared to immediately perfused lung

volume (Tv) nor in airway resistance (R

of perfusion (Tab. 4.10).

Figure 4.33: Time-dependent improvement of tidal volume after thoracic trauma. Rats were exposed to the pressure wave at a distance of 3.5 cm from the nozzle. The rats were sacrificed either 10 minutes (n = 16), 3 hours (n = 7) or 6 hours (n = 7) following the blast. Subsequently lung perfusion was started. The initial tidal volume (Tv) was assessed at standard ventilation pressures (PEEP/ PIP: -2/ -7 cm H2O). Data are expressed as mean ± SD, number of experiments (n). Statistical analysis was performed by One-way ANOVA and Dunnett's Multiple Comparison Test: ** p < 0.01 vs. 10 min. Correlation between time after blast and Tv was performed by linear regression revealing a Pearson r of r = 0.74, p < 0.0001.

74

n of the respiratory mechanics following blast injury, the

achieve 1 ml tidal volume was significantly reduced 3

immediately perfused lungs (Tab. 4.10). Despite of the

(Tv) in lungs perfused three and six hours after trauma

s (10 minutes), no significant difference neither in tidal

A) between the groups could be measured at the end

Results

75

Table 4.10: Lung function assessed ex vivo at different time points after blast injury.


~ -15 min 0 min 150 min 150 min 150 min

time

after blast

PIP

(cm H2O)

Tv

(ml)

RA

(cm H2O*s/ ml)

ΔW

(mg)

10 min 8.36 ± 1.36

(16)

0.66 ± 0.21

(13)

0.47 ± 0.15

(15)

1099 ± 441

(15)

3 h 7.07 ± 0.11

** (8)

0.83 ± 0.31

(7)

0.36 ± 0.07

(7)

826 ± 293

(8)

6 h 7.09 ± 0.08

* (8)

0.76 ± 0.34

(6)

0.38 ± 0.09

(6)

855 ± 311

(8)

Comparison of the entire curves data from the different groups showed a reduced weight gain

3 and 6 hours after blast, as compared to immediately perfused lungs (Fig. 4.34, Tab. 4.11).

However, analysis of the lung weight values assessed at the end of perfusion (Tab. 4.10)

revealed no significant differences between the groups. Whether the decrease of the weight

gain is due to an increased oedema resorption and / or a stop in oedema generation over time

in these lungs can not be determined by the used setting. Nevertheless, the reduced weight

gain did not reach control levels (∆ W: 579 mg ± 184 mg). The macroscopically defined

oedema score correlated with the time after blast exposure (r = - 0.85, p < 0.0001) and was

significantly decreased three and six hours after blast compared to recently injured lungs (3 h:

1.1 ± 1.9, n = 8; 6 h: 0.6 ± 1.5, n = 7; and 10 minutes: 10.1 ± 2.4, n = 16), thus suggesting a

recovery in vivo over time.

Rats were exposed to the pressure wave at a distances of 3.5 cm from the nozzle. The rats were sacrificed either 10 minutes, 3 hours or 6 hours following the blast. Subsequently lung perfusion was started. Inspiratory pressure (PIP), tidal volume (TV), airway resistance (RA), weight gain (ΔW). Data are expressed as mean ± SD, number of experiments (n). One-way ANOVA and Dunnett's Multiple Comparison Test were performed: * p < 0.05, ** p< 0.01 vs. 10min.

Results

4

W

in

w

is

4

D

s

a17

1250

76

ex vivo ΔW

in vivo blast 3 h 6 h

+ 10 min *** ***

+ 3 h *

.4.3 Pharmacological intervention

ith regard to the observed time-dependent “mixed” oedema generation and fluid resorption

vivo after pressure wave exposure, we further examined whether active sodium transport,

hich is reported to be the primary mechanism that activates alveolar fluid clearance in vivo,

involved (reviewed by Matthay and coworkers, 2002, 172).

.4.3.1 The role of the amiloride-sensitive sodium channels after trauma

epending on the species investigated, amiloride–sensitive sodium channels on the apical

ide of alveolar epithelial cells account for 40 to 80% of the sodium uptake 173. Since

miloride, in a species-dependent manner, inhibited 40 - 70% of basal alveolar fluid clearance 2 we used amiloride to examine the role of these sodium channels after trauma.

Figure 4.34, Table 4.11: Pulmonary oedema ex vivo as a function of time. Rats were exposed to the pressure wave at a distances of 3.5 cm from the nozzle. The rats were sacrificed either 10 minutes (n = 15), 3 hours (n = 8) or 6 hours (n = 8) following the blast. Subsequently lung perfusion was started. Data are expressed as mean ± SEM. Lung weight gain ΔW: was assessed over 150 minutes perfusion time. Statistical analysis of the entire curve data from different groups were compared by a Two-way ANOVA design. The p value represents the time dependent recovery after trauma: * p < 0.05 and *** p < 0.0001.

0 20 40 60 80 100 120 140

0

250

500

750

1000trauma: +3htrauma: +6h

trauma: +10min

time [min]

∆ ∆∆∆W

[mg]

Results

P

t

b

c

w

m

b

w

A) B)

Fdlpvf

77

erfusion of blasted rats` lungs with amiloride inhibited the endogenous fluid resorption in

hese lungs, resulting in a further increase of oedema formation, as compared to untreated

lasted lungs (Fig. 4.35 A, Tab. 4.12+13). These data suggest an involvement of sodium

hannels in fluid resorption during trauma. Similarly, tidal volume (Tv), airway resistance (RA)

ere deteriorated upon amiloride treatment (Fig. 4.35 B+C, Tab. 4.12+13). This aggravation is

aybe related to the massive oedema formation. The software only allows to calculate data

etween a defined minimal and maximal value. Therefore, the increase in airway resistance

as only recordable up to 80 minutes after start of perfusion (Fig. 4.35 C).

0 20 40 60 80 100 120 140

0

500

1000

1500

2000

2500

3000

3500trauma (+10min)trauma (+10min) + amiloride

***

time [min]

∆ ∆∆∆ W

[mg]

0 20 40 60 80 100 120 140

0.2

0.4

0.6

0.8

1.0

trauma (+10min)trauma (+10min) + amiloride

***

time [min]

Tv/T

v 0m

in

0 20 40 60 80 100 120 140

1.0

1.2

1.4

1.6

1.8

2.0

trauma (+10min)trauma (+10min) + amiloride

***

C)

time [min]

RA/R

A0m

in

igure 4.35: Role of sodium-channels after trauma. Rats were exposed to the pressure wave at a istance of 3.5 cm from the nozzle. The rats were sacrificed 10 minutes (n = 24) after the blast. The

ungs were treated either with (n = 8) or without (n = 16) amiloride (10-4 M) 4 minutes after start of erfusion. Data are expressed as mean ± SEM. A) ΔW: weight gain, B) Tv/ Tv0min: normalized tidal olume, and C) RA/ RA0min: normalized airway resistance. Statistical analysis of the entire curve data rom two groups were compared by a Two-way ANOVA design: *** p < 0.0001 (Tab. 4.13).

Results

78

4.4.3.2 The role of 2-adrenergic receptors after trauma

Stimulation of β2-adrenergic receptors in intact lungs, in vivo and ex vivo by synthetic agonists

(terbutaline, salmeterol, isoproterenol) and endogenous or exogenous epinephrine

respectively, was reported to increase fluid clearance 7,10,11. In addition, this effect is reported

to occur rapidly after administration of the substances and is completely blocked by unspecific

(propranolol) or specific β2-receptor antagonists 7,22. Therefore, we used the β2-adrenergic

receptor agonists terbutaline and formoterol to examine on the one hand the role of β2-

adrenergic receptors after blunt chest trauma and on the other hand the possible therapeutic

potential of these substances.

In these experiments, thoracic trauma-related deterioration of the lung parameters was

markedly improved by terbutaline and formoterol with respect to tidal volume (Tv), airway

resistance (RA) and weight gain (∆W) (Fig. 4.36 A-C, Tab. 4.12+13). With regard to the weight

gain of blasted rat lungs during perfusion, the terbutaline-induced reduction reached control

levels of non-traumatized lungs (∆W: 559 ± 368 mg and 579 ± 184 mg for traumatized rat

lungs plus terbutaline treatment and sham controls, respectively) (Tab. 4.12).

Taken together, these data indicate that β2-agonist treatment improves the lung weight and

the lung function parameters after blast exposure.

The terbutaline-induced amelioration in tidal volume (Tv) and airway resistance (RA) was

completely blocked by the unspecific β-receptor antagonists propranolol (Tab. 4.12+13). The

oedema formation was even more impaired than in non-treated blasted rat lungs (Tab.

4.12+13).

Amiloride blocked the terbutaline-related decrease in lung weight (Tab. 4.12+13), indicating

that β2-receptor related stimulation is dependent on amiloride-sensitive sodium transport, as

also proposed by several publications 8,22. However, statistical analysis by Two-way ANOVA

yielded also a significant difference between terbutaline plus amiloride versus amiloride

treatment alone, suggesting the implication of an amiloride-insensitive sodium transport.

As expected, and in contrast to the weight gain, amiloride did not influence the terbutaline-

induced improvement of the tidal volume (Tv), airway resistance (RA) (Tab. 4.12+13) which

thus may be considered as sodium channel independent β2-adrenergic effects.

Results

Ffw(4Ar*

A) B)

79

igure 4.36: Modulation of the trauma-induced pulmonary dysfunction by terbutaline and ormoterol. Rats were exposed to the pressure wave at a distances of 3.5 cm from the nozzle. The rats ere sacrificed 10 minutes (n = 37) following the blast. The lungs were perfused either untreated

n = 16) or treated with terbutaline (100 µM ) (n = 9) and formoterol (1 nM) (n = 12), respectively minutes after start of perfusion. Data are expressed as mean ± SEM. ) ΔW: weight gain, B) Tv/ Tv0min: normalized tidal volume and C) RA/ RA0min: normalized airway

esistance were monitored over 150 minutes perfusion time. Two-way ANOVA analysis: ** p < 0.0001 vs. trauma (+ 10 min) (Tab. 4.13).

0 20 40 60 80 100 120 140

0

250

500

750

1000

1250

1500trauma (+10min)trauma (+10min) + terbutalinetrauma (+10min) + formoterol

***

***

time [min]

∆ ∆∆∆W

[mg]

0 20 40 60 80 100 120 140

0.4

0.6

0.8

1.0

trauma (+10min)trauma (+10min) + terbutalinetrauma (+10min) + formoterol

***

***

time [min]

Tv/T

v 0m

in

0 20 40 60 80 100 120 140

0.8

1.0

1.2

1.4

1.6

1.8

trauma (+10min)trauma (+10min) + terbutalinetrauma (+10min) + formoterol

***

***

C)

time [min]

R A/R

A0m

in

Results

80

Table 4.12: Pharmacological intervention: Role of amiloride-sensitive sodium channels and 2-adrenergic receptor modulation.



treatment Tv

(ml)

RA

(cm H2O*s/ ml)

ΔW

(mg)

sham control 1.60 ± 0.15

(7)

0.33 ± 0.05

(7)

579 ± 184

(6)

0.66 ± 0.21

(13) §§§

0.47 ± 0.15

(15) §

1099 ± 441

(15) §

+ amiloride 0.56 ± 0.09

(4)

1.60 ± 0.46

(5) *** ⊗

2429 ± 2265

(8) *

+ terbutaline 1.08 ± 0.26

(9) ***

0.33 ± 0.04

(9) **

559 ± 368

(8) **

+ formoterol 1.12 ± 0.34

(12) ***

0.32 ± 0.04

(12) **

827 ± 381

(12)

+ terbutaline

+ propranolol

0.89 ± 0.51

(3)

0.40 ± 0.12

(3) #

3378 ± 2779

(5) ***, #

trauma

+ terbutaline

+ amiloride

1.09 ± 0.61

(4) *

0.35 ± 0.12

(4)

1534 ± 799

(4) #

Rats were exposed to the pressure wave at a distances of 3.5 cm from the nozzle. The rats were sacrificed 10 minutes (n = 55) following the blast. After preparation of the lungs, the blood-free perfusion was started and the lung function parameters were recorded. The lungs were perfused either untreated (n = 16) or treated with amiloride (10-4 M) (n = 8), terbutaline (10-4 M ) (n = 9), formoterol (1 nM) (n = 12), terbutaline + propranolol (10-4 M) (n = 5) and terbutaline + amiloride (n = 5) respectively 4 minutes after start of perfusion. Non-traumatized animals served as sham controls (n = 7). Data are expressed as mean ± SD and represent the values assessed after t = 150 minutes perfusion time (except ⊗: value at t = 80 minutes). Data were analysed by an unpaired t-test: § p < 0.05 and §§§ p < 0.001 vs. sham control; * p < 0.05, ** p < 0.01 and *** p < 0.001 vs. trauma; # p < 0.05, ## p < 0.01 and ### p < 0.001 vs. trauma + terbutaline.

Results

81

Table 4.13: Statistical analysis of different pharmacological interventions.

Tv

(ml)

RA

(cm H2O*s/ ml)

ΔW

(mg)

treatment tra

uma

+ te

rbut

alin

e

+ am

ilorid

e

traum

a

+ te

rbut

alin

e

+ am

ilorid

e

traum

a

+ te

rbut

alin

e

+ am

ilorid

e

+ amiloride *** *** ***

+ terbutaline *** *** ***

+ formoterol *** *** *** *** *** ***

+ terbutaline

+ propranolol ns *** ns *** *** ***

+ terbutaline

+ amiloride *** ns *** *** ns *** *** *** **

Statistical analysis of the entire curve data from two groups were compared by a Two-way ANOVA design: The analysed condition was the pharmacological modulation during perfusion: ** p < 0.01, *** p < 0.0001 and p > 0.05 was considered to be not significantly different.

Results

82

Table 4.14 illustrates the role of amiloride-sensitive sodium channels and 2-adrenergic

receptor modulation during trauma.



treatment Tv

(ml)

RA

(cm H2O*s/ ml)

ΔW

(mg)

sham control ↔ ↔ ↔

+ amiloride ↔

+ terbutaline

+ formoterol

+ terbutaline

+ propranolol ↔ ↔

trauma

+ terbutaline

+ amiloride ↔ ↔

Table 4.14: ↔ no change, upregulation, downregulation (compared to sham

control or trauma without treatment respectively).

Discussion

83

5. Discussion

The present study consists of three parts, with the primary aim to optimise a blast wave injury

model in order to study primary lung injury in terms of lung physiology, biochemistry and

immunology. The blast trauma model is intended to reproduce the clinical spectrum of injuries

seen in blast victims and is therefore used for studying the pathophysiology and the potential

treatment of thorax trauma in rats. Three approaches for studying the correlation between

clinically relevant physical blast properties and its biological impact were taken.

In the first approach, we assessed and characterized the functional consequences of in vivo

thorax trauma by means of combining the in vivo trauma model with the ex vivo isolated

perfused rat lung model. These studies included the investigation of pharmacological

interventions as to their therapeutic potential.

In the second approach, we examined the pathophysiological consequences including

biochemical responses of the lung following blast injury in vivo over time.

The third approach addresses immunodeficiency after thoracic trauma, which results in a high

number of infections. Due to its key role as modulator of pulmonary inflammation, the function

of the alveolar macrophage was assessed in vitro as to its phagocytic capacity, respiratory

burst and mediator release. These different approaches will be discussed separately

(summarized in Figure 5).

in vivo blast thorax trauma• standardization• monitoring• pressure wave intensity injury

ex vivo lung perfusion• lung function• weight gain• therapeutical

interventionin vivometabolism

in vitroimmune status

primary alveolar macrophage function

pathophysiological consequences

5.1

5.2 5.35.4

in vivo blast thorax trauma• standardization• monitoring• pressure wave intensity injury

ex vivo lung perfusion• lung function• weight gain• therapeutical

interventionin vivometabolism

in vitroimmune status

primary alveolar macrophage function

pathophysiological consequences

5.1

5.2 5.35.4

Discussion

84

5.1 A laboratory model for studying primary blast injury

Recently, the management of blast injured patients has become a problem not only for civilian

trauma surgeons, but also for military physicians, due to both an increase in the number of

accident-related blunt trauma and an increase in terrorist activity. Blast wave injury leads to a

unique but poorly understood spectrum of injuries 119. One of the cardinal clinical problems in

the early posttraumatic phase are the missing clinical signs of injury 114,115 and therefore the

outcome at this stage is unpredictable. Thus, some patients may present acute damage, while

others appear uninjured, but develop respiratory failure after 12 - 24 hours 128. Damage

caused by the pressure wave alone (primary blast injury) is not commonly seen after air-blast,

since most of the casualties are additionally affected by secondary and tertiary injuries.

However, primary blast injuries are estimated to contribute to 47 - 57% of the injuries in

survivors and to 86% of fatal injuries 99. Therefore, studying primary blast injury implies the

following demands: 1. discharge of high energy pressure waves comparable with high energy

chemical explosives, 2. practicable in a safe laboratory environment, 3. high reproducibility

and 4. reproducing the spectrum of injuries seen in blast victims.

5.1.1 Rationale for the development of the blast thorax trauma rat lung model

As already described in the chapter “Introduction”, the lungs are almost always affected by

blast injuries 103,107-109. Furthermore, it has been widely shown that the involvement of the

thorax in polytrauma patients is a major negative determinant for the long-term consequences

after trauma 93,104,112. This emphasizes the need for further basic understanding of the

pathogenesis especially of the lung. Therefore, we used the primary blast injury model

previously developed by Irwin and coworkers 117 to study primary lung injury (the blast wave

generator is described in detail in the chapter “Materials and Methods”). This model develops

a mechanical explosive that simulates blast pressure waves from high explosives and

produces the types of injuries commonly seen in blast victims.

5.1.2 Standardization and monitoring of the blast pressure wave

Attempts to further standardize the pressure wave-induced injury were successfully

undertaken to improve the reproducibility.

Different wave types, characterized by the rate of energy application at the impact side and

the degree of compression produced, have different capacities to injure internal organs 106.

Thus, we further modified the blast wave generator to guarantee a standardized, fast and

complete rupture of the diaphragm and therefore a fast discharge of the blast pressure wave.

Discussion

85

This was achieved by two modifications: First, by using an additional compressed air bottle

located as close as possible to the pressure reservoir, and second by pressing the diaphragm

along a predetermined breaking point between the pressure reservoir and the nozzle, instead

of only fixing it with screws.

We further installed pressure transducers at different positions, including the pressure

reservoir, at the right and left side at the rat thorax level. This Setup allowed a thorough

monitoring and evaluation of the key physical blast wave properties and a comparison of the

data over time. Our blast waves are comparable in form and duration to blast waves from

other groups using such a Setup 116,117,124,174. However, our pressure wave monitoring at the

rat thorax level revealed a complex wave rather than a simple wave form (Friedlander wave

form) possibly due to reflections from the rat surface. Furthermore, a comparison of the data

from the two sensors situated on the right and left side of the rat suggests that the type of

turbulences of the pressure wave may be due to the confinement of the blast wave by

reflective surfaces of the nozzle. In addition, the achieved pressure peaks at comparable

distances and the similar spectrum of injuries are lower as compared to those obtained by

others 116,117,124,174. Because the exact position of the pressure transducer is not mentioned in

these studies, we can only speculate on another position than the one directly about the

thorax e.g. in the nozzle. Since complex wave forms are usually expressed by their impulse

(integral of pressure change over time) rather than by their maximum peak overpressure

values alone 102, we calculated the area under the pressure wave curve. The pressure

measured at the rat thorax level decayed rapidly as it expanded from the nozzle. Our

measured mean overpressure, i.e. the duration of the impact and the area under the curve

correlated significantly with the distance. This allows an exposure of the rats to different blast

intensities by simply varying the distance between the blast wave generator and the animals 116,117,124,174. The necessity of such a thorough monitoring becomes further apparent when

comparing the pressure wave data from survivors versus non-survivors, blasted at the same

distance from the nozzle. In the non-survivor group the rupture pressure of the diaphragm was

significantly higher than in the survivor group, resulting in a higher impact on the animal. Since

the Mylar polyester films show differences in their physical characteristics (supplier

information) and since different lots of the Mylar polyester films were used in the present

study, the observed variances in the rupture pressures may be attributed to these material

differences.

Discussion

86

5.1.3 Standardization of the blast thorax trauma in rats

Our rats were inhalation-anaesthetized and non-intubated during the blast pressure wave

exposure. This Setup is not only easy to handle but also displays the clinical situation and

avoids additional ventilation-related modulations during blast exposure. In addition, the

vagolytic response from the blast wave injury can be investigated. The disadvantage is the

lack of respiratory control and therefore the risk of prolonged apnoea and probably death.

With regard to a possible involvement of the breathing situation in the extent of the induced

lung injury, the animals were all blasted at the top of expiration. This was achieved by an

automatic triggering of the valve during spontaneous breathing at halothane anaesthesia.

However, the anesthetic has been shown to affect the cardiopulmonary physiology 175 and a

modulation of the autonomous nervous system cannot be ruled out. In some animals,

anaesthesia could not be achieved maybe because of stressing of the animals in advance.

Nevertheless, no correlation between the mortality and the reduced pre-blast breathing

frequency, with regard to a prolonged time of apnoea was observed. Mortality occurring

immediately after the blast pressure wave exposure was mostly due to vessel rupture.

Furthermore, air embolism is thought to be the predominant cause of the immediate death

after the blast wave injury 104,119,120. Because of the rapid air absorption in the capillaries, air

embolism is difficult to detect but was seen histologically in some cases. Recently it was

shown that air embolism presenting in blast victims is not a mere ventilation-induced artefact 119. Blast exposure was usually followed by an immediate period of apnoea. In line with the

findings of others our rats had a low breathing frequency immediately after the blast pursued

by fast shallow breathing 116,117,122,124,125,174. The characteristic triad of apnoea, bradycardia,

and hypotension has been attributed to a vagal reflex 124,125.

Taken together, despite technical variances we succeeded in establishing a reproducible,

standardized and well-monitored pressure wave exposure. Thus, this blast wave generator

allows a safe method to study primary blast injury without contact with the thoracic cage.

5.2 Combination of the in vivo thorax trauma model with ex vivo lung perfusion

The direct functional consequences of an in vivo thoracic trauma were assessed in our ex vivo

isolated perfused rat lung model, which has, to the best of our knowledge, not been described

previously. This model is a standardized and well-established tool in our laboratory 1,157,176. In

order to study primary lung injury, the isolated perfused rat lung represents a system much

less complicated than the whole animal, while preserving most of the integrity of the organ. As

an experimental approach it stands between in vivo and in vitro experiments.

Discussion

87

5.2.1 Match of physical and biological impact

The most striking pathological changes were pulmonary haemorrhage and oedema,

sometimes called “blast lung” 177, as determined by BAL red blood cell counts and ex vivo

weight gain or lung wet to lung body weight ratio and confirmed by histological examinations.

These two characteristics of “blast lung” were macroscopically quantified by lung injury scores

(haemorrhage and oedema score) that have been shown to posses high validity. Because the

air in the alveoli is easily compressible, blast pressure waves disrupt the alveolar capillary

membrane and blood is driven from the capillaries into the alveolar space by two mechanisms:

direct compression of the lung and transfer of the blast wave through the body, causing injury

by spalling, implosion and inertia 105,106. Alveolar haemorrhage and oedema formation are

common findings 117,118,122,174 and occurred regularly in our model with increasing intensities

with higher pressure wave intensities.

The ex vivo perfusion of the blasted rat lungs showed that thoracic trauma deteriorates the

lung function with respect to the airway parameters and leads to oedema formation,

depending on the distance from the nozzle and therefore on the pressure wave intensity. This

included not only the tidal volume measured immediately after the onset of the ventilation, but

also an increase of the airway resistance and the weight gain over the perfusion time. The

closer the animals were positioned to the nozzle, e.g. 3.5 cm versus 5 cm, the higher was the

necessary inspiratory pressure (PIP) to ventilate these lungs in order to achieve a vital tidal

volume of at least 1 ml. However, despite the correlation between the distance and the initial

tidal volume (Tv), at the end of the perfusion no distance-dependency could be measured.

Thus, we cannot exclude an involvement of the ex vivo ventilation in the observed changes

during perfusion. Our ventilatory strategy was based on the hypothesis that a minimal vital

tidal volume of 1 ml is necessary for survival of the animals. One has to take into account that

the isolated perfused rat lung model has been optimised for uninjured lungs, in terms of both

negative pressure ventilation (overpressure ventilation during surgical procedure) and

constant pressure perfusion 1,157. This is of importance, since the mechanical ventilation of an

injured lung may induce a secondary injury called “ventilator-associated lung injury”.

Mechanical ventilation is one of the ultimate life-supporting technologies in the therapy of

critically ill patients with respiratory failure. However, data accumulated over the past decade

provide strong evidence that ventilatory strategies associated with excessive end-inspiratory

stretch and/ or collapse/ recruitment of lung units can cause further injury and perhaps

promote the development of multi-organ failure 178-180. Forms of the ventilator-associated lung

injury are barotrauma, volutrauma, atelectotrauma, and biotrauma. In a clinical trial it has been

shown that mechanical ventilation with a lower tidal volume than the one traditionally used

results in decreased mortality and increases the number of days without ventilator use 98.

Discussion

88

Different ventilatory strategies (high PEEP versus low PEEP, conventional mechanical

ventilation versus high frequency ventilation) have been investigated by many authors in

different lung injury models 181-183. However, especially the interpretation of results from ex

vivo perfused lung models must take into account the fact that the degree of collapse and

reopening is magnified, because there is no chest wall and hence at a PEEP of 0,

transpulmonary pressure is also 0. This is not comparable with the in vivo situation.

Taken together, the correlation of pressure wave monitoring together with the injury score, the

BAL red blood cell counts and the ex vivo data collected in the isolated traumatized and

perfused lung indicate that physical and biological impact may be matched

5.2.2 Time-dependence of the lung dysfunction

In addition to the distance-injury correlation, we could also show a time-dependence of

oedema generation. Isolation and perfusion of the lung immediately after blast led to much

more weight gain than 3 and 6 hours after blast exposure. These results provide evidence that

the alveolar capillary barrier membrane is at least in part intact and that oedema resorption

and/ or stop of oedema generation took place in these lungs (discussed in detail in 5.3.4).

Interestingly, no further improvement could be seen between 3 and 6 hours after blast,

indicating that the main resorption capacity and/ or repair occurred within the first 3 hours.

Indeed, many experimental studies indicate that the alveolar epithelium is remarkable

resistant to injury, particularly compared to the adjacent lung endothelium. Even when mild to

moderate alveolar epithelial injury occurs e.g. hyperosmolar injury 184 or leukotriene B4-

induced PMN infiltration 185, the capacity of the alveolar epithelium to transport salt and water

is often preserved.

Also, the tidal volume (Tv) measured immediately after the onset of ventilation was

significantly improved 3 and 6 hours following the blast, reaching almost control levels of non-

traumatized animals. This improvement is accompanied by lower necessary inspiration

pressure (PIP) to achieve 1 ml (see 5.2.1) and may be due to the reduced weight gain.

Despite gentle standard ventilation of these lungs (3 and 6 hours after blast) there was also a

measurable decline of the lung mechanics over time. This observation provided further

evidence for a higher susceptibility of injured lungs towards ventilator-associated lung injury,

even at standard ventilation-pressures. This is important since the success or failure of the

clinical treatment of primary blast injury may depend not only on the judicious use of

resuscitative fluids but also on the respiratory support 128.

Discussion

89

5.2.3 Pharmacological modulation of the injury

The present study suggests that even after severe mechanical injury such as blast injury, the

alveolar-capillary barrier capacity together with the alveolar epithelial fluid transport is at least

in part maintained.

5.2.3.1 Role of the amiloride-sensitive sodium channels

Several studies have shown that transport of sodium is a primary mechanism driving alveolar

fluid clearance (reviewed by Matthay and coworkers 5,172, Sartory and Matthay, 6), although

further work is needed to determine the role of chloride in vectorial transport across the

alveolar epithelium 16. Thus, the objective of the present study was to test the effect of the

known sodium transport inhibitor amiloride on alveolar liquid clearance in our primary blast

injury model. Amiloride was given into the perfusate in a commonly used concentration of 10-4

M 7,22,186. The results of our amiloride studies demonstrate a more than double increase in

oedema formation during perfusion in blasted lungs perfused with amiloride, as compared to

non-treated blasted lungs. Thus, even if limited, a partial integrity of the sodium channels and

therefore a fluid resorption capacity after trauma has been suggested. Such an amiloride-

sensitive fluid resorption could be the underlying mechanism for the reduced weight gain

observed ex vivo 3 and 6 hours after blast exposure (discussed in detail in 5.2.2 and 5.3.4).

The presence of regulable, amiloride-sensitive sodium channels leading to fluid resorption in

the intact mammalian alveolar epithelium has been reported by many authors using different

models and different species 7,8,21. Inhibition of the endogenous fluid resorption in traumatized

rat lungs by amiloride furthermore resulted in a marked deterioration of the airway mechanics.

This may be due to the massive oedema formation rather than to a negative effect of amiloride

itself.

5.2.3.2 Role of ββββ-adrenergic receptors

Upon infusion of terbutaline (10-4 M) or formoterol (1 nM) in the perfusate of blasted rat lungs,

the weight gain during perfusion was significantly reduced. Upon terbutaline treatment, almost

control levels of non-traumatized rat lungs were reached. In rats the response to β-adrenergic

agonists included an increase in sodium unidirectional flux out of the alveoli and stimulation of

fluid absorption 21,186,187. These observations are in line with others, that report that terbutaline

increased sodium transport in alveolar type II cell monolayers 188,189 as well as in vivo 8.

Terbutaline must have stimulated the transport through amiloride-sensitive pathways, since

amiloride reduced the terbutaline-related fluid resorption in this setting. This result is in

accordance with prior studies of the effect of amiloride on the terbutaline-induced increase in

alveolar liquid clearance 7,8,22. With regard to the increased but not complete fluid resorption

Discussion

90

and/ or stop in oedema generation in vivo within the first six hours after blast exposure (5.2.2),

these data further indicate that the fluid transport capacity after traumatic injury may be

upregulated by β-adrenergic agonists. However, infusion of amiloride plus terbutaline resulted

in decreased oedema formation in this injury model than amiloride treatment alone. The most

reasonable explanation for this finding is that terbutaline stimulates sodium transport through

both amiloride-sensitive and insensitive pathways. Similar findings were reported by Goodman

and colleagues 21 and confirmed by others 7,8. In the isolated perfused rat lung, amiloride

alone reduced sodium transport by 30%, terbutaline increased it by 59%, but terbutaline plus

amiloride decreased clearance only by 14%. Another explanation could be a permeability

decreasing action of terbutaline, leading to a reduction of oedema generation after trauma. It

has been reported that isoproterenol attenuates the thrombin-induced increase in endothelial

permeability in vitro 190.

Thoracic trauma-related deterioration of the airway mechanics was markedly reduced by β-

adrenergic agonist treatment, with respect to tidal volume (Tv) and airway resistance (RA). The

reduced airway resistance is probably due to the β2-receptor-mediated airway smooth muscle

relaxation. Improved pulmonary mechanics upon β2-agonist treatment have also been shown

in fetal lambs 191. Particularly, the “longer”-acting β2-agonists such as salmeterol and

formoterol represent a therapeutic advance in the management of asthma 192, especially in

combination with inhaled corticosteroids 193. Furthermore, the positive influence on the airway

compliance (data not shown) and therefore on the tidal volume observed in these studies may

be due to β-agonist-related modulation of the pulmonary surfactant synthesis 191,194,195. In

contrast to the weight gain, amiloride did not affect the terbutaline-related effects on airway

mechanics, which therefore may be considered as sodium-channel independent β2-agonist

effects.

The experiments with the unspecific β-receptor antagonist propranolol indicate that the effect

of terbutaline on airway mechanics in blasted rat lungs is mediated through β-receptors.

However, oedema formation and vascular resistance (data not shown) were even more

impaired upon terbutaline plus propranolol treatment, as compared to non-treated blasted rat

lungs. The block of β2-receptors causes the well known side effects, e.g. vasoconstriction,

delayed response to hypoglycemia in diabetic patients, and bronchoconstriction (reviewed by

Pozzi, 196), which could be involved in the observed effects. A propranolol-related

vasoconstriction in turn might be a possible explanation for the observed increase in oedema

formation. Because propranolol also blocks possible endogenous epinephrine-related-effects

during trauma, this might at least in part explain the observed deterioration. We did not

specifically attempt to block possible trauma-related epinephrine effects, thus we can only

speculate about this. However, the basal β-adrenergic tone in anaesthetized sheep 7 and rats

Discussion

91

8 is probably not an important factor in the normal fluid clearance, because propranolol alone

did not have an effect on lung liquid clearance.

Taken together, these results demonstrate that the integrity of the sodium channels is at least

in part maintained after blast injury in this model. Thus, increased alveolar fluid clearance

might be the underlying mechanism for the observed fluid resorption in vivo within the first six

hours after trauma (discussed in detail in 5.2.2 and 5.3.4). Furthermore, β2-adrenergic therapy

increased not only the oedema resorption and/ or the oedema generation after trauma, but

also improved the airway mechanics by direct β2-adrenergic effects independent of β-receptor-

related fluid resorption. Thus β2-adrenergic substances like terbutaline or formoterol appear to

be beneficial after primary blast injury in this model.

5.3 Thorax trauma-related pathophysiological changes in vivo over time

After an initial non-fatal injury, thoracic trauma is considered to play a pivotal role in the

development of single-or multi-organ failure 93,112, probably making the patient more

susceptible to infections 95. However, the aetiology of pulmonary contusions following

exposure to blast is not clear 118. In the present study, we investigated the sequence of

alveolar injury in vivo over time, with regard to pathological findings and molecular responses

to injury as e.g. inflammatory mediator release.

5.3.1 Structural integrity of the lung tissue after blast injury

At autopsy, none of our rats showed external signs of injury 116,117,122,124. The pulmonary

contusion produced after blast exposure was evident as areas of haemorrhage on gross

inspection of the lungs immediately after blast, with increasing intensities with higher pressure

wave intensities (5.2.1). Clinically, these contusions may be difficult to detect in the absence of

overt symptoms such as breathlessness, haemoptysis or a decreased PaO2, and the lung may

continue to accumulate extravascular lung haemorrhagic or oedema fluid, that may lead to

deterioration in lung function and in the condition of the patient. The implication of these

clinical sequelae is that haemorrhage may continue within the injured lung and that areas of

the lung initially apparently undamaged may later bleed and show other signs of damage 118.

However, our experiments showed that after an initial massive blast-induced alveolar

bleeding, the bronchoalveolar lavage (BAL) red blood cell counts continuously decreased over

time and were no longer detectable after 96 hours. Together with the time-dependent

resolution of the trauma-related alveolar protein content, these results suggested an

enormous clearance capacity and/ or repair of the lung, despite of the mechanically induced

Discussion

92

cell damage in this setting. In this context, Kim and coworkers 166 reported a net absorption of

intact albumin via saturable receptor-mediated transcellular endocytotic processes in primary

cultured rat alveolar epithelial cell monolayers. The resolution of the alveolar haemorrhage is

also in line with clinical situations of alveolar haemorrhage 167 and may be related to the

increasing amounts of alveolar macrophages detected in this “later” phase after blast

exposure (discussed in detail in 5.4.1). Indeed, after haemorrhage, large quantities of

haemoglobin and ingested red blood cells can be detected in alveolar macrophages 197, in

accordance with our observations on the microscopic level (data not shown). Furthermore,

haemosiderin-laden alveolar macrophages are a common finding in patients with alveolar

bleeding 197,198.

Markers of endothelial injury such as the von Willebrand`s factor (vWF) antigen have been

used to reflect systemic endothelial cell activation and/ or damage in patients with acute lung

injury 199. Because of the lack of an available commercial kit for the detection of rat von

Willebrand`s factor (vWF), in the present study the profound release of the cytosolic enzyme

lactate dehydrogenase (LDH) into the BAL was used as a marker for the cellular integrity after

blast. Thanks to the long half life of the enzyme, LDH release may be a suitable marker for

trauma-induced cell lysis. A recent study demonstrated a strong haemoglobin

immunoreactivity of the oedema fluid within the alveolar space of blast victims 119.

Erythrocytes are known to express the LDH-1 isoform and thus may also be considered as a

source for the observed LDH release. Sustained elevated LDH levels might also be an

indication for a delayed cell death after trauma. Morphologic changes from haematoma to

necrosis to gangrene have been observed after primary non-perforative intestinal blast injuries

in rats 200.

Based on the determined wet to dry weight ratios, maximal oedema formation was detected

within the first 90 minutes after blast exposure. The reduced wet to dry weight ratios assessed

from 3 hours after blast exposure on might suggest oedema resorption and/ or a stop in

oedema generation. In contrast, although not significant, our calculated lung wet to body

weight ratios (LWR) and thus the corresponding injury quotient (Qi) indicated a secondary

deterioration between 6 and 24 hours after blast. This was attributed to a secondary increase

in lung wet weight resulting in an increased LWR and Qi. This increase may be explained by

increasing amounts of infiltrating neutrophils during this time (discussed in 5.3.3). Thus, we

can not discriminate between an increased oedema resorption and an increase in pulmonary

infiltrates (discussed in detail in 5.3.4).

Discussion

93

5.3.2 Mediator release in response to primary blast injury

A generalized host inflammatory response is necessary to orchestrate the maintenance or

recovery of tissue repair and immune competence, following severe injury or infection.


patients is strongly associated with a post-traumatic imbalance of the immunologic system 146.

For the outcome of patients with multiple injuries it seems very important to gain an

inflammatory mediator homeostasis as fast as possible. If the pro-inflammatory potential

predominates, systemic inflammatory response syndrome (SIRS) may be the consequence; if

the counter-regulatory, anti-inflammatory mediators predominate (CARS), immunosuppression

(“immunoparalysis”) with ineffective eradication of microorganisms and septic complications

may follow 142.

There were three major reasons for us to study the involvement of mediator release after

primary blast injury in this setting: First, to gain more knowledge about the pathogenesis of the

injury; second, to determine whether the balance between pro- and anti-inflammatory mediator

release is maintained in this setting; and third, to identify markers, either in the early or in the

“late” posttraumatic phase, that may be suitable predictors of the overall outcome.

Since measurement of mediators directly in the lungs has been shown to be more valuable

than measurement in plasma or serum 201, we determined the release of different mediators in

the bronchoalveolar lavage (BAL) fluid of traumatized rat lungs.

5.3.2.1 Thorax trauma-related eicosanoid release

Brückner and colleagues 134,138 reported a pattern-related release of prostanoids which was

rather pronounced in polytrauma associated with damage of the lung. A similar pattern has

been found in patients who underwent thoracotomy plus additional surgical damage to the

lung 136. Therefore, these authors suggested the prostanoid release to be specific for the

involvement of the lung. In line with these results, we detected a significant immediate release

of both thromboxane and prostacycline within the first 10 minutes after blast exposure,

followed by a marked decrease reaching control levels within 90 minutes. Prostanoids are

synthesized from arachidonic acid by the constitutively expressed cyclooxygenase (COX)-1

and the inducible COX-2 enzymes. Among others, the activity of phospholipase A2 regulates

prostanoid production in many cell types 202. Elevated plasma levels of type II a phospholipase

A2 (PLA2) have been found in inflammatory diseases and trauma and are suggested to bear

prognostic value for outcome 137,138. Thromboxane and prostacycline play an important role in

the control of haemodynamics. Prostacycline is predominately released from endothelial cells

as a result of endothelial injury or mediator stimulation. In contrast to prostacycline, the

release of thromboxane from alveolar macrophages, platelets or endothelial cells is believed

Discussion

94

to require specific stimulation 203,204. Thromboxane-induced platelet aggregation and

vasoconstriction are antagonized by prostacycline, and thus the ratio of thromboxane to

prostacycline is of importance. In addition, thromboxane acts as a potent bronchoconstrictor.

Thus, increasing release of thromboxane in polytrauma patients with thorax involvement has

been suggested to be a predictive marker for the development of pulmonary dysfunction after

trauma 136. Despite of pressure wave-induced lung injury, the crucial metabolic activities of the

lung, such as thromboxane degradation as well as extrapulmonary clearance of 6-keto PGF1α

seemed unaltered in the present study, which is in accordance with other reports 136. The

metabolic clearance occurs via two steps: 1. selective, carrier-mediated prostaglandin uptake

across the plasma membrane and 2. non-selective oxidation inside the cell. However, only

TXA2, PGE1, PGE2, PGD2, and PGF2α are transported across the cell membrane. There is

substantially less clearance of PGA2 and essentially none of PGI2 or 6-keto PGF1α (reviewed

by Schuster, 205). With regard to the lack of COX-2 induction within the first 6 hours after blast

(RT-PCR analysis, data not shown), a secondary increase of these mediators in the “late”

posttraumatic phase in this setting appeared rather unlikely.

5.3.2.2 Thorax trauma-related cytokine release

Trauma, shock, and infection initiate a complex inflammatory response in which the pro-

inflammatory cytokines tumor necrosis factor (TNF)-α, interleukin (IL-1)-β, and interleukin (IL-)

6 are thought to play a pivotal role. Overwhelming synthesis of pro-inflammatory cytokines

after trauma has been correlated with poor outcome 168,169.

In the present study IL-6 levels in bronchoalveolar lavage (BAL) fluids were significantly

elevated 6 hours after trauma, as compared to non-traumatized control lungs, returning to

baseline between 12 and 96 hours. This early increase is in accordance with plasma IL-6

levels obtained from polytrauma patients 135,142. IL-6 production is induced in part by TNF-α

and IL-1β, and it has been proposed that IL-6 “integrates” signals produced early in the

inflammatory response 206. Some data support the concept that the production of TNF-α and

IL-6 is regulated independently 168. In contrast to TNF, IL-6 has been reported to be a marker

of the intensity of the injury, bearing some prognostic value for the survival in the early

posttraumatic phase 135. Recent reports support a crucial anti-inflammatory role for IL-6 by

controlling the levels of other pro-inflammatory cytokines 150 and by inducing significant

release/ formation of soluble tumor necrosis factor receptors (sTNFR) 151.

In comparison to non-traumatized control lungs, the BAL TNF-α levels of traumatized lungs

were significantly elevated from 3 hours after trauma on and returned to basal values after 96

hours. Most likely, the TNF-α found in BAL was produced by activated pulmonary

macrophages. A major function of TNF-α in the lung appears to be upregulation and induction

Discussion

95

of endothelial adhesion molecules ICAM-1 and E-selectin, respectively, both of which are

critical for the early steps required in neutrophil adhesion to the activated endothelium and

eventually in transmigration of neutrophils into the alveolar compartiment 69. Indeed, a marked

increase in neutrophil infiltration into the alveoli was observed 6 and 12 hours after blast

exposure in this setting (discussed in 5.3.3).

Simultaneous determination of TNF-α in plasma showed a striking difference from levels in

BAL fluids; plasma TNF-α was not detectable within 96 hours after trauma neither by ELISA

nor by the WEHI-bioassay (data not shown). This suggests that TNF-α is secreted locally in

the lung by pulmonary macrophages. The lack of any increase in plasma TNF-α levels is in

line with clinical data obtained from trauma patients suffering from different traumatic injuries 135. Even in patients with ARDS after trauma, shock or sepsis, plasma cytokines were neither

increased in most patients nor did they correlate with the development or severity of ARDS 169.

Since experimental haemorrhage and blunt trauma do not lead to increased circulating TNF-α

levels in pigs, the authors of this study have suggested that TNF-α plays no role in the

pathogenesis of shock in their model 141. In contrast to sepsis, where bacteremia and

endotoxemia result in the systemic appearance of cytokines 207, blunt trauma may cause the

release of cell-associated TNF-α only at the local side of trauma. However, other animal 208 as

well as clinical studies 168 reported that the induction of TNF-α release after trauma is

significantly enhanced when haemorrhagic shock is also present. These authors suggested

that shock is the main factor leading to the production and release of TNF-α by macrophages.

Nevertheless, the levels of secreted BAL TNF-α after primary blast injury in this setting were

relatively moderate. Some publications have provided evidence for the presence of 55- and

75-k Da soluble tumor necrosis factor receptors (sTNFRs) in injured patients in the absence of

measurable plasma TNF-α, with a strong correlation with mortality 148,149,209. Although it has

been reported that sTNFRs affect TNF functions even at low TNF concentrations by stabilizing

and augmenting some of its activities by a kind of controlled release 210, it is assumed that

shedding of sTNFRs in response to high TNF concentrations could serve as a mechanism for

binding and inhibiting overwhelming TNF levels and localizing the inflammatory response 211.

In the present study, sTNFR (p55 and p75) levels in BAL fluids of traumatized rat lungs were

significantly elevated within 90 minutes after trauma. This early release is in accordance with

plasma sTNFR levels obtained in polytrauma patients 142,149. This suggests that thoracic

trauma initiates the rapid release of preformed receptors on the cell surface of injured tissues

or activated leukocytes, or both. Similarly to the measured BAL TNF levels, also sTNFR levels

were comparatively low as compared to plasma sTNFR levels measured in trauma patients 142,148,209. This may be in part explained by a dilution effect due to the BAL method (for

cytokines by about 100-fold, 206). In contrast to other studies 142,148, data from the present

Discussion

96

study showed that receptor levels return to non-traumatized control levels, indicating a peak-

like profile of sTNFR instead of sustained elevation. Since clinical studies reported declined

sTNFR levels in patients who recovered from injury 149, the decrease in BAL sTNFR levels in

the present study may be correlated with a recovery from primary blast injury. The correlation

between each p55 and p75 with IL-6 release within 96 hours after blast supports the

hypothesis that IL-6 may be involved in the regulation of both sTNFR p55 and p75 levels. This

hypothesis is supported by the report of Tilg and coworkers 151, who observed elevated

sTNFR p55 levels in cancer patients after administration of recombinant IL-6. In the present

study the 55-kDa and 75-kDa sTNFR levels were both significantly elevated within 90 minutes

after trauma, which is in agreement with clinical data obtained from severely injured patients

described by Tan and colleagues 212. In contrast, other authors reported an early shift towards

sTNFR p55 release 142. The factors that may be responsible for increased or prolonged

release of sTNFR after trauma have not been determined yet, although proteases and IL-6

have been discussed 151,152. Neutrophil-derived proteases have been shown to be elevated,

predominantly in patients with isolated thorax trauma 135. With regard to our results, we can

only speculate about an involvement of locally produced IL-6 in the early appearance of

sTNFR subtypes in the injured lung after primary blast injury. Whether sTNFR levels in the

BAL fluids may be the response to local TNF production and/ or the reason why BAL TNF

levels were relatively moderate and plasma levels were even not measurable, can not be

explained in the present study.

Furthermore, we measured IL-10 as a counterregulatory cytokine that inhibits cytokine

production by stimulated macrophages 213. IL-10 is detectable in ARDS BAL fluids, but the

concentrations are very low (10 - 20 pg/ ml), as compared to other cytokines 206. In agreement

with this, our data demonstrated low and even significantly decreased BAL IL-10 levels at any

measured time point within 24 hours after blast exposure, as compared to base line levels of

non-traumatized control lungs, and with recovery only after 96 hours. Such low IL-10 levels

may favour cytokine production in the alveolar environment in this setting. In contrast,

Henseler and colleagues 142, showed elevated plasma IL-10 values within 3 hours after

trauma in patients suffering from multiple injuries. It has also been reported that the lung

perpetuates a constitutive level of anti-inflammatory cytokine IL-10, which is not enhanced

after aerosol exposure of endotoxin, whereas systemically, intraperitoneal delivery of

endotoxin resulted in an increase in circulating IL-10 150. Initiation of coagulation by tissue

factor (TF) has been shown to be a powerful regulator of local inflammatory responses 214.

With regard to this, Miller and colleagues 215 reported that blockade of TF-FVIIa complex

protects the lung from injury by LPS in part by reducing local expression of pro-inflammatory

cytokines and local elaboration of interleukin (IL)-1beta, IL-6, and IL-10. Furthermore, it has

Discussion

97

been shown that trauma induces TF expression on monocytes 216 which may be one reason

for the observed decline in BAL IL-10 after blast injury in this model.

5.3.2.3 Thorax trauma-related NO release

Following blast exposure we could detect nitric oxide related nitrite and nitrate with a

significant maximum 12 hours after trauma. However, RT-PCR analysis revealed no induction

of the inducible nitric oxide synthase (iNOS) within 12 hours (data not shown). These results

indicate an initial trauma-induced NO release due to the constitutive endothelial NOS rather

than due to an involvement of the iNOS during the first four days following blast exposure in

this setting. Whether this increase reflects injury or a signal mechanism to initiate repair

process remains to be established.

5.3.3 Thorax trauma-related chemokine release and inflammatory cell infiltration

Thoracic trauma resulted in a pronounced CINC-3 production and neutrophil infiltration into

the alveolar space after 6 hours, whereas these hallmarks of inflammation were absent in the

bronchoalveolar lavage (BAL) fluids of non-traumatized control lungs. Alveolar macrophages

are thought to be the major source of α-chemokines such as CINC-3 in the airspaces 170,206

and produce CINC in response to LPS, both in vitro 217 and in vivo 218,219. Thus, CINC has

been reported to contribute to the influx of neutrophils and their activation during acute lung

injury 170,220.

Likewise, matrix metalloproteinase (MMP)-9 release was detectable in BAL fluids 6 hours after

blast exposure. MMP-9 is preferentially expressed by inflammatory cells 171 and thus may be

correlate with the observed neutrophil infiltration at this time. In addition, the release and

activation of MMP-2 was measurable within 24 hours after blast, whereas no MMP-2 was

detectable after 96 hours and in BAL fluids from non-traumatized rats. MMP-2 has been

reported to be preferentially secreted by fibroblasts and epithelial cells 171 and may therefore

be released after trauma-induced cell damage and/ or participate in tissue remodeling

associated with pathological situations such as thoracic injury. Recently, matrix

metalloproteinases (MMPs) and the specific tissue inhibitors of metalloproteinases (TIMPs)

have been shown to participate in both parenchymal destruction and repair process resulting,

in extracellular matrix (ECM) remodeling 161,221-223.

Discussion

98

5.3.4 Oedema generation and resorption after thorax trauma: Comparison of the ex vivo and in vivo results

Based on the increasing wet to dry weight ratios, maximal oedema formation was detected

within the first 90 minutes after blast exposure. The reduced wet to dry weight ratios

determined from 3 hours after trauma on might suggest, but were not inevitably due to, an

oedema resorption and thus may result from pulmonary infiltrates, protein sediments,

membrane regeneration and others. Although not significant, the calculated lung wet to body

weight ratios (LWR) and therefore the corresponding injury quotient (Qi) indicated a secondary

deterioration between 6 and 24 hours after blast, arguing against an increased oedema

resorption during this time. This is further supported by an increased infiltration of neutrophils

into the alveolar space starting from 6 hours after trauma on. However, our ex vivo results

demonstrated that the alveolar capillary fluid resorption capacity is at least in part maintained

after blast pressure wave exposure since β2-adrenergic therapy stimulated and amiloride

treatment inhibited the fluid resorption in this model. The reduced weight gain assessed 3 and

6 hours after blast, as compared to the immediately perfused lungs can therefore be related to

both an increased repair, a stop in oedema generation and elevated oedema resorption.

Whether these effects are related to possible endogenous epinephrine-induced actions can

not be explained because we did not specifically attempt to block possible trauma-related

epinephrine effects in this setting.

5.4 Thorax trauma-related impairment of alveolar macrophage function in vitro

5.4.1 Alveolar macrophage population after trauma

The number of alveolar macrophages (AM) recruited from the alveolar space after thoracic

trauma was significantly different when compared to non-traumatized controls. Whereas the

number of AM was decreased within the first 24 hours after trauma, there was an up to 1.5-

fold increase after 96 hours. Since we measured elevated LDH levels after trauma (discussed

in 5.3.1), the decrease in the AM number may be the consequence of increased cell death

due to blast exposure. Furthermore, increased trauma-related adhesion of the AM to the

endothelium/ epithelium would lead to lower amounts of AM in isolated BAL fluids. With regard

to the increasing amounts of AM isolated in the “later” post-traumatic phase, it has been

shown that an increased number of mononuclear phagocytes, characteristic for the chronic 58

and acute inflammatory 224,225 state, can result from an increased rate of recruitment of these

cells from the blood 225, from a decreased rate of efflux (including cell death) from the lung, or

from local replication 58,225.

Discussion

99

5.4.2 Phagocytic capacity, microbial killing and TNF release

Although it has been reported that severe trauma contributes to an increase in infectious

complications through an impairment of host defence mechanisms 163, the cause of such

impairment remains unknown 164. The alveolar macrophage, mobile representative of the

mononuclear phagocyte system, is believed to be the central modulator, both as regulator and

effector of inflammation/ anti-inflammation within the alveolar space 48.

In the present study, blast pressure wave exposure resulted in an immediate decrease in the

alveolar macrophage (AM) phagocytosis rate of Escherichia coli (E. coli) in vitro with recovery

within 96 hours after blast. Although not statistically significant, the phagocytic capacity for

Staphylococcus aureus (S. aureus) particles seemed to be reduced initially after thoracic

trauma. Disorders of the phagocytic process in alveolar macrophages have been reported

after thermal injury 226,227. These studies showed a primary decrease occurring immediately

after the burn. The discrepancy between the different phagocytosis rates of both bacterial

stimuli may be explained by experimental problems rather than by strains differences.

Whereas the E. coli particles were suspended homogenously, the S. aureus particles formed

aggregates despite of sonicating them and this may affect particle counts and therefore the

read-out of the experiment. The use of different lots of fluorescence labelled particles may

also contribute to the controversial results. Nevertheless, as macrophages are the primary

resident phagocyte of the resting human lung, it is intuitive that any functional impairment

leads to infection by micro-organisms. Thus, despite recovery within 4 days after trauma,

these data suggest an initially impaired phagocytosis capacity due to thoracic trauma, that

may result in a higher susceptibility to infections. It is possible that trauma-induced factors or

mediators may either bind to the macrophage surface or suppress macrophage activity,

thereby down-regulating phagocytosis. It is also conceivable that impairment of the pulmonary

surfactant during lung injury contributes to a decreased phagocytosis capacity. Indeed,

surfactant protein A 71 and D 72 are able to precoat bacteria and particles in order to facilitate

macrophage phagocytosis and bactericidal activities. The trauma-related, deteriorated tidal

volume and pulmonary compliance we assessed ex vivo in our lung perfusion model support

such a trauma-induced loss in surfactant.

The production of reactive oxygen species (ROS) (“respiratory burst”) upon receptor-mediated

phagocytosis was investigated to determine the microbicidal activity of the AM in vitro after in

vivo thorax trauma. Thoracic trauma yielded a marked decrease in the production of

superoxide within the first 24 hours after trauma, both with or without the addition of E. coli or

S. aureus, with recovery only after 96 hours. The respiratory burst results from the assembly

and activation of the NADPH oxidase 76. Decreased levels of superoxide can be attained

Discussion

100

either by a decrease in its production or by an increase in its elimination, be it through

interfering with one or more steps in the signal transduction pathway at either the cell

membrane level or subsequent intracellular stage such as dismutase, protein kinase (PK)-C

activation or through scavenging by other radicals, such as nitric oxide. With regard to this, it

has been shown that LPS-related oxidative burst and TNF-release by human monocytes can

be modulated by IL-9 through an upregulation of TGF-β which in turn inactivates the

extracellular signal-regulated kinases (ERK) 1/ 2 84. TGF-β is involved in bronchial wound

repair 228.

The difference between trauma and control in both the resting and the activated state, could

therefore arise from both a decreased activity of one or more oxidases and an increased

ability of the cells to eliminate superoxide, due to cellular damage and released NO (discussed

in 5.3.2.3).

Since overnight culturing of the isolated control AM reduced the spontaneous superoxide

production to about 30% of its initial value, the increased release of superoxide in non-

traumatized controls may be attributed to the activation of NADPH oxidase during the

macrophage isolation procedure. In contrast, the superoxide anion levels produced upon E.

coli stimulation were not affected by overnight culturing in these cells and thus may be related

to the bacterial stimulus (data not shown). In order to preserve the in vivo conditions after

trauma in terms of both macrophage environment and time-dependent AM behaviour, we

abstained form overnight culturing before stimulation.

Similarly, trauma-related initial decrease of superoxide generation was independent of the

presence of lipopolysaccharide (LPS) or lipoteichoic acid (LTA). In contrast to the stimulation

with whole dead bacteria, upon LPS/ LTA stimulation the superoxide levels of non-traumatized

controls were reached already at 24 hours, and were increased 96 hours after the blast. Since

the levels of superoxide produced by the non-stimulated controls were different between the

experiments, the higher respiratory burst capacity observed upon LPS/ LTA stimulation as

compared to E. coli/ S. aureus stimulation, may be due to an increased basal stimulation,

rather than to their particular specificity. In contrast to this stimulus-independent effect, the

earlier recovery together with the increased oxidative burst after 96 hours, suggests higher

susceptibility of the AM towards LPS and LTA as compared to E. coli and S. aureus

stimulation in this setting. However, the most likely explanation for this would be that the

amount of surface-bound endotoxin on E. coli and LTA on S. aureus particles is not

comparable with the amount of LPS and LTA used in these experiments. This explanation is

further supported by the fact that both whole bacteria and endotoxin, act on the same

receptors on the macrophage surface. Recent studies have identified multiple receptors for

endotoxin which has accelerated the study of signalling in macrophages (reviewed by Monick

and Hunninghake, 229, Wright, 66). Two classes of receptors, the CD18 antigens (leukocyte

Discussion

101

integrins) and the scavenger receptor (acetyl low-density lipoprotein), recognize particulate

LPS directly. Both these receptors may be involved in catabolism of LPS, because neither

receptor initiates strong secretory responses e.g. TNF synthesis to LPS. A third receptor,

CD14, recognizes LPS complexed with the serum lipoprotein binding protein (LBP) 62. CD14

appears to participate in both the ingestion of, and the responses to LPS, because a blockade

of CD 14 with monoclonal antibodies (mAbs) strongly inhibits both uptake of bacteria and

secretion of TNF by human mononuclear cells 62. A recent study suggested that the

recognition sites of CD14 for LPS and LTA are distinct, with a partial overlap 230.

Furthermore, leukocytes express three different receptors for complement protein C3, and 10

distinct receptors for the Fc region of IgG 60. With regard to possible regulatory mechanisms of

phagocytosis and oxidative burst in the present study, Pricop and Salmon 77 reported that

FcγR activation induces the NADPH oxidase and thus increases ROS formation. These

authors further demonstrated that ROS generated in an inflammatory milieu act in an

autocrine and paracrine manner in order to rapidly amplify the effector potential of FcγR on

quiescent phagocytes by altering signal transduction.

We further investigated macrophage TNF release after experimental thorax trauma, alone and

in response to either endotoxin/ LTA or whole dead bacteria. The TNF release of AM from

traumatized rat lungs was increased 96 hours after blast as compared to non-traumatized

controls, indicating that the AM are primed to release TNF following thoracic trauma. This

result is in line with increasing TNF levels measured in BAL fluids of traumatized rat lungs in

vivo over time (discussed in 5.3.2.2). However, when compared to the early TNF release

measured in BAL fluids from 3 hours after trauma on, the TNF release by isolated AM in vitro

is delayed. Additional stimulation of the AM with either endotoxin/ LTA or bacteria induced a

marked increase in TNF release 24 hours after blast exposure, whereas initially after blast the

response of the AM towards both stimuli was comparable or even decreased (E. coli

stimulation) with that from AM of non-traumatized controls. We conclude that thoracic trauma

primes AM to release TNF which can be augmented by an additional bacterial stimulus. With

regard to TNF release, AM from sham controls as well as from traumatized rats were more

susceptible to LPS than to LTA (up to 8-fold and 7-fold, respectively at the highest used

concentration) and more to E. coli than to S. aureus stimulation (up to 4.7-fold and 6.5-fold,

respectively). This is in close correlation to reports from Morath and colleagues 231, who

described LTA as a weak inducer of cytokine release in human blood compared to LPS.

Many experimental models of trauma have demonstrated a change in macrophage function as

manifested by a decreased cytokine production 155,165,232, superoxide production 155,156,232,

phagocytosis capacity 156, antigen presentation 155,156, and Candida killing 156,232 immediately

Discussion

102

after injury. The causes of abnormal macrophage function appear to be multifactorial but are

not well understood. Injury causes a stress response which in turn has been recognized as a

source of immunosuppression. Thus, physical or even psychological stresses trigger the

hypothalamic-pituitary-adrenal (HPA) axis, resulting in a marked rise in serum levels of many

hormones e.g. glucocorticoids 130. Receptors for glucocorticoids are present on macrophages 233 and have been shown to downregulate macrophage functions 232. Furthermore, not only

stress but also pro-inflammatory cytokines such as IL-1, IL-6 and TNF cause via the release of

corticotropin the release of glucocorticoids (reviewed in 130). Thus, glucocorticoid-mediated

immunosuppression following injury may represent dysfunction of normal, inhibitory feedback

pathways. With regard to this, Cech and colleagues 232 reported increased plasma

corticosterone levels one day after femure fracture in mice. In this study femur fracture

resulted also in a reduction of 1. Macrophage production of superoxide anion, 2. C. albicans

killing, 3. macrophage synthesis of IL-6 and TNF and 4. an increase in PGE2 synthesis. These

authors further demonstrated that pre-treatment with the glucocorticoid receptor antagonist,

mifepristone (RU 486) significantly prevented or reduced the observed suppression of several

macrophage functions, including oxidative burst capacity and C. albicans killing but did not

alter the increased PGE2 synthesis. However, the initially suppressed macrophage function

was supposed to be at least in part due to the trauma-related PGE2 production 155. The results

from our study demonstrated a recovery from an initially suppressed E. coli phagocytosis and

oxidative burst within 24 hours and 96 hours, respectively. Such a rebound phenomenon has

also been described for peritoneal macrophages after laparotomy 156. We further identified a

hyperactive TNF response of AM to endotoxin/ LTA and whole bacteria stimulation that is

seen 24 hours after blast exposure. This is in line with McCarter and colleagues 155 who found

increased TNF, IL-6 and H2O2 secretion of splenic macrophages in response to endotoxin or

PMA stimulation 7 days after trauma. These authors further demonstrated that inflammatory

hypersecretion does not necessary translate into improved immunologic protection, as

demonstrated by the functional impairment of macrophage antigen presentation. It is likely

that the impairment of phagocytic function and microbicidal activity after thoracic trauma

observed in this setting could contribute to a depressed antigen presentation. However,

preliminary findings showed that Keyhole Limped Hemocyanin (KLH)-sensitised rats

developed a cutaneous anergy during about four days after severe trauma, which underlines

that the intensity of the investigated trauma has systemic repercussions and seems to include

aspects of immunosupression.

The search of the signalling endotoxin receptor led to the discovery of Toll receptors (reviewed

by Monick and Hunninghake, 229): TLR 4 is the LPS-specific receptor, whereas TLR 2 has

been linked to LTA from Gram-positive bacteria. Both TLR 2 and TLR 4 have been found on

alveolar macrophages, providing the entry point for a complex cascade of LTA/ LPS signalling.

Discussion

103

Downstream of the TLR 4 complex, a number of signalling pathways are activated. These

include the MAP kinases, phosphatidylinositol (PI) 3-kinase pathway and sphingolipid

metabolites that regulate LPS-induced gene expression. Protein kinase (PK) C and PI3-kinase

have been shown to be not only involved in the phagocytosis of E.coli particles, but also in the

signalling pathway of LPS-induced TNF release. In contrast, anti-inflammatory

phosphodiesterase inhibitors prevented LPS-induced TNF release but had no effect on the

beneficial phagocytosis of E. coli by macrophages 234.

5.4.3 Involvement of trauma-related alveolar haemorrhage in impaired macrophage functions

As already discussed in 5.3.1, alveolar haemorrhage was the most frequent and obvious

injury detected in our rat blast model, thus it was of interest to examine whether the observed

initial alveolar macrophage (AM) dysregulation was due to the accompanying alveolar

bleeding. In these experiments, the removal of the red blood cells (RBCs) resulted in a

marked increase in the phagocytosis of E. coli particles and superoxide anion production (with

or without E. coli stimulation) by AM from traumatized rat lungs. Although other factors may be

implicated in the downregulation of the macrophage function, the initial suppression of

phagocytosis and oxidative burst capacity by AM may be caused at least in part by the

trauma-related alveolar bleeding. In contrast, the absence of RBC’s did not significantly alter

the E. coli-induced TNF release, thus the alveolar haemorrhage seems to play only a minor

role in the trauma-related TNF release. We could further demonstrate that different blood

components are involved. Whereas the phagocytosis capacity of control AM was significantly

reduced in the presence of intact blood cells (whole blood and washed RBC’s), the oxidative

burst and the TNF release remained unaffected. In contrast, haemolytic blood decreased the

superoxide production and the TNF release but had no effect on the phagocytosis capacity of

E. coli particles. Since previous experiments revealed no difference in superoxide production

with or without the presence of a bacterial stimulus, the plasma-related decrease in

superoxide production upon E. coli stimulation may be explained by a plasma contamination

with intracellular components, due to centrifugation during the experimental procedure. These

results suggest an anti-oxidative blood cell component that may quench the superoxide anion

production. As only haemolytic blood and neither RBC haemolysate nor plasma altered the

oxidative burst (at least without E. coli) and TNF release, one hypothesis would be that an

intracellular component and a plasma component may act together, e.g. glutathione and

glutathione peroxidase while quenching the superoxide. Because both oxidative burst and

TNF release were not affected by either whole blood or RBC’s, a steric interference between

E. coli and AM as the cause for the diminished phagocytosis seemed rather unlikely. Thus

Discussion

104

there are two possible explanations: 1. Intact blood cells bind on the macrophage surface,

thereby down-regulating the phagocytosis of E. coli particles and/ or 2. a preferential

phagocytic engulfment of RBC’s as compared to E. coli. The latter hypothesis is supported by

the finding that after haemorrhage large quantities of haemoglobin, haemosiderin and

ingested red blood cells can be detected in alveolar macrophages 197,198. However, further

work is needed to differentiate these possibilities.

In summary, the in vitro study has demonstrated a phasic response in alveolar macrophage

response to Gram-positive and Gram-negative stimuli that is a function of time after thoracic

trauma. The paradoxical combination of suppressed macrophage function such as

phagocytosis and oxidative burst, and hypersecretion of inflammatory mediators such as TNF,

may simultaneously render the host susceptible to both infectious complications and the

immune-mediated sequelae of SIRS and MOF.

Summary

105

6. Summary

Despite of variations in injury, due to animal as well as technical variabilities, we could

establish a reproducible in vivo model of blast thorax trauma. This includes not only a

standardized discharge of the blast pressure wave, but also a thorough monitoring of the blast

pressure wave parameters and an exact triggering depending on the respiratory cycle, to

monitor key blast wave properties and thus to be able to compare data over time. Our blast

wave generator develops a mechanical explosive that simulates blast pressure waves from

high explosives and produces pulmonary injuries in rats which tend to reproduce the spectrum

of injuries seen in blast victims and previous animal experiments. This model allows to study

primary blast injury without direct contact with the thoracic cage in a safe laboratory

environment.

The direct functional consequences of in vivo thoracic trauma were assessed in our ex vivo

isolated perfused rat lung model which has, to the best of our knowledge, not been described

previously. The correlation of pressure wave monitoring together with a validated injury score,

the BAL red blood cell counts and the ex vivo data collection in the isolated, traumatized and

perfused rat lung indicate that physical and biological impact may be matched in this setting.

Our present results illustrate that thoracic trauma deteriorates lung function, leads to important

oedema formation due to a loss of alveolar-capillary barrier function, depending on the trauma

intensity. The observation that especially the airway mechanics of traumatized rat lungs

appear to be more susceptible towards ex vivo mechanical ventilation and perfusion, provided

evidence for the possible development of a secondary injury called “ventilator-associated lung

injury”. This is of importance, since mechanical ventilation is one of the ultimate life-supporting

technologies in the therapy of critically ill patients with respiratory failure.

Pharmacological modulation during ex vivo lung perfusion indicated that the integrity of the

sodium channels is at least in part maintained after blast injury in this model. Thus, an

increased alveolar fluid clearance might be an underlying mechanism for the postulated fluid

resorption in vivo within the first six hours after trauma and assessed as decreased ex vivo

weight gain. Furthermore, β2-adrenergic therapy increased not only the oedema resorption

and/ or the oedema generation after trauma, but also improved the airway mechanics by direct

β2-adrenergic effects, possible independent of β-receptor-related fluid resorption. Thus β2-

adrenergic substances like terbutaline or formoterol appear to be beneficial after primary blast

injury in this model.

The sequelae of important alveolar haemorrhage and protein accumulation followed by a time-

dependent resolution, reflects not only a disrupted alveolar-capillary barrier induced by the

blast pressure wave, but also an enormous clearance capacity and/ or repair of the lung. One

Summary

106

biochemical response of the lung to the blast pressure wave was an immediate release of the

eicosanoids thromboxane and prostacycline, which have been reported in clinical studies to

be rather specific for the involvement of the lung in polytrauma patients. Thoracic trauma

resulted not only in an increased TNF and IL-6 release after 6 hours but also in a pronounced

CINC-3 production and neutrophil infiltration into the alveolar space. Likewise, matrix

metalloproteinase (MMP)-9/ –2 release and myeloperoxidase activity were detectable during

this time. The low IL-10 levels measured up to 24 hours after blast may favour cytokine

production in the alveolar environment in this setting. The appearance of increasing amounts

of the soluble TNF receptors (p55 and p75) in BAL fluids within the first 12 hours suggest an

early anti-inflammatory counterregulation, which in turn could be the cause for the “delayed”

and moderate TNF release after blast exposure.

With regard to the post-traumatic immune status, our in vitro results demonstrate a phasic

response of alveolar macrophages to endotoxin and whole bacteria stimulation, in terms of

both initially suppressed phagocytosis capacity, superoxide production (independent of

stimulation) and a delayed increase in TNF secretion. The dysregulation was shown to be at

least in part due to the accompanying alveolar haemorrhage and more precisely to different

blood components. However, also with respect to the in vivo results, this response may

represent a state of dysregulation, rather than a purely suppressed or hyperactive state.

Although the rats in our study appear to recover from the injury - confirmed by the parameters

we investigated in vivo, ex vivo and in vitro - the observed dysregulation in the lung, especially

in macrophage function may in fact increase their susceptibility to an infectious challenge or to

the development of multi-organ failure.

In conclusion, the present study demonstrates that the trauma model gives valid and

reproducible results, which in turn underscore the importance to address further questions on

the modulation of this injury in view of the major impact of trauma on especially young patients

and thus on the society.

Deutsche Zusammenfassung

107

7. Deutsche Zusammenfassung

Ein vorrangiges Ziel dieser Studie war die Erzeugung einer reproduzierbaren, isolierten in vivo

Druckwellenverletzung („Blast Injury“) des Thorax der Ratte. Trotz materialbedingter

Variationen und Schwankungen, die auf die verwendeten Tiere zurückgeführt wurden, ist uns

dies weitgehend gelungen. Die standardisierte Generierung der Druckwelle zusammen mit

einer reproduzierbaren Erfassung der Druckwellenparameter unter Berücksichtigung der

Atemlage des Tieres waren hierfür die Voraussetzung. Das für diese Versuche gewählte

Rattenmodell, welches ursprünglich zur Untersuchung von Explosionstraumen entwickelt

wurde, erlaubt die experimentelle Erzeugung einer isolierten Druckwellenverletzung der

Lunge, ohne direkten Kontakt mit den Strukturen des Brustkorbes. Die hierbei entstehende

Schädigung entspricht den beim Menschen auftretenden Verletzungen nach Exposition einer

Druckwelle, und ist vergleichbar mit zahlreichen tierexperimentellen Untersuchungen.

Durch die Verknüpfung zweier Modellsysteme, der in vivo erzeugten Druckwellenverletzung

des Thorax und der anschliessenden ex vivo Perfusion der isolierten Rattenlunge, konnten

erstmals die direkten Auswirkungen der Druckwelle auf die Physiologie und den Metabolismus

der Lunge untersucht werden. Hierbei konnte ein deutlicher Zusammenhang zwischen der

Intensität des auf die Ratte einwirkenden Druckes, also den physikalischen Eigenschaften und

dem daraus resultierenden Ausmass der Lungenschädigung gezeigt werden. Die

zugrundeliegenden Schädigungsparameter waren der Gehalt an roten Blutkörperchen im

Alveolarraum, der makroskopisch berechnete und validierte „Injury score“ als auch die ex vivo

erfasste Lungenfunktion. Basierend auf den Ergebnissen der ex vivo Perfusion konnte eine

direkte pulmonale Funktionsstörung nach Druckwelleneinwirkung gezeigt werden. Diese

beinhaltet eine Verschlechterung der Atemparameter gemessen an einem reduzierten

Atemzugvolumen und einem erhöhten Atemwegswiderstand der Lunge. Des weiteren kann,

aufgrund der Flüssigkeitseinlagerung (Ödembildung) in der Lunge, eine alveloär-kapilläre

Schrankenstörung postuliert werden. Der zur initialen Ödembildung vergleichsweise geringe

ex vivo Gewichtsanstieg zu einem späteren Zeitpunkt nach Trauma (drei und sechs Stunden),

legte die Vermutung einer Flüssigkeitsresorption und/ oder Verminderung der Ödemgenese in

vivo nach Thoraxverletzung nahe. Die Beobachtung, dass die Atemparameter traumatisierter

Tiere sich gegenüber derer „gesunder“ Lungen aus Tieren der Kontrollgruppe innerhalb der

Perfusion verschlechtern, lässt auf eine zusätzliche, ventilationsbedingte Verletzung

schliessen. Dies ist insofern von Bedeutung, als dass die mechanische Beatmung eine häufig

angewendete, lebenserhaltende klinische Maßnahme bei Patienten mit respiratorischer

Insuffizienz darstellt. Mit Hilfe ex vivo eingesetzter pharmakologischer Substanzen konnte

gezeigt werden, dass die Integrität der für die Flüssigkeitsresorption notwendigen Natrium-


108

Kanäle, nach einem Trauma zumindest teilweise besteht. Diese Erkenntnis führte zu der

Hypothese, dass die bereits erwähnte, postulierte Flüssigkeitsresorption in vivo,

möglicherweise auf die Aktivierung solcher Kanäle zurückzuführen ist. Des weiteren konnte

gezeigt werden, dass sich der Einsatz beta-2-adrenerger Substanzen nicht nur positiv auf

Ödemresorption und/ oder die Hemmung der Ödemgenese auswirkt, sondern auch beta-

rezeptor-abhängig die Atemparameter der Lunge nach Trauma verbessern. Beta-2-adrenerge

Substanzen wie Terbutalin und Formoterol, können daher im Modell der ex vivo

Lungenperfusion nach in vivo Druckwellenverletzung als therapeutisch wirksam angesehen

werden. Die traumabedingten Folgeerscheinungen wie die starke Einblutung und

Proteinakkumulation in der Lunge machen eine Zerstörung der Lungenarchitektur deutlich.

Dennoch konnte eine zeitabhängige Verbesserung dieser initialen Schadensparameter

innerhalb von 96 Stunden nach Druckwelleneinwirkung beobachtet werden. Aufgrund dessen

kann von einer enormen Regenerationsleistung der Lunge ausgegangen werden.

Die Analyse biochemischer Parameter nach einem in vivo Thoraxtrauma, ergab eine direkte

Freisetzung der Eikosanoide Thromboxan und Prostazyklin. Eine derartige Freisetzung dieser

Metabolite wird in der Literatur als relativ spezifisch für die Beteiligung der Lunge im

Polytrauma angesehen. Weiterhin konnten ansteigende Mengen der Mediatoren Tumor-

Nekrose Faktor (TNF)-α und Interleukin (IL)-6, innerhalb von sechs Stunden nach Trauma

gemessen werden. Zusammen mit einer Freisetzung des chemotaktisch wirksamen

Metaboliten CINC-3 und einer Infiltration neutrophiler Granulozyten in die Lunge, deutete dies

klar auf eine traumabedingte Entzündungsreaktion hin. Unterstützt wurde diese Vermutung

durch die Messung von Enzymen, wie den Matrixmetalloproteinasen (MMP)-2/ -9, welche als

charakteristisch für eine Einwanderung bzw. das Vorhandensein neutrophiler Granulozyten

angesehen werden können. Die geringen Mengen an messbarem Interleukin (IL)-10, einem

anti-entzündlichen Mediator, könnten das Gleichgewicht zwischen pro-und anti-entzündlichen

Mediatoren nach Druckwellenverletzung zugunsten der pro-entzündlichen Metabolite

verschoben haben. Im Gegensatz dazu, konnten innerhalb der ersten zwölf Stunden

steigende Mengen der löslichen Tumor-Nekrose Faktor (sTNF)-Rezeptoren in der broncho-

alveolären Lavegeflüssigkeit (BAL) gemessen werden, was auf eine anti-inflammatorische

Gegenregulation nach einem Thoraxtrauma hindeutet. Über einen möglichen Zusammenhang

zwischen dem Auftauchen löslicher TNF Rezeptoren, von welchen eine TNF neutralisierende

Wirkung bekannt ist, und der verzögerten (TNF)-α Freisetzung nach einem Thoraxtrauma

kann spekuliert werden.

Im Hinblick auf eine mögliche post-traumatische Immunsuppression wurden isolierte, primäre

Alveolarmakrophagen aus traumatisierten Rattenlungen auf ihre Funktionalität hin untersucht.

Entscheidend für die Beurteilung waren die Fähigkeit zur Phagozytose, zur Bildung reaktiver


109

Sauerstoffmetabolite, wie Superoxidanionen als auch die Freisetzung des pro-entzündlichen

Metaboliten TNF. In diesen in vitro Versuchen konnte eine phasische Antwort der

Alveolarmakrophagen sowohl auf die Stimulation mit bakteriellen Bestandteilen

(Lipopolyssacharid (LPS) und Lipoteichonsäure (LTA)) als auch auf ganze Gram-positive und

Gram-negative Bakterien (Escherichia Coli/ Staphylococcus aureus) hin beobachtet werden:

Initial nach einem Trauma war sowohl die Phagozytosekapazität als auch die Fähigkeit zur

Generierung von Superoxidanionen deutlich vermindert. Im Gegensatz dazu waren diese

Funktionen zu späteren Zeitpunkten nach Trauma wieder vergleichbar mit denen der

Kontrolltiere, wohingegen die TNF Freisetzung der Alveolarmakrophagen traumatisierter Tiere

24 Stunden nach Setzen des Traumas deutlich erhöht war. In der vorliegenden Arbeit konnte

gezeigt werden, dass die beobachteten Veränderungen der Alveolarmakrophagen Funktionen

zumindest teilweise auf die traumainduzierte, massive, alveoläre Einblutung zurückgeführt

werden kann. Weiterführende Untersuchungen lassen unterschiedliche Blutkomponenten als

Ursache für die Störung der jeweiligen Funktionen vermuten. Unter Berücksichtigung der oben

beschriebenen metabolischen Konsequenzen nach einer Druckwellenverletzung der Lunge,

deuten diese in vitro Ergebnisse eher auf eine Trauma bedingte Fehlregulation, als auf eine

generelle Suppression bzw. Hyperreaktion hin.

Die Resultate dieser Arbeit machen deutlich, dass alle Parameter, gemessen in den

verschiedenen Modellsystemen (in vivo, ex vivo und in vitro), sich zeitabhängig nach einer

Druckwellenverletzung der Lunge normalisierten und auch keine offensichtlichen Anzeichen

einer Spätletalität zu verzeichnen waren. Dennoch ist eine erhöhte Empfindlichkeit gegenüber

bakterieller Infektionen, also auch die Entwicklung eines Multi-Organ Versagens (MOF) im

Zustand einer eingeschränkten Abwehr, wie die hier beobachtete durchaus vorstellbar.

Zusammenfassend konnte in der vorliegenden Studie gezeigt werden, dass das Modell des in

vivo erzeugten Thoraxtrauma ein geeignetes Modell zur Untersuchung des primären

Lungenschadens der Ratte nach Druckwelleneinwirkung darstellt. Die erzielten Ergebnisse

tragen nicht nur zum Verständnis der post-traumatischen Schädigung bei, sondern

unterstreichen auch in Anbetracht der Häufigkeit traumabedingter Todesfälle in unserer

Gesellschaft die Notwendigkeit weiterer Untersuchungen.

References

110

8. References

1 Uhlig S, Wollin L. An improved setup for the isolated perfused rat lung. J Pharmacol Toxicol Methods 1994; 31: 85-94.

2 Staub NC. Pulmonary edema. Physiol Rev 1974; 54: 678-811. 3 Folkesson HG, Norlin A, Baines DL. Salt and water transport across the alveolar

epithelium in the developing lung: Correlations between function and recent molecular biology advances (Review). Int J Mol Med 1998; 2: 515-531.

4 Matthay MA, Flori HR, Conner ER, Ware LB. Alveolar epithelial fluid transport: basic mechanisms and clinical relevance. Proc Assoc Am Physicians 1998; 110: 496-505.

5 Matthay MA et al. Alveolar epithelial barrier. Role in lung fluid balance in clinical lung injury. Clin Chest Med 2000; 21: 477-490.

6 Sartori C, Matthay MA. Alveolar epithelial fluid transport in acute lung injury: new insights. Eur Respir J 2002; 20: 1299-1313.

7 Berthiaume Y, Staub NC, Matthay MA. Beta-adrenergic agonists increase lung liquid clearance in anesthetized sheep. J Clin Invest 1987; 79: 335-343.

8 Jayr C et al. Alveolar liquid and protein clearance in anesthetized ventilated rats. J Appl Physiol 1994; 76: 2636-2642.

9 Sakuma T et al. Alveolar fluid clearance in the resected human lung. Am J Respir Crit Care Med 1994; 150: 305-310.

10 Sakuma T et al. Beta-adrenergic agonist stimulated alveolar fluid clearance in ex vivo human and rat lungs. Am J Respir Crit Care Med 1997; 155: 506-512.

11 Fukuda N, Folkesson HG, Matthay MA. Relationship of interstitial fluid volume to alveolar fluid clearance in mice: ventilated vs. in situ studies. J Appl Physiol 2000; 89: 672-679.

12 Smedira N et al. Alveolar and lung liquid clearance in anesthetized rabbits. J Appl Physiol 1991; 70: 1827-1835.

13 Basset G, Crone C, Saumon G. Fluid absorption by rat lung in situ: pathways for sodium entry in the luminal membrane of alveolar epithelium. J Physiol 1987; 384: 325-345.

14 Matalon S, Benos DJ, Jackson RM. Biophysical and molecular properties of amiloride-inhibitable Na+ channels in alveolar epithelial cells. Am J Physiol 1996; 271: L1-22.

15 Sznajder JI, Olivera WG, Ridge KM, Rutschman DH. Mechanisms of lung liquid clearance during hyperoxia in isolated rat lungs. Am J Respir Crit Care Med 1995; 151: 1519-1525.

16 Jiang X, Ingbar DH, O'Grady SM. Adrenergic stimulation of Na+ transport across alveolar epithelial cells involves activation of apical Cl- channels. Am J Physiol 1998; 275: C1610-1620.

17 Stutts MJ, Rossier BC, Boucher RC. Cystic fibrosis transmembrane conductance regulator inverts protein kinase A-mediated regulation of epithelial sodium channel single channel kinetics. J Biol Chem 1997; 272: 14037-14040.

18 Dobbs LG et al. Highly water-permeable type I alveolar epithelial cells confer high water permeability between the airspace and vasculature in rat lung. Proc Natl Acad Sci U S A 1998; 95: 2991-2996.

19 Matthay MA, Folkesson HG, Verkman AS. Salt and water transport across alveolar and distal airway epithelia in the adult lung. Am J Physiol 1996; 270: L487-503.

20 Song Y et al. Role of aquaporins in alveolar fluid clearance in neonatal and adult lung, and in oedema formation following acute lung injury: studies in transgenic aquaporin null mice. J Physiol 2000; 525 Pt 3: 771-779.

21 Goodman BE, Kim KJ, Crandall ED. Evidence for active sodium transport across alveolar epithelium of isolated rat lung. J Appl Physiol 1987; 62: 2460-2466.

22 Saldias F et al. Modulation of lung liquid clearance by isoproterenol in rat lungs. Am J Physiol 1998; 274: L694-701.

References

111

23 Berthiaume Y et al. Alveolar liquid and protein clearance from normal dog lungs. J Appl Physiol 1988; 65: 585-593.

24 Garat C, Carter EP, Matthay MA. New in situ mouse model to quantify alveolar epithelial fluid clearance. J Appl Physiol 1998; 84: 1763-1767.

25 Goodman BE, Anderson JL, Clemens JW. Evidence for regulation of sodium transport from airspace to vascular space by cAMP. Am J Physiol 1989; 257: L86-93.

26 Carstairs JR, Nimmo AJ, Barnes PJ. Autoradiographic visualization of beta-adrenoceptor subtypes in human lung. Am Rev Respir Dis 1985; 132: 541-547.

27 Borok Z et al. Effects of EGF on alveolar epithelial junctional permeability and active sodium transport. Am J Physiol 1996; 270: L559-565.

28 Guery BP et al. Keratinocyte growth factor increases transalveolar sodium reabsorption in normal and injured rat lungs. Am J Respir Crit Care Med 1997; 155: 1777-1784.

29 Folkesson HG, Pittet JF, Nitenberg G, Matthay MA. Transforming growth factor-alpha increases alveolar liquid clearance in anesthetized ventilated rats. Am J Physiol 1996; 271: L236-244.

30 Olivera W, Ridge K, Wood LD, Sznajder JI. Active sodium transport and alveolar epithelial Na-K-ATPase increase during subacute hyperoxia in rats. Am J Physiol 1994; 266: L577-584.

31 Pittet JF, Brenner TJ, Modelska K, Matthay MA. Alveolar liquid clearance is increased by endogenous catecholamines in hemorrhagic shock in rats. J Appl Physiol 1996; 81: 830-837.

32 Pittet JF et al. Stimulation of lung epithelial liquid clearance by endogenous release of catecholamines in septic shock in anesthetized rats. J Clin Invest 1994; 94: 663-671.

33 Lane SM, Maender KC, Awender NE, Maron MB. Adrenal epinephrine increases alveolar liquid clearance in a canine model of neurogenic pulmonary edema. Am J Respir Crit Care Med 1998; 158: 760-768.

34 Fukuda N et al. Mechanisms of TNF-alpha stimulation of amiloride-sensitive sodium transport across alveolar epithelium. Am J Physiol Lung Cell Mol Physiol 2001; 280: L1258-1265.

35 Elia N et al. Identification of the Alveolar edema reabsorption activity of murine Tumor Necrosis Factor. Am J Respir Crit Care Med 2003 Jul 3 (electronic publication), in press.

36 Wright PE, Carmichael LC, Bernard GR. Effect of bronchodilators on lung mechanics in the acute respiratory distress syndrome (ARDS). Chest 1994; 106: 1517-1523.

37 Campbell AR et al. Alveolar epithelial fluid clearance persists in the presence of moderate left atrial hypertension in sheep. J Appl Physiol 1999; 86: 139-151.

38 Garat C et al. Alveolar epithelial fluid clearance mechanisms are intact after moderate hyperoxic lung injury in rats. Chest 1997; 111: 1381-1388.

39 Eickelberg O et al. Ligand-independent activation of the glucocorticoid receptor by beta2-adrenergic receptor agonists in primary human lung fibroblasts and vascular smooth muscle cells. J Biol Chem 1999; 274: 1005-1010.

40 Masclans JR et al. Salbutamol reduces pulmonary neutrophil sequestration of platelet-activating factor in humans. Am J Respir Crit Care Med 1996; 154: 529-532.

41 Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000; 342: 1334-1349.

42 Yi ES et al. Keratinocyte growth factor ameliorates radiation- and bleomycin-induced lung injury and mortality. Am J Pathol 1996; 149: 1963-1970.

43 Wang Y et al. Alveolar epithelial fluid transport can be simultaneously upregulated by both KGF and beta-agonist therapy. J Appl Physiol 1999; 87: 1852-1860.

44 Reynolds HY. Integrated host defence against infections. In Crystal, R. G., West, J. B. (eds.): The Lung: Scientific Foundations. New York, Raven Press 1991: pp. 1899-1911.

45 Reynolds HY. Pulmonary host defenses. State of the art. Chest 1989; 95: 223S-230S.

References

112

46 Kaltreider HB. Immune defenses in the lung. In Sande, M. A., Root, R. K., Hudson, L. D. (eds.): Respiratory Infections. Contemporary Issues in Infectious Diseases. New York, Churchill Livingstone 1986; 5: pp. 47-70.

47 Murray JF, Nadel JA. Textbook of Respiratory Medicine. W.B. Saunders Company 1994; 1: 1-1243.

48 Johnston RB, Jr. Current concepts: immunology. Monocytes and macrophages. N Engl J Med 1988; 318: 747-752.

49 Bezidcek P, Crystal RG. Pulmonary macrophages. In: Crystal, R. G., West, J. B., Weibel, E. R., Barnes, P. J., editors. The Lung. Philadelphia: Lippincott-Raven Publishers 1997: 859-875.

50 Nathan CF. Secretory products of macrophages. J Clin Invest 1987; 79: 319-326. 51 Hogg JC. Neutrophil kinetics and lung injury. Physiol Rev 1987; 67: 1249-1295. 52 Crapo JD et al. Cell number and cell characteristics of the normal human lung. Am

Rev Respir Dis 1982; 126: 332-337. 53 Lehnert BE. Pulmonary and thoracic macrophage subpopulations and clearance of

particles from the lung. Environ Health Perspect 1992; 97: 17-46. 54 Musson RA, Henson PM. Phagocytic cells. In Bienenstock, J. (eds.): Immunology of

the Lung and Upper Respiratory Tract. New York, McGraw-Hill 1984: 119-138. 55 Warner AE, Molina RM, Brain JD. Uptake of bloodborne bacteria by pulmonary

intravascular macrophages and consequent inflammatory responses in sheep. Am Rev Respir Dis 1987; 136: 683-690.

56 Blusse van Oud Alblas A, van Furth R. The origin of pulmonary macrophages. Immunobiology 1982; 161: 186-192.

57 Sibille Y, Reynolds HY. Macrophages and polymorphonuclear neutrophils in lung defense and injury. Am Rev Respir Dis 1990; 141: 471-501.

58 Bitterman PB et al. Alveolar macrophage replication. One mechanism for the expansion of the mononuclear phagocyte population in the chronically inflamed lung. J Clin Invest 1984; 74: 460-469.

59 Standiford TJ et al. Interleukin-8 gene expression from human alveolar macrophages: the role of adherence. Am J Respir Cell Mol Biol 1991; 5: 579-585.

60 Wright SD, Detmers PA. Receptor-mediated phagocytosis. In Crystal, R. G., West, J. B. (eds.): The Lung: Scientific Foundations. New York, Raven Press 1991: pp. 539-551.

61 Simmons DL et al. Monocyte antigen CD14 is a phospholipid anchored membrane protein. Blood 1989; 73: 284-289.

62 Wright SD et al. CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 1990; 249: 1431-1433.

63 Dentener MA et al. Involvement of CD14 in lipopolysaccharide-induced tumor necrosis factor-alpha, IL-6 and IL-8 release by human monocytes and alveolar macrophages. J Immunol 1993; 150: 2885-2891.

64 Kielian TL et al. Lipopolysaccharide modulation of a CD14-like molecule on porcine alveolar macrophages. J Leukoc Biol 1995; 57: 581-586.

65 Haugen TS, Nakstad B, Skjonsberg OH, Lyberg T. CD14 expression and binding of lipopolysaccharide to alveolar macrophages and monocytes. Inflammation 1998; 22: 521-532.

66 Wright SD. Multiple receptors for endotoxin. Curr Opin Immunol 1991; 3: 83-90. 67 Holtzman MJ. Arachidonic acid metabolism. Implications of biological chemistry for

lung function and disease. Am Rev Respir Dis 1991; 143: 188-203. 68 Johnson KJ, Ward PA. Acute immunologic pulmonary alveolitis. J Clin Invest 1974;

54: 349-357. 69 Ward PA. Phagocytes and the lung. Ann N Y Acad Sci 1997; 832: 304-310. 70 Peters-Golden M, Coffey M. Role of leukotrienes in antimicrobial defense of the lung.

J Lab Clin Med 1998; 132: 251-257. 71 Tino MJ, Wright JR. Surfactant protein A stimulates phagocytosis of specific

pulmonary pathogens by alveolar macrophages. Am J Physiol 1996; 270: L677-688.

References

113

72 Restrepo CI et al. Surfactant protein D stimulates phagocytosis of Pseudomonas aeruginosa by alveolar macrophages. Am J Respir Cell Mol Biol 1999; 21: 576-585.

73 Watford WT, Smithers MB, Frank MM, Wright JR. Surfactant protein A enhances the phagocytosis of C1q-coated particles by alveolar macrophages. Am J Physiol Lung Cell Mol Physiol 2002; 283: L1011-1022.

74 Jiang JX et al. Intrapulmonary expression of scavenger receptor and CD14 and their relation to local inflammatory responses to endotoxemia in mice. World J Surg 2003; 27: 1-9.

75 Obot CJ et al. Surface components of airborne particulate matter induce macrophage apoptosis through scavenger receptors. Toxicol Appl Pharmacol 2002; 184: 98-106.

76 Yu BP. Cellular defenses against damage from reactive oxygen species. Physiol Rev 1994; 74: 139-162.

77 Pricop L, Salmon JE. Redox regulation of Fc gamma receptor-mediated phagocytosis: implications for host defense and tissue injury. Antioxid Redox Signal 2002; 4: 85-95.

78 Fernandez ML et al. Role of distinct subpopulations of peritoneal macrophages in the regulation of reactive oxygen species release. Free Radic Biol Med 1999; 27: 797-809.

79 Forman HJ, Torres M. Signaling by the respiratory burst in macrophages. IUBMB Life 2001; 51: 365-371.

80 West MA, Manthei R, Bubrick MP. Autoregulation of hepatic macrophage activation in sepsis. J Trauma 1993; 34: 473-479; discussion 479-480.

81 Turner NC, Wood LJ. Superoxide generation by guinea-pig peritoneal macrophages is inhibited by rolipram, staurosporine and mepacrine in an agonist-dependent manner. Cell Signal 1994; 6: 923-931.

82 Maridonneau-Parini I, Tringale SM, Tauber AI. Identification of distinct activation pathways of the human neutrophil NADPH-oxidase. J Immunol 1986; 137: 2925-2929.

83 Ouadrhiri Y et al. Effect of IgA on respiratory burst and cytokine release by human alveolar macrophages: role of ERK1/2 mitogen-activated protein kinases and NF-kappaB. Am J Respir Cell Mol Biol 2002; 26: 315-332.

84 Pilette C et al. IL-9 inhibits oxidative burst and TNF-alpha release in lipopolysaccharide-stimulated human monocytes through TGF-beta. J Immunol 2002; 168: 4103-4111.

85 Gozal E, Forman HJ, Torres M. ADP stimulates the respiratory burst without activation of ERK and AKT in rat alveolar macrophages. Free Radic Biol Med 2001; 31: 679-687.

86 Schoonbroodt S et al. Crucial role of the amino-terminal tyrosine residue 42 and the carboxyl-terminal PEST domain of I kappa B alpha in NF-kappa B activation by an oxidative stress. J Immunol 2000; 164: 4292-4300.

87 Murray HW. Interferon-gamma, the activated macrophage, and host defense against microbial challenge. Ann Intern Med 1988; 108: 595-608.

88 Bingisser RM, Holt PG. Immunomodulating mechanisms in the lower respiratory tract: nitric oxide mediated interactions between alveolar macrophages, epithelial cells, and T-cells. Swiss Med Wkly 2001; 131: 171-179.

89 Hahn EL, Gamelli RL. Prostaglandin E2 synthesis and metabolism in burn injury and trauma. J Trauma 2000; 49: 1147-1154.

90 Stephan RN et al. Prostaglandin E2 depresses antigen-presenting cell function of peritoneal macrophages. J Surg Res 1988; 44: 733-739.

91 Miller-Graziano CL et al. Role of elevated monocyte transforming growth factor beta (TGF beta) production in posttrauma immunosuppression. J Clin Immunol 1991; 11: 95-102.

92 Riese U et al. Catecholamines induce IL-10 release in patients suffering from acute myocardial infarction by transactivating its promoter in monocytic but not in T-cells. Mol Cell Biochem 2000; 212: 45-50.

References

114

93 Kshettry VR, Bolman RM, 3rd. Chest trauma. Assessment, diagnosis, and management. Clin Chest Med 1994; 15: 137-146.

94 Hudson LD, Milberg JA, Anardi D, Maunder RJ. Clinical risks for development of the acute respiratory distress syndrome. Am J Respir Crit Care Med 1995; 151: 293-301.

95 Volk HD. Immunodepression in the surgical patient and increased susceptibility to infection. Crit Care 2002; 6: 279-281.

96 Woiciechowsky C et al. Sympathetic activation triggers systemic interleukin-10 release in immunodepression induced by brain injury. Nat Med 1998; 4: 808-813.

97 Baker CC et al. Epidemiology of trauma deaths. Am J Surg 1980; 140: 144-150. 98 Ventilation with lower tidal volumes as compared with traditional tidal volumes for

acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000; 342: 1301-1308.

99 Mayorga MA. The pathology of primary blast overpressure injury. Toxicology 1997; 121: 17-28.

100 Stapczynski JS. Blast injuries. Ann Emerg Med 1982; 11: 687-694. 101 Stuhmiller JH, Chuong CJ, Phillips YY, Dodd KT. Computer modeling of thoracic

response to blast. J Trauma 1988; 28: S132-139. 102 Elsayed NM. Toxicology of blast overpressure. Toxicology 1997; 121: 1-15. 103 Weiler-Ravell D, Adatto R, Borman JB. Blast injury of the chest. A review of the

problem and its treatment. Isr J Med Sci 1975; 11: 268-274. 104 Phillips YY. Primary blast injuries. Ann Emerg Med 1986; 15: 1446-1450. 105 Schardin H. The physical principles of the effects of a detonation. In: German Aviation

Medicine, World War II, US Government Printing Office, Washington, DC 1950; 2: 1207-1224.

106 Cooper GJ, Taylor DE. Biophysics of impact injury to the chest and abdomen. J R Army Med Corps 1989; 135: 58-67.

107 Coppel DL. Blast injuries of the lungs. Br J Surg 1976; 63: 735-737. 108 Huller T, Bazini Y. Blast injuries of the chest and abdomen. Arch Surg 1970; 100: 24-

30. 109 McCaughey W, Coppel DL, Dundee JW. Blast injuries to the lungs. A report of two

cases. Anaesthesia 1973; 28: 2-9. 110 Argyros GJ. Management of primary blast injury. Toxicology 1997; 121: 105-115. 111 Menkin M, Schwartzman RJ. Cerebral air embolism. Report of five cases and review

of the literature. Arch Neurol 1977; 34: 168-170. 112 Nast-Kolb D et al. [Biochemical factors as objective parameters for assessing the

prognosis in polytrauma]. Unfallchirurg 1992; 95: 59-66. 113 Johnson JA, Cogbill TH, Winga ER. Determinants of outcome after pulmonary

contusion. J Trauma 1986; 26: 695-697. 114 Fulton RL, Jones CE. The cause of post-traumatic pulmonary insufficiency in man.

Surg Gynecol Obstet 1975; 140: 179-186. 115 Hooker DR. Physiologic effects of air concussion. Amer J Physiol 1924; 67: 219. 116 Irwin RJ et al. Cardiopulmonary physiology of primary blast injury. J Trauma 1997;

43: 650-655. 117 Irwin RJ et al. Global primary blast injury: a rat model. J Okla State Med Assoc 1998;

91: 387-392. 118 Brown RF, Cooper GJ, Maynard RL. The ultrastructure of rat lung following acute

primary blast injury. Int J Exp Pathol 1993; 74: 151-162. 119 Tsokos M et al. Histological, immunohistochemical and ultra-structural findings in

human blast lung injury. in press July 3, 2003. 120 Clemedson CJ, Hultman HJ. Air embolism and the causes of death in blast injury.

Milit Surg 1954; 114: 424-437. 121 Mellor SG, Cooper GJ. Analysis of 828 servicemen killed or injured by explosion in

Northern Ireland 1970-84: the Hostile Action Casualty System. Br J Surg 1989; 76: 1006-1010.

122 Guy RJ, Kirkman E, Watkins PE, Cooper GJ. Physiologic responses to primary blast. J Trauma 1998; 45: 983-987.

References

115

123 Clemedson CJ. An experimental study on air blast injuries. Acta Physiol Scand 1949; 18: 1.

124 Irwin RJ et al. Shock after blast wave injury is caused by a vagally mediated reflex. J Trauma 1999; 47: 105-110.

125 Ohnishi M, Kirkman E, Guy RJ, Watkins PE. Reflex nature of the cardiorespiratory response to primary thoracic blast injury in the anaesthetised rat. Exp Physiol 2001; 86: 357-364.

126 Guy RJ, Watkins PE, Edmondstone WM. Electrocardiographic changes following primary blast injury to the thorax. J R Nav Med Serv 2000; 86: 125-133.

127 Fulda GJ et al. An evaluation of serum troponin T and signal-averaged electrocardiography in predicting electrocardiographic abnormalities after blunt chest trauma. J Trauma 1997; 43: 304-310; discussion 310-302.

128 Guy RJ, Glover MA, Cripps NP. The pathophysiology of primary blast injury and its implications for treatment. Part I: The thorax. J R Nav Med Serv 1998; 84: 79-86.

129 Cobb JP, Buchman TG, Karl IE, Hotchkiss RS. Molecular biology of multiple organ dysfunction syndrome: injury, adaptation, and apoptosis. Surg Infect (Larchmt) 2000; 1: 207-215.

130 Woiciechowsky C et al. Mechanisms of brain-mediated systemic anti-inflammatory syndrome causing immunodepression. J Mol Med 1999; 77: 769-780.

131 Gorbunov NV et al. Air blast-induced pulmonary oxidative stress: interplay among hemoglobin, antioxidants, and lipid peroxidation. Am J Physiol 1997; 272: L320-334.

132 Elsayed NM et al. Antioxidant depletion, lipid peroxidation, and impairment of calcium transport induced by air-blast overpressure in rat lungs. Exp Lung Res 1996; 22: 179-200.

133 Armstrong KL, Cooper MF, Williams MT, Elsayed NM. Vitamin E and lipoic acid, but not vitamin C improve blood oxygenation after high-energy IMPULSE noise (BLAST) exposure. Biochem Biophys Res Commun 1998; 253: 114-118.

134 Brückner UB, Rosch M, Kelbel MW, Gebhard F. [Prostanoid release and lipid peroxidation in patients with thoracic trauma]. Zentralbl Chir 1999; 124: 144-158.

135 Gebhard F, Strecker W, Brückner UB, Kinzl L. Untersuchungen zur systemischen posttraumatischen Inflammation in der Frühphase nach Trauma. Hefte zur Zeitschrift "Der Unfallchirurg". Springer Verlag Heidelberg, New York, Barcelona 2000; 276.

136 Gebhard F, Marzinzig M, Hartel W, Brückner UB. [Systemic release of prostanoids after surgically-induced injury of lung tissue]. Langenbecks Arch Chir 1997; 382: 243-251.

137 Nevalainen TJ, Haapamaki MM, Gronroos JM. Roles of secretory phospholipases A(2) in inflammatory diseases and trauma. Biochim Biophys Acta 2000; 1488: 83-90.

138 Brückner UB et al. Prognostische Bedeutung präklinisch erfaßter biochemischer Mediatoren beim Polytrauma. Zentralbl Chir 1999; 124: 303-310.

139 Cernak I et al. Leukotrienes in the pathogenesis of pulmonary blast injury. J Trauma 1996; 40: S148-151.

140 Zunic G, Cernak I, Malicevic Z, Savic J. Inhibition of leukotriene formation by diethylcarbamazine modifies the acid-base balance in the rabbits with blast injuries of the lungs. Vojnosanit Pregl 1999; 56: 243-247.

141 Stylianos S, Wakabayashi G, Gelfand JA, Harris BH. Experimental hemorrhage and blunt trauma do not increase circulating tumor necrosis factor. J Trauma 1991; 31: 1063-1067.

142 Hensler T et al. Association between injury pattern of patients with multiple injuries and circulating levels of soluble tumor necrosis factor receptors, interleukin-6 and interleukin-10, and polymorphonuclear neutrophil elastase. J Trauma 2002; 52: 962-970.

143 Slutsky AS. Lung injury caused by mechanical ventilation. Chest 1999; 116: 9S-15S. 144 Nakano H et al. Efficacy of intraportal injection of anti-ICAM-1 monoclonal antibody

against liver cell injury following warm ischemia in the rat. Am J Surg 1995; 170: 64-66.

References

116

145 Rossaint R et al. Inhaled nitric oxide for the adult respiratory distress syndrome. N Engl J Med 1993; 328: 399-405.

146 Guirao X, Lowry SF. Biologic control of injury and inflammation: much more than too little or too late. World J Surg 1996; 20: 437-446.

147 Davies MG, Hagen PO. Systemic inflammatory response syndrome. Br J Surg 1997; 84: 920-935.

148 Ertel W et al. Release of anti-inflammatory mediators after mechanical trauma correlates with severity of injury and clinical outcome. J Trauma 1995; 39: 879-885; discussion 885-877.

149 Cinat ME et al. Trauma causes sustained elevation of soluble tumor necrosis factor receptors. J Am Coll Surg 1994; 179: 529-537.

150 Xing Z et al. IL-6 is an antiinflammatory cytokine required for controlling local or systemic acute inflammatory responses. J Clin Invest 1998; 101: 311-320.

151 Tilg H et al. Interleukin-6 (IL-6) as an anti-inflammatory cytokine: induction of circulating IL-1 receptor antagonist and soluble tumor necrosis factor receptor p55. Blood 1994; 83: 113-118.

152 Porteu F et al. Human neutrophil elastase releases a ligand-binding fragment from the 75-kDa tumor necrosis factor (TNF) receptor. Comparison with the proteolytic activity responsible for shedding of TNF receptors from stimulated neutrophils. J Biol Chem 1991; 266: 18846-18853.

153 Fukushima R et al. Time course of production of cytokines and prostaglandin E2 by macrophages isolated after thermal injury and bacterial translocation. Circ Shock 1994; 42: 154-162.

154 Miller-Graziano CL et al. Mechanisms of altered monocyte prostaglandin E2 production in severely injured patients. Arch Surg 1988; 123: 293-299.

155 McCarter MD et al. Trauma-induced alterations in macrophage function. Surgery 1998; 123: 96-101.

156 Redmond HP et al. Effects of laparotomy on systemic macrophage function. Surgery 1992; 111: 647-655.

157 Uhlig S, Heiny O. Measuring the weight of the isolated perfused rat lung during negative pressure ventilation. J Pharmacol Toxicol Methods 1995; 33: 147-152.

158 Stossel TP. Phagocytosis. Prog Clin Biol Res 1977; 13: 87-102. 159 Tan AS, Berridge MV. Superoxide produced by activated neutrophils efficiently

reduces the tetrazolium salt, WST-1 to produce a soluble formazan: a simple colorimetric assay for measuring respiratory burst activation and for screening anti-inflammatory agents. J Immunol Methods 2000; 238: 59-68.

160 Bergmeyer HU. Methods of Enzymatic Analysis. Verlag Chemie, Weinheim 1984. 161 Suga M et al. Characteristic elevation of matrix metalloproteinase activity in idiopathic

interstitial pneumonias. Am J Respir Crit Care Med 2000; 162: 1949-1956. 162 Cooper GJ, Townend DJ, Cater SR, Pearce BP. The role of stress waves in thoracic

visceral injury from blast loading: modification of stress transmission by foams and high-density materials. J Biomech 1991; 24: 273-285.

163 Angele MK, Faist E. Clinical review: immunodepression in the surgical patient and increased susceptibility to infection. Crit Care 2002; 6: 298-305.

164 Polk HC. Supportive therapy in burn care. Consensus summary on infection. J Trauma 1979; 19: 894-896.

165 Zellweger R et al. Effect of surgical trauma on splenocyte and peritoneal macrophage immune function. J Trauma 1995; 39: 645-650.

166 Kim KJ et al. Absorption of intact albumin across rat alveolar epithelial cell monolayers. Am J Physiol Lung Cell Mol Physiol 2003; 284: L458-465.

167 Sherman JM et al. Time course of hemosiderin production and clearance by human pulmonary macrophages. Chest 1984; 86: 409-411.

168 Roumen RM et al. Cytokine patterns in patients after major vascular surgery, hemorrhagic shock, and severe blunt trauma. Relation with subsequent adult respiratory distress syndrome and multiple organ failure. Ann Surg 1993; 218: 769-776.

References

117

169 Suter PM et al. High bronchoalveolar levels of tumor necrosis factor and its inhibitors, interleukin-1, interferon, and elastase, in patients with adult respiratory distress syndrome after trauma, shock, or sepsis. Am Rev Respir Dis 1992; 145: 1016-1022.

170 Shanley TP et al. Requirement for C-X-C chemokines (macrophage inflammatory protein-2 and cytokine-induced neutrophil chemoattractant) in IgG immune complex-induced lung injury. J Immunol 1997; 158: 3439-3448.

171 Murphy G, Docherty AJ. The matrix metalloproteinases and their inhibitors. Am J Respir Cell Mol Biol 1992; 7: 120-125.

172 Matthay MA, Folkesson HG, Clerici C. Lung epithelial fluid transport and the resolution of pulmonary edema. Physiol Rev 2002; 82: 569-600.

173 Canessa CM et al. Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature 1994; 367: 463-467.

174 Jaffin JH et al. A laboratory model for studying blast overpressure injury. J Trauma 1987; 27: 349-356.

175 Schaefer CF et al. Choice of anesthetic alters the circulatory shock pattern as gauged by conscious rat endotoxemia. Acta Anaesthesiol Scand 1987; 31: 550-556.

176 Uhlig S et al. Functional and fine structural changes in isolated rat lungs challenged with endotoxin ex vivo and in vitro. Am J Pathol 1995; 146: 1235-1247.

177 Rutherford RB, Valenta J. An experimental study of "traumatic wet lung&quot. J Trauma 1971; 11: 146-166.

178 Ranieri VM, Giunta F, Suter PM, Slutsky AS. Mechanical ventilation as a mediator of multisystem organ failure in acute respiratory distress syndrome. Jama 2000; 284: 43-44.

179 Slutsky AS. Basic science in ventilator-induced lung injury: implications for the bedside. Am J Respir Crit Care Med 2001; 163: 599-600.

180 Held HD, Boettcher S, Hamann L, Uhlig S. Ventilation-induced chemokine and cytokine release is associated with activation of nuclear factor-kappaB and is blocked by steroids. Am J Respir Crit Care Med 2001; 163: 711-716.

181 Corbridge TC et al. Adverse effects of large tidal volume and low PEEP in canine acid aspiration. Am Rev Respir Dis 1990; 142: 311-315.

182 Kolton M et al. Oxygenation during high-frequency ventilation compared with conventional mechanical ventilation in two models of lung injury. Anesth Analg 1982; 61: 323-332.

183 Muscedere JG, Mullen JB, Gan K, Slutsky AS. Tidal ventilation at low airway pressures can augment lung injury. Am J Respir Crit Care Med 1994; 149: 1327-1334.

184 Folkesson HG, Kheradmand F, Matthay MA. The effect of salt water on alveolar epithelial barrier function. Am J Respir Crit Care Med 1994; 150: 1555-1563.

185 Martin TR et al. Effects of leukotriene B4 in the human lung. Recruitment of neutrophils into the alveolar spaces without a change in protein permeability. J Clin Invest 1989; 84: 1609-1619.

186 Saumon G, Basset G, Bouchonnet F, Crone C. cAMP and beta-adrenergic stimulation of rat alveolar epithelium. Effects on fluid absorption and paracellular permeability. Pflugers Arch 1987; 410: 464-470.

187 Crandall ED, Heming TA, Palombo RL, Goodman BE. Effects of terbutaline on sodium transport in isolated perfused rat lung. J Appl Physiol 1986; 60: 289-294.

188 Goodman BE, Fleischer RS, Crandall ED. Evidence for active Na+ transport by cultured monolayers of pulmonary alveolar epithelial cells. Am J Physiol 1983; 245: C78-83.

189 Cott GR, Sugahara K, Mason RJ. Stimulation of net active ion transport across alveolar type II cell monolayers. Am J Physiol 1986; 250: C222-227.

190 Minnear FL et al. Isoproterenol antagonizes endothelial permeability induced by thrombin and thrombin receptor peptide. J Appl Physiol 1993; 75: 1171-1179.

191 Warburton D et al. Effects of beta-2 agonist on tracheal fluid flow, surfactant and pulmonary mechanics in the fetal lamb. J Pharmacol Exp Ther 1987; 242: 394-398.

References

118

192 Johnson M. Pharmacology of long-acting beta-agonists. Ann Allergy Asthma Immunol 1995; 75: 177-179.

193 Lalloo UG et al. Budesonide and formoterol in a single inhaler improves asthma control compared with increasing the dose of corticosteroid in adults with mild-to-moderate asthma. Chest 2003; 123: 1480-1487.

194 Young SL, Silbajoris RA. Type II cell response to chronic beta adrenergic agonist and antagonist infusions. J Pharmacol Exp Ther 1985; 233: 271-276.

195 Brown LA, Longmore WJ. Adrenergic and cholinergic regulation of lung surfactant secretion in the isolated perfused rat lung and in the alveolar type II cell in culture. J Biol Chem 1981; 256: 66-72.

196 Pozzi R. [True and presumed contraindications of beta blockers. Peripheral vascular disease, diabetes mellitus, chronic bronchopneumopathy]. Ital Heart J 2000; 1: 1031-1037.

197 Finley TN, Aronow A, Cosentino AM, Golde DW. Occult pulmonary hemorrhage in anticoagulated patients. Am Rev Respir Dis 1975; 112: 23-29.

198 Grissom CK, Albertine KH, Elstad MR. Alveolar haemorrhage in a case of high altitude pulmonary oedema. Thorax 2000; 55: 167-169.

199 Ware LB, Conner ER, Matthay MA. von Willebrand factor antigen is an independent marker of poor outcome in patients with early acute lung injury. Crit Care Med 2001; 29: 2325-2331.

200 Tatic V et al. Morphologic characteristics of primary nonperforative intestinal blast injuries in rats and their evolution to secondary perforations. J Trauma 1996; 40: S94-99.

201 Nelson S et al. Compartmentalization of intraalveolar and systemic lipopolysaccharide-induced tumor necrosis factor and the pulmonary inflammatory response. J Infect Dis 1989; 159: 189-194.

202 Mitchell MD, Trautman MS. Molecular mechanisms regulating prostaglandin action. Mol Cell Endocrinol 1993; 93: C7-10.

203 Gryglewski RJ, Korbut R, Ocetkiewicz A. Generation of prostacyclin by lungs in vivo and its release into the arterial circulation. Nature 1978; 273: 765-767.

204 Morioka K et al. Leukocyte and platelet depletion with a blood cell separator: effects on lung injury after cardiac surgery with cardiopulmonary bypass. J Thorac Cardiovasc Surg 1996; 111: 45-54.

205 Schuster VL. Molecular mechanisms of prostaglandin transport. Annu Rev Physiol 1998; 60: 221-242.

206 Martin TR. Lung cytokines and ARDS: Roger S. Mitchell Lecture. Chest 1999; 116: 2S-8S.

207 Hesse DG et al. Cytokine appearance in human endotoxemia and primate bacteremia. Surg Gynecol Obstet 1988; 166: 147-153.

208 Ayala A et al. Differential alterations in plasma IL-6 and TNF levels after trauma and hemorrhage. Am J Physiol 1991; 260: R167-171.

209 Keel M et al. [Increased plasma level of Type I (p55) and Type II (p75) TNF-receptors following trauma]. Swiss Surg 1995: 241-244.

210 Aderka D et al. Stabilization of the bioactivity of tumor necrosis factor by its soluble receptors. J Exp Med 1992; 175: 323-329.

211 Mohler KM et al. Soluble tumor necrosis factor (TNF) receptors are effective therapeutic agents in lethal endotoxemia and function simultaneously as both TNF carriers and TNF antagonists. J Immunol 1993; 151: 1548-1561.

212 Tan LR et al. Trauma causes early release of soluble receptors for tumor necrosis factor. J Trauma 1993; 34: 634-638.

213 Fiorentino DF et al. IL-10 inhibits cytokine production by activated macrophages. J Immunol 1991; 147: 3815-3822.

214 Welty-Wolf KE et al. Tissue factor in experimental acute lung injury. Semin Hematol 2001; 38: 35-38.

References

119

215 Miller DL et al. Extrinsic coagulation blockade attenuates lung injury and proinflammatory cytokine release after intratracheal lipopolysaccharide. Am J Respir Cell Mol Biol 2002; 26: 650-658.

216 Utter GH et al. Injury induces increased monocyte expression of tissue factor: factors associated with head injury attenuate the injury-related monocyte expression of tissue factor. J Trauma 2002; 52: 1071-1077; discussion 1077.

217 Dong W, Lewtas J, Luster MI. Role of endotoxin in tumor necrosis factor alpha expression from alveolar macrophages treated with urban air particles. Exp Lung Res 1996; 22: 577-592.

218 Yamasawa H, Ishii Y, Kitamura S. Cytokine-induced neutrophil chemoattractant in a rat model of lipopolysaccharide-induced acute lung injury. Inflammation 1999; 23: 263-274.

219 Kang JL et al. Genistein prevents nuclear factor-kappa B activation and acute lung injury induced by lipopolysaccharide. Am J Respir Crit Care Med 2001; 164: 2206-2212.

220 Baggiolini M, Loetscher P. Chemokines in inflammation and immunity. Immunol Today 2000; 21: 418-420.

221 Fukuda Y et al. Localization of matrix metalloproteinases-1, -2, and -9 and tissue inhibitor of metalloproteinase-2 in interstitial lung diseases. Lab Invest 1998; 78: 687-698.

222 Ricou B et al. Matrix metalloproteinases and TIMP in acute respiratory distress syndrome. Am J Respir Crit Care Med 1996; 154: 346-352.

223 Finlay GA et al. Matrix metalloproteinase expression and production by alveolar macrophages in emphysema. Am J Respir Crit Care Med 1997; 156: 240-247.

224 Hayatdavoudi G et al. Pulmonary injury in rats following continuous exposure to 60% O2 for 7 days. J Appl Physiol 1981; 51: 1220-1231.

225 Blusse van Oud Alblas A, van der Linden-Schrever B, Van Furth R. Origin and kinetics of pulmonary macrophages during an inflammatory reaction induced by intra-alveolar administration of aerosolized heat-killed BCG. Am Rev Respir Dis 1983; 128: 276-281.

226 Schmidt K, Bruchelt G, Kistler D, Koslowski L. Phagocytic activity of granulocytes and alveolar macrophages after burn injury measured by chemiluminescence. Burns Incl Therm Inj 1983; 10: 79-85.

227 Vazina IR, Pylaeva SI, Vasil'chuk OA. [Disorder of the phagocytosis process in alveolar macrophages in thermal injury]. Biull Eksp Biol Med 1984; 97: 542-544.

228 Howat WJ, Holgate ST, Lackie PM. TGF-beta isoform release and activation during in vitro bronchial epithelial wound repair. Am J Physiol Lung Cell Mol Physiol 2002; 282: L115-123.

229 Monick MM, Hunninghake GW. Activation of second messenger pathways in alveolar macrophages by endotoxin. Eur Respir J 2002; 20: 210-222.

230 Hermann C et al. Cytokine induction by purified lipoteichoic acids from various bacterial species--role of LBP, sCD14, CD14 and failure to induce IL-12 and subsequent IFN-gamma release. Eur J Immunol 2002; 32: 541-551.

231 Morath S, Geyer A, Hartung T. Structure-function relationship of cytokine induction by lipoteichoic acid from Staphylococcus aureus. J Exp Med 2001; 193: 393-397.

232 Cech AC, Shou J, Gallagher H, Daly JM. Glucocorticoid receptor blockade reverses postinjury macrophage suppression. Arch Surg 1994; 129: 1227-1232.

233 Werb Z, Foley R, Munck A. Interaction of glucocorticoids with macrophages. Identification of glucocorticoid receptors in monocytes and macrophages. J Exp Med 1978; 147: 1684-1694.

234 Schwöbel F. Execution and elimination of apoptotic cells. Dissertation, Universität Konstanz 2002.

Abbreviations

120

9. Abbreviatons

AM alveolar macrophage ARDS acute respiratory distress syndrome AUC area under the curve BAL bronchoalveolar lavage BW body weight CINC chemokine-induced neutrophil chemotactic factor CL dynamic lung compliance COX cyclooxygenase E. coli Escherichia coli Fpre, Fpost pre-and post breathing frequency IL interleukin LDH lactate dehydrogenase LPS lipopolyssacharide LTA lipoteichoic acid LWR lung wet weight weight to body weight ratio LWW lung wet weight MMPs matrix metalloproteinases MPO myeloperoxidase NO nitric oxide PEEP positive end-expiratory pressure PGF1a prostacycline PIP positive inspiratory pressure PMN polymorphonuclear leukocytes Pp pressure peak Qi injury quotient RA airway resistance RBCs red blood cells Rp cracking pressure S. aureus Staphylococcus aureus SEM/ SD standard error of means/ standard deviation sTNFRs soluble tumor necrosis factor receptors t duration TNF tumor necrosis factor TXA2 thromboxane A2

∆W weight gain

Thorax trauma-induced experimental lung injury

Documents

Transcript of Thorax trauma-induced experimental lung injury