A

ipwpCpgscsime©

K

1

Mdg

Rnf

U

0d

Available online at www.sciencedirect.com

Vaccine 25 (2007) 8611–8621

Chemokine and cytokine gene expression profiles in chickens inoculatedwith Mycoplasma gallisepticum strains Rlow or GT5

Javed Mohammed a,c, Salvatore Frasca Jr. b,c, Katharine Cecchini b,c,Debra Rood a,c, Akinyi C. Nyaoke b,c,

Steven J. Geary b,c, Lawrence K. Silbart a,c,d,∗a Department of Animal Science, University of Connecticut, Storrs, CT 06269, United States

b Department of Pathobiology and Veterinary Science, University of Connecticut, Storrs, CT 06269, United Statesc Center of Excellence for Vaccine Research, University of Connecticut, Storrs, CT 06269, United States

d Department of Allied Health Sciences, University of Connecticut, Storrs, CT 06269, United States

Received 31 July 2007; received in revised form 20 September 2007; accepted 25 September 2007Available online 16 October 2007

bstract

Mycoplasma gallisepticum infection in chickens leads to tracheitis, airsacculitis, poor feed conversion and reduced egg production, result-ng in considerable economic hardship on the poultry industry. The chemokines and cytokines responsible for recruitment, activation androliferation of leukocytes in affected tissues have not been described. In the current study, chemokine and cytokine gene expression profilesere investigated in tracheas of chickens inoculated with M. gallisepticum strains Rlow (pathogenic) and GT5 (attenuated) at days 1, 4 and 8ost-inoculation. Expression of lymphotactin mRNA was higher in Rlow-inoculated chickens than GT5- or PBS-inoculated chickens, whileXCL13/BCA1 mRNA expression level was higher in both GT5- or Rlow-inoculated chickens than in PBS-inoculated controls on day 1ost-inoculation. However, both Rlow and GT5 strains induced a down-regulation in mRNA expression of CCL20, IL-1�, IL-8 and IL-12p40enes, with CCL20 and IL-12 mRNA levels remaining lower on days 4 and 8 post-inoculation. On day 4, Rlow-inoculated chickens exhibitedignificantly higher tracheal lesion scores and higher levels of lymphotactin, CXCL13, CXCL14, RANTES, MIP-1�, IL-1� and IFN-� mRNAompared to PBS-inoculated controls. The mRNA levels of these genes were also higher in Rlow-inoculated chickens that had moderate toevere tracheal lesion scores on day 8 post-inoculation. These results reflect the importance of lymphocyte and monocyte chemotactic factors

n the development of tracheal lesions in chickens inoculated with M. gallisepticum strain Rlow. Our data also suggest that M. gallisepticum

ay modulate the host response causing dramatic decreases in CCL20, IL-8 and IL-12 mRNA levels in GT5- or Rlow-inoculated chickens asarly as one day post-inoculation.

2007 Elsevier Ltd. All rights reserved.

lisepticu

cro

eywords: Vaccine; Mucosal; Live-attenuated; Chicken; Mycoplasma; Gal

. Introduction

Significant losses to the poultry industry from

ycoplasma gallisepticum infection occur primarily

ue to reduced egg production, hatchability and down-rading of carcasses [1]. M. gallisepticum infection in

∗ Corresponding author at: Department of Allied Health Sciences, CANR,oom 227 Koons Hall; Unit 2101, 358 Mansfield Road, University of Con-ecticut, Storrs, CT 06269-2101, United States. Tel.: +1 860 486 0028;ax: +1 860 486 4191.

E-mail address: Lawrence.Silbart@uconn.edu (L.K. Silbart).RL: http://ceh.uconn.edu (L.K. Silbart).

Mmeomiaca

264-410X/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2007.09.057

m; Gene expression profile; Infection

hickens causes inflammatory lesions in the respiratory andeproductive tracts involving trachea, air sacs, lung andviduct [2]. Studies examining early interactions between. gallisepticum and the chicken trachea indicate thatycoplasma adhere and colonize the surfaces of ciliated

pithelial cells [3]. This interaction results in the releasef mucus from goblet cells and a catarrhal exudate withucofibrinous plugs [4]. Various other cytopathic changes

nclude deciliation, rounding and exfoliation of ciliatednd non-ciliated epithelial cells followed by lysis, causingellular organelles and cilia to intermix with the mucus thatccumulates in the tracheal lumen [5]. Perhaps the most

8 accine

nothoir[

pamtiitccsohfiUcpa

scWtwRcGcloCtiwoiMfcer

2

2

c

wwp

2

LiaiwobstCc

2

AmftfvaoatAi

2

zmwcipw2tdTic

612 J. Mohammed et al. / V

otable pathological finding occurring after colonizationf the respiratory epithelium is an increase in mucosalhickness due to cellular infiltrates consisting mainly ofeterophils [5], macrophages and lymphocytes [6,7]. A vig-rous adaptive immune response accompanies inflammatorynfiltrates; however, M. gallisepticum is believed to evade thisesponse through phase variation of the VlhA gene family8–11].

Although it is widely recognized that the crux of theathogenesis of M. gallisepticum involves the recruitmentnd activation of large numbers of inflammatory cells inucosal tissues, the underlying mechanisms responsible for

his recruitment have not been described. Chemokines aremplicated in recruiting leukocytes to sites of mycoplasmanfection and cytokines in the activation and proliferation ofhese cells [12–14]. This has not been examined in detail inhickens, largely due to the lack of specific antibodies againsthicken chemokines and cytokines. However, the successfulequencing of the chicken genome [15] and the availabilityf genomic information on chemokines and cytokines [16]as provided the basis for examining gene expression pro-les of specific genes of interest, at least at the mRNA level.tilizing these genomic tools, several research groups have

haracterized the chemokine and cytokine genes involved inathogenic mechanisms of avian pathogens such as Eimeriand Salmonella species [17–21].

In the current study, differential mRNA expression ofeveral chemokines and cytokines was examined in tra-heal tissue following inoculation of specific pathogen-free

hite Leghorn chickens with either a pathogenic M. gallisep-icum strain, Rlow, or an attenuated strain, GT5. Strain GT5as originally produced by reconstitution of high-passagestrain, Rhigh (lacking the cytadhesin molecule GapA and

ytadherence-related molecule A, CrmA) with GapA [22].T5 was shown to be avirulent as well as protective in

hickens following challenge [23]. As anticipated, inocu-ation with the virulent strain resulted in the up-regulationf several key cytokine and chemokine genes, includingXCL13 (B cell-activating chemokine, BCA-1), lympho-

actin, and RANTES (CCL5), each of which may play anmportant role in the development of inflammatory lesionsithin the trachea. Conversely, the expression of severalther genes was dramatically down-regulated in chickensnoculated with both the virulent and attenuated strains of

. gallisepticum. These include CCL20, a chemoattractantor immature dendritic cells and effector/memory T and Bells, IL-12 and IL-8, and may reflect immunomodulatoryffects exerted by M. gallisepticum on the host immuneesponse.

. Materials and methods

.1. Strains and media

M. gallisepticum strains Rlow and GT5 and their growthonditions have been previously described [23]. All strains

ic1t

25 (2007) 8611–8621

ere cultured at 37 ◦C in Hayflick’s medium supplementedith 10% horse serum and 5% yeast extract, or on Hayflick’slates.

.2. Animals and immunization

Five-week-old, specific-pathogen-free, female Whiteeghorn chickens (SPAFAS, North Franklin, CT) were used

n this study. Non-medicated feed (Blue Seal, Waltham, MA)nd water was provided ad libitum. Chickens were separatednto three groups (N = 15) and were inoculated intratracheallyith 100 �l of either 1 × 107 cfu M. gallisepticum strain Rlowr 3 × 108 cfu M. gallisepticum strain GT5 or phosphate-uffered saline (PBS). Five animals from each group wereacrificed on days 1, 4 and 8 post-inoculation. All animal pro-ocols were approved in advance by the Institutional Animalare and Use Committee (IACUC), University of Connecti-ut, Storrs, CT.

.3. Necropsy

Necropsy was performed as previously described [7].nnular tissue samples of ∼5 mm width from the cranial,iddle, and caudal trachea were fixed in 10% neutral buffered

ormalin, routinely processed for paraffin embedding, sec-ioned at 4 �m, and stained with hematoxylin and eosinollowing standard histological protocols [24]. Each trans-erse sections was subsequently examined microscopicallynd inflammatory cell infiltrates, consisting of combinationsf heterophils, lymphocytes and macrophages, were scored inblinded fashion as described previously [7]. The remaining

racheal tissue was immersed in RNAlater solution (Ambion,ustin, TX), tightly capped and stored for subsequent RNA

solation.

.4. RNA isolation from tissues and cDNA synthesis

Total RNA was isolated from chicken trachea using TRI-ol reagent (Invitrogen, Carlsbad, CA) according to theanufacturer’s instructions. Homogenized tissue samplesere treated with TRIzol reagent followed by phenol-

hloroform phase separation. RNA was precipitated usingsopropyl alcohol, washed with 75% ethanol, dried and resus-ended in 50 �l of diethylpyrocarbonate (DEPC)-treatedater. RNA was quantified by measuring absorbance at60 nm on a spectrophotometer, and purity assessed usinghe 260:280 nm ratio. The quality of each RNA sample wasetermined by electrophoresis on formaldehyde agarose gels.otal RNA was treated with DNase I to remove contaminat-

ng DNA using the DNA-freeTM kit (Ambion, Austin, TX).DNA synthesis of the RNA samples was performed using the

ScriptTM cDNA synthesis kit (Bio-Rad Laboratories, Her-ules, CA) as per the manufacturer’s recommendation using�g of DNase-treated RNA from individual samples during

he reaction.

J. Mohammed et al. / Vaccine 25 (2007) 8611–8621 8613

Table 1Primers to various chicken chemokine and cytokine genes. Primers were designed using LightCycler Probe Design Software 2.0 (Roche Applied Science,Indianapolis, IN)

Gene Accession numbera Primer sequence location Amplicon size

GAPDH K01458 FP: 199-219, RP: 330-351 153 bpLymphotactin AF006742 FP: 250-277, RP: 384-409 160 bpCXCL13/BCA1 XM 420474 FP: 267-284, RP: 403-423 157 bpIL-16 AJ508678 FP: 1247-1263, RP: 1395-1413 167 bpRANTES AI982103 FP: 117-139, RP: 252-268 152 bpCCL4/MIP-1� AJ243034 FP: 177-193, RP: 304-330 154 bpCXCL14 NM 204712 FP: 860-876, RP: 1023-1044 185 bpCCL20/MIP-3� NM 204438 FP: 199-216, RP: 354-377 179 bpIL-1� Y15006 FP: 550-568, RP: 740-757 208 bpIL-2 AF000631 FP: 275-304, RP: 410-431 157 bpIL-6 AJ309540 FP: 439-459, RP: 572-590 152 bpIL-8 AJ009800 FP: 177-200, RP: 336-354 178 bpIL-12p40 AY262752 FP: 227-245, RP: 357-378 152 bpIL18 AJ277865 FP: 199-216, RP: 355-375 177 bpTNF-� AY765397 FP: 368-385, RP: 523-541 174 bpTGF-�4 M31160 FP: 881-902, RP: 1048-1066 186 bpIFN-� AF424744 FP: 120-146, RP: 271-291 172 bp

FP: forward primer, RP: reverse primer.

2

uApCGcomSawpAf(aacfitcrptdrdose

gsmrgc

2

fwmwotBS

3

3

i(em

a GenBank accession numbers.

.5. Quantification of cDNA by real-time PCR

Primers specific for chicken genes (Table 1) were designedsing the LightCycler probe design software 2.0 (Rochepplied Science, Indianapolis, IN). Amplification of cDNAsrepared from tracheal tissues was measured using a Light-ycler 2.0 instrument (Roche Diagnostics, Mannheim,ermany). Real-time PCR was performed using LightCy-

ler capillaries with a total volume of 20 �l containing 50 ngf cDNA, 10 �M forward and reverse primers and masterix prepared from LightCycler FastStart DNA MasterPLUS

YBR Green I kit (Roche Diagnostic Corporation, Indi-napolis, IN). Annealing temperatures for each primer setere standardized and amplification of specific products waserformed in keeping with the manufacturer’s instructions.mplification was performed by incubating samples at 95 ◦C

or 5 min, followed by 45 cycles of 95 ◦C for 10 s, 58–62 ◦Cbased on specific primer annealing temperatures) for 10 snd 72 ◦C for 10 s. The fluorescent intensity was measuredt the end of the elongation phase of each cycle. A meltingurve analysis was then performed to confirm specific ampli-cation of a single product by incrementally increasing the

emperature at a rate of 0.1 ◦C/s from 65 to 95 ◦C. Ampli-on size was confirmed on a 1.5% agarose gel with lengthsanging from 152–208 bp. A standard curve for individualrimer sets was prepared using serial cDNA dilutions plot-ed against the crossing point (CP, threshold cycle), whichenotes the cycle at which the fluorescence of the sampleises above background levels, with the slope of the stan-

ard curve indicating primer efficiency. The absolute amountf a particular mRNA sample was calculated based on thetandard curve. To calculate fold-change over control chick-ns treated with PBS, the ratio of concentration of target

mOio

ene to a reference gene (housekeeping gene, GAPDH) inamples from PBS-inoculated controls (calibrator) was nor-alized. The normalized ratio was then compared to the

atio of the concentration of the target gene to the referenceene in individual samples from GT5- and Rlow-inoculatedhickens.

.6. Statistical analysis

Comparison between the experimental groups was per-ormed using an unpaired Student’s t-test. Tukey’s testas done in instances where multiple comparisons wereade following ANOVA. Log transformation was performedhen the data was not normally distributed. A p valuef <0.05 was considered statistically significant. All sta-istical analyses were performed using the InStat Instantiostatistics Version 3.05 software program (GraphPad,an Diego, CA).

. Results

.1. Development of lesions in Rlow infected chickens

Tracheal lesion scores were significantly higher (p < 0.01)n Rlow-inoculated chickens on day 4 post-inoculationTable 2). Rlow-inoculated chickens had infiltrates of het-rophils, lymphocytes and macrophages in the trachealucosa, while GT5- and PBS-inoculated chickens had

inimal lesions, with negligible heterophilic infiltration.n day 8, two of five Rlow-inoculated chickens exhib-

ted moderate lesions, represented by a lesion scoref 2.0.

8614 J. Mohammed et al. / Vaccine

Table 2Tracheal lesion scores

Days post-inoculation PBS GT5 Rlow

1 0.5 ± 0.2 0.4 ± 0.1 0.6 ± 0.24 0.4 ± 0.1 0.3 ± 0.1 1.4 ± 0.3**

8 0.4 ± 0.2 0.9 ± 0.1 1.1 ± 0.4

Chickens were inoculated intratracheally with 1 × 107 cfu M. gallisepticumstrain Rlow or 3 × 108 cfu strain GT5 or PBS, and tracheal lesions werees

3p

sstermcGtHRtDs(tfasc

1cGIBRcwgfaac

3p

t

RioiFa(atRRt(o

fir(vwe(

3p

RfmitmsIill

RoGins

4

l

valuated at days 1, 4 and 8 post-inoculation. Tracheal lesion scores wereignificantly higher in Rlow-inoculated chickens on day 4.** p < 0.01 (n = 5).

.2. Chemokine and cytokine gene profile on day 1ost-inoculation

Total RNA from chicken trachea was isolated and cDNAynthesized. Real time RT-PCR was performed on cDNAamples using specific chemokine and cytokine gene primerso quantitatively measure the fold-difference in mRNA lev-ls between the Rlow- and GT5-inoculated chickens withespect to the control chickens. Lymphotactin and CXCL13RNA levels were three-fold higher in Rlow-inoculated

hickens while MIP-1� and CXCL13 mRNA levels in theT5-inoculated chickens were 7- and 3.5-fold higher, respec-

ively, than the PBS-inoculated control chickens (Fig. 1A).owever, CCL20 and IL-8 mRNA were 10-fold lower inlow-inoculated chickens and 3.5- and 4-fold lower, respec-

ively, in GT5-inoculated chickens than in controls (Fig. 1A).own-regulation of IL-8 in Rlow-inoculated chickens was

ignificant when compared to GT5-inoculated chickensp < 0.05, Fig. 1A). A comparison of normalized ratios of thearget and reference (GAPDH) genes revealed significant dif-erences in CCL20 mRNA levels between controls and Rlow-nd GT5-inoculated chickens (Fig. 1B). IL-8 mRNA wasignificantly different between controls and Rlow-inoculatedhickens (Fig. 1B).

The mRNA levels of IL-2, IL-6 and TNF-� were 2.0-,.8- and 1.8-fold higher, respectively, in Rlow-inoculatedhickens compared to PBS-inoculated controls (Fig. 2A).T5-inoculated chickens had >2-fold higher IL-6 and

FN-� mRNA levels compared to controls (Fig. 2A and). The mRNA levels of IL-18 and TGF-� in GT5- andlow-inoculated chickens did not vary with respect to theontrols (Fig. 2A). However, IL-1� and IL-12p40 mRNAere down-regulated in chickens inoculated with either M.allisepticum strains GT5 or Rlow (Fig. 2A and B). A two-old decrease in IL-1� mRNA in Rlow-inoculated chickensnd a 20-fold decrease in IL-12p40 mRNA (p < 0.001) werepparent in both GT5- or Rlow-inoculated chickens whenompared to PBS-inoculated controls (Fig. 2B).

.3. Chemokine and cytokine gene profile on day 4

ost-inoculation

There was up-regulation of chemokine genes, lympho-actin (four-fold), CXCL13 (17-fold), CXCL14 (five-fold),

ttoc

25 (2007) 8611–8621

ANTES (two-fold) and MIP-1� (two-fold) in chickensnfected with the Rlow strain (Fig. 3A). The mRNA levelsf lymphotactin and CXCL13 gene were significantly highern Rlow-inoculated chickens compared to controls (p < 0.05,ig. 3B). GT5-inoculated chickens had moderate increases inbsolute levels of lymphotactin and CXCL13 mRNA levels1.8- and 2-fold, respectively) compared to controls (Fig. 3And B). IL-16, a cytokine with lymphocyte chemoattrac-ant function, was moderately up-regulated (>1.5-fold) inlow-inoculated chickens. The level of IL-8 mRNA in thelow-inoculated group remained similar to that of the con-

rols, while it was five-fold lower in GT5-inoculated chickensp < 0.05, Fig. 3B). CCL20 mRNA was 10-fold lower in GT5r Rlow-inoculated chickens (p < 0.01, Fig. 3B).

Chickens within the Rlow-inoculated group had three-old higher IL-1� and IFN-� mRNA levels and a moderatencrease in IL-2, IL-6 and IL-18 mRNA (2-, 1.8- and 1.7-fold,espectively) when compared to PBS-inoculated controlsFig. 4A). The mRNA levels of TNF-� and TGF-�4 did notary between the three groups (Fig. 4A). IL-12p40 mRNAas down-regulated by four-fold in Rlow inoculated chick-

ns (p < 0.05), but by 10-fold in GT5-inoculated chickensp < 0.001; Fig. 4).

.4. Chemokine and cytokine gene profile on day 8ost-inoculation

Chickens with a tracheal lesion score of 2.0 within thelow-inoculated group had more than 2.5-fold higher mRNA

or lymphotactin, RANTES and MIP-1� and a 45-fold higherRNA for CXCL13 gene than controls (Fig. 5A). Chickens

noculated with GT5 had a modest increase in lympho-actin mRNA (1.7-fold) and a four-fold decrease in CXCL13

RNA while mRNA levels of other chemokines remainedimilar to that of controls (Fig. 5A and B). CCL20 andL-8 mRNA remained down-regulated in both GT5- or Rlow-noculated chickens (Fig. 5A). CCL20 mRNA was 3.5-foldower in Rlow-inoculated chickens (p < 0.05) and two-foldower in GT5-inoculated chickens (Fig. 5B).

IL-1� mRNA was elevated by more than two-fold inlow-inoculated chickens exhibiting a tracheal lesion scoref 2.0 (Fig. 6A). IL-12p40 mRNA was down-regulated inT5- as well as Rlow-inoculated chickens, although signif-

cant differences between the three groups (Fig. 6B) wereot detected. The mRNA levels of other cytokines remainedimilar between the three groups.

. Discussion

Infection of chickens with M. gallisepticum strain Rloweads to inflammation of the trachea, air sacs and lungs while

he GT5 strain, which is attenuated in its ability to cytadhereo tracheal epithelial cells, does not cause significant pathol-gy [22,23,25]. Virulent M. gallisepticum strains adhere tohicken tracheal epithelial cells, initiating a cascade of cyto-

J. Mohammed et al. / Vaccine 25 (2007) 8611–8621 8615

Fig. 1. (A) Fold-change of chemokine mRNA in GT5- or Rlow-inoculated chickens over the controls in tracheal tissue on day 1. (B) Normalized ratios ofchemokine mRNA between the three groups. The average normalized ratio of PBS-inoculated chickens was adjusted to 1. CCL20 and IL-8 mRNA levelsw ars repr*

phmsmrtcmcre

olawaBi

ere significantly down regulated in GT5- or Rlow-inoculated chickens. B**p < 0.001.

athic effects with subsequent sub-epithelial infiltration byeterophils, macrophages and lymphocytes into respiratoryucosa [5]. Infiltration and activation of lymphocytes at the

ite of infection is critical to the pathogenesis of chronicycoplasmal respiratory diseases [6,7]. The goal of the cur-

ent study was to investigate the molecular events that leado inflammation by elucidating some of the cytokines andhemokines that might play a role in the progression of

ycoplasmal lesions by mediating infiltration of mononu-

lear cells to the site of inflammation (chemokines) andegulating the influx, activation and proliferation of immuneffector cells (cytokines).

tcvo

esent the mean ± standard error of the mean, n = 5. *p < 0.05, **p < 0.01,

Within 24 h of Rlow-inoculation, dramatic up-regulationf lymphotactin, a chemokine known to play a role inymphocyte chemotaxis [26], and CXCL13/BCA1, a B-cellctivating chemokine, was observed. Chickens inoculatedith a 30-fold higher dose of the attenuated, GT5 strain

lso showed an increase in CXCL13 mRNA (Fig. 1A and). The transcript levels of these chemokines were elevated

n Rlow-inoculated chickens through day 4, with lympho-

actin mRNA levels significantly higher when compared toontrols (Fig. 3B). These data are consistent with our pre-ious studies indicating substantial B- and T-cell infiltrationf the trachea in Rlow-inoculated chickens [7,23]. There was

8616 J. Mohammed et al. / Vaccine 25 (2007) 8611–8621

F hickenc f PBS-w e mean

aimRtmamtnR21c

Bl

kvtgmom

ig. 2. (A) Fold-change of cytokine mRNA in GT5- or Rlow-inoculated cytokine mRNA between the three groups. The average normalized ratio oere down-regulated in GT5- or Rlow-inoculated chickens. Bars represent th

lso up-regulation of CXCL14, RANTES and MIP-1� genesn Rlow-inoculated chickens. Up-regulation of chemokine

RNA correlated with the severity of tracheal lesions inlow-inoculated chickens at day 4 (Table 2), in keeping with

he role of these chemokines in the inflammatory process. Aodest increase in mRNA for these chemokines, leading tomild lymphocytic infiltration in GT5-inoculated chickens,ay be part of a more controlled adaptive immune response

hat correlates with the establishment of prophylactic immu-ity as demonstrated in earlier studies [7,23]. By day 8,

low-inoculated chickens with lesion scores of 2.0 showed a.5-fold up-regulation of lymphotactin, RANTES and MIP-� mRNA and 45-fold up-regulation of CXCL13 mRNAompared to PBS-inoculated control chickens (Fig. 5A and

iolg

s over the controls in tracheal tissue on day 1. (B) Normalized ratios ofinoculated chickens was adjusted to 1. IL-1� and IL-12p40 mRNA levels± standard error of the mean, n = 5. ***p < 0.001.

). GT5-inoculated chickens exhibited a mild increase inymphotactin mRNA (1.7-fold) compared to controls.

Taken together, it is clear that up-regulation of severaley chemokine genes occurs in chickens inoculated with airulent strain of M. gallisepticum. These results are consis-ent with the theory that up-regulation of specific chemokineenes plays an important role in leukocyte recruitment intoucosal tissues, a key component in the immunopathol-

gy associated with mycoplasmal disease. Furthermore, theinimal to mild up-regulation of these genes in chickens

noculated with an attenuated strain suggests that high levelsf leukocyte infiltration are unnecessary to confer prophy-actic immunity. Such information can provide importantuidance when assessing panels of attenuated mutants as

J. Mohammed et al. / Vaccine 25 (2007) 8611–8621 8617

Fig. 3. (A) Fold-change of chemokine mRNA in GT5- or Rlow-inoculated chickens over the controls in tracheal tissue on day 4. (B) Normalized ratios ofchemokine mRNA between the three groups. The average normalized ratio of PBS-inoculated chickens was adjusted to 1. One chicken with an 88-fold increase inmRNA levels of CXCL13 in the GT5-inoculated group was considered to be an outlier and was excluded from the analysis (n = 4). Lymphotactin and CXCL13mRNA levels were significantly up-regulated in Rlow-inoculated chickens. CCL20 and IL-8 mRNA levels are down regulated in GT5- or Rlow-inoculatedchickens. Bars represent the mean ± standard error of the mean, n = 5. *p < 0.05, **p < 0.01.

8618 J. Mohammed et al. / Vaccine

Fig. 4. (A) Fold-change of cytokine mRNA in GT5- or Rlow-inoculatedchickens over the controls in tracheal tissue on day 4. (B) Normalizedratios of cytokine mRNA between the three groups. The average normal-ized ratio of PBS-inoculated chickens was adjusted to 1. IL-1� and IFN-�mRNA levels were up-regulated in chickens with higher lesion scores withinthe Rlow-inoculated group. IL-12p40 mRNA levels remained significantlydown-regulated in GT5- or Rlow-inoculated chickens. Bars represent themean ± standard error of the mean, n = 5. *p < 0.05, ***p < 0.001.

pmtaatopPstutcio

(hshripgwoSacRt[spt

acdccmtogn“bImRaRhr

25 (2007) 8611–8621

otential vaccine candidates, especially when the specificutation is known (e.g. isogenic mutants generated through

ransposon mutagenesis [27]). In these instances, a directssociation may be made between the specific deleted genend the ensuing host immune response. In the current study,he GT5 strain induced a modest increase in mRNA levelsf chemokines with lymphocyte chemotactic activity (lym-hotactin and CXCL13) at days 1 and 4 post-inoculation.rotection following challenge in these chickens has beenhown to be mediated by induction of humoral components ofhe adaptive immune response [7]. Such an approach could besed to pick and choose genes to be deleted or over-expressed,hereby allowing the “rational design” of attenuated vac-ines tailored to induce only favorable elements of the hostmmune response, while avoiding those that induce pathologyr immune dysregulation.

Mycoplasmal lipid-associated membrane proteinsLAMP’s) have well-known immunomodulatory effects onost cells [13,14,28]. Unfortunately, the molecules respon-ible for the immunomodulatory effect of M. gallisepticumave not yet been identified. There was a significant down-egulation of CCL20, IL-8 and IL-12p40 genes in chickensnoculated with GT5 and Rlow strains as early as day 1ost-inoculation (Fig. 1A and B). Down-regulation of theseenes in chickens inoculated with the attenuated GT5 strain,hich is derived from strain R [22], suggests a possible rolef mycoplasmal cellular components in mediating this effect.uppression of IL-8 upon M. gallisepticum treatment haslso been reported in vitro in Marek’s disease lymphoblastoidell line (MSB-1) [29]. Inoculation of chickens with strainlow but not GT5 results in heterophilic infiltration of the

racheal mucosa, contributing to the severity of the lesion23]. This data suggests that unlike Salmonella entericaerovar typhimurium infection [18], IL-8 probably does notlay a dominant role in heterophilic infiltration within therachea of Rlow-inoculated chickens, at least at day 1.

Mycoplasmal lipoproteins have also been shown to playrole in macrophage activation [12,30,31]. M. gallisepticumauses infiltration of macrophages into tracheal lesions atay 1 post-inoculation as evidenced by immunohistochemi-al staining of tracheal tissue harvested from Rlow-inoculatedhickens (unpublished results). Since macrophages are aajor source of IL-12, the results presented here are consis-

ent with the notion that part of the immunomodulatory effectf M. gallisepticum is mediated by down-regulation of theene encoding IL-12 in this cell population. At the same time,o significant differences were observed in mRNA levels ofclassical” pro-inflammatory cytokines primarily producedy activated monocytes/macrophages, such as TNF-� andL-6; the exception was IL-1�, which was up-regulated byore than two-fold in chickens with higher lesions withinlow-inoculated group on day 4 and 8. IFN-� mRNA was

lso up-regulated in chickens with higher lesion scores inlow-inoculated chickens on days 4 and 8. A significantlyigher infiltration of CD4+ and CD8+ T-cells was previouslyeported during acute stages of M. gallisepticum infection of

J. Mohammed et al. / Vaccine 25 (2007) 8611–8621 8619

Fig. 5. (A) Fold-change of chemokine mRNA in GT5- or Rlow-inoculated chickens over the controls in tracheal tissue on day 8. (B) Normalized ratios ofchemokine mRNA between the three groups. The average normalized ratio of PBS-inoculated chickens was adjusted to 1. One chicken with a 28-fold increasein mRNA levels of CXCL13 in GT5-inoculated group was considered to be an outlier and was excluded from the analysis (n = 4). Lymphotactin, CXCL13 andRANTES mRNA levels were up- regulated in Rlow-inoculated chickens with higher tracheal lesion scores. CCL20 and IL-8 mRNA levels were down-regulatedin GT5- or Rlow-inoculated chickens. Bars represent the mean ± standard error of the mean, n = 5. *p < 0.05.

8620 J. Mohammed et al. / Vaccine 25 (2007) 8611–8621

Fig. 6. (A) Fold-change of cytokine mRNA in GT5- or Rlow-inoculated chickens over the controls in tracheal tissue on day 8. (B) Normalized ratios of cytokinem oculatedc p40 mRB

ccmT1totlotcTtt“

alaweais

ttasAudcc

A

dWtBg

RNA between the three groups. The average normalized ratio of PBS-inhickens with higher lesion scores within the Rlow-inoculated group. IL-12ars represent the mean ± standard error of the mean, n = 5.

hickens [7]. Although Th1 cells, CD8+ T cells and/or NKells are the major source of IFN-�, the induction of IFN-�RNA in these chickens appears to be independent of IL-12.he mRNA expression levels of Th2 cytokines, IL-4 and IL-3, within the tracheal tissue could not be measured as theseranscripts were present at very low levels in all three groupsf chickens. In addition, we had only partial success in quan-ifying IL-10 mRNA levels in tracheal tissue due to the veryow abundance of this transcript. Although somewhat equiv-cal, our results indicate a moderate up-regulation of IL-10ranscripts in chickens inoculated with Rlow (data not shown),onsistent with other in vitro and in vivo studies [32–34].aken together, it appears that macrophages infiltrating the

racheal lesion subsequent to Rlow-inoculation down-regulatehe IL-12 gene and do not produce high levels of severaltraditional” pro-inflammatory cytokines.

The interplay of chemokines and cytokines is a complexnd dynamic process in the development of inflammatoryesions. In this study, transcript levels of selected chemokinend cytokine genes from tracheal mucosa upon inoculationith virulent and attenuated M. gallisepticum strains in chick-

ns were identified and compared. Modulations in chemokinend cytokine expression profiles, such as those reportedn this study, may alter the types, numbers and activationtatus of cells that subsequently comprise the inflamma-

R

chickens was adjusted to 1. The IL-1� mRNA level was up-regulated inNA levels remained down-regulated in GT5- or Rlow-inoculated chickens.

ory tracheal infiltrate. Future studies aimed at identifyinghe cell types that are major sources of these chemokinend cytokine genes will provide important insights into theequence of events that mediate the pathogenic process.dditional studies are needed to elucidate the molec-lar mechanisms of mycoplasmal immunomodulation toetermine critical molecules responsible for triggering spe-ific cellular responses (e.g. signaling pathways) in thehicken.

cknowledgements

We thank Martha Gladd and Vickie Weidig for assistanceuring necropsy procedures and Ione Jackman and Deniseoodward for tissue processing and histological prepara-

ions. We also thank Kevin Kavanagh, Roger Zemek and Donessette for animal care. This work was supported by USDArant 58-1940-0-007.

eferences

[1] Ley DH. Mycoplasma gallisepticum Infection. Ames: Iowa State Press;2003.

accine

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

J. Mohammed et al. / V

[2] Nunoya T, Tajima M, Yagahashi T, Sannai S. Natural case of salpingitisapparently caused by Mycoplasma gallisepticum in chickens. AvianPathol 1997;26:391–8.

[3] Tajima M, Nunoya T, Yagihashi T. An ultrastructural study on theinteraction of Mycoplasma gallisepticum with the chicken trachealepithelium. Am J Vet Res 1979;40(7):1009–14.

[4] Hod I, Yegana Y, Herz A, Levinsohn S. Early detection of trachealdamage in chickens by scanning electron microscopy. Avian Dis1982;26(2):450–7.

[5] Dykstra MJ, Levisohn S, Fletcher OJ, Kleven SH. Evaluation ofcytopathologic changes induced in chicken tracheal epithelium byMycoplasma gallisepticum in vivo and in vitro. Am J Vet Res1985;46(1):116–22.

[6] Gaunson JE, Philip CJ, Whithear KG, Browning GF. Lymphocytic infil-tration in the chicken trachea in response to Mycoplasma gallisepticuminfection. Microbiology 2000;146(Pt 5):1223–9.

[7] Javed MA, Frasca Jr S, Rood D, Cecchini K, Gladd M, GearySJ, et al. Correlates of immune protection in chickens vaccinatedwith Mycoplasma gallisepticum strain GT5 following challengewith pathogenic M. gallisepticum strain Rlow. Infect Immun2005;73(9):5410–9.

[8] Glew MD, Browning GF, Markham PF, Walker ID. pMGA phe-notypic variation in Mycoplasma gallisepticum occurs in vivo andis mediated by trinucleotide repeat length variation. Infect Immun2000;68(10):6027–33.

[9] Levisohn S, Rosengarten R, Yogev D. In vivo variation of Mycoplasmagallisepticum antigen expression in experimentally infected chickens.Vet Microbiol 1995;45(2–3):219–31.

10] Markham PF, Glew MD, Browning GF, Whithear KG, Walker ID.Expression of two members of the pMGA gene family of Mycoplasmagallisepticum oscillates and is influenced by pMGA-specific antibodies.Infect Immun 1998;66(6):2845–53.

11] Papazisi L, Gorton TS, Kutish G, Markham PF, Browning GF, NguyenDK, et al. The complete genome sequence of the avian pathogenMycoplasma gallisepticum strain Rlow. Microbiology 2003;149(Pt9):2307–16.

12] Razin S, Yogev D, Naot Y. Molecular biology and pathogenicity ofmycoplasmas. Microbiol Mol Biol Rev 1998;62(4):1094–156.

13] Rottem S. Interaction of mycoplasmas with host cells. Physiol Rev2003;83(2):417–32.

14] Chambaud I, Wroblewski H, Blanchard A. Interactions betweenmycoplasma lipoproteins and the host immune system. Trends Micro-biol 1999;7(12):493–9.

15] Hillier LW, Miller W, Birney E, Warren W, Hardison RC, PontingCP, et al. Sequence and comparative analysis of the chickengenome provide unique perspectives on vertebrate evolution. Nature2004;432(7018):695–716.

16] Kaiser P, Poh TY, Rothwell L, Avery S, Balu S, Pathania US, et al. Agenomic analysis of chicken cytokines and chemokines. J InterferonCytokine Res 2005;25(8):467–84.

17] Smith CK, Kaiser P, Rothwell L, Humphrey T, Barrow PA, Jones MA.Campylobacter jejuni-induced cytokine responses in avian cells. Infect

Immun 2005;73(4):2094–100.

18] Withanage GS, Kaiser P, Wigley P, Powers C, Mastroeni P, Brooks H,et al. Rapid expression of chemokines and proinflammatory cytokinesin newly hatched chickens infected with Salmonella enterica serovartyphimurium. Infect Immun 2004;72(4):2152–9.

25 (2007) 8611–8621 8621

19] Withanage GS, Wigley P, Kaiser P, Mastroeni P, Brooks H, PowersC, et al. Cytokine and chemokine responses associated with clearanceof a primary Salmonella enterica serovar Typhimurium infection inthe chicken and in protective immunity to rechallenge. Infect Immun2005;73(8):5173–82.

20] Hong YH, Lillehoj HS, Lee SH, Dalloul RA, Lillehoj EP. Analysis ofchicken cytokine and chemokine gene expression following Eimeriaacervulina and Eimeria tenella infections. Vet Immunol Immunopathol2006.

21] Laurent F, Mancassola R, Lacroix S, Menezes R, Naciri M. Analysisof chicken mucosal immune response to Eimeria tenella and Eime-ria maxima infection by quantitative reverse transcription-PCR. InfectImmun 2001;69(4):2527–34.

22] Papazisi L, Troy KE, Gorton TS, Liao X, Geary SJ. Analysisof cytadherence-deficient, GapA-negative Mycoplasma gallisepticumstrain R. Infect Immun 2000;68(12):6643–9.

23] Papazisi L, Silbart LK, Frasca S, Rood D, Liao X, Gladd M, et al. Amodified live Mycoplasma gallisepticum vaccine to protect chickensfrom respiratory disease. Vaccine 2002;20(31–32):3709–19.

24] Sheehan DC HB. Nuclear and cytoplasmic stains. In: Theory and prac-tice of histotechnology. Columbus, OH: Battelle Press, 1980.

25] Levisohn S, Dykstra MJ, Lin MY, Kleven SH. Comparison of in vivoand in vitro methods for pathogenicity evaluation for Mycoplasmagallisepticum in respiratory infection. Avian Pathol 1986;15:233–46.

26] Kelner GS, Kennedy J, Bacon KB, Kleyensteuber S, Largaespada DA,Jenkins NA, et al. Lymphotactin: a cytokine that represents a new classof chemokine. Science 1994;266(5189):1395–9.

27] Hudson P, Gorton TS, Papazisi L, Cecchini K, Frasca Jr S, Geary SJ.Identification of a virulence-associated determinant, dihydrolipoamidedehydrogenase (lpd), in Mycoplasma gallisepticum through in vivoscreening of transposon mutants. Infect Immun 2006;74(2):931–9.

28] You XX, Zeng YH, Wu YM. Interactions between mycoplasma lipid-associated membrane proteins and the host cells. J Zhejiang Univ Sci2006;7(5):342–50.

29] Lam KM. Mycoplasma gallisepticum-induced alterations in cytokinegenes in chicken cells and embryos. Avian Dis 2004;48(1):215–9.

30] Muhlradt PF, Kiess M, Meyer H, Sussmuth R, Jung G. Isolation, struc-ture elucidation, and synthesis of a macrophage stimulatory lipopeptidefrom Mycoplasma fermentans acting at picomolar concentration. J ExpMed 1997;185(11):1951–8.

31] Muhlradt PF, Kiess M, Meyer H, Sussmuth R, Jung G. Structureand specific activity of macrophage-stimulating lipopeptides fromMycoplasma hyorhinis. Infect Immun 1998;66(10):4804–10.

32] Rawadi G, Roman-Roman S, Castedo M, Dutilleul V, Susin S,Marchetti P, et al. Effects of Mycoplasma fermentans on themyelomonocytic lineage. Different molecular entities with cytokine-inducing and cytocidal potential. J Immunol 1996;156(2):670–8.

33] Pietsch K, Ehlers S, Jacobs E. Cytokine gene expression in thelungs of BALB/c mice during primary and secondary intranasal infec-tion with Mycoplasma pneumoniae. Microbiology (Reading, England)1994;140(Pt 8):2043–8.

34] Sun X, Jones HP, Hodge LM, Simecka JW. Cytokine and chemokine

transcription profile during Mycoplasma pulmonis infection in sus-ceptible and resistant strains of mice: macrophage inflammatoryprotein 1beta (CCL4) and monocyte chemoattractant protein 2 (CCL8)and accumulation of CCR5+ Th cells. Infect Immun 2006;74(10):5943–54.