of June 20, 2018.This information is current as

Severity of Lyme ArthritisPathways Associated with Differential Gene Expression Profiling Reveals Unique

and Janis J. WeisWooten, James F. Zachary, John H. Weis, Robert B. Weiss Hillary Crandall, Diane M. Dunn, Ying Ma, R. Mark

http://www.jimmunol.org/content/177/11/7930doi: 10.4049/jimmunol.177.11.7930

2006; 177:7930-7942; ;J Immunol 

MaterialSupplementary http://www.jimmunol.org/content/suppl/2006/11/16/177.11.7930.DC1

Referenceshttp://www.jimmunol.org/content/177/11/7930.full#ref-list-1

, 29 of which you can access for free at: cites 58 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2006 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on June 20, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

Gene Expression Profiling Reveals Unique PathwaysAssociated with Differential Severity of Lyme Arthritis1

Hillary Crandall,* Diane M. Dunn,† Ying Ma,* R. Mark Wooten,‡ James F. Zachary,§

John H. Weis,* Robert B. Weiss,† and Janis J. Weis2*

The murine model of Lyme disease provides a unique opportunity to study the localized host response to similar stimulus, Borreliaburgdorferi, in the joints of mice destined to develop severe arthritis (C3H) or mild disease (C57BL/6). Pathways associated withthe response to infection and the development of Lyme arthritis were identified by global gene expression patterns using oligo-nucleotide microarrays. A robust induction of IFN-responsive genes was observed in severely arthritic C3H mice at 1 wk ofinfection, which was absent from mildly arthritic C57BL/6 mice. In contrast, infected C57BL/6 mice displayed a novel expressionprofile characterized by genes involved in epidermal differentiation and wound repair, which were decreased in the joints of C3Hmice. These expression patterns were associated with disease state rather than inherent differences between C3H and C57BL/6mice, because C57BL/6-IL-10�/� mice infected with B. burgdorferi develop more severe arthritis than C57BL/6 mice and displayedan early gene expression profile similar to C3H mice. Gene expression profiles at 2 and 4 wk postinfection revealed a commonresponse of all strains that was likely to be important for the host defense to B. burgdorferi and mediated by NF-�B-dependentsignaling. The gene expression profiles identified in this study add to the current understanding of the host response to B.burgdorferi and identify two novel pathways that may be involved in regulating the severity of Lyme arthritis. The Journal ofImmunology, 2006, 177: 7930–7942.

L yme disease is a multisystem disorder caused by infectionwith the tick-borne spirochete Borrelia burgdorferi (1).Signs of infection include a bull’s-eye rash at the site of

the tick bite, termed erythema migrans, followed by disseminationof bacteria to various tissues resulting in neurological abnormali-ties, myocarditis, and arthritis (2). Lyme arthritis occurs in �60%of individuals not treated with antibiotics at the time of the tickbite, is associated with the presence of B. burgdorferi in joint tis-sue, and resolves with successful antibiotic treatment (3, 4). Asmall percentage of individuals with subacute arthritis progress toa chronic treatment-resistant arthritis that is no longer associatedwith bacteria in joint tissue and is postulated to be autoimmune-mediated (5, 6).

Infection-associated Lyme arthritis has been studied in themouse, where arthritis develops 3–4 wk following intradermal in-oculation and is histopathologically similar to Lyme arthritis inhumans (7, 8). The severity of arthritis is genetically regulated,with C3H mice developing severe arthritis whereas C57BL/6 mice

develop mild to moderate disease (8, 9). Although this differenceis not dependent on MHC alleles, it has been linked to quantitativetrait loci (QTL)3 on chromosomes 1, 4, 5, 11, and 12 (10, 11).Interestingly, infected C3H and C57BL/6 mice harbor similarnumbers of bacteria in joint tissues, indicating that differences inarthritis severity are not due to differences in host defense, butrather reflect different abilities to regulate the localized inflamma-tory response (9). B. burgdorferi lipoprotein interaction with TLR2results in the production of proinflammatory cytokines and che-mokines, several of which have been implicated in modulating thedevelopment of arthritis (12–15). Furthermore, C57BL/6 micelacking the potent anti-inflammatory cytokine IL-10 (C57BL/6 IL-10�/�) develop more severe arthritis than wild-type C57BL/6mice while more effectively controlling bacterial growth (16–18).

The mouse model of Lyme arthritis provides a unique opportu-nity to study contrasting responses to similar bacterial stimuli inmice developing severe or mild arthritis (8, 9, 19). Localized re-sponses to B. burgdorferi were assessed by global gene expressionanalysis in whole joint tissue from C3H, C57BL/6, and C57BL/6-IL-10�/� mice during the progression of disease development.This analysis revealed the activation of two unexpected and diver-gent pathways in response to infection in mice destined to developarthritis of different severities, and suggested that an early com-mitment to a gene expression phenotype in infected joint tissuecould determine the severity of subsequent Lyme arthritis.

Materials and MethodsMice

Female C3H/HeNCr (C3H) and C57BL/6NCr (C57BL/6) mice were ob-tained from the National Cancer Institute, whereas female B6.129P2-IL-10tm1Cgn/J (C57BL/6-IL-10�/�) mice on the closely related C57BL/6Jmouse were obtained from The Jackson Laboratory. Mice were housed in

*Department of Pathology, and †Department of Human Genetics, University of Utah,Salt Lake City, Utah 84112; ‡Department of Medical Microbiology and Immunology,Medical University of Ohio, Toledo, Ohio 43614; and §Department of Pathobiology,University of Illinois at Urbana-Champaign, Urbana, Illinois 61802

Received for publication June 13, 2006. Accepted for publication September12, 2006.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by National Institutes of Health Grants (AI-32223 toJ.J.W. and J.F.Z.; AR-43521 to J.J.W.; AI-24158 to J.H.W.; and HL-072903 toR.B.W.), National Institutes of Health/National Institute of Diabetes and Digestiveand Kidney Diseases Training Grant (DK07115 to H.C.), the American Heart Asso-ciation (Grant 0335148N to R.M.W.), and by funds from Associated Regional Uni-versity Pathologists.2 Address correspondence and reprint requests to Dr. Janis J. Weis, Department ofPathology, University of Utah School of Medicine, 15 North Medical Drive East,Room 2100, Salt Lake City, Utah 84112-5650. E-mail address: janis.weis@path.utah.edu

3 Abbreviations used in this paper: QTL, quantitative trait loci; SAM, significanceanalysis of microarray; F, forward; R, reverse; KO, knockout; NC, not changed;MMP, matrix metalloproteinase; TIMP, tissue inhibitor of MMP.

The Journal of Immunology

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

the Animal Resource Center at the University of Utah Health Science Cen-ter according to the guidelines of the National Institutes of Health for thecare and use of laboratory animals.

B. burgdorferi culture and infection

Mice were infected by intradermal injection at 6 wk of age with 2 � 103

B. burgdorferi clone N40 (provided by S. Barthold, University of Califor-nia, Davis, CA) that had been cultured for 4 days in Barbour-Stoenner-Kelly II medium containing 6% rabbit serum (Sigma-Aldrich).

Assessment of infection status and arthritis severity

Infection of mice was confirmed by culture of spirochetes from bladders,production of B. burgdorferi-specific Abs, and detection of B. burgdorferirecA in ear tissues by quantitative PCR (7, 20). Ankle swelling was usedas a relative indicator of arthritis development in the actual tissue collectedfor microarray analysis and was determined from measurements made ofthe rear ankle joints with a metric caliper. Increases in ankle measurementwere similar to those in previous studies, where complete histological as-sessment of arthritis severity was performed (9, 16).

Isolation of RNA

Total RNA was isolated from tissues and cells using acid guanidine ex-traction (21). Skin was removed from the rear ankles, and tissue extending�5 mm in each direction was collected from infected and control mice atthe indicated times. Joint tissues were flash frozen and homogenized incold acid guanidine using an Ultra-Turrax disperser (IKA Works), andRNA was separated by cesium chloride cushion centrifugation. RNA wasrecovered by ethanol precipitation and applied to a RNeasy kit (Qiagen).

Gene expression analysis

Equal amounts of total RNA from the more swollen ankle of five individualmice of each genotype from each time point were pooled into a singlesample that was prepared for Affymetrix array hybridization according tothe manufacturer’s instructions (Affymetrix) (22). cDNA was synthesizedfrom 8 �g of total RNA and biotin-labeled using the One-Cycle TargetLabeling Kit (Affymetrix). Each sample was hybridized to triplicateGeneChip arrays; either GeneChip Mouse Expression Array 430A (C3H/HeN) or GeneChip Mouse Genome 430 2.0 Array (C57BL/6 and C57BL/6IL-10�/�) (Affymetrix) depending on availability, and then stained andwashed in the Affymetrix Fluidics Station 450 with program EukGE-WS2v4. Arrays were scanned with either the GeneArray 2500 Scanner(C3H/HeN) or GeneChip Scanner 3000 (C57BL/6 and C57BL/6 IL-10�/�)laser confocal slide scanner. Data from all GeneChips were preprocessedusing the affy and gcRMA (Robust Multiarray Average) packages in R(23–25). Statistical analysis was performed using significance analysis ofmicroarrays (SAM), and data were filtered based on a significant p value( p � 0.05 as determined by SAM) (26). Transcripts with changes meetingsignificant p values were filtered based on fold change, and those with achange of 2-fold or greater were considered differentially expressed,whereas other transcripts were considered not changed (NC). The functionof gene products was primarily determined with the National CancerInstitute/Center for Information Technology microarray database (NCI/CIT mAdb, �http://www.cit.nih.gov/home.asp�) and with Gene Ontologyannotations according to Mouse Genome Informatics (27). In some cases,gene function was inferred based on the function of orthologous genesand/or phenotype of other gene family members as determined from publicdatabases such as Ensembl and Mouse Genome Informatics (27, 28). Mi-croarray data from this manuscript may be accessed at the GEO database�http://www.ncbi.nlm.nih.gov/geo/� under accession no. GSE6055.

Real-time quantitative RT-PCR

RT-RCR on 5 �g of total RNA was performed using random primers andM-MLV Reverse Transcriptase (Invitrogen Life Technologies). QuantitativePCR was performed using LightCycler SYBRPlus MasterMix on the Light-Cycler (Roche Applied Science). The oligonucleotide primers used to detect�-actin were bactin.F (forward) (5�-GTAACAATGCCATGTTCAAT-3�) andbactin.R (reverse) (5�-CTCCATCGTGGGCCGCTCTAG-3�); Cxcl13 werecxcl13.F (5�- -3�) and cxcl13.R (5�- -3�); Elovl4 were elovl4.F (5�-TACTATGGGCTGACTGCGTTCG-3�) and elovl4.R (5�-GACTGCTTCGGCTCATTGTATGTC-3�); Flg were flg.F (5�-CAATGAAGACTGGGAGGCAAGC-3�) and flg.R (5�-TGACTGGAGATGGTTTGGAGTGG-3�); Hrnrwere hrnr.F (5�-GCAACAAGATGCCTAAACTCCTGG-3�) and hrnr.R(5�-GCTGGTGACTGTGATTTTTCTGC-3�); Igtp were igtp.F (5�-TAGAGCAGACCCACAGAGTTCAGG-3�) and igtp.R (5�-CAGCAGTCATAGATTTAGACCACGG-3�); Iigp1 were iigp1.F (5�-GTAGTGTGCTCAATGTTGCT

GTCAC-3�) and iigp1.R (5�-TACCTCCACCACCCCAGTTTTAGC-3�); Il1bwere il1b.F (5�-TCCCAAGCAATACCCAAAGAGAA-3�) and il1b.R (5�-TGGGGAAGGCATTAGAAACAGTC-3�); Irf7 were irf7.F (5�-TGTGACCCTCAACACCCTAATACC-3�) and irf7.R (5�-CAATAGCCAGTCTCCAAACAGCAC-3�); Mmp3 were mmp3.F (5�-TTGTGTGCTCATCCTACCCATTG-3�) and mmp3.R (5�-TTCCTCCATTTTGGCGAACC-3�); and Stat1 werestat1.F (5�-CGTGGGAACGGAAGCATTTG-3�) and stat1.R (5�-ACGAGACATCATAGGCAGCGTG-3�). Primers for B. burgdorferi 16S rRNA were5�GGTCAAGACTGACGCTGAGTCA and 5�GGCGGTCCACTTAACACGTTAG, as described previously (29).

FIGURE 1. Gene expression profiles as they relate to arthritis develop-ment. A, Ankle swelling was determined for each infected and uninfectedmouse as tissues were collected for expression analysis, as described inMaterials and Methods. B, RT-PCR with primers for Borrelia 16S rRNAwas performed on the individual RNA samples used for microarray toestimate relative levels of B. burgdorferi in tissues. Values for C3H andC57BL/6 mice were not significantly different at 1 and 2 wk of infection,whereas the difference at 4 wk was significant (p � 0.05, Student’s t test).Differences between C57BL/6 and IL-10�/� mice were significant at 2 and4 wk of infection. C, Lines representative of gene expression reflect theaverage fold change of five genes selected from the IFN-responsive profile(Igtp, Iigp2, Iigp1, Tgtp, and Ifi1) and the epidermal differentiation profile(D) (Flg, Krt2–1, Hrnr, Sprr1b, and Lor.

7931The Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Design of microarray experiment

Gene expression profiling using Affymetrix GeneChip microarrays wasperformed during the development of arthritis of differing severities inC3H, C57BL/6, and C57BL/6-IL-10�/� mice infected with B. burgdorferi.Expression profiles in joint tissues were determined at 1, 2, and 4 wk ofinfection, and compared with mock-infected mice, age matched with the1-wk time point. Previous studies indicated that by 1 wk of infection, B.burgdorferi had spread to the joint tissue, but that inflammatory cell infil-trate was not yet detectable, thus providing an opportunity to capture theearly response of endogenous cells of the joint tissue to invading bacteria(4, 7). By 2 wk of infection, bacterial number was at its greatest and therewas robust inflammatory cell infiltrate, whereas the 4-wk time point wasdesigned to coincide with the peak of arthritis development in C3H mice(4, 7). Expression profiles were performed on C57BL/6-IL-10�/� mice,which develop more severe arthritis than C57BL/6 mice, to help iden-tify changes in gene expression that are common to the development ofarthritis (16).

Because technical and biological variation were expected, RNA wasisolated from the joint tissue of five mice at each time point and pooledbefore microarray analysis, which was performed in triplicate (30). Resultswere confirmed for selected transcripts by quantitative RT-PCR with theindividual samples used to generate the pool, as well as confirmationof selected transcripts from individual animals infected in a secondexperiment.

ResultsOverview of gene expression profiles

B. burgdorferi infection of C3H, C57BL/6, and C57BL/6-IL-10�/� mice was followed over time, with measurements of rearankle swelling taken as an approximation of the progression ofarthritis in the actual joint tissues pooled for microarray analysis(Fig. 1A). Although ankle measurements provide a limited assess-ment of arthritis, the results reported in Fig 1A are similar to thosein our previous report that included histological assessment of ar-thritis development in C57BL/6-IL-10�/� mice (16). To gain anapproximation of B. burgdorferi levels in the individual ankle tis-sues collected for microarray analysis, we performed RT-PCR us-ing primers for the flaB gene and for 16S rRNA of Borrelia (29,31). Results were similar with the two primer sets and are shownfor the more sensitive 16S rRNA (Fig. 1B). Similar levels of 16SrRNA were present in ankle tissues from C3H and C57BL/6 miceat 1 and 2 wk of infection, whereas at 4 wk of infection 16S rRNAlevels were greater in C57BL/6 than C3H joint tissues. 16S rRNAwas not detected in samples from uninfected mice, and levels of16S rRNA were much lower at all time points in C57BL/6-IL-10�/� mice than in either wild type, consistent with previous re-ports using quantitative PCR detection of DNA in tissues (16).

Affymetrix GeneChip array analysis revealed alterations in ex-pression of discrete subsets of genes in RNA collected from thejoint tissue of infected mice. At 1 wk of infection, transcripts from156 genes were increased by �2-fold in C3H mice, whereas inC57BL/6 mice 119 transcripts were increased; this result is basedon the 22,000 unique transcripts included on GeneChip 430A ofthe mouse (Table I). Interestingly, there were only three transcripts

changed by �2-fold that were common to C3H and C57BL/6 miceat 1 wk of infection (Figs. 2, 3A). Tissue from infected C57BL/6-IL-10�/� mice revealed 419 gene transcripts increased at 1 wk ofinfection and displayed clear overlap with those increased in C3Hmice (Fig. 3A). At 2 wk postinfection, significant recruitment ofleukocytes to joint tissues is evident by histological analysis (7, 9),which was supported by the increase in a large number of genes inall three strains of mice associated with host defense and inflam-matory cell infiltrate (Fig. 3B). The 4-wk postinfection time pointrepresents the peak of arthritis as well as full development of theadaptive immune response, consistent with the identification of agroup of genes induced only at this late time point. The number ofgenes changed at 4 wk was less than at 2 wk in both C3H andC57BL/6 mice, possibly reflecting resolution of infection and inflam-matory response (Table I and Fig. 3C). In contrast, C57BL/6 IL-10�/� mice displayed an exaggerated and uncontrolled response toinfection, with the number of genes changed �2-fold still increasingat the 4-wk time point. This is consistent with the potent ability ofIL-10 to regulate acute inflammatory processes (Table I) (32).

B. burgdorferi infection induces an early proinflammatoryresponse in arthritis-susceptible C3H mice

At 1 wk of infection, C3H mice displayed a robust induction ofgenes indicative of a strong proinflammatory response by endog-enous cells. In particular, the most highly increased genes werethose known to be inducible by type I (IFN-� and IFN-�) and/ortype II (IFN-�) IFN (Table II). In fact, 36% (56 of 156) of thegenes induced �2-fold in C3H mice were clearly annotated as IFNresponsive, and 67 of the 100 most highly induced genes wereeither IFN responsive or could be linked to the induction and reg-ulation of IFN responses (Fig. 2 and Supplement 1)4 (27). Addi-tional genes increased in C3H mice could be indirectly linked toIFN by their role in regulation and development of the immuneresponse, such as Parp14, Bst2, and Nfil3 (Table II). The genes ofthe C3H profile were highly induced, up to 120-fold comparedwith uninfected joint tissue. It is difficult to assign responsibilityfor this robust IFN response because changes were not detected ineither type I or type II IFN transcripts in C3H or C57BL/6 mice atany time, and most of the genes listed in Table II can be inducedby either type I or type II IFN (including the highly induced Iigp1,Gbp1, and Tap1 (27, 33)). The increased transcript levels for thesignaling molecules Stat1, Irf1, Irf7, and Irf8 also did not providestrong evidence for the selective presence of type I or type II IFN(Table II) (33).

In striking contrast, none of the IFN-inducible gene transcriptswas increased �2-fold at 1 wk postinfection in the ankle joints ofC57BL/6 mice (Table II and Fig. 2). Although some IFN-inducibletranscripts were increased by 2 wk postinfection, the magnitude of

4 The on-line version of this article contains supplemental material.

Table I. Number of unique gene transcripts changed �2-fold in the joints of C3H, C57BL/6, and C57BL/6IL-10�/� mice at 1, 2, and 4 wk postinfectiona

ArthritisPhenotypeb

Borreila no.in Jointc 1 wk 2 wk 4 wk

C3H/HeN Severe High 1156 2401 1695 2647 1456 2222C57BL/6 Mild High 1119 26 1594 223 1340 221IL-10 KO Intermediate Low 1419 1231 11332 2769 11423 2870

a Gene expression in ankle joint tissue at 1, 2, and 4 wk postinfection was compared to uninfected ankle joint tissue.Affymetrix probe sets identifying the same gene transcript were counted as a single gene transcript.

b As described in Ref. 16 and Fig 1A.c As described in Ref. 16 and Fig 1B.

7932 NOVEL PATHWAYS IN B. burgdorferi INFECTION

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

change was generally much less than that seen in C3H mice, sug-gesting that ankle tissues of C57BL/6 mice have a temporally de-layed and diminished IFN response to B. burgdorferi infection(Table II, Fig. 1C, and Supplement 2). The association with ar-thritis was further strengthened by findings with C57BL/6-IL-10�/� mice, which displayed robust induction of many of the IFN-responsive genes characteristic of the expression profile foundwith severely arthritic C3H mice (Table II, Fig. 3A, and Supple-ment 3). Although the IFN-induced profile was delayed in IL-10�/� mice, the levels of IFN-inducible genes remained elevatedthroughout the course of infection rather than returning to baselineas seen with C3H mice (Table II and Fig. 1C) (17, 32). Thesefindings suggest that early production of IFN in the environment ofthe C3H mouse joint is associated with inflammatory events ofarthritis development and that its absence from C57BL/6 mice isprotective.

The expression levels of several highly changed transcripts in-cluding Iigp1 and Igtp and the signaling molecules Stat1 and Irf7were assessed in individual mice using quantitative RT-PCR (Fig.4A–D). The trends of expression observed by microarray analysiswere confirmed by RT-PCR for all genes tested, and selected ex-pression patterns were confirmed with samples from a second in-fection experiment. This result provides important documentationthat the microarray analysis reflected the contribution of individualmice rather than the dominant expression of an aberrant sample inthe pool and indicate that the robust induction of IFN-responsivegenes in joint tissue is a universal feature of the early response ofC3H mice to B. burgdorferi infection that is absent in the earlyresponse of C57BL/6 mice.

A novel group of genes is selectively increased in infectedC57BL/6 mice

Joint tissue from C57BL/6 mice displayed elevation of an entirelydifferent set of genes than C3H mice at 1 wk postinfection. In fact,

of the �100 genes increased at 1 wk postinfection, some up to40-fold, only two genes were shared with C3H mice, serum amy-loid A 3 (Saa3) and the chemokine Ccl12 (Fig. 2). In C57BL/6mice, a large number (55%; 65 of 119) of transcripts increased�2-fold, and 75 of the 100 most highly induced transcripts at 1 wkpostinfection were involved in epidermal differentiation, cell ad-hesion, and cell-cell interaction or wound repair (Table III andSupplement 2) (27, 34). Increased expression of genes involved inepidermal development was surprising because all skin was re-moved from the joint tissue before collection, and expression waslow in joint tissue collected from uninfected C57BL/6 mice. Theincreased expression of genes involved in wound repair, such asKrt-2a, and cell-cell interactions, such as Ly6d and Dsp, could besuggestive of a response to disseminating bacteria (35, 36).

Of note, most of the 119 genes which increased �2-fold at 1 wkpostinfection in the joints of C57BL/6 mice were decreased inC3H mice, and comprised a subset of the 401 genes decreased by�2-fold in C3H mice (Figs. 1D and 3A, Table III, and Supple-ments 1and 2). Seventy-six of the genes with reduced expressionin C3H mice were also reduced in C57BL/6 IL-10�/� mice (TableIII, Fig. 3A), suggesting that regulation of this response could berelated to the inflammatory status of the tissue. Differential expres-sion of this epidermal differentiation/wound repair profile wasmost evident at 1 wk postinfection; however, elevation in C57BL/6mice and reduction in C3H and C57BL/6-IL-10�/� mice wasmaintained at 2 and 4 wk postinfection with B. burgdorferi (Fig.1D). Interestingly, not all genes induced during epidermal differ-entiation were altered during B. burgdorferi infection, includingIvl, Ppl, and Sprr2f (Table III) (27).

RT-PCR was used to assess the transcriptional pattern for se-lected epidermal differentiation/wound repair gene transcripts fromTable III, including Flg, Hrnr, and Elovl4. Changes observed bymicroarray analysis in pooled samples were reproducibly presentin the joints of five individual mice comprising the pools for eachtime point (Fig. 4, E–G) and in tissues from a second infectionexperiment (data not shown). These data demonstrate that the in-creased expression of the epidermal transcripts in C57BL/6 miceand reduced expression in C3H and IL-10�/� mice was a featureof the response to B. burgdorferi infection that correlated with theseverity of arthritis.

Great care was taken in the removal of skin from the ankle jointsused in this study, and histological assessment of many rear ankletissues prepared in the same manner as for microarray has notrevealed differences in the inclusion of skin tissue in the joints ofany mouse genotype. To determine whether altered expression ofthe epidermal profile was a generalized response to infection of theskin or whether it was a response unique to the joint tissue, RT-PCR was performed on RNA prepared from the ears of infected

FIGURE 2. Distinct gene expression profiles at 1 wk postinfection inC3H/HeN and C57BL/6 mice. Of the 100 most highly increased transcriptsin C3H mice, 67 were related to the IFN response, whereas 75 of the 100most increased transcripts in C57BL/6 mice were related to epidermal dif-ferentiation and wound repair. Only 2 of the 100 most highly increasedtranscripts were shared in the two mouse strains, Saa3 and Ccl12.

FIGURE 3. The number of transcripts changed �2-fold in the joint tissue of C3H/HeN, C57BL/6, and C57BL/6 IL-10�/� mice at 1 wk (A), 2 wk (B),and 4 wk (C) postinfection. The number of nonredundant transcripts with significant p values and changes �2-fold were detected by microarray analysisas described in Materials and Methods, with arrows indicating induction or reduction relative to joint tissue from uninfected mice. Transcripts with changedvalues that were shared among strains are placed within the appropriate common areas of the Venn diagrams.

7933The Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Tab

leII

.E

xpre

ssio

nle

vels

ofse

lect

edIF

N-r

espo

nsiv

ege

netr

ansc

ript

sin

crea

sed

at1

wk

post

infe

ctio

nin

join

tsti

ssue

ofC

3Hm

icea

Prob

eSe

tID

Gen

eT

itle

Gen

eSy

mbo

l

1w

k2

wk

4w

k

C57

BL

/6C

3H/H

eNIL

-10

KO

C57

BL

/6C

3H/H

eNIL

-10

KO

C57

BL

/6C

3H/H

eNIL

-10

KO

IFN

-ind

ucib

lege

nes

1417

141_

atIF

N-�

-ind

uced

GT

Pase

Igtp

NC

128

NC

7.1

1649

NC

3.6

4914

1779

3_at

IFN

-ind

ucib

leG

TPa

se2

Iigp

2N

C12

32.

44.

812

35N

C2.

638

1419

042_

atIF

N-i

nduc

ible

GT

Pase

1Ii

gp1

NC

113

2.5

1216

622.

45.

386

1449

009_

atT

cell-

spec

ific

GT

Pase

Tgt

pN

C42

2.5

7.5

1472

2.1

5.3

7714

3867

6_at

Mac

roph

age

activ

atio

n2

like

Mpa

21N

C36

NC

4.1

3.6

31N

CN

C36

1418

825_

atIF

N-i

nduc

ible

prot

ein

1Ifi

1N

C35

NC

3.6

3.3

32N

CN

C34

1419

714_

atC

D27

4A

gC

d274

NC

33N

C5.

35.

153

NC

NC

5714

1601

6_at

Tra

nspo

rter

1,A

TP-

bind

ing

cass

ette

,su

bfam

ilyB

(MD

R/T

AP)

Tap

1N

C15

2.0

6.4

5.8

342.

12.

635

1418

240_

atG

uany

late

nucl

eotid

e-bi

ndin

gpr

otei

n2

Gbp

2N

C12

2.8

124.

172

2.6

2.1

7114

5069

6_at

Prot

eoso

me

(pro

som

e,m

acro

pain

)su

buni

t,�

type

9(l

arge

mul

tifun

ctio

nal

prot

ease

2)Ps

mb9

NC

11N

C6.

57.

940

2.4

3.6

45

1453

196_

a_at

2�–5

�ol

igoa

deny

late

synt

heta

se-l

ike

2O

asl2

NC

4.4

4.2

2.8

NC

29N

CN

C33

Sign

alin

gm

olec

ules

1450

034_

atSi

gnal

tran

sduc

eran

dac

tivat

orof

tran

scri

ptio

n1

Stat

1N

C23

2.1

3.0

5.2

24N

CN

C39

1448

436_

a_at

IFN

regu

lato

ryfa

ctor

1Ir

f1N

C4.

5N

C2.

72.

515

NC

NC

1614

1724

4_a_

atIF

Nre

gula

tory

fact

or7

Irf7

NC

4.0

2.1

2.2

NC

18N

CN

C16

1416

714_

atIF

Nre

gula

tory

fact

or8

Irf8

NC

3.8

NC

4.9

4.0

152.

5N

C15

1418

131_

atSA

Mdo

mai

nan

dH

Ddo

mai

n,1

Sam

hd1

NC

4.7

NC

4.0

6.0

24N

C2.

521

1418

265_

s_at

IFN

regu

lato

ryfa

ctor

2Ir

f2N

CN

CN

CN

CN

CN

CN

CN

C2.

1O

ther 1451

564_

atPo

ly(A

DP-

ribo

se)

poly

mer

ase

fam

ily,

mem

ber

14Pa

rp14

NC

7.3

NC

2.9

3.2

15N

CN

C16

1429

947_

a_at

Z-D

NA

-bin

ding

prot

ein

1Z

bp1

NC

5.3

NC

4.9

3.4

40N

C2.

147

1438

855_

x_at

TN

F,�

-ind

uced

prot

ein

2T

nfai

p2N

C4.

5N

CN

C2.

05.

1N

CN

C6.

214

2492

1_at

Bon

em

arro

wst

rom

alce

llA

g2

Bst

2N

C3.

8N

C2.

72.

111

NC

NC

13

aG

ene

expr

essi

onin

ankl

ejo

int

tissu

eat

1,2,

and

4w

kpo

stin

fect

ion

was

com

pare

dto

unin

fect

edan

kle

join

ttis

sue.

Num

bers

indi

cate

fold

chan

ge.

Cha

nges

�2-

fold

are

desi

gnat

ed“N

C.”

7934 NOVEL PATHWAYS IN B. burgdorferi INFECTION

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

and uninfected mice. Transcript levels for filaggrin were notchanged in ear tissue from infected C3H or C57BL/6 mice, sug-gesting that this differential response to infection was specific forthe ankle tissue (data not shown).

B. burgdorferi infection results in expression of genes involvedin host defense

In addition to the very distinct gene expression profiles seen earlyduring infection that correlate with arthritis severity, all threestrains of mice shared a common response to B. burgdorferi thatwas evident at 2 and 4 wk postinfection (Fig. 3, B and C). Therecruitment of PMNs is a hallmark of Lyme arthritis (8, 15), andby 2 wk of infection modest increases (5-fold) in the PMN-recruit-ing chemokines Cxcl1, Cxcl2, and Cxcl5 were seen in both C3Hand C57BL/6 mice as well as increased levels of PMN gene prod-ucts such as Mpo, Ncf1, and Ncf4 (Table IV and Supplements 1–3).Additionally, increased transcription of several mononuclear cell-recruiting chemokines, such as Ccl2, Ccl3, Ccl7, and Ccl12, wasobserved (37). The IFN-responsive T cell-recruiting chemokineCxcl9 was increased in both C3H and C57BL/6 mice over thecourse of infection, whereas Cxcl10 was also highly induced injoints of C57BL/6 and C57BL/6-IL-10�/� mice (Table IV) (37).Additional T cell-recruiting chemokines were increased during in-fection in both C3H and C57BL/6 mice, including Cxcl16, Ccl2,and Ccl7, (37). Transcript levels for the B cell chemokine Cxcl13were dramatically increased in the joints of both C57BL/6 andC3H mice by 2 wk infection, whereas joints of C57BL/6-IL-10�/�

mice displayed a less robust induction of Cxcl13 (Table IV). Infact, Cxcl13 was the only chemokine with greater induction inC57BL/6 mice than in C57BL/6 IL-10�/� mice, an interestingfinding in light of the clinical association between Cxcl13 expres-sion and spirochete load in neuroborreliosis (38) and the reducedpresence of B. burgdorferi in the joints of IL-10�/� mice (Fig. 1B).Expression profiles for many of the chemokines were confirmed byRT-PCR in individual mice, with the unique profile for Cxcl13shown in Fig. 4H.

Genes whose expression is associated with host defense were alsoincreased at 2 wk of infection, including additional markers for neu-trophils and macrophages, Mpo, Cd14, Tlrs, Fc receptors, and argi-nase. Complement component C1qa, C1qb, C1qg, C1r, C2, and C3transcripts were modestly increased in the joints of both C3H andC57BL/6 mice, further suggesting the presence of activated mac-rophages (Table IV). Transcripts for markers of APCs and Ag-processing machinery, including the MHC, Cd1, Tap1/2, andproteosome subunits indicated the presence of an active ac-quired host defense in joint tissues at 2 and 4 wk of infection,as did the increased expression of Ig genes (Table IV) (27).

Borrelia and its lipoproteins are known to activate NF-�B andinduce the production of many cytokines including IL-6, IL-1�,TNF-�, and IL-12 in vitro (39, 40). Surprisingly, IL-1� was theonly hallmark inflammatory cytokine to be increased in either C3Hor C57BL/6 ankle tissue (Table IV and Fig. 4I). In contrast, tran-scripts for many cytokines were robustly induced at 2 and 4 wk ofinfection in the IL-10�/� mice (including IL-1b, IFN-�, IL-6, IL-7,

FIGURE 4. Quantification of selected genetranscripts in the joints of individual mice. Expres-sion level of IFN-related transcripts Iigp1 (A), Igtp(B), Stat1 (C), and Irf7 (D); epidermal differenti-ation-related transcripts Flg (E), Hrnr (F), andElovl4 (G); transcripts associated with the host de-fense Cxcl13 (H) and Il1b (I); and a transcript as-sociated with chondrocyte activation Mmp3 (J)were determined using quantitative RT-PCR. Foldchange was determined based on the differencebetween the five individual mice used to generatepooled samples and the average of uninfected con-trols. In all cases, trends observed by microarraywere confirmed by RT-PCR analysis.

7935The Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Tab

leII

I.E

xpre

ssio

nle

vels

ofse

lect

edep

ider

mal

diffe

rent

iati

on/w

ound

repa

ir-a

ssoc

iate

dge

netr

ansc

ript

sin

crea

sed

at1

wk

post

infe

ctio

nin

join

tti

ssue

ofC

57B

L/6

mic

ea

Prob

eSe

tID

Gen

eT

itle

Gen

eSy

mbo

l

1w

k2

wk

4w

k

C57

BL

/6C

3H/H

eNIL

10�

/�C

57B

L/6

C3H

/HeN

IL10

�/�

C57

BL

/6C

3H/H

eNIL

10�

/�

Stru

ctur

alm

olec

ules

1427

268_

atFi

lagg

rin

Flg

47�

12�

2324

�7.

2N

C66

�16

�17

1422

481_

atK

erat

inco

mpl

ex2,

basi

c,ge

ne1

Krt

2-1

38�

124

�41

16�

3.7

NC

67�

131

�23

1451

613_

atH

orne

rin

Hm

r36

�8.

6�

1831

�5.

7�

5.3

90�

10�

1714

2715

4_at

Ker

atin

com

plex

2,ba

sic,

gene

17K

rt2-

1724

�86

�6.

817

�6.

7N

C40

�90

�6.

814

2267

2_at

Smal

lpr

olin

e-ri

chpr

otei

n1B

Sprr

1b20

�7.

0�

1314

�5.

9�

2.1

74�

8.9

�13

1420

183_

atL

oric

rin

Lor

8.5

�18

�2.

75.

6�

5.0

NC

10�

24�

6.3

1422

667_

atK

erat

inco

mpl

ex1,

acid

ic,

gene

15K

rt1-

156.

7N

CN

C3.

7N

CN

C5.

9�

2.1

NC

1422

222_

atIn

volu

crin

Iv1

NC

NC

NC

NC

NC

NC

NC

NC

NC

1460

732_

a_at

Peri

plak

inPp

1N

CN

CN

CN

CN

CN

CN

CN

CN

C14

4983

3_at

Smal

lpr

olin

e-ri

chpr

otei

n2F

Sprr

2fN

CN

CN

CN

CN

CN

CN

CN

CN

CC

ell

adhe

sion

/cel

l-ce

llin

tera

ctio

ns14

1693

0_at

Lym

phoc

yte

Ag

6co

mpl

ex,

locu

sD

Ly6

d15

�7.

8�

3.1

8.5

�6.

5N

C25

�9.

2�

2.9

1435

493_

atD

esm

opla

kin

Dsp

15�

7.3

�5.

38.

4�

4.5

NC

24�

9.4

�5.

014

3519

1_at

Cor

neod

esm

osin

Cds

n14

�4.

9�

5.0

7.0

�3.

3N

C24

�5.

9�

7.5

1418

799_

a_at

Proc

olla

gen,

type

XV

II,

alph

a1

Col

17a1

7.3

�3.

6�

2.0

3.4

�3.

2N

C7.

0�

4.9

�3.

5W

ound

resp

onse

rela

ted

1421

752_

a_at

Seri

ne(o

rcy

stei

ne)

prot

eina

sein

hibi

tor,

clad

eB

,m

embe

r5

Serp

inb5

11�

4.4

�4.

17.

1�

2.2

NC

23�

3.4

�3.

6

1421

092_

atSe

rine

(or

cyst

eine

)pr

otei

nase

inhi

bito

r,cl

ade

A(�

-1A

ntip

rote

inas

e,an

titry

psin

),m

embe

r12

Serp

ina1

24.

7�

3.1

�4.

83.

1�

2.7

NC

10�

3.4

�5.

5

1422

784_

atK

erat

inco

mpl

ex2,

basi

c,ge

ne6a

Krt

2-6a

3.7

�2.

2N

C3.

9�

2.1

6.6

8.2

�2.

2N

CO

ther 1450

633_

atC

alm

odul

in4

Cal

m4

29�

86�

1016

�7.

7N

C49

�13

4�

1314

2430

6_at

Elo

ngat

ion

ofve

rylo

ngch

ain

fatty

acid

s(F

EN

1/E

lo2,

SUR

4/E

lo3,

yeas

t)-l

ike

4E

lov1

414

�4.

5N

C7.

0�

4.1

NC

15�

5.7

�3.

0

1450

645_

atM

etal

loth

ione

in4

Mt4

20�

22�

126.

5�

4.4

NC

36�

30�

9.0

aG

ene

expr

essi

onin

ankl

ejo

int

tissu

eat

1,2,

and

4w

kpo

stin

fect

ion

was

com

pare

dto

unin

fect

edan

kle

join

ttis

sue.

Num

bers

indi

cate

fold

chan

ge.

Num

bers

prec

eded

bya

“_”

are

redu

ced

com

pare

dto

unin

fect

edan

kle

join

ttis

sue.

Cha

nges

�2-

fold

are

desi

gnat

ed“N

C.”

7936 NOVEL PATHWAYS IN B. burgdorferi INFECTION

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Tab

leIV

.E

xpre

ssio

nle

vels

ofge

nes

invo

lved

inth

eim

mun

ere

spon

seth

atar

ech

ange

d�

2-fo

lda

Prob

eSe

tID

Gen

eT

itle

Gen

eSy

mbo

l

1w

k2

wk

4w

k

C57

BL

/6C

3H/H

eNIL

10�

/�C

57B

L/6

C3H

/HeN

IL10

�/�

C57

BL

/6C

3H/H

eNIL

10�

/�

Che

mok

ines

1420

380_

atC

hem

okin

e(C

-Cm

otif

)lig

and

2C

cl2

NC

NC

2.1

226.

742

3.7

2.8

6614

1956

1_at

Che

mok

ine

(C-C

mot

if)

ligan

d3

Ccl

3N

CN

CN

C14

NC

7.5

2.4

NC

1114

1726

6_at

Che

mok

ine

(C-C

mot

if)

ligan

d6

Ccl

6N

CN

CN

C2.

9N

C7.

0N

CN

C5.

214

2122

8_at

Che

mok

ine

(C-C

mot

if)

ligan

d7

Ccl

7N

CN

CN

C13

1128

3.4

4.6

4914

1968

4_at

Che

mok

ine

(C-C

mot

if)

ligan

d8

Ccl

8N

C2.

9N

C7.

615

484.

99.

784

1417

936_

atC

hem

okin

e(C

-Cm

otif

)lig

and

9C

cl9

NC

NC

NC

4.0

3.3

3.4

2.0

NC

3.0

1419

282_

atC

hem

okin

e(C

-Cm

otif

)lig

and

12C

cl12

2.5

2.6

NC

412.

017

3.5

NC

3214

1920

9_at

Che

mok

ine

(C-X

-Cm

otif

)lig

and

1C

xcl1

NC

NC

3.4

5.9

5.6

343.

45.

779

1449

984_

atC

hem

okin

e(C

-X-C

mot

if)

ligan

d2

Cxc

l2N

CN

C2.

25.

7N

C17

NC

NC

2014

1972

8_at

Che

mok

ine

(C-X

-Cm

otif

)lig

and

5C

xcl5

NC

NC

NC

NC

6.7

13N

C5.

627

1418

480_

atC

hem

okin

e(C

-X-C

mot

if)

ligan

d7

Cxc

l7N

CN

C2.

4N

CN

CN

CN

CN

CN

C14

1865

2_at

Che

mok

ine

(C-X

-Cm

otif

)lig

and

9C

xcl9

NC

173.

160

1059

612

3.6

811

1418

930_

atC

hem

okin

e(C

-X-C

mot

if)

ligan

d10

Cxc

l10

NC

2.6

2.7

11N

C14

12.

3N

C18

914

4882

3_at

Che

mok

ine

(C-X

-Cm

otif

)lig

and

12C

xcl1

2N

C�

2.0

NC

NC

NC

2.1

NC

2.2

2.6

1417

851_

atC

hem

okin

e(C

-X-C

mot

if)

ligan

d13

Cxc

l13

2.3

NC

NC

401

377.

812

145

2014

1845

7_at

Che

mok

ine

(C-X

-Cm

otif

)lig

and

14C

xcl1

4N

CN

CN

CN

C5.

4N

CN

C2.

8N

C14

4919

5_s_

atC

hem

okin

e(C

-X-C

mot

if)

ligan

d16

Cxc

l16

NC

NC

NC

6.4

7.9

154.

33.

322

1415

803_

atC

hem

okin

e(C

-X3-

Cm

otif

)lig

and

1C

x3cl

1N

CN

CN

CN

CN

CN

CN

CN

C3.

0C

hem

okin

ere

cept

ors

1419

609_

atC

hem

okin

e(C

-Cm

otif

)re

cept

or1

Ccr

1N

CN

CN

C2.

5N

C3.

1N

CN

C2.

414

2118

7_at

Che

mok

ine

(C-C

mot

if)

rece

ptor

2C

cr2

NC

NC

2.4

4.0

NC

8.8

2.1

NC

18C

ytok

ines

1425

947_

atIF

N-�

Ifng

NC

NC

NC

NC

NC

16N

CN

C22

1450

330_

atIL

-10

Il10

NC

NC

NC

2.2

NC

6.2

NC

NC

6.4

1418

219_

atIL

-15

Il15

NC

NC

NC

NC

NC

2.8

NC

NC

3.4

1449

399_

a_at

IL-1

�Il

1bN

CN

CN

C4.

03.

113

2.9

2.5

4214

2137

0_a_

atIL

-1fa

mily

,m

embe

r5

(�)

Il1f

52.

6N

CN

CN

CN

CN

C6.

5�

2.1

NC

1451

798_

atIL

-1re

cept

oran

tago

nist

Il1m

NC

NC

NC

9.4

NC

143.

0N

C22

1421

034_

a_at

IL-4

rece

ptor

,�

Il4r

aN

CN

CN

CN

C3.

34.

8N

CN

C5.

514

5029

7_at

IL-6

Il6

NC

NC

NC

NC

NC

13N

CN

C72

1436

861_

atIL

-7Il

7N

CN

CN

CN

CN

CN

CN

CN

C2.

114

1913

5_at

Lym

phot

oxin

BL

tbN

CN

CN

CN

CN

C4.

3N

CN

C4.

214

1834

5_at

TN

F(l

igan

d)su

perf

amily

,m

embe

r13

Tnf

sf13

NC

NC

NC

NC

NC

NC

NC

2.1

2.3

1460

255_

atT

NF

(lig

and)

supe

rfam

ily,

mem

ber

13b

Tnf

sf13

bN

CN

CN

C4.

3N

C6.

63.

2N

C10

Inna

teim

mun

ere

spon

se14

1954

9_at

Arg

inas

e1,

liver

com

plem

ent

com

pone

nt1,

qsu

bcom

pone

nt,

�A

rg1

2.1

�4.

2N

C63

NC

4716

�2.

832

1417

381_

atPo

lype

ptid

eco

mpl

emen

tco

mpo

nent

1,q

subc

ompo

nent

,�

C1q

aN

CN

CN

C2.

92.

34.

82.

32.

95.

2

1417

063_

atPo

lype

ptid

eco

mpl

emen

tco

mpo

nent

1,q

subc

ompo

nent

,�

C1q

bN

CN

CN

C4.

23.

36.

72.

93.

57.

7

1449

401_

atPo

lype

ptid

eC

1qg

NC

NC

NC

3.2

3.1

5.5

2.6

2.6

5.9

1417

009_

atC

ompl

emen

tco

mpo

nent

1,r

subc

ompo

nent

C1r

NC

NC

NC

NC

2.6

6.8

NC

2.5

7.6

1416

051_

atC

ompl

emen

tco

mpo

nent

2(w

ithin

H-2

S)C

2N

CN

CN

CN

C2.

26.

2N

CN

C8.

814

2395

4_at

Com

plem

ent

com

pone

nt3

C3

NC

NC

NC

3.4

3.9

102.

52.

111

1460

334_

atD

rebr

in-l

ike

Dbn

1N

CN

CN

CN

CN

C3.

5N

CN

C3.

814

2554

8_a_

atL

euko

cyte

spec

ific

tran

scri

pt1

Lst

1N

CN

CN

C7.

416

162.

54.

512

(Tab

leco

ntin

ues)

7937The Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Tab

leIV

.(C

onti

nued

)

Prob

eSe

tID

Gen

eT

itle

Gen

eSy

mbo

l

1w

k2

wk

4w

k

C57

BL

/6C

3H/H

eNIL

10�

/�C

57B

L/6

C3H

/HeN

IL10

�/�

C57

BL

/6C

3H/H

eNIL

10�

/�

1455

346_

atM

anna

n-bi

ndin

gle

ctin

seri

nepr

otea

se1

Mas

p1N

CN

CN

CN

CN

C�

3.7

NC

NC

�4.

614

1596

0_at

Mye

lope

roxi

dase

Mpo

NC

NC

4.3

�2.

2N

C6.

2�

2.0

NC

NC

1425

609_

atN

eutr

ophi

lcy

toso

licfa

ctor

1N

cf1

NC

NC

NC

4.8

NC

9.6

NC

NC

1214

4856

1_at

Neu

trop

hil

cyto

solic

fact

or2

Ncf

2N

CN

CN

C4.

33.

55.

8N

CN

C5.

114

1846

5_at

Neu

trop

hil

cyto

solic

fact

or4

Ncf

4N

CN

CN

C6.

77.

66.

73.

23.

26.

414

5227

9_at

Prop

erdi

nfa

ctor

,co

mpl

emen

tPf

cN

CN

CN

C6.

57.

011

NC

2.7

6.3

1416

625_

atSe

rine

(or

cyst

eine

)pr

otei

nase

inhi

bito

r,cl

ade

G,

mem

ber

1Se

rpin

g1N

CN

CN

CN

CN

CN

CN

CN

C2.

1A

dapt

ive

imm

une

resp

onse

1449

289_

a_at

�-2

-mic

rogl

obul

inB

2mN

CN

CN

C2.

73.

94.

6N

C2.

84.

914

4913

0_at

CD

1d1

Ag

Cd1

d1N

C2.

0N

CN

CN

CN

CN

CN

CN

C14

1759

7_at

CD

28A

gC

d28

NC

NC

NC

NC

NC

2.9

NC

NC

2.6

1426

324_

atH

isto

com

patib

ility

2,D

regi

onlo

cus

1H

2-D

1N

C2.

2N

C2.

52.

37.

8N

CN

C8.

714

2774

6_x_

atH

isto

com

patib

ility

2,K

1,K

regi

onH

2-K

1N

C2.

12.

53.

76.

115

2.1

2.9

1714

2538

5_a_

atIg

Hch

ain

1a(s

erum

IgG

2a)

Igh-

1aN

CN

CN

CN

CN

C2.

44.

8N

C13

014

5163

2_a_

atIg

Hch

ain

1a(s

erum

IgG

2a)

Igh-

1aN

CN

CN

CN

CN

CN

CN

C17

NC

1425

247_

a_at

IgH

chai

n4

(ser

umIg

G1)

Igh-

4N

CN

CN

CN

CN

C2.

0N

C5.

75.

014

2532

4_x_

atIg

Hch

ain

4(s

erum

IgG

1)Ig

h-4

NC

NC

NC

NC

NC

NC

NC

9.0

2.9

1427

756_

x_at

IgH

chai

n4

(ser

umIg

G1)

Igh-

4N

CN

CN

CN

CN

CN

CN

C7.

02.

814

2430

5_at

Igjo

inin

gch

ain

Igj

NC

NC

NC

NC

NC

NC

326.

028

1425

519_

a_at

Ia-a

ssoc

iate

din

vari

ant

chai

nIi

NC

NC

NC

3.6

2.7

102.

8N

C13

1422

962_

a_at

Prot

eoso

me

(pro

som

e,m

acro

pain

)su

buni

t,�

type

8(l

arge

mul

tifun

ctio

nal

prot

ease

7)Ps

mb8

NC

7.4

2.2

6.0

9.4

432.

64.

546

1450

696_

atPr

oteo

som

e(p

roso

me,

mac

ropa

in)

subu

nit

�ty

pe9

(lar

gem

ultif

unct

iona

lpr

otea

se2)

Psm

b9N

C11

NC

6.5

7.9

402.

43.

645

1417

056_

atPr

otea

som

e(p

roso

me,

mac

ropa

in)

28su

buni

t,�

Psm

e1N

C2.

2N

CN

C2.

65.

0N

CN

C4.

214

1601

6_at

Tra

nspo

rter

1,A

TP-

bind

ing

cass

ette

,su

bfam

ilyB

(MD

R/

TA

P)T

ap1

NC

152.

06.

45.

834

2.1

2.6

35

1453

913_

a_at

Tra

nspo

rter

2,A

TP-

bind

ing

cass

ette

,su

bfam

ilyB

(MD

R/

TA

P)T

ap2

NC

4.0

NC

2.3

NC

12N

CN

C12

Cel

lsu

rfac

ere

cept

ors

1417

268_

atC

D14

Ag

Cd1

4N

CN

CN

C6.

45.

74.

93.

23.

85.

114

6025

1_at

Fas

(TN

Fre

cept

orsu

perf

amily

mem

ber)

Fas

NC

NC

NC

NC

NC

2.4

NC

NC

3.0

1418

340_

atFc

rece

ptor

,Ig

E,

high

-affi

nity

I,�

poly

pept

ide

Fcer

1gN

CN

CN

C5.

46.

09.

52.

53.

38.

414

1787

6_at

Fcre

cept

or,

IgG

,hi

gh-a

ffini

tyI

Fcgr

1N

CN

C2.

214

8.0

483.

33.

051

1451

941_

a_at

Fcre

cept

or,

IgG

,lo

w-a

ffini

tyII

bFc

gr2b

NC

�3.

0N

C4.

42.

47.

32.

12.

69.

214

4862

0_at

Fcre

cept

or,

IgG

,lo

w-a

ffini

tyII

IFc

gr3

NC

NC

NC

4.9

5.7

6.9

2.7

3.2

7.4

1422

903_

atL

ymph

ocyt

eA

g86

Ly8

6/M

D-1

NC

NC

NC

6.3

8.7

113.

84.

112

1449

874_

atL

ymph

ocyt

eA

g96

Ly9

6/M

D-2

NC

NC

NC

NC

NC

3.5

NC

NC

4.3

1449

184_

atPe

ptid

ogly

can

reco

gniti

onpr

otei

n1

Pgly

rp1

NC

NC

NC

NC

NC

3.2

NC

NC

NC

1417

860_

a_at

Spon

din

2,ex

trac

ellu

lar

mat

rix

prot

ein

Spon

2N

CN

CN

CN

CN

CN

CN

C2.

14.

714

4904

9_at

TL

R1

Tlr

1N

CN

CN

C13

NC

203.

4N

C23

1419

132_

atT

LR

2T

lr2

NC

NC

NC

6.1

4.3

7.4

3.4

3.2

1314

2278

1_at

TL

R3

Tlr

3N

CN

CN

CN

CN

C2.

7N

CN

C4.

614

1816

3_at

TL

R4

Tlr

4N

CN

CN

CN

CN

C2.

1N

CN

C2.

614

2135

2_at

TL

R6

Tlr

6N

CN

CN

C2.

1N

C2.

8N

CN

C3.

714

4947

3_s_

atT

NF

rece

ptor

supe

rfam

ily,

mem

ber

5T

nfrs

f5N

CN

CN

C2.

12.

46.

1N

CN

C10

(Tab

leco

ntin

ues)

7938 NOVEL PATHWAYS IN B. burgdorferi INFECTION

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

IL15, and TNF family members), indicating that B. burgdorferidoes induce these cytokines in the joint but that in the presence ofIL-10 their levels are quite low (Table IV and Supplement 3). Infact, RT-PCR, which is much more sensitive than microarray anal-ysis, did reveal the presence of extremely low levels of transcriptsfor IL-6, TNF, IFN-�, and IFN-� in the joint tissues of infectedC3H and C57BL/6 mice, with levels below detection in uninfectedtissues (data not shown). These results may explain the failure todetect significant changes by microarray. Taken together, they sug-gest that whereas the contribution of proinflammatory cytokinesappears to be tightly controlled in wild-type mice, it is not a dom-inant feature of the localized response to infection in joint tissue.Interestingly, RT-PCR analysis of spleen samples from infectedmice also revealed very low levels of these proinflammatory cy-tokines and of the IFN-inducible genes Igtp and Iipt (data notshown).

B. burgdorferi infection activates chondrocytes and triggersreactive responses in C3H mice

The most severe manifestation of arthritis in C3H mice includesevidence of reactive processes such as the formation of foci of newchondrocytes and new bone formation at 4 wk of infection (4, 9).Microarray revealed modest increases in gene products associatedwith chondrocytes from the joints of 4 wk postinfection C3Hmice, including several collagen genes, such as Ctsk and Dspg3(Table V). Quantification of Col1�2 by RT-PCR demonstrated arange of expression, and in this case only a portion of the individ-ual samples (3 of 5) showed increased expression of this gene latein disease. Future studies will be required to determine whetherassociations between gene induction and clinical markers of dis-ease can be made. Interesting, many of these genes associated withend-stage arthritis are also induced in rodent models of rheumatoidarthritis (41).

In agreement with findings reported by others (42, 43), tran-scripts for Mmp3, Mmp9, Mmp13, and Mmp23, as well as tissueinhibitor of matrix metalloproteinase 1 (Timp1) were induced inthe joints of C3H mice, whereas transcripts for Mmp3, Mmp8, andTimp1 were increased in C57BL/6-IL-10�/� mice (Table V).Mmp3 expression was also increased in joints of C57BL/6 mice,but to a lesser extent than in C3H or C57BL/6-IL-10�/� mice.Expression profiles of several of the metalloproteinases were con-firmed by RT-PCR with individual mice, with results for Mmp3shown in Fig. 4J.

DiscussionThe murine model of Lyme disease initially developed by Bartholdet al. (8) has provided a unique opportunity to study the localizedresponse to a similar infectious challenge in tissues that will ulti-mately develop distinct severities of arthritis (9, 19). The responseto invading bacteria at 1 wk of infection precedes the influx ofinflammatory cells and revealed a dramatic difference in the local-ized response. Although C3H mice displayed a robust responsedominated by IFN-inducible transcripts, these transcripts were ab-sent at 1 wk from infected C57BL/6 mice (Table II and Figs. 1 and2). Instead, the C57BL/6 mouse displayed an increase in an en-tirely distinct and unexpected group of genes previously associatedwith epidermal differentiation and wound repair (Table III). TheC57BL/6-IL-10�/� mouse shared expression profiles with theC3H mouse, implicating the IFN response in arthritis developmentrather than in a strain-specific response to infection, and suggestingthat suppression of the IFN response in C57BL/6 mice was crucialto suppression of arthritis (Figs. 1 and 3).

A second surprising feature of the early C3H profile was thedown-regulation of a large number of genes, many of which wereT

able

IV.

(Con

tinu

ed)

Prob

eSe

tID

Gen

eT

itle

Gen

eSy

mbo

l

1w

k2

wk

4w

k

C57

BL

/6C

3H/H

eNIL

10�

/�C

57B

L/6

C3H

/HeN

IL10

�/�

C57

BL

/6C

3H/H

eNIL

10�

/�

Sign

alin

gm

olec

ules

1423

048_

a_at

Tol

l-in

tera

ctin

gpr

otei

nT

ollip

NC

NC

3.3

NC

NC

NC

NC

NC

2.0

1419

272_

atM

yelo

iddi

ffer

entia

tion

prim

ary

resp

onse

gene

88M

yd88

NC

NC

NC

NC

NC

4.5

NC

NC

4.9

1417

856_

atA

vian

retic

uloe

ndot

helio

sis

vira

l(v

-rel

)on

coge

ne-r

elat

edB

Rel

bN

CN

CN

CN

CN

C3.

2N

CN

C3.

814

1811

0_a_

atIn

osito

lpo

lyph

osph

ate-

5-ph

osph

atas

eD

Inpp

5dN

CN

CN

C3.

43.

65.

5N

CN

C4.

614

2293

2_a_

atV

av1

onco

gene

Vav

1N

CN

CN

C7.

42.

45.

92.

2N

C6.

3C

ell-

cell

inte

ract

ions

/lect

ins

1422

013_

atC

-typ

ele

ctin

dom

ain

fam

ily4,

mem

ber

a2C

lec4

a2N

CN

CN

C15

1412

4.2

4.3

9.5

1420

804_

s_at

C-t

ype

lect

indo

mai

nfa

mily

4,m

embe

rd

Cle

c4d

NC

NC

NC

1316

798.

08.

486

1420

330_

atC

-typ

ele

ctin

dom

ain

fam

ily4,

mem

ber

eC

lec4

eN

CN

CN

C3.

72.

265

NC

NC

6614

2595

1_a_

atC

-typ

ele

ctin

dom

ain

fam

ily4,

mem

ber

nC

lec4

nN

CN

CN

CN

CN

C12

2.1

2.8

1614

1962

7_s_

atC

-typ

ele

ctin

dom

ain

fam

ily4,

mem

ber

nC

lec4

nN

CN

CN

CN

C4.

77.

02.

05.

210

1421

366_

atC

-typ

ele

ctin

dom

ain

fam

ily5,

mem

ber

aC

lec5

aN

CN

CN

C2.

1N

C4.

2N

CN

C5.

714

1969

3_at

Col

lect

insu

bfam

ilym

embe

r12

Col

ec12

NC

NC

NC

NC

NC

�2.

3N

CN

CN

C14

2213

3_at

Sial

opho

rin

Spn

NC

NC

NC

2.1

NC

3.6

NC

NC

4.6

Imm

une

deve

lopm

ent

1417

976_

atA

deno

sine

deam

inas

eA

daN

CN

CN

CN

C4.

24.

2N

C2.

54.

014

2492

3_at

Seri

ne(o

rcy

stei

ne)

prot

eina

sein

hibi

tor,

clad

eA

,m

embe

r3G

Serp

ina3

gN

C6.

62.

18.

42.

111

9N

CN

C11

414

4936

1_at

T-b

ox21

Tbx

21/T

bet

NC

NC

NC

NC

NC

2.7

NC

NC

2.4

aG

ene

expr

essi

onin

ankl

ejo

int

tissu

eat

1,2,

and

4w

kpo

stin

fect

ion

was

com

pare

dto

unin

fect

edan

kle

join

ttis

sue.

Num

bers

indi

cate

fold

chan

ge.

Num

bers

prec

eded

bya

“_”

are

redu

ced

com

pare

dto

unin

fect

edan

kle

join

ttis

sue.

Cha

nges

�2-

fold

are

desi

gnat

ed“N

C.”

7939The Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

in the same group of epidermal genes that were increased inC57BL/6 mice. More striking was the fact that expression of thesegenes was also reduced in C57BL/6-IL-10�/� mice, similar toC3H mice, and again arguing that regulation of expression of theepidermal profile was related to the development of inflammatoryarthritis, not due to strain-specific differences between C3H andC57BL/6 mice (Table III and Figs. 1 and 3). Expression of severalof the epidermal profile genes, including filaggrin, loricrin, andcytokeratins, can be down-regulated by endogenous peptide ligandsfor nicotinic acetylcholine receptors, providing precedent for reducedexpression following signaling pathway activation (44).

It cannot be determined from our results which IFN is respon-sible for the intense induction of this group of transcripts in C3Hmice, or even the possible role of another pleotropic cytokine suchas TNF-� (45). Studies with mice lacking IFN-� indicate that it isnot required for arthritis development (46); however, it is possiblethat there are compensatory pathways acting in the gene ablationmodel. Based on the known inflammatory potential of B. burgdor-feri, the production of either type I IFN or IFN-� by the milieu ofthe joint tissue could certainly occur by 1 wk of infection (17, 47,48). Alternatively, the early presence the dendritic cell-specifictranscript Oasl2 (Table II) in C3H mice may be highly relevantdue to the importance of dendritic cells as a sources of type I IFN(49, 50). In clinical trials, treatment with type I IFN has resulted intransient arthritis in patients with multiple sclerosis and hepatitis Cinfection, providing precedent for involvement of IFN in inflam-matory arthritis development (51, 52). Additional experiments willbe required to identify the cellular source and identity of the cy-tokine responsible for the IFN-inducible profile.

Additional experiments will also be required to detail the inter-play between the epidermal profile and inflammatory responses.Several published reports provide some insight as to a connectionbetween altered epidermal gene expression profiles in the joints ofB. burgdorferi-infected mice and regulation of the inflammatoryresponse. Normal differentiation of the epidermis during develop-ment requires the I�B kinase IKK1 (53), a key component of theclassical and alternative NF-�B-signaling pathways, implicatingIKK1 as a common link between regulation of epidermal differ-entiation genes and inflammation. Additionally, mice with a kera-tinocyte-specific deletion of the inflammation-associated transcrip-tion factor AP-1 displayed altered expression of epidermal

differentiation genes and the development of both localized andsystemic inflammatory lesions including psoriasis and psoriatic ar-thritis (54). Because B. burgdorferi achieve high levels in the skinof infected mice, it is also possible that dissemination through theskin promotes increased trafficking of cells that are not normallypresent in the joint tissue, and that this trafficking influences theoverall inflammatory status of the joint (55, 56). Finally, in rheu-matoid arthritis, citrullinated proteins such as filaggrin have beenidentified as targets for autoantibody; however, it is only the cit-rullinated species that are recognized by rheumatoid arthritis se-rum, and unmodified filaggrin is actually not expressed in jointtissue (57).

The 2-wk profile revealed unexpected similarities in genes in-creased in C3H and C57BL/6 mice (Table IV and Fig. 3). Numer-ous chemokines were induced in both wild-type strains of mice,without correlation with the greater inflammatory cell infiltrateseen in C3H mice (8, 15). In contrast, the robust production ofNF-�B-dependent inflammatory cytokines seen with cultured cellsin vitro and during infection of animals and patients was absentfrom localized response in joint tissues of C3H and C57BL/6 mice(2, 12, 48). The lack of differentially regulated genes downstreamof the TLR2, MyD88, and NF-�B-driven response to B. burgdor-feri was unexpected and suggests that activation of this pathwayoccurs independently from the development and regulation ofLyme arthritis. The possibility that the NF-�B pathway is a betterindicator of the host response to infection rather than a determinantof arthritis severity is consistent with the heightened expression ofcytokines and more effective host defense in IL-10�/� mice (Fig.1B and Table IV). These findings should also be interpreted in lightof previous studies with TLR2-deficient mice: these mice harborhigh levels of spirochetes in tissues and develop severe arthritiseven though isolated cells are defective in production of cytokinesin response to B. burgdorferi (58). The lack of differential expres-sion of the NF-�B-dependent cytokines in B. burgdorferi-infectedC3H and C57BL/6 mice is consistent with the observed develop-ment of Lyme arthritis in TLR2�/� C3H. The possible implicationof type I IFN in arthritic responses is also consistent with Lymearthritis development in the TLR2�/� mouse, because type I IFNis not a target of TLR2 signaling and its production should proceednormally in the absence of TLR2 (50, 58, 59).

Table V. Expression levels of gene transcripts indicative of reactive processes and/or activation of chondrocytes by B. burgdorferia

Probe Set ID Gene TitleGene

Symbol

2 wk 4 wk

C57BL/6 C3H/HeN IL10�/� C57BL/6 C3H/HeN IL10�/�

MMPs and TIMPs1416136_at MMP 2 Mmp2 NC NC NC NC 3.3 NC1418945_at MMP 3 Mmp3 7.7 18 34 5.3 13 451449366_at MMP 8 Mmp8 NC NC 4.9 NC NC 2.11448291_at MMP 9 Mmp9 NC 3.9 NC NC NC NC1417256_at MMP 13 Mmp13 NC 10 NC NC 13 3.01417282_at MMP 23 Mmp23 NC 2.4 NC NC 2.6 NC1460227_at TIMP 1 Timp1 NC 10 4.0 NC 5.8 3.8

Chondrocyte related1423606_at Periostin, osteoblast-specific factor Postn NC 6.3 NC NC 6.0 NC1418365_at Cathepsin H Ctsh 2.5 6.2 5.7 2.1 3.7 5.91450652_at Cathepsin K Ctsk NC 2.7 NC NC 3.8 NC1421114_a_at Dermatan sulphate proteoglycan 3 Dspg3 NC 2.8 NC NC 3.7 NC1423607_at Lumican Lum NC 3.7 NC NC 6.6 NC1423110_at Procollagen, type I, � 2 Col1a2 NC 2.1 NC NC 2.3 NC1427884_at Procollagen, type III, � 1 Col3a1 NC 8.8 NC NC 7.9 NC1450625_at Procollagen, type V, � 2 Col5a2 NC 3.7 NC NC 4.6 NC

a Gene expression in ankle joint tissue at 2 and 4 wk postinfection was compared to uninfected ankle joint tissue. Numbers indicate fold change. Changes �2-fold aredesignated “NC.”

7940 NOVEL PATHWAYS IN B. burgdorferi INFECTION

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Global gene expression analysis led to the identification of pre-viously unrecognized differences in expression profiles in thejoints of C3H and C57BL/6 mice in response to the same stimulus:invading B. burgdorferi. This justifies our strategy of assessingexpression in the complex joint tissue and provides new insightinto the development of this inflammatory pathology. The differ-ences in gene expression between the arthritis-susceptible (C3Hand IL-10�/�) and arthritis-resistant (C57BL/6) strains was evi-dent by 1 wk of infection, suggesting that a very early discrimi-nating event occurs in each strain that determines the outcome ofinfection. One interpretation of these results is that in the presenceof high levels of chemokines, the IFN response of C3H mice pro-motes leukocytic infiltration resulting in inflammation and arthri-tis; whereas in C57BL/6 mice, some aspect of the epidermal/wound profile prevents leukocytic infiltration and inflammation,thus limiting arthritis. These early initial events determine the out-come of arthritis severity even in the presence of similarly highlevels of bacteria, chemokine production, and proinflammatory cy-tokine production.

Future investigations will be directed toward a mechanistic under-standing of the involvement of these novel pathways in regulation ofthe development of Lyme arthritis. Additionally, experiments de-signed to assess the participation of QTL regulating the severity ofmurine Lyme arthritis in control of these distinct responses mayprovide further insight into the genetic and biochemical regulationof arthritis development (10, 11).

AcknowledgmentsWe acknowledge the helpful comments of Liming Yang from the NationalInstitutes of Health, and Joseph Holden, Lorise Gharing, Scott Rogers, andall the members of the Weis laboratory at the University of Utah.

DisclosuresThe authors have no financial conflict of interest.

References1. Burgdorfer, W., A. G. Barbour, S. F. Hayes, J. L. Benach, E. Grunwaldt, and

J. P. Davis. 1982. Lyme disease: a tick-borne spirochetosis? Science 216:1317–1319.

2. Steere, A. C., J. Coburn, and L. Glickstein. 2004. The emergence of Lyme dis-ease. J. Clin. Invest. 113: 1093–1101.

3. Nocton, J. J., F. Dressler, B. J. Rutledge, P. N. Rys, D. H. Persing, andA. C. Steere. 1994. Detection of Borrelia burgdorferi DNA by polymerase chainreaction in synovial fluid from patients with Lyme arthritis. N. Engl. J. Med. 330:229–234.

4. Yang, L., J. H. Weis, E. Eichwald, C. P. Kolbert, D. H. Persing, and J. J. Weis.1994. Heritable susceptibility to severe Borrelia burgdorferi-induced arthritis isdominant and is associated with persistence of large numbers of spirochetes intissues. Infect. Immun. 62: 492–500.

5. Steere, A. C., R. T. Schoen, and E. Taylor. 1987. The clinical evolution of Lymearthritis. Ann. Intern. Med. 107: 725–731.

6. Gross, D. M., T. Forsthuber, M. Tary-Lehmann, C. Etling, K. Ito, Z. A. Nagy,J. A. Field, A. C. Steere, and B. T. Huber. 1998. Identification of LFA-1 as acandidate autoantigen in treatment-resistant Lyme arthritis. Science 281:703–706.

7. Barthold, S. W., D. H. Persing, A. L. Armstrong, and R. A. Peeples. 1991.Kinetics of Borrelia burgdorferi dissemination and evolution of disease afterintradermal inoculation of mice. Am. J. Pathol. 139: 263–273.

8. Barthold, S. W., D. S. Beck, G. M. Hansen, G. A. Terwilliger, and K. D. Moody.1990. Lyme borreliosis in selected strains and ages of laboratory mice. J. Infect.Dis. 162: 133–138.

9. Ma, Y., K. P. Seiler, E. J. Eichwald, J. H. Weis, C. Teuscher, and J. J. Weis. 1998.Distinct characteristics of resistance to Borrelia burgdorferi-induced arthritis inC57BL/6N mice. Infect. Immun. 66: 161–168.

10. Roper, R. J., J. J. Weis, B. A. McCracken, C. B. Green, Y. Ma, K. S. Weber,D. Fairbairn, R. J. Butterfield, M. R. Potter, J. F. Zachary, et al. 2001. Geneticcontrol of susceptibility to experimental Lyme arthritis is polygenic and exhibitsconsistent linkage to multiple loci on chromosome 5 in four independent mousecrosses. Genes Immun. 2: 388–397.

11. Weis, J. J., B. A. McCracken, Y. Ma, D. Fairbairn, R. J. Roper, T. B. Morrison,J. H. Weis, J. F. Zachary, R. W. Doerge, and C. Teuscher. 1999. Identification ofquantitative trait loci governing arthritis severity and humoral responses in themurine model of Lyme disease. J. Immunol. 162: 948–956.

12. Wooten, R. M., and J. J. Weis. 2001. Host-pathogen interactions promoting in-flammatory Lyme arthritis: use of mouse models for dissection of disease pro-cesses. Curr. Opin. Microbiol. 4: 274–279.

13. Hirschfeld, M., C. J. Kirschning, R. Schwandner, H. Wesche, J. H. Weis,R. M. Wooten, and J. J. Weis. 1999. Cutting edge: inflammatory signaling byBorrelia burgdorferi lipoproteins is mediated by Toll-like receptor 2. J. Immunol.163: 2382–2386.

14. Aliprantis, A. O., R. B. Yang, M. R. Mark, S. Suggett, B. Devaux, J. D. Radolf,G. R. Klimpel, P. Godowski, and A. Zychlinsky. 1999. Cell activation and ap-optosis by bacterial lipoproteins through Toll-like receptor-2. Science 285:736–739.

15. Brown, C. R., V. A. Blaho, and C. M. Loiacono. 2003. Susceptibility to exper-imental Lyme arthritis correlates with KC and monocyte chemoattractant pro-tein-1 production in joints and requires neutrophil recruitment via CXCR2. J. Im-munol. 171: 893–901.

16. Brown, J. P., J. F. Zachary, C. Teuscher, J. J. Weis, and R. M. Wooten. 1999.Dual role of interleukin-10 in murine Lyme disease: regulation of arthritis se-verity and host defense. Infect. Immun. 67: 5142–5150.

17. Ganapamo, F., V. A. Dennis, and M. T. Philipp. 2000. Early induction of �interferon and interleukin-10 production in draining lymph nodes from mice in-fected with Borrelia burgdorferi. Infect. Immun. 68: 7162–7165.

18. Ganapamo, F., V. A. Dennis, and M. T. Philipp. 2001. CD19 cells produceIFN-� in mice infected with Borrelia burgdorferi. Eur. J. Immunol. 31:3460–3468.

19. Brown, C. R., and S. L. Reiner. 1998. Clearance of Borrelia burgdorferi may notbe required for resistance to experimental Lyme arthritis. Infect. Immun. 66:2065–2071.

20. Morrison, T. B., Y. Ma, J. H. Weis, and J. J. Weis. 1999. Rapid and sensitivequantification of Borrelia burgdorferi-infected mouse tissues by continuous flu-orescent monitoring of PCR. J. Clin. Microbiol. 37: 987–992.

21. Chirgwin, J. M., A. E. Przybyla, R. J. MacDonald, and W. J. Rutter. 1979.Isolation of biologically active ribonucleic acid from sources enriched in ribo-nuclease. Biochemistry 18: 5294–5299.

22. GeneChip Expression Analysis Technical Manual. 2005. Affymetrix Incorpo-rated, Santa Clara.

23. Bolstad, B. M., R. A. Irizarry, M. Astrand, and T. P. Speed. 2003. A comparisonof normalization methods for high density oligonucleotide array data based onvariance and bias. Bioinformatics 19: 185–193.

24. Irizarry, R. A., B. M. Bolstad, F. Collin, L. M. Cope, B. Hobbs, and T. P. Speed.2003. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res.31: e15.

25. Gautier, L., L. Cope, B. M. Bolstad, and R. A. Irizarry. 2004. affy: analysis ofAffymetrix GeneChip data at the probe level. Bioinformatics 20: 307–315.

26. Tusher, V. G., R. Tibshirani, and G. Chu. 2001. Significance analysis of microar-rays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 98:5116–5121.

27. Eppig, J. T., C. J. Bult, J. A. Kadin, J. E. Richardson, J. A. Blake,A. Anagnostopoulos, R. M. Baldarelli, M. Baya, J. S. Beal, S. M. Bello, et al.2005. The Mouse Genome Database (MGD): from genes to mice: a communityresource for mouse biology. Nucleic Acids Res. 33: D471–D475.

28. Hubbard, T., D. Andrews, M. Caccamo, G. Cameron, Y. Chen, M. Clamp,L. Clarke, G. Coates, T. Cox, F. Cunningham, et al. 2005. Ensembl 2005. NucleicAcids Res. 33: D447–D453.

29. Ornstein, K., and A. G. Barbour. 2006. A reverse transcriptase-polymerase chainreaction assay of Borrelia burgdorferi 16S rRNA for highly sensitive quantifi-cation of pathogen load in a vector. Vector Borne Zoonotic Dis. 6: 103–112.

30. Kendziorski, C., R. A. Irizarry, K. S. Chen, J. D. Haag, and M. N. Gould. 2005.On the utility of pooling biological samples in microarray experiments. Proc.Natl. Acad. Sci. USA 102: 4252–4257.

31. Hodzic, E., S. Feng, K. J. Freet, and S. W. Barthold. 2003. Borrelia burgdorferipopulation dynamics and prototype gene expression during infection of immu-nocompetent and immunodeficient mice. Infect. Immun. 71: 5042–5055.

32. Moore, K. W., R. de Waal Malefytqq, R. L. Coffman, and A. O’Garra. 2001.Interleukin-10 and the interleukin-10 receptor. Annu. Rev. Immunol. 19: 683–765.

33. Platanias, L. C. 2005. Mechanisms of type-I- and type-II-interferon-mediatedsignalling. Nat. Rev. Immunol. 5: 375–386.

34. Mack, J. A., S. Anand, and E. V. Maytin. 2005. Proliferation and cornificationduring development of the mammalian epidermis. Birth Defects Res. C EmbryoToday 75: 314–329.

35. Wojcik, S. M., D. S. Bundman, and D. R. Roop. 2000. Delayed wound healingin keratin 6a knockout mice. Mol. Cell. Biol. 20: 5248–5255.

36. Brakenhoff, R. H., M. Gerretsen, E. M. Knippels, M. van Dijk, H. van Essen,D. O. Weghuis, R. J. Sinke, G. B. Snow, and G. A. van Dongen. 1995. The humanE48 antigen, highly homologous to the murine Ly-6 antigen ThB, is a GPI-anchored molecule apparently involved in keratinocyte cell-cell adhesion. J. CellBiol. 129: 1677–1689.

37. Esche, C., C. Stellato, and L. A. Beck. 2005. Chemokines: key players in innateand adaptive immunity. J. Invest. Dermatol. 125: 615–628.

38. Narayan, K., D. Dail, L. Li, D. Cadavid, S. Amrute, P. Fitzgerald-Bocarsly, andA. R. Pachner. 2005. The nervous system as ectopic germinal center: CXCL13and IgG in Lyme neuroborreliosis. Ann. Neurol. 57: 813–823.

39. Wooten, R. M., V. R. Modur, T. M. McIntyre, and J. J. Weis. 1996. Borreliaburgdorferi outer membrane protein A induces nuclear translocation of nuclearfactor-�B and inflammatory activation in human endothelial cells. J. Immunol.157: 4584–4590.

40. Sellati, T. J., D. A. Bouis, R. L. Kitchens, R. P. Darveau, J. Pugin, R. J. Ulevitch,S. C. Gangloff, S. M. Goyert, M. V. Norgard, and J. D. Radolf. 1998. Treponema

7941The Journal of Immunology

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

pallidum and Borrelia burgdorferi lipoproteins and synthetic lipopeptides acti-vate monocytic cells via a CD14-dependent pathway distinct from that used bylipopolysaccharide. J. Immunol. 160: 5455–5464.

41. Ibrahim, S. M., D. Koczan, and H. J. Thiesen. 2002. Gene-expression profile ofcollagen-induced arthritis. J. Autoimmun. 18: 159–167.

42. Behera, A. K., E. Hildebrand, J. Scagliotti, A. C. Steere, and L. T. Hu. 2005.Induction of host matrix metalloproteinases by Borrelia burgdorferi differs inhuman and murine Lyme arthritis. Infect. Immun. 73: 126–134.

43. Gebbia, J. A., J. L. Coleman, and J. L. Benach. 2001. Borrelia spirochetes up-regulate release and activation of matrix metalloproteinase gelatinase B (MMP-9)and collagenase 1 (MMP-1) in human cells. Infect. Immun. 69: 456–462.

44. Arredondo, J., A. I. Chernyavsky, D. L. Jolkovsky, R. J. Webber, and S. A. Grando.2006. SLURP-2: a novel cholinergic signaling peptide in human mucocutaneousepithelium. J. Cell. Physiol. 208: 238–245.

45. Gallagher, J., J. Howlin, C. McCarthy, E. P. Murphy, B. Bresnihan, O. FitzGerald,C. Godson, H. R. Brady, and F. Martin. 2003. Identification of Naf1/ABIN-1 amongTNF-�-induced expressed genes in human synoviocytes using oligonucleotide mi-croarrays. FEBS Lett. 551: 8–12.

46. Brown, C. R., and S. L. Reiner. 1999. Experimental Lyme arthritis in the absenceof interleukin-4 or � interferon. Infect. Immun. 67: 3329–3333.

47. Ma, Y., K. P. Seiler, K. F. Tai, L. Yang, M. Woods, and J. J. Weis. 1994. Outersurface lipoproteins of Borrelia burgdorferi stimulate nitric oxide production bythe cytokine-inducible pathway. Infect. Immun. 62: 3663–3671.

48. Kang, I., S. W. Barthold, D. H. Persing, and L. K. Bockenstedt. 1997. T-helper-cell cytokines in the early evolution of murine Lyme arthritis. Infect. Immun. 65:3107–3111.

49. Tiefenthaler, M., R. Marksteiner, S. Neyer, F. Koch, S. Hofer, G. Schuler,M. Nussenzweig, R. Schneider, and C. Heufler. 1999. M1204, a novel 2�,5�oligoadenylate synthetase with a ubiquitin-like extension, is induced during mat-uration of murine dendritic cells. J. Immunol. 163: 760–765.

50. Honda, K., H. Yanai, A. Takaoka, and T. Taniguchi. 2005. Regulation of the typeI IFN induction: a current view. Int. Immunol. 17: 1367–1378.

51. Strueby, L., B. Nair, A. Kirk, and R. M. Taylor-Gjevre. 2005. Arthritis andbursitis in multiple sclerosis patients treated with interferon-�. Scand. J. Rheu-matol. 34: 485–488.

52. Wilson, L. E., D. Widman, S. H. Dikman, and P. D. Gorevic. 2002. Autoimmunedisease complicating antiviral therapy for hepatitis C virus infection. Semin. Ar-thritis Rheum. 32: 163–173.

53. Li, Q., Q. Lu, J. Y. Hwang, D. Buscher, K. F. Lee, J. C. Izpisua-Belmonte, andI. M. Verma. 1999. IKK1-deficient mice exhibit abnormal development of skinand skeleton. Genes Dev. 13: 1322–1328.

54. Zenz, R., R. Eferl, L. Kenner, L. Florin, L. Hummerich, D. Mehic, H. Scheuch,P. Angel, E. Tschachler, and E. F. Wagner. 2005. Psoriasis-like skin disease andarthritis caused by inducible epidermal deletion of Jun proteins. Nature 437:369–375.

55. Sellati, T. J., S. L. Waldrop, J. C. Salazar, P. R. Bergstresser, L. J. Picker, andJ. D. Radolf. 2001. The cutaneous response in humans to Treponema pallidumlipoprotein analogues involves cellular elements of both innate and adaptive im-munity. J. Immunol. 166: 4131–4140.

56. Norgard, M. V., B. S. Riley, J. A. Richardson, and J. D. Radolf. 1995. Dermalinflammation elicited by synthetic analogs of Treponema pallidum and Borreliaburgdorferi lipoproteins. Infect. Immun. 63: 1507–1515.

57. Baeten, D., I. Peene, A. Union, L. Meheus, M. Sebbag, G. Serre, E. M. Veys, andF. De Keyser. 2001. Specific presence of intracellular citrullinated proteins inrheumatoid arthritis synovium: relevance to antifilaggrin autoantibodies. ArthritisRheum. 44: 2255–2262.

58. Wooten, R. M., Y. Ma, R. A. Yoder, J. P. Brown, J. H. Weis, J. F. Zachary,C. J. Kirschning, and J. J. Weis. 2002. Toll-like receptor 2 is required for innate,but not acquired, host defense to Borrelia burgdorferi. J. Immunol. 168:348–355.

59. Wang, G., Y. Ma, A. Buyuk, S. McClain, J. J. Weis, and I. Schwartz. 2004.Impaired host defense to infection and Toll-like receptor 2-independent killing ofBorrelia burgdorferi clinical isolates in TLR2-deficient C3H/HeJ mice. FEMSMicrob. Lett. 231: 219–225.

7942 NOVEL PATHWAYS IN B. burgdorferi INFECTION

by guest on June 20, 2018http://w

ww

.jimm

unol.org/D

ownloaded from