UREA RETENTION MECHANISMS IN THE BRANCHIAL EPITHELniTM OF A

MARINE ELASMOBRANCH, THE S P D N DOGFISH (SQUAXUS ACANTHL4S)

A Thesis

Presented to

The Facdty of Graduate Studies

of

The University of Guelph

by

GLENN ALEXANDER FINES

In partial fiilfillment of requirernents

for the degree of

Master of Science

July, 2000

O Glenn A. Fines, 2000

National Library 1+1 of,", Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Senrices services bibliographiques

395 Wellington Street 395. rue Wellington Ottawa ON KI A ON4 Ottawa ON K1A ON4 Canada Canada

Your rile Votre rëferenœ

Our iile Noire refdrence

The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distriibute or sell reproduire, pGtcr, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/fïlm, de

reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fkom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.

ABSTRACT

T m MECECANISMS OF UREA RETENTION IN BRANCHIAL EPITHELIUM OF ELASMOBRANCHS

Glenn Alexander Fines University of Guelph, 2000

Advisors: Dr. J.S. Baliantyne and Dr. PA. Wright

The retention of hi& concentrations of urea in the tissues of marine

elasmobranchs is the key to their osmoregulatory strategy. The rnechanisms responsible

for the low urea permeability of the gill epithelium nom the sphy dogfïsh (Squalus

acanthias) were investigated using enriched basolateral membrane vesicles (BLMV).

Urea uptake &=IO IIM) was sodium dependent and inhibited by phloretin (Iso=0.08

mM) and urea analogs (e.g. thiourea and methylurea). Much of the impermeability of the

impermeability of the BLMV may be due to the very high cholesterol content apparent

fiom the high cholesterol to phospholipid ratio (3.68). The high phosphatidylcholine and

low polyunsaturated fatty acid Levels are also likely to confer increased order to the

bilayer membrane, making it less fluid and less permeable.

Taken together, these findings indicate that low urea permeability in the dogfish

gill is primarily due to an active urea transporter that returns urea to the blood against the

concentration gradient and a unique lipid composition that minimizes diffusion urea

across the basolateral membrane.

ACKNOWLEDGEMENTS

1 gratefidly achowledge the f3nancial support of the Natural Sciences and

Engineering Research Council and the Huntsman Marine Science Centre. I would like to

thank my advisors, Dr. Jim BalIantyne and Dr. Pat Wright for their support and gïving me

the chance to work on this great project I a h want to thank Dr. Glen Van der Kraak for

serving on my advisory cornmittee and all of his helpful comments on my thesis. 1 thank

dl of my labmates (Jason, Natasha, Andy, Marc, J o ~ , Andrea, and Dave) for listening to

my incessant chattering about urea transporters and throwing great parties. 1 wouid like

to thank my family for their support. Finally, 1 wodd Like to thank Jennifer for her

support and encouragement.

TABLE OF CONTENTS

........................................................................................................ Açknowledgements i

.. Table of Contents .......................................................................................................... 11 ... ............................................................................................................... List of Tables iir

.............................................................................................................. List of Figures iv

.............................................. General Introduction .. .................................................. 1

Chapter 1 Active urea transport and an unusual basolateral membrane composition in the gills of a marine elasmobranch .......................................................... 11

........................................................................................................ Introduction -12

........................................................................................ Materials and Methods 15

................................................................................................................ Results -23

........................................................................................................... Discussion 45

Chapter 2 Lipid Composition of the Basolateral Membrane in Gill Epithelium of the ..................................................................... Spiny Dogfish (Squalus acanthias) 56

........................................................................................................ Introduction S7

........................................................................................ Materials and Methods 59

................................................................................................................. Results 61

........................................................................................................... Discussion 65

...................................................................................................... General Discussion 71

.................................................................................................................... References 79

................................................................................................................... Appendix I 95

.................................................................................................................. Appendiv II 96

LIST OF TABLES

Table 1.1. Marker enzyme specific activities and magnitude of purification of basolateral membrane ............................. ,.. ................................................. -24

Table 1.2. Total activity of marker enzymes, percent recovery, and percent contamination in the final basolateral membrane vesicle preparation ............. 25

Table 1.3. Percentage of the basolateral membrane fraction as resealed and the orientation of the vesicles ............................................................................... -26

Table 1.4. Percentage of phospholipid types and total phospholipid and cholesteroi in the basolateral membrane of gill epithelium fiom the spiny dogfish,

........................................ SquaItcs acanthias

Table 1 S. Cho1esterol:phospholipid ratios of representative species fkom difTerent .............................................................................................................. phyla.. .43

Table 1.6 Cornparison of cholesterol to protein ratios in the basolateral membrane of the gill epithelium fiom the spiny dogfïsh and a marine and fieshwater

............................................................................................................. teleost. -44

Table 2.1. Cumulative percentages of individuai fatty acids in gill basolaterd plasma membrane fiom Squalur acanthias ..................................................... 63

Table 2.2. Percentages of individual fatty acids in the major phospholipids fiom the ................................. gill basolateral plasma membrane of Squalus acanthias 64

iii

LIST OF FFGURES

Figure 1.1 Filter optirniration for autofluorescence and non-specinc binding of 14c-

urea for use in urea transport rapid filtration experiments. Blanks involve filtering the radioactive mixture only through the filter. Controls involve adding radioactive mixture to diluted vesicles and immediately filtering through the filter. (Mean t SE, n=2) ............................................................... 28

Figure 1.2 a. Rates of urea uptake at variable urea concentrations, by BLMV fiom the gill of the dogfkh, SquaZus acanthias. (Mean t SE, n=8). b. Expansion of the low end of the [urea] range fiom a. The regression is y = 0a07741n(x) + 0.01 l6 ,Z = 0.9064 (Mean & SE, n=7) c. Lineweaver-Burke transformation of the effect of variable [urea] on uptake rate by BLMV. The regression is y=2.96 12t29.8 lx, 2=0.9778 (Mean f SE, n=7) ................. 30

Figure 1.3 Dose dependent phloretin inhibition of urea uptake in BLMV fiom the gill of the dogfish, Squalus acanîhias (Ise = 0.08 mM) (Mean t SE, n=5) ...... 33

Figure 1.4 Inhibition of urea uptake in BLMV by the urea analogues, acetamide, thiourea, N-methylurea, and NPTU (3 70 mM). (Mean & S .E., n=3). ............ ..3 5

Figure 1.5 ATP (10 mM) stimulation of urea uptake and effects of ouabain (1 mM) and NEM (1 mM) in BLMV fiom the gill of the dogfkh, Squalur acanthias (Mean + SE, n=8). * significant merence fiom control (paired t-test,

............................................................................................................. pcO.05) 37

Figure 1.6 Rate of urea uptake in BLMV from the gill of the dogfish, Squalur acanthias, in the presence of physiologically oriented sodium (225 mM outwardly directed) or potassium gradients (225 mM inwardly directed) (Mean f SE, n=6). * sipifkant dserence fiom control (paired t-test, pcO.005) ** significant difference between sodium and potassium (paired t-

................................................................................................... test, pa).OO5) 39

Figure 1.7 Schematic representation of the proposed ~a+-cou~led, active urea transporter present in the basolateral membrane of the dogfkh gill

...................................................................................................... epithelium. -48

Fig. 1.8 A mode1 for the mechanism of reduced urea permeability of the dogfish gill basolateral membrane due to cholesterol. The cholesterol (red) contributes to a more tightly packed phospholipid bilayer membrane (yellow) thereby

... physically reducing the permeability of the basolateral membrane to urea.. 54

GENERAL INTRODUCTION

The vertebrate class Chondrichthyes (the cartilaginous fishes) is an ancient

lineage that contains two extant subclasses, the Elasmobranchii (sharks, skates, and rays)

and the Holocephali (ratfïsh or chimaeras) (Pough et al. 1996). The Elasmobranchii

evolved fiom the early chondrichthyans, which first appeared during the early Devonian,

approximately 400 million years ago (mya). The most recent radiation of elasmobranchs

appeared by the early Triassic (245 mya), with living families having evolved by the

Jurassic (208 mya) and extant genera appearing in the Cretaceous (144 mya) (Pough et al.

1996). Modem elasmobranchs typically occur in marine environments, dthough some

ascend rivers beyond tidal influence and some are permanent inhabitants of fieshwater

(i.e. Potomoîrygon spp.). The most fascinating characteristic of the elasmobranchs,

however, is the presence of elevated bIood levels of urea (350 to 600mM) (Robertson

1989). In most vertebrates urea is a waste product of amino acid catabolism and is

excreted fiom the body as rapidly as it is produced. Retention of such levels of urea in

elasmobranchs is therefore of considerable interest and has been under investigation for

over a century.

Discoverv of urea in elasmobranchs

Stadeler and Frérichs (1858) were the fist to discover the presence of "colossal

quantities" of urea in extracts fiom the muscle of Raja batis, R. clmata, and the dogfïsh

Scyllurn canicula (=Scylorhinus canicula). Their examination of other animals,

including teleosts and lamprey (Peh.omyzon sp.) revealed only traces of urea. Stadeler

(1 859) extended this observation to the spiny dogfish Spinm acanthias (=Squalus

acanthias) and the torpedos Torpedo marmorata and T. ocellata, and this was later

codbmed by Schultze (1 861) and Rabuteau and Papillon (1873). Knikenberg (1 88 1,

1 886, 1887, 1888) was the fïrst to undertake a systematic examination of the distribution

of urea in vertebrates and demonstrated that large amounts of urea were present not only

in the Selachii and Batoidei (=ElasmobranchÜ), but in the Chimaeroidei (=Holocephali)

as weU, although not in the lungfish (Neoceradohcs). This was an important observation

because of the close relation between the Elasmobranchii and Holocephali. Although the

lungfkh was thought to be extremely primitive, it was more closely affiliated with the

higher bony fishes. This suggested that the presence of elevated blood urea levels was an

evolutionary adaptation shared by ail Chondrichthyans. However, the function of urea in

these organisms and the physiological mechanisms by which the observed uraemic state

(pathological elevation of blood urea in mammals) is brought about, remained to be

described.

Earlv investigations of the function of urea in elasrnobranchs

The involvement of urea in the production of the electric discharge of the electric

organ of Torpedo ocellata was the first proposed function for urea in an elasrnobranch

(Gréhant and Jolyet 1891). They had observed that an increase in the urea content of the

electric organ tissue accompanied the electric discharge. Two years later Rohmann

(1893) reexamined the Gndings of Gréhant and Jolyet (189 1) and concluded that neither

urea nor any other nitrogenous compounds were involved in the production of electricity.

Baglioni (1906% 19 l7b) later confïrmed this by analysing various compounds in the

muscle, electric organ and senun of torpedo. His analyses revealed great difiibility of

urea and its nearly uniform distribution throughout the body of T. ocellata, including the

electric organ.

Another theory for the role of urea in elasmobranchs that was popdar for over 20

years at the tum of the century was that urea was essential for cardiac hct ion. Straub

(1901) found that 3.4% NaCl (the same osmotic pressure as elasmobranch blood) could

not keep the heart beating in good condition. Baglioni (1905) confinned Straub's

observation, having examined the effects of various NaCl solutions upon the heart and

concluded that urea was absolutely essential for cardiac activity. Baglioni (1905)

inferred that all elasmobranch tissues required urea for proper h c t i o n with the ideal

mixture being 2% NaCl and 2.2% urea. Further observations by Baglioni (1 9O6b)

demonstrated that the urine of elasmobranchs contained only low levels of urea, which he

believed to confirm his fïnding that urea was essential to life, by neutralizing the harmfùi

effects of high NaCl levels in the blood. Botazzi (1906), however, cnticized Baglioni's

experiments based on the fact that the red blood cells of elasmobranchs haemolysed in a

urea solution isotonic with 1.3-1.6% NaCl. He argued that a 3.4% NaCl solution was

hypertonie for the elasmobranch heart if the heart was equally permeable to urea because

the total osmolarity of the heart would be made up of both ionic and urea components,

not just ionic components. The key to a urea fiee perfusion medium eluded discovery

(Fühner 1908, De Meyer 19 10, Bornpiani 19 l3), and Baglioni (1 9 Ua) reasserted his

original view that urea was necessary for the maintenance of physiological activity in the

heart and other tissues. He even went so far as to classi@ urea as a hormone, according

to the definition of Bayliss and Starling (1902).

Frédéricq (1922) was the fist to attempt to disprove Baglioni's theory. He began

by confïrmïng that 3.5% NaCl was incapable of maintaining contractility in the ScyZZium

heart, whereas 2% NaCl and 2% urea was. However, he ccntinued by experimenting

with weaker salt solutions and found a urea-fkee solution would maintain a heart beat for

a long t h e when potassium and calcium was present in definite proportions. Frédéricq

concluded that urea plays an important role with regard to the osmotic pressure of the

blood, but is not important osmotically or chernically with regard to the tissues, which are

fkeely permeated by i t Simply stated the current thinking of the time held that urea

fünctioned in an osmotic role across the branchial lamellae, which are Unpermeable to it,

while within the organism it is essentially inert (Smith 1936).

Osmotic role of urea in elasmobranchs

The theory that urea was used for an osmotic hc t ion was hrst alluded to during

the early studies of elasmobranch physiology (Rodier 1 899, 1900, Frédéricq 19O4), but

was not M y embraced until much later (Frédéricq 1922, Duvall and Portier 1923, and

Smith 193 1). It was revealed that elasmobranch blood is normally hypertonie to seawater

@uvall and Portier 1923) and Duvall(1925) estimated that on average, urea is

responsible for 44% of the total osmotic pressure of the blood, and therefore contributes

significantly to the maintenance of the hypertonicity of the blood. This lead to the

conclusion that urea is osmotically important with regard to the extemal environment

(Duvall 1925). The analysis of the blood of fieshwater elasmobranchs revealed that the

chIoride content was 25% lower than in marine elasmobranchs (Thorson et al. 1 967).

Urea content is also rnuch lower, accounting for half of the 50% reduction in total

osmotic pressure of the blood. This discovery led to the conclusion that urea is

physiologically involved in osmotic adaptation, particularly to a marine habitat (Smith

193 1). This has been the accepted function for urea ever since and research has focused

on the mechanisms responsible for enabling the use of urea in this way.

Phvsiolo@cal mechanisms of urea retention

Znitially it was believed that the "rnechanism" responsible for the elevated blood

urea Zevels found in elasmobranchs was rend insufïiciency, or the inability of the kidney

to remove urea fiom the blood (von Schroeder 1890). This theory was short-lived,

however, and was discarded once it was realized that the elevated blood urea levels

played a physiological role (the exact role had yet to be elucidated). Thus began the

search for the mechanisms responsible for urea retention in the elasmobranchs.

It was quickiy determined that the elasmobranch kidney played an active role in

urea retention. Baglioni (1906~) reported that the urea content of elasmobranch urine

rarely rose above one third of that of blood. It was later calculated that approximately

90% of the urea filtered by the glomedus is reabsorbed by the kidney tubule (Clarke and

Smith 1932). The exact site within the kidney and the mechanism responsible for this

reabsorption are uncertain due to conflicting and incomplete evidence for active andor

passive urea reabsorption. Evidence for an active mode of urea reabsorption in the

elasmobranch kidney indudes: 1) the fractional excretion of urea is 0.5% under normal

conditions (similar to active glucose reabsorption) (Kempton 1953), 2) urea reabsorption

is iso-osmotic (Kempton 1953), 3) 95% of filtered urea is reabsorbed, whereas only 35%

of the urea analog thiourea is reabsorbed (Boylan 1967), 4) ~ a + and urea are reabsorbed

at a fked ratio of 1.6 moles of urea per mole of ~ a ' (Schmidt-Nielsen et al. 1972), and 5)

both phloretin and chromate (inhibitors of urea transport in the toad bladder) inhibit urea

reabsorption in the dogfish kidney (Hays et al. 1977). These finding have led to the

proposai of an active urea reabsorption mechanism (Smith 193 1, Kempton 1953,

Schmidt-Nielsen et al. 1972) with loop II of the elasmobranch nephron being implicated

as a possible site for active sodium-linked urea reabsorption (Stolte et al. 1977).

However, there is conflicting evidence which suggests the possibility for the

involvement of a passive urea transport mechanism in the reabsorption of urea in the

elasmobranch kidney, including: 1) failure to detect a transport maximum for urea despite

markedly raising plasma urea concentrations (Kempton 1953), 2) the urea andogs methyl

urea and acetamide do not saturate the urea carrier mechanism (Schmidt-NieIsen and

Rabinowitz 1964), and 3) probenicid, a substance which blocks active urea secretion in

the fiog, does not affect urea reabsorption in the dogfish (Forster and Berglund 1957).

Based on these results, an alternative passive model for urea reabsorption has been

proposed that involves iso-osmotic reabsorption of sodium and water, together with

tubular impermeability to urea in either or both loop II and III of the nephron (E3oyla.n

1972). This model is supported by cytological and histological evidence (Lacey et al.

1975, Endo 1984).

Detailed study of the elasmobranch nephron ultrastructure has revealed a complex

rend counter-current multiplier system that may be involved in fluid regdation and the

passive reabsorption of urea (Lacey et al. 1985). Further evidence supporting the passive

model is the molecular characterization of a facilitated urea transporter fiom the dogfish

kidney, which has a 66% identity to the- rat facilitated urea transporter protein UT-A2

(Smith and Wright 1999). Despite this new molecular data, definitive evidence for the

involvement of either active or passive urea reabsorption in the elasmobranch kidney is

stiil lacking. However, it appears most likely that a combination of counter-current

exchange and facilitated transport is responsible for the highiy efficient reabsorption of

urea by the elasmobranch kidney.

Urea retention bv the elasmobranch d l

Although urea reabsorption by the kidney has been the primary focus of most

researchers, the early studies of the kidney provided the diagnostic tools necessary to

facilitate the examination of the mechanisms involved in the retention of urea by the giU,

an important environmental interface. It was acknowledged early on that the gill was

involved in the retention of urea and that it must have special properties in order to

prevent excess loss of urea across its large s d a c e area. Homer W. Smith (1936) fïrst

proposed that the physicai properties of the gill were such that the gill was rendered

"highly serni-penneable" to urea. Boylan and his colleagues (Boylan and Antkowiak

1962, Boylan et al. 1963) performed the fist detailed examination of the permeability of

the elasmobranch gill. Using their gin-perfüsion technique, Johnson et al. (1964)

measured the rate of loss of urea and water and Farber et al. (1965) measured the rate of

loss of sodium across the dogfish gill and calculated their permeability coefficients

(urea=7.5x 1 ~ - ~ c m / s , &0=7.6~ 1 o 6 c d s , ~ a + = l . 9x 1 O-' cmk). They concluded that the

membranes of the dogfish gill epithelium are generally tighter to each species of solute

than the toad bladder and frog skin. The difference in urea permeability of the dogfish

gdl and the amphibian tissues was much greater than for either H20 or ~ a + . They

concluded that this large difference in urea permeability was due an active transport

process specific for urea. Boylan and Lockwood (1962) tested this hypothesis by

infusing thiourea into dogfish and measuring urea and thiourea loss across the gill.

Previously it had been demonstrated that the kidney rejects thiourea (37% reabsorption

vs. 95% urea reabsorption) (Boylan and Lockwood 1 962). If an active process specific

for urea was present in the gill, they proposed that thiourea should be lost at a greater

rate. However, their measured loss of urea vs. thiourea after thiourea infusion was not

signi£icautly Werent, leading to the conclusion that another process must be at work in

the gill. The effects of urea loading was determined (Boylan and Antkowiak 1962,

Boylan et al. 1963) and it was demonstrated that urea loss increased out of proportion to

the increased gradient. This evidence led these researchers to conclude that a physical

property of the structure of the gill membrane, as opposed to a transport system, was

directly affecthg its permeability. They reasoned that if a transport system was involved,

the loss of urea across the gill should have been proportional to the gradient once the

experimentally increased urea Ievels saturated the transport system. Farber et al. (1 965)

M e r ruled out the likeiihood of an active transport system when they tested the effects

of ouabain and chloromorodrh on the rate of urea loss at the giU and found no

diBeremes fkom the controls. The final conclusion of Boylan (1967) was that the relative

impermeability of the dogfïsh gil1 was due to special characteristics of the membrane

structure and probably not an active transport system.

Focus once again retumed to the elasmobranch kidney and for the next 25 y e m

essentially no M e r research was done to characterize the urea retention mechanisms in

the gill. Wood et al. (1 995) retumed to the gill and demonstrated that contrary to the

earlier studies (Boylan and Antkowiak 1962, Boylan et al. 1963) the infusion of

acetamide and thiourea ~ i g n ~ c a n t l y increased urea efflwc across the gill while the

infusion of urea did not. These results led to the conclusion that urea retention by the gill

was more complex than once thought, and probably involved a transport system (Wood et

al. 1995) in addition to the structural properties of the membrane as proposed by Boylan

(1 967)- This hypothesis was examhed M e r by Part et al. (1 998) with measurements of

the permeability coefficients for water and urea. These values were comparable to those

of Boylan (1 967) and they also demonstrated that phloretin, a non-competitive inhibitor

of urea transport, signifïcantly increased urea efflux at the gills without affecthg water

flux. Part et al. (1998) interpreted these results as providing evidence for a "back

transport" system that is present in the basolateral membrane of the gill epithelial cells

and that the permeability of the apical membrane to urea is very low. The properties of

this transport system and the membrane structure were not examined. Additional

circumstantial evidence for the presence of a urea transport system in the dogfish gill was

provided by Smith and Wright (1 999), when they performed low stringency Northem

analysis of gill tissue showing some cross-reactivity with the dog£ïsh kidney urea

transporter (S hUT).

These studies have provided evidence for the presence of a urea transporter in the

dogfish gill warranting M e r direct detection and characterization of such a system

using vesicles of isolated basolateral membrane. in chapter one 1 describe the methods

used for the isolation of the basolateral membrane fkom the gill epitheliurn of a

representative elasmobranch, the spiny dogfish shark (SquaZus acanthias), the kinetic

characterization of an active urea transporter in the basolateral membrane, and analysis of

the cholesterol content. In chapter two the basolateral membrane preparation was

analyzed for the phospholipid and fatty acid composition using chromatographie rnethods

(thin layer chromatography and gas chromatography). Taken together these studies

contribute to a more complete picture of the mechanisms responsible for the retention of

urea by the elasmobranch giIL

CHAPTER 1: Active urea transport and elevated cholesterol Ievels: Mechanisms

responsible for conferring Iow urea permeability to the elasmobranch gill.

INTRODUCTION

Urea is a molecule found in many organisms with a range of hctions, fiom

acting as a nitrogen source in prokaryotes to a prime waste product in many vertebrates.

In 1858, Stadeler and Frérichs first discovered the presence of "colossal quantities" of

urea in muscle tissue of marine elasmobranch fish (sharks, skates, and rays), sparking

research that has continued for over 140 years. At levels fiom 350 - 600 mM (Robertson

1989), urea functions as an osmolyte in the elasmobranchs, balancing the osmotic

pressure of seawater (Smith 1936). An important unsolved question concems how

elasmobranchs maintain these elevated levels of urea in theu body fluids and tissues,

against an essentially W t e gradient with seawater. Homer W. Smith k t provided

indirect evidence that the elasrnobranch kidney efficiently reabsorbs urea, preventing loss

via the urine (Smith 1936). The giU, with its huge surface area, is consequently the most

important site of urea loss to the environment (Wood et al. 1995). The elasmobranch gill

is relatively impermeable to urea compared to the gills of most teleost fishes (Smith

1936, Boylan 1967), but the mechanism(s) responsible for this impermeability is stiil

unknown.

One possible mechanism that may confer low urea permeabiiity to the

elasmobranch gill is the incorporation of a carrier-mediated urea transport system into the

basolateral membrane. In this case, a urea transport protein would have to be oriented to

return urea to the blood fiom the gill epithelium against the gradient. A variety of urea

transporters have been described fkom tissues of different species including mamrnalian

kidney (Sands et al. 1997), amphibian skin (Ehrenfeld 1998), toadnsh gill (Walsh et al. in

press) and dogfish kidney (Smith and Wright 1999), where thcy perfom various

functions. An experiment using the isolated pemised dogfish head preparation revealed

that phloretin increases urea efflux across the gill (Part et al. 1998), providing

physiological evidence for the presence of a gill urea transporter. In addition, low

stringency Northem analysis of dogfish gili tissue revealed a possible homolog to the

dogfish kidney phloretin-sensitive urea transporter (ShUT) (Smith and Wright 1999).

Taken together, these previous studies provided circumstantial evidence for a specialized

giU urea transport protein.

A second possible mechanism that may confer low urea permeability to the

elasmobranch gill is the incorporation of cholesterol into the phospholipid bilayer

membranes of the gill epithelial celis. Cholesterol is directly correlated with the

permeability of biological membranes to solutes such as urea (Pugh et al. 1989). The low

sodium and water permeability of elasmobranch gills, relative to teleost gills (Boylan

1967), supports the involvement of a general mechanism in conferring an overall

impermeability to the gill, such as the incorporation of cholesterol into the membranes.

The purpose of this study was two-fold: 1) examine the dogfish gill in order to

determine whether an active urea transport system is present in the basolateral membrane

of the gill epithelial ceUs and then characterize its properties if present, 2) examine the

Lipid composition of the dogfish gill basolateral membrane to determine if modifications

to the composition are consistent with the observed extremely low urea permeability.

Rates of urea uptake were measured using a rapid filtration method and resealed vesides

prepared fkom purifïed basolateral membrane. Validation of the purification procedure of

Perry and Flik (1988) for use with elasmobranch gill epithelial tissue was accomplished

by measuring the activities of the enzyme markers for various cellular membranes in the

initiai homogenate and the final basolateral membrane preparation. These measurements

dowed the determination ofthe purification, contamination, resealing, and orientation of

the BLMV. Vesicular volume was rneasured using a radioisotopic method (Brand 1995),

allowing additional c o b a t i o n of vesicle resealing. Urea transport in the BLMV was

estabfished and characterized wing I4c-urea and a rapid filtration technique (Perry and

Flik 1988). Cornpetitive and non-cornpetitive inhibitors of urea transport were tested to

characterize the mode of transport. The energy dependence of urea transport in BLMV

was determined by measuring urea uptake in the presence of ATP, with and without

inhibitors (ouabain and NEM) present. Finally, the ion specincity of urea transport in the

BLMV was determined by incubating vesicles with modified compositions of the intemal

and extemal mediums. The lipid composition and cholesterol content of the BLMV were

detennined using chromatographic and spectrophotometric methods, respectively.

MATERIALS AND METHODS

Experimental Animals

Dogfkh (Squalus acanthias) were obtained by otter trawl in Passamaquoddy Bay,

New Brunswick between mid-July and the end of August 1999 and maintained at the

Huntsrnan Marine Science Centre in 1000 L outdoor tanks under natural photopenod and

supplied with filtered seawater. Dogfish do not feed in captivity and were thus held for

no more than 10 days prior to use. Arctic char (Salvelinus a(pinus) were obtained fiom

stock maintained in the Hagen Aqualab at the University of Guelph. Winter flounder

(Pleuronectes americanus) were O btained b y otter trawl in Passamaquoddy Bay, New

.Brunswick at the end of August 1999, transported to the University of Guelph and

maintained in the Hagen Aqualab.

Gill basolateral plasma membrane vesicles

BLMV were prepared using the method of Perry and Flik (1988), with some

modifications. All steps were performed at W°C. Aduit dogfish were killed by a blow

to the head and the gill arches were removed without being perfûsed. The soft tissue of

the gill arches was scraped nom the cartilaginous tissue aod homogenized with a Dounce

homogenizer first with a loosely and then with a tightly fitting pestle (30 strokes each) in

15 ml of hypotonie homogenization b&er containing (in mM) 25 NaCl, 1 dithiothreitol,

0.5 disodium ethylenedinitrilo tetra-acetic acid (EDTA), 1 N-2-hydroxyethylpiperazine-

N'-2-ethanesulfonic acid (HEPES), 1 Tris[hydroxymethyl]-aminomethane hydrochloride

(Tris-HCl) (pH 8.0), plus 100U/ml Aprotinin. Following homogenization, the volume

was adjusted to a final volume of 50 ml with the same buffer. A sarnple (0.5 ml) was

saved for later enzyme analysis. This homogenate was divided into two centrifuge tubes

and centrifuged at 550 g for 15 min to remove nuclei and cellular debris (Pd. The

supernatant fkom the &st spin CS,) was decanted into a clean centrifuge tube and then

centrifuged at 50000 g for 1 h producing a pellet with a light portion (plasma membranes)

and a dark portion (mitochondria), instead of at 100000 g for 45 min. The light portion

of the pellet was shaken loose with 15 ml (instead of 60 ml) of sucrose b a e r containing

(in mMJ 250 sucrose, 5 MgC12, 5 HEPES, 5 Tris (pH 7.4) and homogenized (Dounce

homogenizer, tight pestle, 100 strokes). This second homogenate was centrifuged at

1000 g for 10 min and then 10000 g for 10 min producing a pellet containing the

remaining contamînating membranes. The supernatant was decanted into a clean

centrifuge tube and centnfuged at 30000 g for 45 min (instead of 30 min) to produce a

linal pellet of enriched basolateral membranes. This final pellet was re-suspended in 0.5

ml of suspension medium containing (in mM) 5 MgC12, 150 NaCl, 20 HEPES, 20 Tris-

HCl (pH 7.4) and was used immediately for enzyme analysis, vesicle volume

deterrninations, urea transport assays, protein concentration determinations, cholesterol

assays, and phospholipid analysis.

Marker e n m e s

Enzyme assays (n=6 individual dogfish) were performed on the initial

homogenate and the final pellet of enriched BLMV to determine the relative purity of the

final preparation and the relative contamination of the &al preparation by other cellular

membranes. Using published methods, N~+,K+-ATP~s~ (McCormick 1993), glucose-6-

phosphatase (Stio et al. 1988). cytochrome C oxidase (Blier and Guderley 1988), and

NMN-adenylyltransferase (Ruggieri et al. 1990) were used as marker enzymes for the

basolateral membrane, endoplasmic reticulurn, iiiner mitochondrial membrane, and

nuclear membrane, respectively. AU measurements were made in duplicate at 25OC in a

temperature controil ed Perkin Elmer Lambda 2 spectrophotometer (Perkin-Elmer Corp.,

Norwallc, CT).

Vesicle resealing and orientation

The extent of resealing and the orientation of the basolateral membranes were

determined using a previously descnbed method (van Heeswijk et al. 1984). The vesicles

were assayed for N~',K+-ATP~s~ activity in the presence and absence of the detergent

digitonin (0.04%) to determine the percentage of reseaied vesicle. To unmask inside-out

(IO) oriented reseaied vesicles, N~+,K+-ATP~s~ activity was determined in the presence

of 0.5 mM ouabain and the difference in K+ stimulated (10 mM KCl) ATP hydrolysis in

untreated and digitonin treated vesicles was determined. Vesicles assayed in EX1

containing medium were pre-incubated with 10 rnM KCI for 20 min on ice. The

percentage of right side-out ( ' O ) vesicles was then calculated as the difference between

resealed and IO vesicles. (See Appendix 1 for more details.) Vesicle resealing was also

determined in BLMV preparations that were fiozen at -80°C, thawed and passed through

a 23-gauge needle and 1 mi syringe ten times to determine the viability of fiozen BLMV

preparations.

BLMV volume measurement

BLMV v o l ~ e s were determined using a previously described method @rand

1995). In brief, 40 pl of BLMV (-0.1 mg of protein) were incubated in 1 ml of medium

containing 10 pI(1 pCi) of 3 ~ z ~ and 10 pl (0.1 pCi) of 1 4 ~ - ~ ~ ~ - 4 0 0 0 for 2 min at 37°C.

The BLMV were sedimented by centrifugation at 12000g for 4 minutes. The supernatant

(500 pi) was then transferred to a scintiilation vial and 15 ml of Scintisafe Econo F

scintillation cocktail (Fisher Scientific, Fair Lawn, NJ) was added. The remainder of the

supernatant was discarded, the pellet was re-suspended in 40p1 of 20% Triton X-100

(v/v) and mixed by vigorous vortex mixing. The base of the centrifuge tube was then cut

off, placed in a scintillation vial with 15 ml of scintillation cocktail and the radioactivity

counted with a liquid scintillation spectrometer (Mode1 121 1 Rack Beta, LKB-Wallac).

Vesicle volume was calculated in pVmg protein as ( 3 ~ z 0 space - 1 4 ~ - ~ ~ ~ - 4 0 0 0

space)/mg protein added (See Appendix II for more details.)

Urea trans~ort assaw

Measurement of the transport of 14c-urea into the BLMV was performed in

duplicate at 10°C by a rapid fi1tration technique as previously described (Perry and Flik

1988). Freshiy prepared BLMV pellets were resuspended at a protein concentration of

0.5 m g h l in resuspension buffer containing (in ) 300 NaCl, 5.2 KC1, 2.7 MgSQ, 5

CaC12, 370 D-mannitol, and 15 Tris-HC1 (pH 7.4) and allowed to equilibrate for one hou

on ice. The BLMV were then collected by centrifugation at 30 000 g for 45 min and

again resuspended at a protein concentration of approximately 6 mg/ml. Thorough

mixing was achieved by passage through a 23-gauge needle (10 times). Transport

experiments were initiated by mixing 10-pl aliquots of BLMV with 40-ul aliquots of the

radioactive elasmobranch incubation medium ( E i î ) (containing 50 pCi/ml 14c-urea),

vortexed and incubated for 15 sec at 10°C. EIM contained (in mM) 300 NaCl, 5.2 KCl,

2.7 MgS04, 5 CaCl*, 370 urea, and 15 Tris-HC1 @H 7.4). Using D-mannitol as an

osmotic substitute, EIM solutions containing urea concentrations of 1 mM to 370 mM

were prepared. The detailed composition of the EIM used for each incubation is

described in the corresponding figure for each experirnent. Incubations were terminated

by the addition of 1 ml of ice-cold stop solution (EIM containing 370 m . urea). The

diluted mixtures were immediately filtered through pre-wetted Millipore Isopore filters

(Millipore, 0.4 p m HTTP type). Filters were washed with two 3 ml aliquots of ice-cold

stop solution and placed in a via1 with 15 ml of Scintisafe Econo F scintillation fluid. The

radioactivity al l each vial was then counted in a liquid scintillation spectrometer (Mode1

12 1 1 Rack Beta, LKB- Wallac or Beckman 6400). Controls for each experiment involved

diiuting 10 pl of BLMV with 1 ml of ice-cold stop solution first, then adding 40 pl of

radioactive mixture and immediately filtering this through pre-wetted Isopore filters.

These filters were then treated as above.

Experiments

Filter control test. Six types of filters were tested for autofluorescence and non-

specific binding of 14c-urea. These were 0.8 p glass fibre (Whatman), 0.45 p PVDF

Durapore (Millipore), 0.6 p. mixed cellulose (Millipore), 0.45 p nylon (Magna), and 0.4 p

polycarbonate Isopore (Mïllipore) filters. Filters were tested for autofluorescence by pre-

wetting them with non-radioactive EIM and then counting as described above. Non-

specific binding of 14c-urea was determined by adding an aliquot of radioactive mixture

to a tube, diluting with 1 ml of ice cold stop solution and then filtering this through each

filter type followed by two 3 ml rinses of ice cold stop solution. The flters were then

counted as described above.

Concentration dependence. Urea uptake was measured over a range of urea

concentrations (1 -3 70 mM). The EIM solutions containhg concentrations of urea less

than 370 mM contained appropriate concentrations of D-mannitol, which fünctions as an

osmotic replacement, thereby elirninating the eEects of osmotic differences. Urea uptake

was measured as described above using each individual EIM solution.

Inhibition assays. Cornpetitive and non-cornpetitive inhibition of urea transport was

examined to M e r define the properties of the transporter. The urea analogues,

acetamide, N-methylurea, thiourea and nitrophenylthiourea were tested. BLMV

were prepared in the same manner as above but individual EIM solutions were made up

by substituting each of these compounds at a concentration of 370 mM in place of urea

and used in place of the mannitol containing re-suspension buffer. The vesicles

containing the respective analogue were then incubated with a radioactive mixture

containing 2.5 mM urea and 3 pCi of 14c-urea and treated as descnbed above. Stock

solutions of the non-competitive inhibitors phloretin (0.03-0.16 mM in ethanol),

amiloride (0.1 mM), bumetanide (0.1 mM in ethanol), N-ethylmaleirnide (NEM) (0.1

mM in ethanol), p-chloromercuriphenylsulfonic acid (pCMPS) (0.1 mM), and

phenylmethylsulfonyl fluoride (PMSF) (0.1 mM in ethanol) were prepared in ethanol or

water at a concentration of 250 mM. These stock solutions were diluted to give the final

concentrations in parentheses above ( h a l ethanol concentrations were less than 0.1%).

Prior to the addition of the radioactive mixture, 10 ul of each inhibitor was added to the

BLMV and vortexed. Then urea uptake was measured as described above.

ATP dependence. ATP dependence of urea uptake was determined by measuring urea

uptake in 15 mM urea EIM containing ATP (10 mM), ATP (10 mM) and ouabain (1

mM), ATP (10 mM) and NEM (1 m m , plus a control. Urea uptake was measured as

described above.

Cation speczficily. Cationic specincity of urea transport was also examhed in the

BLMV, using momed re-suspension buffer and radioactive mixture containing only one

of the following salts NaCl or KCl. BLMV were prepared as descnbed above. The h a l

pellet was resuspended in medium containing (in mM): 15 D-mannitol, 2.7 MgSQ, 5.0

CaC12, 15 Tris-HCl, and 250 NaCl or 25 KCl and 225 N-methyl-D-glucosamine

(NMDG). The incubation medium contained 1 5 mM urea in place of mannitol and 50

pCi/ml 14c-urea. Control incubations used the same medium in which the final pellet

was resuspended. Gradient incubations used (in mM) 25 NaCl and 225 NMDG or 250

KCl for sodium and potassium gradients, respectively.

Protein, phospholipid, and cholesterol rneasurements

The protein concentrations of the BLMV preparations fiom dog£ish, arctic char and

winter flounder were determined by the method of Bradford (1976) using a Bio-Rad kit

(Richmond, CA) with bovine semm albumin as the standard. The basolateral plasma

membrane phospholipids were extracted, separated, and analyzed as previously described

(Bligh and Dyer 1959, Ballantyne et al. 1993). Cholesterol levels of dogfish, arctic char,

and winter flounder BLMV were measured using a commercially available enzymatic

end-point assay kit (Sigma, St. Louis, MO).

Chemicals

Digitonin was obtained fiom The British Drug Houses (Canada) Ltd. (Toronto,

Canada). N-methylurea was obtained fiom Fluka through Sigma-Aldrich Chemicals.

Semicarbazide hydrochioride was obtained fkom Aldrich (Milwaukee, WI). 14c-

polyethylene glycol-4000 was obtained fiom Amersham Life Science (Baie d'Urfé,

Quebec). 3 ~ 2 ~ was obtained fiom Dupont-New England Nuclear (WiImington, DE). 14c-

urea was obtained fiom ICN (Montreal, Quebec). Al1 other chemicals were obtained

fiom either Fisher Scientific (Whitby, ON) or Sigma Chernical (St. Louis, MO) and were

of reagent grade.

Statistical methods

Values are expressed as mean + SE. Statistical cornparisons were made either by

Student's t-test or ANOVA with secondary Student's t-test and considered statistically

significant if pCO.05.

RESULTS

Marker Enzymes

The measurement of the four marker enzymes demonstrated that the basolateral

membrane preparation was highly purifïed and only slightly contaminated by other

membranes (Table 1.1 and Table 1.2). The marker enzyme for basolateral membrane

was enriched 4.12-fold. The specific activities of glucose-6-phosphatase (endoplasmic

reticulum) and cytochrome C oxidase (mitochondria) in the initial homogenate were 0.67

+ 0.25 and 5.72 + 1.98, and 0.07 + 0.07 and 0.17 f 0.06 in the f i a l BLMV preparation,

respectively (Table 1.1). This corresponds to a reduction in endoplasmic reticulum and

inner mitochondrial membrane of 9.49-fold and 34.25-fold, respectively. The nuclear

membrane was siightly enriched in the BLMV preparation with an increase in specific

activity of 1.1 9-fold (Table 1.1). However, the total ac tivity of NMN-adenyl yltransferase

(nuclear membrane) was only 0.55 umol substrate/ h which corresponds to 1.15%

recovery of the initial amount of enzyme and 2.96% contamination (percentage of the

total N~+,K+-ATP~s~ activity (Table 1.2). Endoplasmic reticulum and inner

mitochondrial membrane had recoveries of 0.15% and 0.05% of the initial amount of

enzyme and contributed 1.55% and 4.25% contamination (Table 1.2). Final recovery of

basolateral membrane was 6.15% (Table 1.2), which is similar to other studies (Perry and

F E 1988).

Vesicle Resealing, Orientation. and Volume

The passage of the basolateral membrane preparation through a 23-gauge needle

resulted in the formation of resealed vesicles (Table 1.3), approximately 35% of the

Table 1.1. Marker enzyme specific activities and magnitude of purification of each

membrane

N~+,K+- Glucose-6- Cytochrome C NMN-

ATPase phosphatase Oxidase Adenylyltransferase

umol substrate/h/mgprotein

Homogenate 1.02 k 0.25 0.674 f 0.25 5.72 t 1.98 0.11 kO.10

BLMV 4.20 f 1.71 0.071 + 0.07 0.17 t 0.06 0.13 + 0.10 Magnitude of Purz3cation

+4.12-fold -9.49-fold -33 -65-f01d +1. 194-fold

Values presented as mean t SE (~6).

Table 1.2. Total activity of marker enzymes, percent recovery, and percent contamination

in the final basolateral membrane vesicle preparation

Na-K-ATPase Glucose-6- Cytochrome C NMN-

phosphatase Oxidase Adenyl yltransferase

umol substruteh

Homogenate 302.67 k 35.87 197.39 k 5 1.52 1700.59 -t 481 -92 47.58 + 43.57

BLMV 18.61 t 8.16 0.29 I0 .23 0.79 4 0.47 0.55 + 0.46 Percent Recovery

6.1 5% 0.15% 0.05% 1.16%

Percent Contamination

- 1.56% 4.25% 2.96%

Values presented as mean + SE ( ~ 6 ) .

Table 1.3. Percentage of the basolateral membrane fiaction as resealed vesicles and

vesicle orientation (see text for details).

Percentage of basolateral membrane fiaction

Fresh 65.00 _+ 7.49% 9.41 i 232% 25.59 k 9.83%

Freezefïhaw 87.92 + 3 -92% 6.97 k 3-03% 5-11 -ir 0.89%

Values presented as mean k SE ( ~ 6 ) .

membrane population. This group of resealed vesicles is composed of 9.41% inside-out

and 25.59% right side-out orientation (Table 1.3). Freezing and thawing of the final

BLMV preparation resulted in a decrease in the percentage of resealed vesicles of both

orientations (Table 1.3), despite the passage of the membranes through a 23-gauge needle

&er thawing. The volume of the BLMV was relatively large, 8.60 f 1.78 p l h g protein,

providing additional confirmation of vesicle resealing.

Filter Control Test

None of the nIters tested exhibited any appreciable autofluorescence in the

scintillation cocktail used (Fig. 1.1 ; Blanks). The optimal filter type for use with I4c-urea

was found to be the polycarbonate Isopore filter from Mïllipore (Fig. 1.1). This filter

produced consistently low background counts due to non-specific binding of 14c-urea to

the filter. This accounted for approximately 10% of the total counts in the urea uptake

assays, similar to previous studies (Hopfer et al. 1973).

Concentration Dependence of Urea Uvtake bv BLMV

Urea uptake by BLMV was measured over a range of urea concentrations in the

incubation medium revealing two components of uptake (Fig. 1.2a). At hi&

concentrations (15 - 370 mM) urea uptake is linearily dependent upon the urea

concentration (Fig. 1.2a). However, at low urea concentrations (1 - 15 mM) urea uptake

exhibits saturation-like kinetics (Fig. 1.2b). When these data are transformed using a

Lineweaver-Burke plot, it is revealed that urea uptake at urea concentrations of 1 - 15

mM has a Km of 10.07 mM and V,, of 0.338 poVhr/mg protein (Fig. 1.2~).

Figure 1.1 Filter optimization for autofluorescence (blanks) and non-specific binding of

14c-urea (controls) for use in urea transport rapid filtration experiments. The blanks

represent fïiters only in scintillation cocktail. The controls involve adding radioactive

mixture to pre-diluted vesicles followed by immediate filtration through the filter. (Mean

k SE, n=2).

Figure 1.2 a. Rates of urea uptake at variable urea concentrations, by BLMV fiom the

giii o f the dogfïsh, Squaluî acanthias. (Mean k SE, ~ 8 ) . b. Expansion of the low end of

the [urea] range fiom A. The regression is y = 0.077Ln(x) + 0.0 12,8 = 0 .go6 (Mean + SE, n=7) c. Lineweaver-Burke transformation of the effect of variable [urea] on uptake

rate by B L W . The regression is y=2.96 + 29.8 lx, 1?=0.978 (Mean f SE, ~ 7 ) .

O 50 100 150 200 250 3 0 0 3 5 0 4 0 0

Urea concentration (rriM)

Inhibition of Urea Uptake by BLMV

Urea uptake by BLMV demonstrated sensitivity to the non-competitive inhibitor

phloretin (Fig. 1.3). Phloretin produced a dose-dependent inhibition of urea uptake by

BLMV, with 50 % inhibition occurring at a concentration of 0.08 mM. The use of urea

analogs demonstrated competitive inhibition of urea uptake by BLMV. N-methylurea

and NPTU significantly reduced the rate of urea uptake (pc0.05; Fig. 1.4). Thiourea and

acetamide also tended to reduce the rate of urea uptake although this decrease was not

significantly different fiom the control. The sulfhydryl reagents pCMBS, PMSF, and

NEM did not show any inhibition of urea uptake by BLMV (data not shown).

ATP Deuendence and Inhibition of Urea Uptake by BLMV

Urea uptake was signincantly stimulated by the addition of ATP to the incubation

medium @<0.05; Fig. 1 S). However, upon the addition of ouabain the rate of urea

uptake by BLMV decreased to control levels (Fig. 1.5). The addition of NEM had no

effect on the ATP stimulated urea uptake (Fig. 1.5).

Cation De~endence of Urea Uptake by BLMV

There was no significant difference between urea uptake in medium containing

ody sodium or only potassium ions with no concentration gradient present (Fig. 1.6).

When urea uptake was measured in the presence of a potassium concentration gradient

there was also no significant change in the rate of urea uptake by BLMV. However,

when a sodium concentration gradient was present during urea uptake measurernents, the

rate of urea uptake significantly increased (pC0.0 1 ; Fig. 1.6).

Figure 1.3 Dose dependent phioretin inhibition of urea uptake in BLMV fiom the gill of

the dogtïsh, SquuZus acanthias = 0.08 mM) @dean f SE, n=5).

Phloretin concentration (mM)

Figure 1.4 Inhibition of urea uptake in BLMV by the urea analogues, acetamide, thiourea,

N-methylurea, and NPTU (3 70 mM). (Mean t S E ., n=3)

Figure 1.5 ATP (1 0 mM) stimulation of urea uptake and effects of ouabain (1 mM) and

NEM (1 mM) in BLMV fkom the giII of the dogfish, Squalus ucunthias (Mean * SE, n=8). * significant dserence £iom contrd (paireci t-test, pc0.05)

Figure 1.6 Rate of urea uptake in BLMV fTom the giII of the dogfish, SquuZus acanthim,

in the presence of physiologicdy oriented sodium (225 mM outwardly directed) or

potassium gradients (225 mM inwardly directed) (Mean f SE, n=6). * significant

difference fiom control (paired t-test, p<0.005) ** significant clifference between sodium

and potassium @aired t-test, p<0.005)

j Control i / I Gradient I

Phospholipid composition and choIesterol content

In the BLMV of SquaIus acanthias the dominant phospholipid was PC,

representing 45.6% of the total membrane phospholipids, while PE and PS made up 26.5

and lî.3%, respectively, of the membrane composition (Table 1 -4). Sphingomyelin and

phosphatidylinositol were minor components of the membrane phospholipids. The

mitochondrial phospholipid cardiolipin was present in a smaii amouut, 2.6% of total

membrane phospholipid due to the minor mitochondrial contamination indicated above.

The cholesterol content of BLMV was reiatively high (Table 1.4) resdting in a C: P ratio

of 3.68 & 0.26 (n=8; Table 1.5). The dogfish also had the highest cholesterol to protein

ratio when compared to seawater and fieshwater teleosts (Table 1 A).

Table 1.4. Percentage of phospholipid types and total phospholipid and cholesterol

in the basolateral membrane of gill epithelium fiom the spiny dogfish, SquaZus

Cardiolipin

Phosphatidylcholine

Phosphatidylethanolarnine

Phosphatidylinositol

Phosphatidylserine

Sphingomyelin

PCIPE

TotaI Phospholipid

(nmoVmg protein)

TotaI Cholesterol

( n m o h g protein)

Cholesterol: phospholipid

(moVm01)

Values presented as means & S.E. (n=8). PC, phosphatidylchohe; PE, phosphatidylethanolarnine

Table 1.5 Cholesterol:phospholipid ratios of representative species fkom difEerent

C:P Species Tissue SubceiIular Fraction (mol mol-') Reference

Spiny Dogfish Gill Basolateral Membrane 3.68 i 0.26 Present

Rat (Ratttcs nomegrgreus)

Little Skate (Raja erinacea) IRaUibow Trout (Oncorhynch mykiss)

Squid (Loligo pealei)

Crab (Cancer pagurrrs)

Crab (Carcinus maenus)

Cucumber (Cucu~bitafic@iolia) Fungus (Bohyfis cinerea)

Bacteria (Methylococcus

Epithelium Lens

Intestine

Liver

Intestine

Cerebral and Optic Lobes Steilate Ganglia Leg muscle

Leg muscle

Root

Hypha

Plasma Membrane

Bnish Border Membrane Basolateral Membrane

Basolateral Membrane

Brush Border Membrane

Basolaterai Membrane

Plasma Membrane

Plasma Membrane

Plasma Membrane

Plasma Membrane

Plasma Membrane

Plasma Membrane

Plasma Membrane

mdy B orchman et al. 1989 Molitoris et

al. 1985 Molitoris et al, 1985

Smith and Ploch 199 1

Crockett and Hazel

1995 Crockett

and Hazel 1995

Yamaguchi et al. 1987 Yamaguchi et al, 1987 Cuculescu et al. 1995 Cuculescu et ai. 1995 Bulder et ai. 1991

Kodali et al. 1998 Jahnke

1992 caps ulatus)

9otal stero1:phospholipid ratio

Table 1.6. Cornparison of cholesterol to protein ratios in the basolateral membrane of the

giii epithelium fiom the spiny dogfish and a marine and fieshater teleost.

Fresh Water Arctic Salt Water Arctic Winter Spiny Dogfish

Char ( ~ 8 ) Char (n=8) Flounder (n=6) (n=8)

Mean 0.070 0.050a 0.064~ 0.1 leacd S.E. 0.004 0.004 0.005 0.008

'Signincantly difEerent fkom Fresh Water Arctic Char @<0.005) b~ignificantly different fiom Salt Water Arctic Char @<O.OS) CSignincantly dBerent fiom Salt Water Arctic Char @<0.000005) * ~ i ~ n i n c a n t l ~ different nom Winter Flounder (pc0.0005)

DISCUSSION

Methodolow

The method (Perry and Flik 1988) used to prepare basolateral plasma membranes

Çom the gill epithelium of the spiny dogfish (Squulus acanthias) yielded a specific

enrichment of ~ a f , K'-ATP~s~ indicating selective isolation of basolateral plasma

membranes. Although there was only minor contamination (6%) of membranes fiom

the endoplasrnic reticulum, mitochondria and nucleus, this may have led to a slight

underestimation of urea uptake by the gill BLMV. The final recovery of ~ a * , lC'-~TPase

activity (6.2%) is consistent with previous studies (Perry and Flik 1988), while the

vesicle orientation (IO=9.4%; RO=25.6%) and resealing efficiency (35%) were

somewhat lower than reported values for eel (IO=33%; RO=23%; resealing efficiency

56%) (Flik et al. 1985). The volume measurements c o b e d that the vesicles

successfblly resealed. The decrease in resealing efficiency afier fieezing and thawing of

the final BLMV preparation suggested decreased integrity of the membranes and

therefore onl y fieshly prepared BLMV were used for transport experirnents. Significant

t h e was saved by omitting the g d perfüsion step, used to clear the gills of red blood

cells (Perry and Flik 1988), as there are no urea transporters present in elasmobranch

erythrocyte plasma membranes (Carlson and Goldstein 1997).

Urea Transport

The measurement of urea uptake by enriched BLMV revealed saturation kinetics at

low urea concentrations (KmlO.1 mM, Vm,=0.34 pmoVh/mg protein), suggesting the

presence of carrier-mediated urea transport. The low Km, relative to the urea

concentration in the blood, indicates that the transporter has a relatively high affinity for

urea. This implies that the putative urea transporter acts to "scavenge" intracelldar urea,

actively returning it to the blood and thereby maintaining a low urea concentration within

the gill epithelial cells. The effects of several known inhibitors of urea transport were

examined to M e r characterize the saturable component of urea uptake by the BLMV.

Inhibition by the non-competitive inhibitor phloretin is diagnostic of both facilitated and

secondary active, urea transport systems Wato and Sands 1998, Levine et al. 1973, Smith

and Wright 1999, Walsh et al. 1994). The dose-dependent inhibition of urea uptake in

shark gill BLMV by phloretin in this study is consistent with a previous study on the

isolated perfused dogfish head preparation (Part et al. 1998), which demonstrated that

phloretin &ion signifïcantly increased urea efflux across the gill. The urea analogues

N-methylurea and NPTU also signincantly inhibited urea uptake in shark gill BLMV.

Acetamide and thiourea had a slight, but non-signincant effect on urea uptake rates.

However, these results are inconclusive u t i l the analogue concentrations are optimized,

since studies have shown that under dif5erent conditions urea analogues may or may not

have statistically signincant effects on the branchial urea efflwc in the spiny dogfkh (I?W

et al. 1998, Wood et al. 1995). In mammalian inner medullary collecting duct,

methylurea and thiourea signi6cantly reduced urea permeability, while acetamide did not

(Chou and Knepper 1989). These results demonstrate that urea transporters in difZerent

tissues and animals exhibit dif3erent sensitivities to urea analogues. The inhibition of

urea uptake in shark gill BLMV by phloretin and urea analogues supports previous

hypotheses (Smith and Wright 1999, Wood et al. 1995) for a carrier-mediated urea

transport system in the dogfish shark gill.

Urea uptake is energy dependent in shark gill BLMV. The addition of adenosine

triphosphate (ATP) to the incubation medium sienif?cantly increased the rate of urea

uptake, whereas ouabain, a specific inhibitor of the enzyme N~~,K+-ATP~s~, returned the

rate of ATP-dependent urea uptake to control levels. In the presence of NEM, an

alkylating agent that binds selectively to sulfhydryl groups blocking V-type and P-type

ATPases (e-g. proton pumps) (Ehrenfeld 1998), ATP-dependent urea uptake was

unchanged. This ouabain sensitivity, but NEM insensitivity, of ATP-dependent urea

uptake in shark gill BLMV signifies that urea uptake was indirectly ATP dependent,

coupled to either the sodium or potassium gradients created by N~+,K+-ATP~s~. In the

presence of an outwardly directed sodium gradient, urea uptake was signincantly higher

relative to both the control and potassium gradient experiment. Taken together, these

data confirm that the saturable component of urea uptake in the shark gdi BLMV is due

to a ~ a + - c o u ~ l e d secondary active urea transporter, similar to the type found in the

mammalian kidney (Kato and Sands 1 998).

The dog£ïsh gill urea transporter functions in an antiport fashion (Fig. 1.7), similar

to the transporter described in the rat inner medullary collecting duct, but differs from the

putative sodium-linked urea transporter described fiom the dogfkh kidney (Schmidt-

Nielsen et al. 1972), which is thought to k c t i o n in a symport fashion. By using the

inwardly directed concentration gradient of ~ a + , the dogfkh gill urea transporter actively

pumps urea back into the blood fiom within the cell. This would decrease the

intracellula. urea concentration of the gill epithelial cells, reducing the concentration

gradient for urea across the apical surface of the cells and thus the rate of urea diffusion

or loss across the gill. Depending on the stoichiometry of the exchange, the shark can

Fig. 1.7 Schematic representation of the proposed Na+-coupled, active urea transporter

present in the basolateral membrane of the dogEish gill epithelium. 1 propose that the

active retum of urea to the blood reduces the intracellular urea concentration thereby

reducing the rate of urea diffusion across the apical membrane to the seawater.

(Thickness of the dashed arrows indicate the relative rate of urea diffusion.)

Seawater O mM Urea

Blood 370 mM Urea

Urea

1 I I I

Urea

7

Apical

Basolateral

Urea ~ a + K'

Save up to 5 ATP equivdents (the metabolic cost of synthesizing one urea molecule via

the ornithine-urea cycle including glutamine synthetase) for each irrea molecule returned

to the blood. In the dogfrsh kidney, sodium and urea are reabsorbed at a Itixed ratio of 1.6

moles of urea per mole of ~ a + (Schmidt-Nielsen et ai. 1972). The energetic cost of

transporthg ~ a + via ~a+,lK+-~TPase is 1 ATP for every 3 moles- If a similar ratio of

exchange for sodium and urea were involved in the gill as in the kidney, savings of 4.8

ATP equivdents per molecule of urea retumed to the blood would be achieved. This

means that the metabolic cost associated with urea transport in h e gill is low, but the

metabolic savings are significant.

Gill basolateral membrane composition

In the basolateral membrane of the gill f?om the spiny dogfîsh, SquaZus acanthias,

PC and PE were the main phospholipids, typical of most eukaryotic membranes. The

resulting PC/PE ratio was 1.72. PC stabilizes the membrane as it favors the formation of

a laminar bilayer, while PE destabilizes the membrane by keeping it close to the phase

transition between laminar and hexagonal (HII) phase conformaQions (Thurmond and

Luidblom 1997). In most temperate species the ratio of PC to PE is approxirnately 1.0.

This ratio can be altered according to the enivronmental and physiologica1 conditions

encountered by the organism. For example, cold-acclimated organisms (Arctic char,

molluscs) increase the proportion of PE in their membranes iin order to maintain

membrane fluidity at low temperature (Hazel 1995). This results im a PCPE ratio of less

than one (0.3-0.5). However, in elasmobranchs the cellular membranes face the opposite

problern, increased fluidity due to urea (Barton et al. 1999). Orne adaptation that has

evolved in elasmobranchs to deal with this effect of urea is the presence of

trimethylamine oxide (TMAO). TMAO counteracts the negative effects of urea on both

proteins (Yancey and Somero 1980) and phospholipid membranes (Barton et al. 1999).

Another strategy, which may work in conjunction with TMAO, is increasing the ratio of

PC to PE, which would provide additional stability to the membrane.

The very high cholesterol to phospholipid molar ratio (3.68) reported here is the

highest for native membranes (Table 1.5). Cholesterol decreases the permeability of

biological membranes to urea (Pu& et al. 1989) by inducing an increased order of the

phospholipid molecules that compose the bilayer membrane, allowing them to pack

closer together forming a tighter barrier (Mouritsen and Jmgensen 1994). Cholesterol,

when inserted in the appropriate place in the membrane, also increases membrane

permeability to oxygen (Dumas et al. 1997), therefore the elevated cholesterol levels in

shark gill basolateral membranes would not impair gas exchange. We propose that the

high cholesterol content of shark gill basolateral membrane provide a physical barrier that

retards passive loss of urea at the gill without affecting oxygen permeability (Fig. 1.8).

This hi& C:P molar ratio may also explain the low permeability of the shark gill to water

(Boylan 1967, Part et al. 1998) and sodium (Boylan 1967) relative to teleost fishes. Our

hdings also point to the possibility of similar structural modifications occurring in the

mammalian kidney tubule where urea permeability varies dong the nephron (Sands et al.

1997).

In Perspective

Based on the results of this study we propose that a unique combination of

physiological and structural mechanisms is at least in part responsible for the low urea

permeabiLity of the dofish shark gill. The marine elasmobranch gill is approximately 80

h e s less permeable to urea than the teleost gill resulting in a urea efflux of 270

poVkg/hr (Part et al. 1998). If the elasmobranch gill were as permeable to urea as the

rainbow trout gill, the resulting urea efflux would be immense (10,000 pol/kg/hr) (Part

et al. 1998), because of the enormous blood-to-water gradient. Ushg data from the

present study, it is possible to calculate the relative contribution of active urea transport

to the ciifference between observed and predicted (based on rainbow trout gill urea

permeability) urea efflux rates. The maximal velocity of urea uptake (Vrn~0.34

jmol/h/mg protein) was corrected for vesicle resealing (35%), vesicle orientation (27%

of resealed vesicles), membrane recovery (6.15%), protein level (9 mg/animal), and

animal mass (1.3kg). This results in a total rate of active transport of urea, back into the

blood fiom within the gill epithelium, of 535 ~ o V k g / h . This value is approximately 6%

of the difference in urea permeability between the rainbow bout and elasmobranch gill

(i.e. 10 000 - 270 p o v k g h ) . The remaining 94% may be in part, or in whole, due to

the elevated C:P molar ratio in the basolateral membrane. Thus we envision that the

primary role of the basolateral membrane is to substantially reduce the influx of urea into

the gill epithelial cells, thereby maintaining low intracellular urea concentrations at which

the urea transport system hct ions . The urea transport system actively transports urea

out of the epitheiial cells back into the blood maintaining low intracellular urea

concentrations. The low intracellular urea concentrations achieved by these

conplimentary mechanisms, leads to a reduced diffhion gradient for urea across the

apical membrane and thus a lower effective permeability of the gill to urea. One could

argue, therefore, that high cholesterol levels and active urea transport in the gill

basolateral membranes probably CO-evolved in elasmobranchs enabling the retention of

m a and its use as a key component of their osmoregulatory strategy, while minimising

the energetic cost. However, the hi& non-specific urea uptake by the BLMV suggests

that there are other mechanisms andor structures, particularly the composition of the

apical membrane, which may contribute signifïcantly to the overall low urea permeability

of the elasmobranch gill. Further studies of the elasmobranch gill are thus required to

completely resolve this issue.

Fig. 1.8 A mode1 for the mechanism of reduced urea permeabiiity of the dogfish gill

basolateral membrane due to cholesterol. The cholesterol (dark) contributes to a more

tightly packed phospholipid bilayer membrane (light) thereby physically reducing the

pemeability of the basolateral membrane to urea.

UREA

UREA

CHAPTERZ: A comparison of phospholipid and cholesterol composition of the

basolateral membranes from the gill epithelium of an elasmobranch, the dogfïsh

(SquaZus acan fhias)

INTRODUCTION

Elasmobranchs are unique among marine fish (except for the coelacanth) in that

they retain high leveb of urea (400-600 mM) io their tissues and body fluids for the

purpose of osmoregulation (Smith 193 6). Although urea retention provides an efficient

mechanism for osmutic balance, the chaotropic properties of urea pose another problem,

protein and phospholipid bilayer destabilization. The presence of trirnethylamine oxide

(TMAO) in the blood and tissues counteracts the destabilizing properties of urea on

protein (Yancey and Somero 1980) and membrane integrity (Barton et ai. 1999).

Modification of the phospholipid and fatty acid composition of cellular membranes is

also invoived in maintainhg membrane integrity of fishes under a variety of

environmental conditions (i.e. temperature and salinity) (Hazel and Williams 1990).

Phosphatidylcholine (PC) with a high content of saturated fatty acids (SFA) stabilizes

phospholipid bilayers, while phosphatidylethanolamine (PE) containing more unsaturated

fatty acids, destabilizes phospholipid bilayers in fish (Hazel and Landrey 1988a,b).

Elasmobranch mitochondrial membranes have signincantly higher levels of saturated

fatty acids and signincantly lower levels of polyunsahirated fatty acids (PUFA) relative

to other marine fishes such as the hagfïsh and winter flounder (Glemet and Ballantyne

1996). It has been suggested that these characteristics are adaptations to maintain

mitochondrial membrane i n t e e in the presence of hi& urea concentrations in the

tissues (Glemet and Ballantyne 1996), but other elasmobranch membranes have not been

examined.

The elasmobranch gill, like the gill of al1 fish, is involved in gas exchange,

ionoregulation, and osmoregulation. In addition to these vital fùnctions the elasmobranch

gill must also act as a barrier to urea loss. Compared to the teleost gill, the elasmobranch

gill is approximately 80 times less permeable to urea than the teleost gül (Part et al.

1998). The basolateral plasma membrane of the elasmobranch gli epithelium is in

contact with high levels of urea in blood on the extraceliular side and purportedly low

levels of urea on the intracellular side and may be the main permeabiiity barrier (Wood et

ai. 1995). In chapter one, we identified a ~a+-coupled active urea transport mechaaism

(K,=10 mM) in the basolateral membrane (BLM) of S. acanthiar gills that would act to

scavenge intracehlar urea and retum it to the blood, against the concentration gradient.

In addition, we reported that these same membranes had a cholesterol (C) to phospholipid

(P) molar ratio of 3.68, an extremely hi& value compared to other tissues and organisms

(Table 1.5). A high C:P molar ratio is characteristic of membranes with reduced

permeability to low molecular weight solutes, such as urea (Lande et al. 1995). Taken

together, the urea transport system and hi& C:P molar ratio would facilitate urea

retention in dogfïsh gills. It is also possible that the phospholipid and fatty acid

composition may contribute to the low permeability of the dogfkh gill to urea. The lipid

composition of various epitheiial tissues (Le. mammalian epidermis and trout intestine,

skin, and opercular membrane) has been demonstrated to innuence ~ a + and water

permeability @i Costanzo et al. 1983, Ziboh and Miller 1990, and Ghioni et al. 1997).

The purpose of this study was to characterize the phospholipid and fatty acid

composition of the basolateral plasma membrane fiom the dogfïsh gill epithelium. We

hypothesize that the phospholipid and fatty acid composition of this membrane, coupled

with the high C:P molar ratio, is in part responsible for maintaining membrane integrity

and contributhg to the extremely low permeability of the elasmobranch gill to urea.

METHODS AND MATERIALS

Ex~erimental animals

Spiny dogfish (Squalur acanthiar) were coliected by Otter Trawl in

Passamaquody Bay, New Brunswick fiom mid-July to the end of August 1999 and

maintained at the Huntsman Marine Science Centre in outdoor tanks under natural

photoperiod, and supplied with filtered seawater (lO°C). The dogfish did not feed in

captivity and were heId for a maximum of 10 days prior to experimentation.

Basolateral membrane @LM) preparation

BLM was prepared £kom epithelial tissue scraped fiom the gill arches. Tissue was

homogenized using a Dounce homogenizer in a hypotonie homogenization b a e r

containing (in mM): 25 NaCl, 1 dithiothreitol, 0.5 disodium ethylenedinitrilotetra-acetic

acid (EDTA), 1 N-2-hydroxyethylpiperazine-N'-2-ethanesulc acid (HEPES), 1

Tris [hydroxymethyl] aminomethane hydrochloride (Tris-HCl), pH 8 .O, plus 1 00U/ml

aprotinin. This homogenate was centrifuged at 550 g for 15 min and the supematant

decanted and centnfuged at 50000 g for 1 h. The light portion of the resultuig pellet

(plasma membranes) was homogenized again and centrifûged at 1000 g for 10 min,

10000 g for 10 min and the rernainîng supematant was then centrifuged at 30000 g for 45

min producing a h a 1 pellet of e ~ c h e d basolateral membranes. The BLM were then

fiozen and stored at -80°C for further analysis.

Characterization of basolateral membrane purity

Marker enzymes were measured in both the initial homogenate and in the BLM

preparation. The following were used as markers for cellular membranes: basolateral

membrane - N~+,K+-ATP~s~ (McCormick 1990), endoplasmic reticulum - glucose-6-

phosphatase (Stio et al. 1988), mitochondnal membrane - cytochrome C oxidase @lier

and Guderley 1988), and nuclear membrane - NMN-adenylyltrmsferase (Ruggieri et al.

1990). Al1 measurements were made in duplicate at 25OC.

Analvsis of basolateral membrane phospholipid composition

The BLM phospholipids were extracted, and separated as described by Bligh and

Dyer (1959). Phospholipid separation using thin layer chromatography and fatty acid

analysis using gas chromatography were as described previously (Baliantyne et al. 1993).

Protein measurements

The protein concentrations of the BLM preparations were determined by the

method of Bradford (1976) using a Bio-Rad kit (Richmond, CA) with bovine serum

albumin as the standard.

Chemicals

The lipid standard used (Nu-Chek-Prep., Inc. Elysian, MN) was augmented by the

addition of 15:0, 15:1, and 17:l methylated fatty acids. Solvents were obtained fiom

Fisher Scientific Ltd. (Whitby, Ontario) and were of American Chemical Society-

certified grade. Ail other chemicals were obtained fiom Sigma Chemical Co. (St. Louis,

MO) and were of the highest purity available.

Characterization of basolateral membrane puritv

The BLM was substantially enriched in the basolateral membrane marker N~+,K+-

ATPase, in relation to the marker enzymes for endoplasmic reticulum (glucose-6-

phosphatase), mitochondrial membrane (cytochrome c oxidase), and nuclear membrane

(NMN-adenylyltransferase) (see Table 1.1). This indicated little contamination of the

basolateral membrane with these cellular membranes.

Phos~holipid composition

In the BLM, the dominant phospholipid was PC, representing 46% of the total

membrane phospholipids, while PE and PS made up 27% and 12%, respectively, of the

membrane composition (see Table 1.4). Sphingomyelui and phosphatidylinositol were

minor components of the membrane phospholipids. The phospholipid cardiolipin was

present in a small amount, 2.6% of total membrane phospholipid (see Table 1.4), similar

to the level of mitochondrial contamination indicated by marker enzyme measurements

(see Table 1.2).

Gill basolateral membrane fattv acid composition

The predorninant fatty acids in the cumulative fatty acid composition of

membrane phospholipids were l6:O, 18:0, 18: 1, 20:4n-6, 20:5n-3, 22:6n-3 (Table 2. l),

which accounts for 82% of the cumulative fatty acid composition. Saturated fatty acids

were the prevalent class of fatty acids (38 %) (Table 2.1).

In the BLM, PUFA was the largest class of fatty acids in the phospholipid PE

(Table 2.2). The main individual fatty acids were 20:4n-6, 20511-3, and 22:6n-3 totaling

39% (Table 2.2). PE had the highest levels of an unknown fatty acid (Table 2.2). This

unknown fatty acid was tentatively identïfïed as 16:3 or 16:4 based upon the retention

time relative to stearic acid (Ackman 1962). In PI, SFA and PUFA were the prevalent

classes of fatty acids (Table 2.2). This was due to the large proportions of 18:O and

20:4n-6 (Table 2.2). In PS, SFA and PUFA were the most abundant classes of fatty acids

(Table 2.2) due primarily to 18:0, 20:4n-6 and 22:6n-3 although the monoene 18: 1 was

also prevalent (Table 2.2). In PC, SFA was the dominant fatty acid class (Table 2.2) as a

result of the levels of 16:O (Table 2.2). The monoene 18: 1 was also prevalent in PC

equalling (Table 2.2). Monoenes comprïsed the largest class of fatty acids in SM (Table

2.2). This was due to the levels of 14:l and 24:l (Table 2.2). The saturated fatty acids

14:O and 16:O were also prominent in SM (Table 2.2).

Table 2.1 Cumulative percentages of individual fatty acids in gin basolateral plasma

membrane fiom Squalus acanthiasa

Fatty acid (mol %)

24: 1 1.59 10.19 Total SFA 38.11 I 1.70 Total Monoene 28.02 10.72 Total PUFA 34.88 I 1.61 Total n-3 PUFA 17.86 10.97 Total n-6 PUFA 17.02 k0.80 n-3/n-6 PUFA 1 .O5 I 0.05 Unsaturation indexC 186.48 I 56.58 Mean Chain lengt.hd 18.66 & 0.40

Values presented as mean f S.E.M. (n=8). bvnknown A (9.08 min) eluted between 17: 1 (8.76 min) and l8:O (9.32 min). 'Unsaturation index = Çrnim; where mi is the mole percentage and ni is the number of C-C double bonds in fatty acidi. %lean chain length = Xfci; where 6 is the mole hction and Ci is the number of carbon atoms in fatty acidi.

Table 2.2 Percentages of individual fatty acids in the major phospholipids fiom the gill

basolateral plasma membrane of Squalus ucanthiasa

Fatty Acid Phosphatidyl- Phosphatidyl- Phophatidyl- Phosphatidyl- Sphingomyelin choline ethanolamine inositol serine

1.81 f 0.13 0.40 f 0.11 0.08 I 0.01 0.05 I 0.01 10.68 & 1.63

0.45 I 0-06 0.05 i 0.04 ND 0.50 I 0.25 28.99 rt 2.13 Total SFA 45.91 f 1.20 20.54 I 2.02 45.42 i: 4.87 45.06 I 6.62 38.39 2.34 Total monoenes 29.50 f 0.53 33.28 + 1.41 4.79 I 0.61 18.72 f 2.97 53.78 i: 1.91 Total PUFA 25.16 i: 1-57 47.90 I 2.70 51.15 I 4.75 37.64 k 4.27 8.78 I 1.81 Totaln-3 PUFA 14.48 I 0.87 23.78 I 1.55 14.37 i 1.55 21.27 f 2.65 4.97 i 0.92 Total n-6 PUFA 10.68 f 0.81 24.11 f 1.41 36.78 f 3.98 16.37 k 1.79 3.81 I 1.05 n-3/nd PUFA 1.38 & 0.07 0.99 i- 0.05 0.42 i: 0.05 1.30 1: 0.09 1.56 * 0.27 Unsaturation 144.18*6.38 253.67k11.84 219.31 118.23 193.43f23.09 91.29i8.01 index Mean Chain 17.91 i 0.05 19.33 i 0.11 19.34 I 0.14 19.24 I 0.16 19.27 1: 0.34

values are presented as means f S.E.M. (n=8). ND not detectable. bvnknown A (9.08 min) eluted between 17: 1 (8.76 min) and l8:O (9.32 min).

DISCUSSION

Phospholipid Composition

Zn the basolateral membrane of the dogfkh gili epithelium, PC and PE were the

m a i . phospholipids, typical of vertebrate membranes (Hazel and Williams 1990). These

phospholipids play an important role in regulating the fluidity and permeability of

biological membranes (Hazel and Landrey 1988a). PC stabilizes the membrane as it

favors the formation of a laminar bilayer, while PE destabilizes the membrane by keeping

it close to the phase transition between larninar and hexagonal (HE) phase conformations

(Thurmond and Lindblom 1997). The ratio of PC to PE can be used as an indication of

membrane fluidity and rnay be altered according to the environmental and physiological

conditions encountered by an animd (Hazel and Williams 1990). In rainbow trout this is

clearly demonstrated when the PC/PE ratio dropped fiom 1.71 to 0.78 within 8 hours of

cold acclimation (5OC) and rose fkom 1.3 1 to 2.0 by the second day of warm acclimation

(20°C) (Hazel and Landrey 1988a). Despite inhabiting cold waters (5-10°C), the PC/PE

ratio in dogfish gill basolateral membrane (1.72) is quite high and similar to values seen

in warm acciimated rainbow trout (1.71 @ 20°C) (Hazel and Landrey 1988a) and

European yellow eels (Anguilla anguillu, 1.72 @ 15'C) (Aciemo et al. 1996). The high

PC/PE ratio in the dogfish gill BLM may be a physiological response to the increased

membrane fluidity caused by the hi& levels of urea found within the tissues and body

fluids.

It is also possible that the relatively high PCPE ratio in the BLM contributes to

the low urea pemeability observed in the dogfish gill (Part et al. 1998, Wood et al.

1995). Diet-induced increases of the PC/PE ratio in the bmsh border membranes fkom

rainbow trout can be correlated with decreased sodium permeability, demonstrating that

the capacity for PC molecules to pack closer together than PE molecules results in the

reduction of membrane permeabiiity (Di Costanzo et al. 1983). Active urea transport and

high cholesîerol levels have also been implicated as possible mechanisms involved in

reducing the permeability of the giu to urea (Fines et al. in press). The hi& cholesterol

levels in the BLM of the elasmobranch gill epithelium contributes to the highest reported

cholesterol:phospholipid molar ratio (Table 1.4). Cholesterol decreases the permeability

of biological membranes to urea by induchg increased order of the phospholipid

molecules that comprise the bilayer membrane (Mouritsen and Jmgensen 2994). We

proposed that the high cholesterol levels allow this membrane to firnction as a physical

barrier to the passive loss of urea, appearing as reduced urea permeability. Thus, a high

PC/PE ratio may be complimentary to the high cholesterol levels and also contributes to

the low urea permeability of the dogfish gill.

Fattv Acid Composition

Cumulatively, the most prominent fatty acid class Çom the phospholipids of the

dogfish gill BLM was the saturated fatty acids (38.11%, Table 2.1), similar to the

intestinal mucosa of the marine teleost Anguilla anguilla (39- 17%, Acierno et al. 1996).

Phospholipids containhg saturated fatty acids form tightly packed bilayers with relatively

low membrane fluidity and permeability, while phospholipids containing unsaturated

fatty acids pack less tightly resulting in more permeable membranes (Chen et al. 1971).

The chain length of fatty acids is also important in determining the fluidity and

permeability of a membrane. Longer chain fatty acids (18:O) increase the order of

phospholipid bilayers more than shorter chah fatty acids (1 6:O) thus contributhg more to

overall membrane stability (Chen et ai. 197 1, b e l and Landrey 1988b). Several studies

have demonstrated that fatty acid composition affects membrane permeability including

studies on ~ a + and glucose permeability in the b w h border membrane of trout @i

Costanzo et al. 1983) and channel catfïsh (Houpe et al. 1997), respectively. These studies

clearly demonstrated that modification of the levels of SFA in the phospholipids fiom

these membranes alters the permeability to these two solutes. The high levels of SFA,

both l6:O and 1 8 :O, in the dogfish gill BLM are likely involved in maintaining membrane

stability and the reduction of overall membrane permeability. This is supported by the

fact that phospholipids of mitochondrial membranes fiom fish that do not retain elevated

urea levels (hagfish and flounder) contain significantly higher levels of PUFA when

compared to those of the Little skate, an elasmobranch (Glemet and Ballantyne 1996).

This reflects the adaptation of these membranes to the different intemal solute

environments that occur in each of these fish. This difference between elasmobranchs

and other marine fishes is even more evident when one considers that hagfkh and

flounder occur at the same temperature and salinity as the little skate.

The monoene 18: 1 was cumulatively, the second most prominent fatty acid in the

phospholipids fiom the BLM of the dogfish gill epithelium. This also contributed to the

hi& level of monoenes in the dogfish basolateral membrane, relative to other fishes

(hagfïsh and flounder mitochondria, Glemet and Ballantyne 1996; European eel and sea

bass intestinal mucosa, Aciemo et al. 1996; channel catfish intestinal brush border,

Houpe et al. 1997). The increase in the proportion of monoenes in dogfkh gill(28.02%,

Table 2.1), compared to the eel intestinal mucosa (14.92%, Aciemo et al. 1996),

corresponds to a decrease in PUFA levels (dogfish, 34.88%, Table 2.1; eel, 45.91%,

Acierno et al. 1996). This shift in fatty acid class proportions is primarily due to

decreased 22:6n3 levels (dogfish, 8.27%, Table 2.1; eel, 24.90%, Acierno et al. 1996) and

increased 18:l levels (doash, 18.86%, Table 2.1; eel, 13.00%, Acierno et al. 1996) in

the elasmobranch. The effect of the increased monoene and decreased PUFA ievels is

unclear at this time, but they may contribute to increased membrane stabiiity and reduced

membrane permeability. However, the type of unsaturated fatty acids present in

membranes is of secondary importance to the proportion of saturated fatty acids when

considering the physical properties of a membrane (Hazel and Williams 1990).

The primary PUFA in dogfish gül BLM were 20:4n-6, 20511-3, and 22:6n-3.

Overail, PUFA levels were lower than in the erythrocyte membrane of the little skate

(Barton et al. 1999) and in the liver mitochondrial membranes of the haash , flounder,

and linle skate (Glemet and Ballantyne 1996). PUFA increase the fluidity of biological

membranes due to their "bullcy shape", indicating that the levels are kept low in the

dogfish gill BLM to help maintain membrane stability in the presence of elevated urea

levels. It is important to note that PUFA levels in the mitochondrial membrane of the

little skate liver are higher than those of the dogfkh gill BLM, but are stiil much lower

than in hagfkh or flounder liver mitochondrial membranes (Glemet and Ballantyne

1996). This suggests that in addition to maintaining membrane stability, the very low

PUFA levels of the dogfïsh gili BLM may also be involved in the retention of urea by the

a- Certain PUFA have other roles in membranes. Arachidonic acid (20:4n-6) (AA)

is the principle precursor for the production of eicosanoids (prostaglandins (PG),

leukotrienes, thromboxanes, and hydroxyeicosatetraenoic acids) in fish as it is in

terrestrial mammals (Beil et al. 1986). In teleost fish, the gill is one of the most active

tissues in vitro for PG synthesis (Christ and Van Dorp 1972, Ogata et al. 1978,

Henderson et al. 1985) and PG have been identifïed in gill of carp, trout, plaice, eel,

killifish, Atlantic salmon, and turbot (Ogata et al. 1978, Henderson et al. 1985, Srivastava

and M u d a 1984, Van Praag et al. 1987, Brown et al. 1991, Knight et al. 1995, Bell et

al. 1996). PI is the primary source of AA (Marshall et al. 1982). The dogfish gill is rich

in PI when compared to the mammalian erythrocyte, which is considered to be a rich

source of PI (Simpson and Sargent 1985). Furthermore the PI of the dogfkh has hi&

levels of AA (30.70%, Table 2.2) compared to the gills of rnost teleosts and the

mammalian lung (22.3% and 23.996, respectively; Zabelinskii et al. 1995). Tnus the high

level of PI and AA in the dogfïsh gill suggests that this tissue may be important in PG

biosynthesis as it is in teleosts.

PG play an important role in regulating ion transport in fish. PGE2 decreases ion

transport in opercular skin of the killifish (Van Praag et al. 1987) and PGF2, and PGEi

decrease branchial efflwc of ~ a + and Cl- in seawater adapted mullet without affecting

water efflwc (Pic 1975). The high content of PI rich in AA may indicate that PG are

involved in the regulation of ion and urea in the dogfish gill.

Conclusions

The phospholipid and fatty acid composition of the BLM fiom the dogfish gill

epithelium is characterised by a high PCPE ratio, elevated SFA levels, and reduced

PUFA levels. Ali of these modifications to the BLM lipid composition coupled with the

high cholesterol levels reported eariier (see Table 1.4) would contribute to the

maintenance of membrane stability in the presence of the extremely high urea

concentrations present in elasmobranch tissues. These characteristics also likely result in

decreased permeability to urea in the giil BLM, which is important in minimishg the

rnetabolic cost of replacing urea lost to the extemal environment by diffusion dong the

huge concentration gradient.

GENERAL DISCUSSION

The results reported in the previous chapters indicate that two cornplimentary

mechanisms exist in the elasmobranch gill, active urea transport and a unique lipid

composition, that may contribute to the low urea pemeability. 1 propose that the

modified Lipid composition of the basolateral membrane of the giil epithelium,

particdarly the elevated cholesterol level, increases the stability of the membrane

allowing it to fiinction as the primary barrier to urea diffusion fiom the blood into the

epithelial cells. This property of the basolateral membrane would reduce the rate of urea

diffusion into the epithelial ceiis. Secondarily, the urea transport system wouid function

to scavenge the urea that diffuses into the epithelial cells across the cholesterol-rich

basolateral membrane, transporting it back hto the blood against the concentration

gradient. This system of complimentary mechanisms would result in very low

intracellular urea concentrations in the gill epithelium, which would sigdïcantly reduce

the concentration gradient for urea across the apical membrane of these cells. This would

M e r have the affect of decreasing the rate of urea diffusion across the apical membrane

to the extemal medium, which is ultimately responsible for the low rate of urea loss

observed in the elasmobranch giil. The relatively low Km value (-10 mM) reported for

the urea transporter characterized in this study supports the proposal that the intracellular

urea concentrations of the gill epithelial cells are very low compared to that of the blood.

In this situation, where substrate (urea) levels are proposed to be low, the transporter

would be required to have a high affhity (low Km) for the substrate in order h c t i o n

efficiently. In order to maintain the low intracellular urea concentration, the velocity of

the "back transport" process would be similar to, but ultimately less than the rate of urea

diffusion fiom the blood into the epithelial celIs because the Ioss of urea across the gill is

constant.

The presence of high cholesterol levels in the basolateral membrane of the

elasmobranch gill epithelium is one of the most signincant fïndings of this study. The

interactions between cholesterol and phospholipids are s t i l l not well understood and this

naturally occurring cholesterol-nch membrane provides an ideal mode1 to study the

ability of cholesterol to regulate the fluidity and permeability of membranes. Further

investigations of the organization of cholesterol in these membranes using fluorescent

probes that specifically insert in membrane domains nch in cholesterol, such as lawdan

and prodan, could increase our understanding of how cholesterol exerts different effects

on membrane properties depending on the physiological conditions present.

Previous attempts to identfi the mechanisms responsible for the low urea

permeability of the elasmobranch giil focused ody on the involvement of a urea

transporter (Boylan and Antkowiak 1962, Boylan and Lockwood 1962, Boylan et al.

1963, Payan et al. 1974, Part et al. 1998). The results presented here demonstrate the

importance of utilizing different techniques and approaches to investigate physiological

phenornena. Furthermore, they show that more than one mechanism may be involved in

the design of some physiological processes.

The examination of both membrane proteins and lipids may be usefiil in

investigating other epithelial tissues that exhibit selective permeability to various

electrolytes and solut es (e. g. kidney, bladder, intestine, and gall bladder). Although a

wide range of studies have focused on the protein transporters present in these tissues,

analysis of the phospholipid and fatty acid composition of the membranes may provide

crucial information to the understanding of their selective permeability. For example, the

mammalian kidney tubule, or nephron, is a complex structure composed of many regions

that Vary in their permeability to water, urea, ions, and other solutes. The thin ascending

limb of Henle's loop has relatively high urea permeabiIity mai 1977, Imai and Kokko

1974, Imai et al. 1984) and despite the proposed presence of a urea transporter, none has

been detected. The high urea permeability of this segment is thought to be higher than

cm be explained by simple lipid-phase diffusion, but this is based on calculations using

artificial phospholipid bilayen (Sands et al. 1987). It is possible that in this region of the

nephron there is no transporter present and modifications to the lipid composition of the

cell membranes are solely responsible for the observed urea permeability. Modincations

that would confer high urea permeability include low cholesterol levels, high

phosphatidylethanolamine (PE) levels, low phosphatidylcholine (PC) levels, high

unsaturated fatty acids V A ) levels, low saturated fatty acids (SFA) levels and a shorter

mean fatty acid chah length (Pugh et al. 1989, Hazel and Landrey l988a, 1988b, Hazel

and Williams 1990). In contrast, regions of the nephron that exhibit low urea

pemeability, such as the outer medullary collecting duct (OMCD), the cortical coIlecting

duct, the distal convoluted tubule, the thick ascending limb of Henle's loop, and the thin

descending limb of Henle's loop (reviewed by Gillin and Sands 1993), would likely

exhibit higher cholesterol, PC, and SFA levels and lower PE and UFA levels and a longer

mean fatty acid chah length. These speculations on membrane lipid composition can

only be confirmed by f is t isolating individual nephron segments and directly measuring

the specifïc lipid components. Further investigation of the segments of the nephron could

reveal that a system of structural and transport function, similar to the one 1 have

described for the dogfish gill, operates within the mammalian kidney to control urea

permeability.

Urea transporters are present in many organisms, fiom bacteria to mammals, and

are primarily involved in osmoreguIation and nitrogenou waste management. The

presence of urea transporters in bacteria suggests that these proteins have a long

evolutionary history and have been relatively unrnodi£ïed for hundreds of millions of

years. In some organisms such as bactena, the transporter fimctions to acquire urea fiom

the enviromnent for utilization as a nitrogen source (Jahns 1992, Siewe et al. 1998). In

other organisms the transporters act to recover urea that couid be lost to the environment

by various excretory mechanisms. The facilitated urea transporter present in the

elasmobranch kidney (Smith and Wright 1999) is thought to fiinction in the reabsorption

of urea fkom the urine (Schmidt-Nielsen and Rabinowitz 1964, Schmidt-Nielsen et ai.

1972). In toads and fiogs, active &ansporters are involved in the absorption of urea

through the skin (Garcia-Romeu et al. 1981, Rapoport et al. 1989, Lacoste et al. 1991)

and the reabsorption of urea fkom urine in the bladder to facilitate water conservation

(Levine et al. 1973, Couriaud et al. 1999). On the other hand, the unusual teleost fish,

Opsanu beta (the gulf toadfish) utilize a facilitated transporter to excrete urea in pulses

when stressed (Part et al. 1999). In mammals, a faciiitated transporter in the IMCD

segment of the nephron is believed to contribute to the concentration and excretion of

urea by the kidney (Sands et al. 1987).

Despite the dserent functions and roles of urea transporters in different

organisms, they exhibit remarkable physiological and molecular similarity. Facilitated

and active urea transporters both exhibit dose-dependent inhibition by phloretin and

competitive inhibition by urea analogs such as thiourea, acetamide, and methylurea. The

majority of studies of the molecular characterization of urea transporters have focused on

the facilitated transporters of the mammalian kidney. There are two groups of urea

transporters that have been isolated, types A and B, which are products of two genes

(Sands et al. 1997). The homology of the elasmobranch kidney urea transporter (ShUT)

to both the UTA-2 and UT-B2 urea transporter proteins is approximately 60% (Smith and

Wright 1999). The similarity between the two families (UT-A2 and UT-£32) suggests

that they share a common ancestor. ShUT is not this ancestor because it has a higher

identity with UT-A2 than with UT-B2 due to the lack of residues found only in UT-B2

family members. This led to the suggestion that ShUT is an ancestral form of UT-A2 and

that an ancestral form of UT-B2 may also be present in elasmobranchs but has yet to be

discovered. Regardless, the preliminary molecular evidence for a urea transporter in the

dogfish gill provided by Smith and Wright (1999) has been supported by the

characterization of the urea transporter in this study. These Çidings j u s t e M e r study

of the genes encoding for these proteins in elasmobranch gill tissue in hope of providing

an improved understanding of the evolutionary lineage and expression of urea

transporters in vertebrates.

The results fiom the lipid analysis of the basolateral membrane fiom the dogfkh

gill epithelium has revealed that structural modifications in the membrane, especially the

incorporation of high levels of cholesterol, may be primarily responsible for making the

dogfkh gill relatively impermeable to urea. One may then ask, what came first, the high

levels of urea or the impermeable membrane? Did early elasmobranchs have modified

gill epithelial membranes that facilitated the accumulation of urea within the tissues,

allowing these fish to adapt to the marine environment? Did urea levels increase in

response to saha ter exposure and the necessary changes in the lipid composition of the

membranes evolved in order to adapt to the high levels of urea? I suspect that the fïrst

hypothesis is more likely because the urea permeabiiity of the gills fkom euryhaline

skates (Raja erinacea and Raja radiata) remains constant when the animals are exposed

to dilute (50%) seawater (Payan et al. 1973). Irregardless of which came first, the

beneficial result of this adaptation was that the modifïed membranes assisted in retaining

urea and reduced the energetic costs of this osmoregulatory strategy, making it an

efficient and effective strategy for over 350 million years.

Other aspects of the phospholipid composition of the basolateral membrane fiom

the elasmobranch giil epithelium hint at other functions for this membrane. The

abundance of arachidonic acid (20:4n6) containing phosphatidylinositol (PI) suggests that

the elasmobranch gill plays an important role in the biosynthesis of eicosanoids (e.g.

prostaglandins). In teleostç, prostaglandins are involved in reproduction, particularly

ovulation and spawning behaviou. (Stacey and Goetz 1982), and there is evidence for the

involvement of prostaglandins in regulating osmoregulation (Pic 1 975, Van Praag et al.

1987). It is possible that prostaglandins are involved in regulating osmoregulation and

reproduction in elasmobranchs, which is of great interest due to their unique

osmoregulatory strategy and the range of reproductive strategies in this group. Further

examination of the involvement of the elasmobranch gill in the biosynthesis of

eicosanoids is therefore necessary.

Overall, these investigations have contributed signifïcantly to our understanding

of possible mechanisms responsible for the low urea permeability of the elasmobranch

giu. This is particularly relevant due to the importance of urea in the osmoregdatory

strategy of elasmobranchs and the hi& metabolic cost associated with producing elevated

levels of urea. The metabolic cost of replacing urea that is lost to the environment by

diffusion across the gill is approximately 97.3% less than would be predicted based on

the urea permeability of teleost gills. This translates into savings of 48.65 mm01

ATP/kg/h. Without the retention of urea by the gill due to its low urea permeability, the

metabolic cost of living for elasmobranchs would be almost double (106.35 mm01

ATPlkgh) the estimated total energy requirements of the dogfish (57.70 mm01

ATPlkgh) (Bushnell et al. 1989). However, it should be noted that al l of these values are

calculated and require validation.

The use of purified basolateral membrane vesicles has provided important

information regarding the mechanisms involved in making the elasmobranch gill

relatively impermeable to urea but further characterization of this membrane needs to be

undertaken. Firstly, measwing the urea permeability of protein-fiee vesicles to urea

would provide additional information to support our proposed barrier function of

basolateral membrane phospholipid bilayer. Also effects of temperature, urea

concentration, trimethylamine oxide and other physiological factors on the

fluidity/permeability of this membrane can be determined using flourescent probes.

Probes specifk for cholesterol (laurdan and prodan) could be used to examine the

organization of cholesterol in these membranes in an effort to understand how cholesterol

provides a low urea permeability to the membrane. Secondly, studies should be

conducted to m e r characterize the urea transporter described in chapter one using

vesicles of a single orientation. Lectin affhîty chromatography would pennit the

separation of inside out and right side out vesicles and provide a more precise

detemination of the physiological properties of this transporter (e.g. kinetic parameters,

optimal analogue and inhibitor concentrations, stoichiometry of ~ a + and urea exchange).

Further studies should be undertaken to investigate the intracellular urea

concentration of the gill epithelial ceils. This would help to clar* the mode1 1 have

proposed for the involvement of the basolateral membrane in the retention of urea by the

gill. Measuring the total urea concentration of excised gill filaments and then correcting

for the blood space by measuring hemoglobin content could provide an indirect

determination of intracehlar urea concentration.

Studies on the effect of changes in salinity upon the lipid composition and urea

transport activity in the gill should aiso be conducted as blood urea levels decrease in

dilute seawater and modifications consistent with a homeoviscous response would be

expected. The use of euryhaline elasmobranchs, such as Raja erinacea, and fieshwater

elasmobranchs (Potamot?ygon spp.) may assist in the elucidation of the evolution of

these mechanisms in the elasmobranchs.

Finally, analysis of the lipid composition of the apical membrane would provide

information vital to the complete understanding of the mechanisms involved in

conferring low urea permeability to the elasmobranch gill. This membrane is the last

possible barrier to urea loss across the gill and is very likely to have a modified lipid

composition, not unlike that described for the basolateral membrane in chapter two.

REFERENCES

Aciemo, R., Mas, M., Sicuro, P., Fiammata, L., Rollo, M., Ronzini, L., and Storelli, C.

(1996) Lipid and fatty acid composition of intestinal mucosa of two antarctic teleosts,

Comp. Biochern. Physiol. 1 1 SA, 303-307.

Ackman, RG. (1962) Structure and retention time in the gas-liquid chromatography of

unsaturated fatty acids on polyester substrates, Namre 194,970-97 1.

Baglioni, S. (1905) Die bedeutung des harnstoffes bei den selachiern, Zbl. PhysioL 19,

385-388.

Baglioni, S. (1906a) Vergleichende chemische untersuchmgen an den muskeln, den

elektrischen organen und dem blutserurn von Torpedo ocellata, Beitr. Chem. Physiol.

Path. 8, 456-471.

Baglioni, S. (1906b) Einige daten zur kenntnis der quantitativen nisammensetzung

verschiedener kerrperflüssigkeiten von seetieren (Fischen und einigen wirbellosen),

Beitr. Chem. PhysioZ. Parh. 9,50-66.

Baglioni, S. (1906~) Zur kenntnis des K s t o ~ e c h s e l s der fische (Die bedeutung des

harnstoffes bei den selachiern), Zbl. Physiol. 20, 105-108.

Baglioni, S. (1 9 17a) L'action physiologique de l'urée, Arch. Ital. Biol. 67,49-68.

Baglioni, S. (1917b) Sur la nature des processus physiologiques des organes électriques,

Arch. Ital. Biol. 67, 93-104.

Ballantyne, J.S., Glemet, H.C., Chamberlin, ME., and Singer, T.D. (1993) Plasma

nonesterifïed fatty acids of marine teleost and elasmobranch fishes, Mar. Biol. 116,

47-52.

Barton, K-N., Buhr, MM., and Ballantyne, J.S. (1999) Effects of urea and

trimethyiamine N-oxide on fluidity of liposomes and membranes of an elasmobranch,

Am. J. Phys. 276, R397-R406.

Bayliss, W.M. and Starling, E.H. (1902) The mechanism of pancreatic secretion, J;

PhysiÛl. 28,325-353.

Bell, J.G., Famdale, B.M., Dick, J-R., and Sargent, J.R. (1996) Modification of

membrane fatty acid composition, eicosanoid production, and phospholipase A

activity in atlantic salmon (Salmo salar) giI1 and kidney by d i e w lipid, L i p i h 3 1,

1163-1171.

Bell, M.V., Henderson, RI., and Sargent, J.R. (1986) The role of polyunsaturated fatty

acids in fish, Comp. Biochem. Physiol. 83B, 71 1-7 19.

Blier, P. and Guderley, H. (1988) Metabolic responses to cold acchnation in the

swimming musculature of lake whitefish, Coregonm clupeaformk, J. Exp. 2001. 246,

244-252.

Bligh, E. G. and Dyer, W.J. (1959) A rapid method of total lipid extraction and

purification, Can. J. Biochem. Physiol. 37,9 1 1-9 17.

Bompiani, R. (1913) Sulla sostituibilità deil' urea nelle soluzioni artificiali per il cuore

isolat0 dei selaci, 2. Allg. Physiol. 15,292-3 15.

Borchman, D., Delamere, N-A., McCauley, L.A., and Paterson, C.A. (1989) Studies on

the distribution of cholesterol, phospholipid, and protein in the huma. and bovine lem,

Lens Eye Tox. Res. 6,703-724.

Botazzi, F. (1906) Sulla regolazione della pressione osmotica negli organismi animali.

Nota 2a. Resistenza dei corpuscoli rossi di Scyllium e di SipunmZm a cedere

rispettivamente l'emoglobina e l'emeritrina, Arch. Fr'siol. 3,495-506.

Boylan, J.W. (1 967) Gill permeability in SquaZw acanthias, in Sharks, skates and rays

(Gilbert, P.W., Mathewson, RF., and Rall, D.P., eds.) pp. 197-206, The Johns

Hopkins Press, Baltimore.

Boylan, J.W. (1972) A mode1 for passive urea reabsorption in the elasmobranch kidney,

Comp. Biochern. Physiol. 42A, 27-3 0.

Boylan, J.W., and Antkowiak, D. (1962) Blood-sea water barrïer to urea at the dog£ish

gill: effect of hypothermia and urea loading, Bull. Mt. Desert Is. Biol. Lab. 4, 24.

Boylan, J.W., and Lockwood M. (1962) Urea and thiourea excretion by dogfish kidney

and gill: effect of temperature, Bull. Mt. Desert 1s. Biol. Lab. 4,25.

Boylan, J.W., Feldman, B., and Antkowiak, D. (1963) Factors affecthg gill permeability

in Squalus acanthias, Bull- Mt Desert 1s. BioZ. Lab. 5(1), 29.

Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram

quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72,

248-254.

Brand, M.D. (1995) Measurement of mitochondrial protonmotive force, in Bioenergetics:

A practical approach, (Brown, G.C. and Cooper, C.E., eds.), pp. 44.46, IRL Press,

New York.

Brown, LA., Gray, C.J., Hattersley, G., and Robinson, J. (1991) Prostaglandins in the

kidney, urinary bladder and gills of the rainbow trout and european eel adapted to

fiesh water and seawater, Gen. Comp. Endocrinol. 84, 328-33 5 .

Bulder, H. A. M., Den Mjs, A.P.M., Speek, E.J., v a . Hasselt, P.R., and Kuiper, P.J.C.

(1991) The effect of low root temperature on growth and lipid composition of low

temperature tolerant rootstock genotypes for cucumber, J; Plant Physiol. 138, 661-

666.

BushneU, P.G., Lutz, P.L., and Gruber, S.H. (1989) The metabolic rate of an active,

tropical elasmobranch, the lemon shark (Negaprion brevirostris), E q . Biol. 48, 279-

283.

Carlson, S.R. and Goldstein, L. (1997) Urea transport across the cell membrane of skate

erythrocytes, J: Exp. Zool. 277,275-282.

Chen, L.F., Lund, D.B, and Richardson, T. (1971) Essential fatty acids and glucose

permeability of lecithin membranes, Biochim. Biophys. Acta 225, 89-95.

Chou, C.-L., Sands, J.M., Nonoguchi, H., and Knepper, M.A. (1990) Concentration

dependence of urea and thiourea transport pathway in rat b e r medullary collecting

duct, Am. J: Physiol. 258, F486-F494.

Chou, C.-L. and Knepper, M.A. (1989) Inhibition of urea transport in inner medullary

coliecting duct by phloretin and urea analogues, Am. J Physiol. 257, F3 59-F365.

Christ, E.J. and Van Dorp, D.A. (1972) Comparative aspects of prostaglandin

biosynthesis in animal tissues, Biochim. Biophys. Acta 270, 537-545.

Clarke, R.W. and Smith, H.W. (1932) Absorption and excretion of water and salts by the

elasmobranch fishes. III. The use of xylose as a rneasure of the glomenilar filtrate in

Squalus acanthias, J; CeZZ. Comp. Physiol. 1, 1 3 1 - 143.

Crockett, E. L., and Hazel, J. R (1995) CholesteroI levels explain inverse compensation

of membrane order in b w h border but not homeoviscous adaptation in basolateral

membranes fiom the intestinal epithelia of rainbow trout, J. E'. Biol. 198, 1 105-

11 13.

Couriaud, C., Leroy, C., Simon, M., Siberstein, C., Bailly, P., Ripoche, P., and

Rousselet, G. (1999) Molecular and functional characterization of an amphibian urea

transporter, Biochim. Biophys. Acta 142 1,347-3 52.

Cuculescu, M., Hyde, D., and Bowler, K.J. (1995) Temperature acclimation of marine

crabs: changes in plasma membrane fluidity and lipid composition. Therm. Biol. 20,

207-222.

De Meyer, J. (19 10) Étude sur les altérations du courant d'action du coeur de Scyllium

canicula, Arch. Int. Physiol. 10, 100-129.

Di Costanzo, G., Duportail, G., Florentz, A., and Leray, C. (1983) The brush border

membrane of trout intestine: Influence of its lipid composition on ion permeability,

enzyme activity and membrane fluidity, Mol. Physiol. 4,279-290.

Dumas, D., Muller, S., Gouin, F., Baros, F., Viriot, M., and Stoltz, J. (1 997) Membrane

fluidity and oxygen diffusion in cholesterol - e ~ c h e d erythrocyte membrane, Arch.

Biochem. Biophys. 34 1,34-39.

Duvall, M. (1925) Sur la pression osmotique du milieu intérieur des sélachiens, Ann.

Physiol. f hysicochem. 1,3 12-326.

Duvail, M. and Portier, P. (1923) Imperméabilite à l'urée de divers tissus des poissons

sélaciens, CR. Acad. Sci., Paris, 176,920-921.

Ehrenfeld, J. (1998) Active proton and urea transport by amphibian skin, Comp.

Biochem. Physiol. 1 1 9A, 3 5-45.

Endo, M. (1984) Histological and enzymatic studies on the rend tubules of some marine

elasmobranchs, J. Morphol. 182,63-69.

Farber, J., Gerstein, B., and Boylan, J.W. (1965) Permeability of dogfish gill to sodium,

Bull- M. Desert Is. Bio l. Lab. 5(2), 14.

Fines, GA., Ballantyne, J.S., and Wright, P.A. Active urea transport and an unusual

basolateral membrane composition in the g u s of a marine elasmobranch, Am. J,

Physiol., in press.

Flik, G., Wendelaar Bonga, SE., and Fenwick, J.C. (1985) Active Ca2+ transport in

plasma membranes of branchial epithelium of the North American eel, Anguilla

rostrata Lesueur, Biol. Cell55,265-272.

Forster, R P . and Berglund, F. (1957) Contrasting inhibitory effects of probenecid on

rend tubular excretion of PAH and on the active absorption of urea in the dogfïsh,

Squaius acanthias, J . Cell. Comp. Physiol. 49,28 1-285.

Frédéricq, L. (1904) Sur la concentration moléculaire du sang et des tissus chez les

animaux aquatiques, Arch. Biol. 20, 709-730.

Frédéricq, L. (1922) Pulsations de coeur de ScyZIium catu1u.s en l'absence d'urée, Arch.

Int. Physiol. 19,253-256.

Fühner, H. (1908) Über eine speisungflüssigkeit für selachierherzen, Allg. Physiol. 8 ,

485-49 1.

Garcia-Romeu, F., Masoni, A., and Isaia, J. (1 98 1) Active urea transport through isolated

skins of fiog and toad, Am. J. Physiol. 241, R114-R123.

Ghioni, C., Beii, J.G., Bell, M.V., and Sargent, J.R. (1997) Fatty acid composition,

eicosanoid production and permeability in skin tissues of rainbow trout

(Oncorhynchur mykiss) fed a control or an essentiai fafty acid deficient diet,

Prostaglandins Leukot. Essent Fa@ Acids 5 6,479-48 9.

Gillin, A.G. and Sands, J.M. (1993) Urea transport in the kidney, Sem. Nephrol. 13, 146-

154.

Glemet, K.C. and Ballantyne, J.S. (1996) Cornparison of Iiver mitochondnal membranes

fiom an agnathan (Myxine glutinosa), an elasmobranch (Raja erinacea) and a teleost

fish (Pleuronectes americanm), Mar. BioL 124, 509-5 1 8.

Gréhant and lolyet (1891) Formation de l'urée par la décharge électrique de la torpille,

C R . Soc. BzOl., Paris, 43,687-691.

Hays, R.M., Levine, S.D., Myers, J.D., Henemann, H.O., Kaplan, M.A., Franki, N., and

Berliner, H. (1977) Urea transport in the do@sh kidney, J fip Zool. 199,309-3 16.

Hazel, J.R. (1995) Thermal adaptation in biological membranes: 1s homeoviscous

adaptation the explanation? Annu. Ra? Physiol. 57, 19-42.

Hazel, J.R. and Landrey, S.R (1988a) Time course of themai adaptation in plasma

membranes of bout kidney. 1. Headgroup composition, Am. J. Physiol. 255, R622-

R627.

Hazel, J.R and Landrey, S R (1988b) Time course of thermal adaptation in plasma

membranes of trout kidney. II. Molecdar species composition, Am, J. Physiol. 255,

R628-R634.

Hazel, J.R. and Williams, E.E. (1990) The role of alterations in membrane lipid

composition in enabling physiological adaptation of organisms to their physical

environment, Prog. Lipid Res. 29, 167-227.

Henderson, R J., Bell, M.V., and Sargent, J.R. (1 985) The conversion of polyunsaturated

fatty acids to prostaglandins by tissue homogenates of the turbot, Scophthalmus

mmcimus (L.), J. EXp. Mar. Biol. Ecol. 85,93-99.

Hopfer, U., Nelson, K., Perrotto, J., and Isselbacher, K.J. (1973) Glucose transport in

isolated brush border membrane fkom rat small intestine, J. Biol. Chem. 248,2532.

Houpe, K.L., Malo, C., and Buddington, R.K. (1997) Dietary lipid and intestinal b m h

border membrane phospholipid fatty acid composition and glucose transport of

channel catnsh, Phys. Zool.70,230-236.

Imai, M. (1977) Function of the thin ascending Iimbs of Henle of rats and hamsters

perfused in vitro, Am. J. Physiol. 232, F20 1 -FZO9.

Irnai, M., Hayashi, M., and Araki, M. (1984) Functional heterogeneity of the descending

limbs of Henle's Loop. 1. Intemephron heterogeneity in the hamster kidney, Pjluegers

Arch. 402,3 85-392.

Imai, M. and Kokko, J.P. (1974) Sodium, chloride, urea, and water transport in the thin

ascending limb of Henle, J. Clin. Invest. 53,393-402.

Jahnke, L.L. (1992) The effects of growth temperature on the methyl sterol and

phospholipid fatty acid composition of Methylococncr capsulatus (Bath), FEMS

Microbiology Letiers 93,209-2 12.

Jahns, T. (1992) Urea uptake by the marine bacterium Deleya venustai HG1, J. Gen.

Microbiol. 138, 1815-1820.

Johnson, R., Antkowiak, D., and Boylan, J.W. (1964) Gill permeability in Squalus

acanthias, Bull. Mt. Desert Is. Biol. Lab. 5(1), 55.

Kato, A., and Sands, J.M. (1998) Evidence for sodium-dependent active urea secretion in

the deepest subsegment of the rat inner meduliary collecting duct, J. Clin. Invest. 101,

423428.

Kempton, R.T. (1953) Studies on the elasmobranch kidney: II. Reabsorption of urea by

the smooth dogfkh, Mustelus canis, Biol. Bull. 104,45-56.

Knight, J., Holiand, J.W., Bowden, L.A., Halliday, K., and Rowley, A.F. (1995)

Eicosanoid generating capacities of different tissues fiom the rainbow trout,

Oncorhynchus mykiss, Lipiak 30,451458.

Kondo, Y. and Imai, M. (1987) Effects of glutaraldehyde fixation on reanl tubuiar

function. 1. Preservation of vasopressin-stimulated water and urea pathways in rat

papillary collecting duct, Pfluegers Arch. 408,479483.

Koulali, Y., Fonvieille, J.L., Es-Sgaouri, A., and Dargent, R (1998) Influence of

monensin on the lipid and protein composition of hyphae membranes in Bobytis

cinerea Pers. and Sclerotium rolfsii Sacc, Can. J. Microbiol. 44,937-944.

Knikenberg, C.Fr.W. (1 88 1) Untersuchmg der fleischextrakte verschiedener fische und

wirbeilosen, fitersuch. Physiol. h t i t . Univ. Heidelberg. 4,33-63.

Krukenberg, C.Fr.W. (1886) Gnuidzüge einer vergleichenden phyiologie der contractilen

gewebe, Verglechend-physiologische Vortrage 1, Heidelberg, 273-379.

Krukenberg, C.Fr.W. (1887) Die hamstofietention in den organen der rochen und haie,

Zbl. Med. Wiss. 25,450455.

Knikenberg, C.Fr.W. (1888) La rétention de l'urée chez les sélaciens, Ann. MW. Hist

Nat. Mareille, 3, 3e mémoire.

Lacey, E.R., Schmid-Nielsen, B., Galaske, RG., and Stolte, H. (1975) Configuration of

the skate (Raja erinacea) nephron and ultrasticucture of two segments of the proximal

tubule, Bull. Mt- Desert h l , Biol. Lab. 15, 54-56.

Lacey, EX, Reale, E., SchIusselberg, DaS., Smith, W.K., and Woodward, D. J. (1985) A

rend countercurrent system in marine elasmobranch fish: A computer-assisted

reconstruction, Science 227, 13 5 1 - 13 54.

Lacoste, I., Dunel-Erb, S., Harvey, B.J., Laurent, P., and Ehrenfeld, J. (1991) Active urea

transport independent of H+ and ~ a + transport in fiog skin epithelium, Am. J. Physiol.

261, R898-R.906.

Lande, M.B., Donovan, J.M., and Zeidel, M.L. (1995) The relationship between

membrane nuidity and permeabilities to water, solutes, smmonia, and protons, J. Gen.

Physiol. 106,67-84.

Levine, S., Franki, N., and Hays, RM. (1973) Effect of phloretin on water and solute

movement in the toad bladder, J. Clin. Invest. 52, 143 5-1442.

Marshaii, P.J., Dixon, J.F., and Hokin, L.E. (1 9 82) Evidence that phosphatidylinositol

breakdown releases arachidonic acid, forming prostaglandins which are involved in

stimulus-secretion coupling in the exocrine pancreas, Prog. Lipid Res. 20, 229-232.

McCormick, S.D. (1993) Methods for nonlethal gill biopsy and measurement of Na+,

K+-ATPase activity, Can. J. Fish. Aquat. Sci. 50,656-658.

Moiitoris, B.A., Alfiey, A.C., Harris, R.A., and Simon, F.R. (1985) Rend apical

membrane cholesterof and fiuidity in regdation of phosphate transport, Am. J. Physiol.

249, F 12-F 19.

Mouritsen, O. Ga, and K. Jmgensen. Dynamical order and disorder in lipid bilayers.

Chem. Phys. Lipids 73: 3-25, 1994.

Ogata, Hm, Nomura, T e , and Hata, M. (1978) Prostaglandin biosynthesis in the tissue

homogenates of marine animals, Bull. J q . Soc. Sci. Fish. 44, 1367- 1370.

Part, P., Wright, P.A., and Wood, C.M. (1998) Urea and water permeability in dogfish

(Squalus acanthias) gills, Comp. Biochem. Physiol. 1 1 9A, 1 1 7- 1 23.

Part, P., Wood, C.M., Gilmour, KM., Perry, S.F., Laurent, P., Zadunaisky, J., and Walsh,

P.J. (1999) Urea and water permeabiiity in the ureotelic gulf toadfish (Opsanus beta),

J. Exp. Zool. 283, 1-12.

Payan, P., Goldstein, L., and Forster, R.P. (1973) GiUs and kidneys in ureosmotic

regdation in euryhaline skates, Am. J. Physiol. 224,367-372.

Perry, S. F. and Fm G. (1988) Characterization of branchial transepithelial calcium

fluxes in fieshwater trout, Salmo guirdneri. Am. J. Physiol. 254, R491-R498.

Pic, P. (1975) Effets des prostaglandines (PG El et F2a) sur les échanges branchiaux

d'eau de Na et Cl chez Mugi1 capito adapté à l'eau de mer, J. Physiol. (Paris) 71,

L46A.

Pough, F.H., Heiser, J.B., and McFarland, W.N. (1996) Vertebrate lif, 4" edn., pp. 197-

2 19, Prentice Hall, Upper Saddle River, NJ.

Pugh, E.L., Bittman, R., Fugler, L., and Kates, M. (1989) Comparison of steady-state

fluorescence polarization and urea permeability of phosphatidylcholine and

phosphatidylsulfocholine liposomes as a function of sterol structure. Chem. Phys.

Lei& 50,43-50.

Rabuteau and Papillon, F. (1873) Observations sur quelques liquides de l'organisme des

poissons, des crustacés et des céphalopodes, C.R Acad Sci., Paris, 77(2), 135-138.

Rapoport, J., Chaimovitz, C., and Hays, RM. (1989) Active urea transport in toad skin is

coupled to @ gradients, Am. J. Physiol. 256, F830-F835.

Robertson, J.D. (1989) Osmotic comtituents of the blood plasma and parietal muscle of

Scyliorhinus canicula (L.), Cornp. Biochern- Physiol. 93A, 799-805.

Rodier, E. (1899) Observations et expériences comparatives sur l'eau de mer, le sang et

les liquides internes des animaux marins, Soc. Sei. Slat. Zool. D Arcachon 103- 123.

Rodier, E. (1900) Sur la pression osmotique du sang et des liquides internes chez les

poissons sélaciens, C. R. Acad. Sci., Paris, 13 1(2), 1008-10 10.

Rohmann, F. (1893) a e r den stoffumsatz in dem thatigen elektrischen organ des

zitterrochen nach versuchen an der zoologischen station zu neapel, Arch. Anat.

Physiol. 5,423-482.

Ruggieri, S., Gregori, L., Natalini, P., Vita, A., EmanueLli, M., Raffaelli, N., and Magni,

G. (1990) Evidence for an inhibitory effect exerted by yeast NMN adenylyltransferase

on poly(ADP-ribose) polymerase activity, Biochemishy 29,250 1-2506.

Sands, J.M. and Schrader, D.C. (1991) An independent effect of osmolality on urea

transport in rat terminal IMCDs, J. Clin. Invest. 88, 137- 142.

Sands, J.M., Nonoguchi, H., and Knepper, M.A. (1987) Vasopressin effects on urea and

HzO transport in imer medullary collecting duct subsegments, Am. J Physiol. 253,

F823-F832.

Sands, J.M., Timrner, R.T., and Guan, R.B. (1997) Urea transporters in kidney and

erythrocytes, Am. J. Physiol. 273, F32 1-F339.

Schmidt-Nielsen, B. and Rabinowitz, L. (1964) Methyluea and acetamide: Active

reabsorptioa by elasmobranch rend tubules, Science 146, 1587-1588.

Schmidt-Nielsen, B., Truniger, B., and Rabhowitz, L. (1972) Sodium-linked urea

transport by the rend tubule of the spiny dogfish Squalus acmthias, Comp. Biochem.

Physiol. 42A, 13-25.

Schultze, M. (1861) Ergebnisse einiger die elektrischen organen von Torpedo und das

schwanzorgan von Raja betreffender chemischer untersughungen, J. Prakt. Chern. 82,

1-12.

Siewe, R.M., Weil, B., Burkovski, A., Eggeling, L., Kriimer, R, and Jahns, T. (1998)

Urea uptake and urease activiîy in Corynebacterium glutamiam, Arch- Microbiol.

169,411-416.

Simpson, C.M.F. and Sargent, J.R (1 985) Inosito t lipid turnover and adenosine 3,s cyclic

monophosphate in the salt-secreting rectal gland of the dogfish (Sqdiorhinus

canicula), Comp. Biochem. Physiol. 82B, 781-786.

Smith, C.P. and Wright, P.A. (1999) Molecular characterization of an elasmobranch urea

transporter, Am. J. Physiol. 276, R622-R626.

Smith, D.J. and Ploch, S.A. (199 1) Isolation of Raja erinacea basolateral Iiver plasma

membranes: characterization of lipid composition and fluidity, J. Exp. ZooZ. 258, 189-

195.

Smith, H. W. (193 1) The absorption and excretion of water and salts by the elasmobranch

fishes. II. Marine elasmobranchs, Amer. J. Physiol. 98,296-3 10.

Smith, H.W. (1936) The retention and physiological role of urea in the elasmobrançhii,

Biol. Rev. 1 1,49-82.

Srivastava, K.C. and Mustafa, T. (1984) Arachidonic acid metabolism and prostagiandins

in lower animals, Mol. Physiol. 5,53-60.

Stacey, NE. and Goetz, F.W. (1982) Role of prostaglandins in fish reproduction, Can. J.

Fish. Aquat. Sci. 39, 92-98.

Stadeler, G. (1859) Weitere beobachtungen über das vorkommen von hamstoff in den

organen der plagiostomen, J. Prnkt- Chem. 73,48-55.

Stadeler, G., and Frérichs, F.T. (1858) Über das vorkommen von hamstoff, taurin, and

scyllit in den organen der plagiostomen, J. Prakt. Chem. 73,4845.

Stio, M., Vanni, P., and Pinzauti, G. (1988) A continuous spectrophotometric assay for

the enzymatic marker glucose-6-phosphatase, Anal. Biochem. 174,32-37.

Stolte, K., Galaske, R.G., Eisenbach, G.M., Lechene, C., Schmidt-Nielsen, B., and

Boylan, J.W. (1977) Rend tubule ion transport and collecting duct function in the

elasmobranch Little skate Raja erinacea, J. Exp. Zool. 199,403-4 10.

Straub, W. (1901) Toxikologische untersuchungen an selachierhenen, 2. Biol. 24, 363-

376.

Thorson, T.B., Cowan, C.M., and Watson, D.E. (1967) Potamohygon spp.:

Elasmobranchs with low urea content, Science 158,375-377.

Thurmond, RL. and Lindblom, G. (1997) NMR studies of membrane lïpid properties, in

Current Topics in Membranes (Epand, R., ed.) Vol. 44, pp. 103- 1 16, Academic Press,

New York.

Van Heeswijk, M.P.E., Geertsen, J.A.M., and van Os, CH. (1984) Kinetic properties of

the ATP-dependent Ca2+ pump and the Na+/Ca+ exchange system in basolateral

membranes fiom rat kidney cortex, J Membrane Biol. 79, 19-3 1.

Van Praag, D., Farber, S.J., Minkin, E., and Primor, N. (1987) Production of eicosanoids

by the kdlifïsh g u s and opercular epithelia and their effect on active transport of ions,

Gen Camp- Endocrinol. 67,50-57.

Von Schroeder, W. (1890) Über die harstonbildung der hainsche, Huppe-Seyl. S. 14,

576-598.

Walsh, P., Wood, C.M., Perry, S.F., and Thomas, S. (1 994) Urea transport by hepatocytes

and red blood cells of selected elasmobranch and teleost fishes. J; Exp. Biol. 193,32 1-

335.

Walsh, P.J., Heitz, M., Campbell, CE., Cooper, G.J., Medina, M., Wang, Y.S., Goss,

G.G., Vincek, V., Wood, C.M., and Smith, C.P. Molecdar identification of a urea

transporter in gill of the ureotelic gulftoadfish (Opsanus beta), J. Exp. Biol. in press.

Wood, C.M., Part, P., and Wright, P.A. (1995) Ammonia and urea metabolism in relation

to gill b c t i o n and acid-base balance in a marine elasmobranch, the spiny dogfïsh

(Squalus ucanthias), J. Exp. Biol. 198, 1545- 1558.

Yamaguchi, H., Tanaka, T., Ichioka, T., Stoskopf, M., Kishimoto, Y., and Gould, R.

(1987) Characterization and comparison of lipids in different squid nervous tissues,

Biochirn. Biophys. Acta 922,78-84.

Yancey, P.H. and Somero, G.N. (1980) Methylamine osmoregulatory solutes of

elasmobranch fishes counteract urea inhibition of enzymes, J. Exp. 2001. 212, 205-

213.

Zabelinskii, S.A., Brovtsyna, N.B., Chebotareva, UA., Gorbunova, O.B., and

Krivchenko, A.I. (1995) Comparative investigation of lipid and fatty acid composition

of fish gills and mammalian lungs. A mode1 of the membrane iïpid component areas,

Comp. Biochem. Physiol. 1 1 lB, 127-140.

Ziboh, V.A., and Miller, C.C. (1990) Essential fatty acids and polyunsaturated fatty

acids: simiif?cance in cutaneous biology, Ann. Rev. Nutr. 10,433-450.

APPENDIX 1

Schematic representation of the assay of the percentage of inside-out-oriented vesicles. Assuming that resealed vesicles are impermeable to either ATP or ouabain, then N~+,K+- ATPase activity of untreated vesicles reflects the leaky vesicle fiaction. The detergent digitonin (0.04% w/v) stimulates N~+,K+-ATP~s~ activity by permeabilizing the membrane, providing ATP and ouabain access to the interior of the resealed vesicles. The percent ciifference in activity between untreated and digitonin trerited vesicles represents the percentage of resealed vesicles. The percentage of resealed vesicles with an inside-out orientation is obtained by measuring the stimulation of ATP hydrolysis induced by K+ ions of untreated and digitonin-treated vesicles in medium containhg ~ a + and ouabain. During short incubations, K' ions will equilibrate across the membrane quickly while ouabain does not. The N~+,K+-ATP~s~ protein of inside-out vesicles is inaccessible to ouabain and when digitonin is added the membrane barrier for ouabain is removed. This results in a decrease in N~+,K+-ATP~s~ activity which represents the percentage of inside-out vesicles. The difference between total resealed vesicles and inside-out vesicles equals the percentage of right-side out vesicles.

Right-side Out Ouabain

ATP

APPENDIX II

Schematic representation of the assay for measuring vesicle volume. Vesicles are incubated with a radiolabeled probe, such as 3 ~ f l , and then sedimented. The accessible volume of the pellet is calculated nom the specific activity of the probe in the supematant and the total radioactivity in the ellet. The difference in accessible volume (pellet P space) for penneant probes like H20 and probes like 1 4 ~ - ~ ~ ~ - 4 0 0 0 that do not cross the huer membrane equals the volume of the vesicle.

d.p.m. x supematant volume)/supernatant d.p.m. Vesicle Volume = space - "c-PEG-4000 space)/mgprotein added

Incubation Precipitation