of May 2, 2014.This information is current as

against MelanomaDisease Progression or Associated ProtectionOngoing Priming of Naive CD8 T Cells for Autoimmune Vitiligo Does Not Require the

and Mary Jo TurkKatelyn T. Byrne, Peisheng Zhang, Shannon M. Steinberg

http://www.jimmunol.org/content/192/4/1433doi: 10.4049/jimmunol.1302139January 2014;

2014; 192:1433-1439; Prepublished online 8J Immunol 

MaterialSupplementary

9.DCSupplemental.htmlhttp://www.jimmunol.org/content/suppl/2014/01/07/jimmunol.130213

Referenceshttp://www.jimmunol.org/content/192/4/1433.full#ref-list-1

, 16 of which you can access for free at: cites 31 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

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

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2014 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

at D

artmouth C

ollege on May 2, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Autoimmune Vitiligo Does Not Require the Ongoing Primingof Naive CD8 T Cells for Disease Progression or AssociatedProtection against Melanoma

Katelyn T. Byrne, Peisheng Zhang, Shannon M. Steinberg, and Mary Jo Turk

Vitiligo is a CD8 T cell–mediated autoimmune disease that has been shown to promote the longevity of memory T cell responses to

melanoma. However, mechanisms whereby melanocyte/melanoma Ag-specific T cell responses are perpetuated in the context of

vitiligo are not well understood. These studies investigate the possible phenomenon of naive T cell priming in hosts with

melanoma-initiated, self-perpetuating, autoimmune vitiligo. Using naive pmel (gp10025–33–specific) transgenic CD8 T cells, we

demonstrate that autoimmune melanocyte destruction induces naive T cell proliferation in skin-draining lymph nodes, in an Ag-

dependent fashion. These pmel T cells upregulate expression of CD44, P-selectin ligand, and granzyme B. However, they do not

downregulate CD62L, nor do they acquire the ability to produce IFN-g, indicating a lack of functional priming. Accordingly, adult

thymectomized mice exhibit no reduction in the severity or kinetics of depigmentation or long-lived protection against melanoma,

indicating that the continual priming of naive T cells is not required for vitiligo or its associated antitumor immunity. Despite this,

depletion of CD4 T cells during the course of vitiligo rescues the priming of naive pmel T cells that are capable of producing IFN-g

and persisting as memory, suggesting an ongoing and dominant mechanism of suppression by regulatory T cells. This work reveals

the complex regulation of self-reactive CD8 T cells in vitiligo and demonstrates the overall poorly immunogenic nature of this

autoimmune disease setting. The Journal of Immunology, 2014, 192: 1433–1439.

The autoimmune destruction of melanocytes, known as vit-iligo, has long been recognized as an independent posi-tive prognostic factor for melanoma patients, correlating

with improved overall and tumor-free survival rates (1–4). Ourwork has recently shown that vitiligo is also a key determinant forthe generation of long-lived memory CD8 T cell responses tomelanoma (5). We found that melanocyte Ags, which are liberatedduring the course of autoimmune vitiligo, are required to main-tain nonexhausted and functional memory CD8 T cell responsesagainst melanoma (5). Thus, there exists a causal relationship be-tween tissue-specific autoimmunity and the maintenance of immunityto cancer.Understanding the mechanisms whereby autoimmunity is per-

petuated is now an important component in understanding howantitumor immunity can be optimally maintained. However, theontogeny of melanocyte/melanoma Ag-specific T cells in hostswith vitiligo remains incompletely understood. Although we haveshown that vitiligo maintains populations of melanoma-primedCD8 T cells for many months as memory (5), it remains unclear

whether the ongoing destruction of melanocytes also drives thecontinual priming of new T cells from the naive pool. Such newlyprimed effectors could contribute to the pathogenesis of vitiligoand to melanoma tumor protection.Precedence exists for the recruitment of naive T cells during the

course of ongoing T cell responses against both self and non–self-Ags. After initiation of experimental autoimmune encephalomy-elitis with a single antigenic peptide, CD4 T cells with specificitiesfor additional epitopes have been detected (6, 7). Epitope spread-ing has also been observed during the course of CD8 T cell–mediated antitumor immunity (8–11). The priming of naive CD8T cells occurs during chronic infections involving polyoma virus(12, 13) and persistent MCMV (14), and newly primed effectorT cells are critical for maintaining viral immune surveillance. De-spite this, it has recently been suggested that CD8 T cell–mediatedtissue destruction is self-limiting. This is based on studies in miceexpressing OVA under the control of the rat insulin promoter,wherein pancreatic tissue destruction was initiated by transfer ofOVA-specific CD8 effector T cells (OT-1 cells) (15). The authorsfound that naive OT-I T cells underwent deletional tolerance whenencountering OVA liberated and cross-presented in draining lymphnodes of these mice (15). However, pancreas destruction resolvedwithout overt autoimmune disease (15). Thus, it remains unknownwhether ongoing CD8 T cell–mediated autoimmune disease caninduce the priming of naive self-Ag–specific T cells.The present studies investigate the priming of naive melanocyte/

melanoma Ag-specific T cells in mice with progressive, melanoma-initiated vitiligo. We use a model in which CD8 T cell–mediatedvitiligo is induced by regulatory T cell (Treg) depletion, followedby surgical excision of dermal B16 melanoma tumors (5, 16, 17).We report that naive Ag-specific CD8 T cells are driven to pro-liferate in hosts with ongoing vitiligo. However, these T cellsnever acquire full effector function, nor do they contribute tovitiligo progression or immunity against melanoma. Despite this,the depletion of CD4 T cells during the course of autoimmune

Department of Microbiology and Immunology, Geisel School of Medicine at Dart-mouth, Hanover, NH 03755; and Norris Cotton Cancer Center, Dartmouth-HitchcockMedical Center, Lebanon, NH 03756

Received for publication August 9, 2013. Accepted for publication December 4,2013.

This work was supported by the National Institutes of Health (Grant R01 CA120777-06 to M.J.T.; Grant T32 A107363 to K.T.B.; Grant T32 GM00874 to S.M.S.), theAmerican Cancer Society (Grant ACS RSG LIB-121864 to M.J.T.), the DartmouthImmunology Program (to K.T.B.), and the Joanna M. Nicolay Melanoma Foundation(to K.T.B.).

Address correspondence and reprint requests to Dr. Mary Jo Turk, Geisel School ofMedicine at Dartmouth, 1 Medical Center Drive, Rubin Building 732, Lebanon, NH03766. E-mail address: mary.jo.turk@dartmouth.edu

The online version of this article contains supplemental material.

Abbreviations used in this article: i.d., intradermally; Treg, regulatory T cell.

Copyright� 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1302139

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

disease can rescue the priming of naive CD8 T cells, resulting infunctional effector cells that are maintained as memory. Thesestudies elucidate the poorly immunogenic nature of CD8 T cell–mediated autoimmune vitiligo and illustrate a dominant mecha-nism of suppression that could be therapeutically manipulated inthis setting.

Materials and MethodsMice and tumor cell lines

Animal studies were reviewed and approved by the Dartmouth InstitutionalAnimal Care and Use Committee. All animal studies were in compliancewith the U.S. Department of Health and Human Services Guide for the Careand Use of Laboratory Animals. Male and female mice were used at 6–12wk of age. C57BL/6 mice (5–6 wk old) were obtained from Charles RiverLaboratories or The Jackson Laboratory. Pmel mice expressing a trans-genic TCR specific for gp10025–33 (a melanocyte differentiation Ag foundin melanosomes) in the context of H-2Db (18), on a congenic Thy1.1+

background, were originally a gift from Nicholas Restifo (National CancerInstitute). Pmel mice were also bred onto a Ly5.2 background, and Thy1.1or Ly5.2 congenically marked pmel cells were used interchangeably. OT-1mice (expressing a TCR recognizing OVA257–264 in the context of H-2Kb)were bred onto a congenic Ly5.2+ background. Homozygous C57BL/6-KitW-sh (Wsh) mice, which lack melanocytes (5), were purchased from TheJackson Laboratory and bred in-house. C57BL/6 thymectomized mice hadsurgical excision of the thymus performed at 6 wk of age, at The JacksonLaboratory, and were then shipped to Dartmouth together with age-matchedcontrol C57BL/6 mice.

The B16-F10 (B16) mouse melanoma cell line was originally obtainedfrom Isaiah Fidler (MD Anderson Cancer Center) and passaged intrader-mally (i.d.) in C57BL/6 mice seven times to ensure reproducible growthbefore use in these studies. Cell lines were tested by the Infectious MicrobePCR AmplifiCation Test (IMPACT) and authenticated by the Research andDiagnostics Laboratory at the University of Missouri. Melanoma cells werecultured in RPMI 1640 containing 7.5% FBS, harvested by brief tryp-sinization, and inoculated into mice i.d. Cells were used only if viabilityexceeded 96% upon harvest.

mAbs and peptides

Ab-producing hybridoma cell lines were obtained from American TypeCulture Collection. Depleting anti-CD4 (clone GK1.5) was produced asbioreactor supernatant and administered in doses of 250 mg i.p. More than98% depletion of target T cell populations was confirmed by flow cy-tometry. Peptides (.80% purity) were obtained from New EnglandPeptide: gp10025–33 (EGSRNQDWL) and OVA257–264 (SIINFEKL).

Induction of vitiligo

C57BL/6 mice were inoculated i.d. with 1.2 3 105 B16 cells on day 0 andthen treated with anti-CD4 mAb i.p. on days 4 and 10, as previously de-scribed (17, 19) and outlined in Fig. 1A. Only mice that developed primarytumors (.95%) were used. Primary tumors were surgically excised fromskin, with negative boundaries, on day 12. Spontaneous tumor metastaseswere not observed with this B16 subline, and mice with recurrent primarytumors after surgery (,5%) were removed from the study. After surgery,mice were monitored weekly for signs of overt vitiligo, defined as distinctpatches of white hair growth (Supplemental Fig. 1). As we have previouslyreported, ∼60% of mice develop melanoma-associated vitiligo within∼30 d after surgery, and the remaining ∼40% maintain a virtually unaf-fected appearance (5). Depigmentation was designated “local” if it wasconfined to the right flank from which the primary tumor had been excised(Supplemental Fig. 1), or “disseminated” if it was observed in sites beyondthe right flank.

Adoptive transfer and monitoring of pmel and OT-I T cells

Congenically marked CD8 T cells were isolated from combined lymphnodes and spleens of 6- to 8-wk-old, naive, Thy1.1+ or Ly5.2+ pmel mice, orLy5.2+ OT-I mice. Naive cells were isolated by magnetic purification(Miltenyi Biotec) involving anti–CD44-PE negative selection, followed byanti-CD8 positive selection. In proliferation experiments, cells were firstlabeled with CFSE at a concentration of 3 mM/ml, by incubation for 5–10min at room temperature followed by the addition of cold serum-containingmedium and repeated washes to remove free CFSE.

At various time points, 106 naive, transgenic T cells were adoptivelytransferred into vitiligo-affected, unaffected, naive, and Wsh recipients. Tendays after transfer (or 30 d where indicated), mice were euthanized and

inguinal lymph nodes (or spleens when indicated) were harvested andmechanically dissociated. Cell suspensions were stained with combina-tions of the following Abs: anti–CD8-PerCP (clone 53-6.7; BioLegend),anti–Thy1.1-PE, -allophycocyanin, or –PE-Cy7 (clone H1S51; eBio-science), anti–CD62L-FITC or -PE (clone MEL14; BD Pharmingen), anti–CD69-FITC (clone H1.2F3; BD Pharmingen), anti–CD25-PE (clone PC61;eBioscience), anti–granzyme B-PE (clone 16G6; eBioscience), and anti–CD44-FITC, -allophycocyanin, or –allophycocyanin-Cy7 (clone IM7;BioLegend). For detecting P-selectin ligand, cells were first incubated withanti–P-Sel-L (clone 4RA10; BD Pharmingen) and then stained with anti-Rat IgG-PE (Jackson Immunoresearch). As a positive control for CD69and CD25 staining, naive pmel splenocytes were cultured for 3 d in RPMIcontaining PHA (3 mg/ml final concentration). As a positive control foreffector pmel cells that express P-selectin ligand and produce granzyme B,mice received pmel cells 1 d before B16 tumor inoculation (day 0) andanti-CD4 treatment (days 4 and 7), and pmel responses in tumor-draininglymph nodes were analyzed on day 10. Flow cytometry was performed ona FACSCalibur, FACSCanto (BD Biosciences), or MACSQUANT (Mil-tenyi Biotec), and data were analyzed using FlowJo software (Tree Star).

Intracellular cytokine staining

Ten or 30 d after naive pmel T cell transfer into either naive or vitiligoaffected cohorts of mice, adoptively transferred mice were sacrificed andcells were harvested from lymph nodes. Lymphocytes were aliquoted into96-well plates and incubated for 5 h at 37˚C with mouse gp10025–33 orOVA257–264 (irrelevant) peptide (1 mg/ml), in RPMI containing IL-2 (10U/ml) and brefeldin A (10 mg/ml). After incubation, cells were washed,stained with Abs against CD8 and Thy1.1 or Ly5.2, and then fixed, per-meabilized, and stained with anti–IFN-g–allophycocyanin (clone XMG1.2;BioLegend). Flow cytometry was performed as described earlier. As apositive control for effector pmel cells that produce IFN-g, mice receivedpmel cells 1 d before B16 tumor inoculation (day 0) and anti-CD4 treat-ment (days 4 and 10), and pmel responses in tumor-draining lymph nodeswere analyzed on day 12.

Tumor challenge

A total of 1.2 3 105 live B16 cells was inoculated i.d., in the left flank,30 d after surgical excision of the primary tumor. Tumor diameters weremeasured thrice weekly, and mice were euthanized when tumor diametersreached 10 mm.

Statistical analyses

Statistical differences between groups analyzed by flow cytometry weredetermined by unpaired, Student two-tailed t test. A paired Student t testwas used for comparison between relevant and irrelevant peptide-specificresponses in the intracellular cytokine staining analysis. For experimentsinvolving a comparison between three or more distinct groups, a one-wayANOVA, with Bonferroni posttests was used. Data were considered sig-nificant if p # 0.05. Statistical differences in tumor-free survival andvitiligo incidence were determined by log-rank analysis of Kaplan–Meierdata, pooled over strata.

ResultsMelanocyte Ag drives the proliferation of naive CD8 T cells inhosts with autoimmune vitiligo

To determine whether autoimmune melanocyte destruction wascapable of initiating priming of self-Ag–specific CD8 T cells, thebehavior of naive transgenic T cells specific for gp10025-33 (pmelcells) was assessed in mice with melanoma-initiated vitiligo. CD8T cell–mediated vitiligo was induced by B16 tumor inoculation,followed by treatment with anti-CD4 to eliminate Tregs and surgeryto remove established tumors, as we have previously published (5,17) (Fig. 1A). Seventy-five days after surgery, mice were segregatedinto overtly vitiligo-affected and unaffected groups as described inMaterials and Methods (Supplemental Fig. 1). These mice werethen adoptively transferred with 106 naive (CD442 sorted) pmelcells that had been labeled with CFSE (Fig. 1A). Ten days afteradoptive transfer, pmel cells were identified in skin-draining (in-guinal) lymph nodes by expression of the congenic marker Thy1.1.Indeed, pmel cells transferred into vitiligo-affected hosts un-

derwent several rounds of division, with a significantly larger

1434 VITILIGO DOES NOT INDUCE NAIVE CD8 T CELL PRIMING

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

population of pmel cells dividing in vitiligo-affected mice ascompared with untreated (naive) control mice, identically treatedmice that never developed vitiligo (unaffected), or identicallytreated Wsh mice that lack melanocytes (Fig. 1C). Proliferationwas similar regardless of whether pmel cells were transferred 30,60, or 75 d after surgery (Fig. 1C). Thus, T cell proliferation re-quired the presence of both host melanocytes and progressiveautoimmune disease.Accumulation of proliferating pmel cells was also determined at

each of these time points. In all cases, the proportion of pmel cellsamong total CD8 T cells was significantly elevated in vitiligo-affected hosts as compared with naive hosts (Fig. 1D). However,population sizes were small, suggesting that proliferating pmelcells accumulated to a minimal extent. To determine whether pmelcell proliferation was also occurring systemically, we analyzedpmel cell proliferation in spleens after transfer on day 75. Ascompared with naive, unaffected, and Wsh negative control groups,we detected no significant proliferation of pmel cells in spleens ofvitiligo-affected hosts (Fig. 1E). Thus, naive pmel cells were ca-pable of proliferating and accumulating throughout the course ofvitiligo, but only in skin-draining lymph nodes.Autoimmunity is associated with the liberation of self-Ags, but

also with nonspecific inflammation and cytokine release. To de-termine whether T cell proliferation was melanocyte Ag-specific,

we adoptively transferred vitiligo-affected mice with naive, Ag-irrelevant OT-I cells. OT-I cells did not proliferate to a signifi-cant extent in vitiligo-affected hosts, having a CFSE profile thatwas indistinguishable from that of OT-I cells transferred intonaive hosts (Fig. 1F). Indeed, the proliferation of OT-I cells wassimilar in all groups of hosts, regardless of vitiligo status (Fig.1F). This indicated that the inflammatory environment associ-ated with autoimmune vitiligo was insufficient to drive T cellproliferation. Thus, naive CD8 T cells were driven to proliferatein vitiligo-affected mice specifically as a result of exposure tomelanocyte Ags.

Progressive vitiligo does not initiate the priming of functionalmelanocyte-specific CD8 T cells

We next assessed the phenotype of pmel cells in lymph nodes ofvitiligo-affected mice to determine the extent of functional priming.Ten days after adoptive transfer, ∼40% of pmel cells in vitiligo-affected hosts had acquired an Ag-experienced CD44hi phenotype(Fig. 2A). This was in comparison with naive, unaffected, and Wsh

control mice, all of which had very low proportions of CD44hi

pmel cells (Fig. 2A). Pmel cells acquired CD44 expression whentransferred at multiple time points throughout the course of viti-ligo (Fig. 2B). Within this Ag-experienced CD44hi pmel pop-ulation, we observed significant upregulation of the skin-homing

FIGURE 1. Naive pmel cells prolifer-

ate locally in skin draining lymph nodes

of vitiligo-affected mice, in an Ag-spe-

cific manner. (A) Schematic diagram; vit-

iligo was induced by inoculation of B16

tumors followed by CD4 depletion and

surgery. Mice were stratified into vitiligo-

affected and unaffected groups, which re-

ceived adoptive transfer of CFSE-labeled

naive T cells. (B) Ly5.2+ naive pmel cells

were adoptively transferred into indicated

mice 75 d after surgery, and inguinal lymph

nodes (LN) were analyzed 10 d later; his-

tograms (left) are gated on CD8+Ly5.2+

cells; values in graph (right) are based on

gating strategy shown on histogram at left.

(C) Pmel cells were adoptively transferred

at the indicated time points after surgery,

and flow was performed on LNs 10 d af-

ter each transfer; gating strategy as shown

in (B). (D) Mice were treated as in (B) and

gated on CD8+ cells to determine the pro-

portion of Ly5.2+ pmel cells. (E) Mice were

treated as in (B), and responses were ana-

lyzed 10 d later in spleen; gated on CD8+

Ly5.2+ cells. (F) Mice were treated as in (B)

but received Ly5.2+ OT-1 cells instead of

pmel cells; gated on CD8+Ly5.2+ cells.

Histograms depict representative data,

symbols represent individual mice, and

horizontal lines depict averages. Data

are representative of three to five repeat

experiments, each with three to six mice/

group. Statistical significance was deter-

mined by one-way ANOVA with Bon-

ferroni posttest; *p , 0.05, **p , 0.01,

***p , 0.001.

The Journal of Immunology 1435

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

marker P-selectin ligand, with levels equivalent to that of recentlyactivated effector pmel cells (Fig. 2C). Despite this, CD25 wasnot significantly upregulated, nor was CD62L significantlydownregulated on these cells (Fig. 2D). CD69 was upregulatedby a proportion of CD44hi pmel cells from vitiligo-affected mice,although this was equivalent to CD69 upregulation in pmel cellstaken from naive hosts, indicating no enhancement by vitiligo(Fig. 2D).

To discern the functional status of Ag-experienced pmel cells invitiligo-affected mice, we assessed granzyme B and IFN-g pro-duction. Whereas a significant proportion of CD44hi pmel cellsproduced granzyme B (Fig. 2E), IFN-g production was com-pletely absent from this population (Fig. 2F). Thus, the destructionof melanocytes in mice with vitiligo was a poorly immunogenicprocess, which was incapable of priming new effectors from thenaive repertoire.

FIGURE 2. Pmel cells transferred into mice with vitiligo become partially activated but do not acquire full effector function. (A) Mice were treated as

depicted in Fig. 1A, and received 106 naive Ly5.2+ pmel cells 75 d after surgery. Ten days later, the proportion of CD44hi cells among live CD8+Ly5.2+ cells

was determined in inguinal lymph nodes. Data are representative of five to eight experiments with three to five mice per group. (B) Pmel cells were

transferred into vitiligo-affected mice at indicated time points after surgery, or into naive mice, and the proportion of CD44hi pmel cells among live CD8+

Ly5.2+ cells in lymph nodes was determined 10 d after each transfer. (C–F) Naive hosts, or vitiligo-affected hosts treated as depicted in Fig. 1A, were each

adoptively transferred with 106 naive Thy1.1+ pmel cells (see Materials and Methods for description of positive controls). Ten days after transfer, the

proportion of (C) P-selectin ligand+ or (D) CD25+, CD62Llow, or CD69+ cells among CD8+CD44hi Thy1.1+ cells in the inguinal lymph nodes was de-

termined. Data are representative of two to four experiments, with three to five mice per group. (E) The proportion of granzyme B+ cells among CD8+

CD44highThy1.1+ (pmel) after cells were fixed and permeabilized; gating was set using the naive CD8 T cell population (CD8+CD44lowThy1.12), such that

,1% of events was positive. (F) Lymph nodes were restimulated ex vivo for 5 h with irrelevant peptide (OVA) or cognate peptide (gp100) in the presence of

brefeldin A; gated on CD8+Thy1.1+CD44hi pmel cells. Data are representative of three experiments, with three to six mice/group. Symbols represent

individual mice, and horizontal lines depict averages. Statistical significance was determined by paired t test to irrelevant control within same host, or

unpaired t test as indicated by brackets, *p , 0.05, **p , 0.01, ***p , 0.001.

1436 VITILIGO DOES NOT INDUCE NAIVE CD8 T CELL PRIMING

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

Vitiligo pathology and antimelanoma immunity do not requirethymic output of naive T cells

Whereas gp10025–33–specific pmel cells did not undergo func-tional priming in mice with vitiligo, our published studies haveshown that vitiligo-affected mice also maintain functional CD8T cell responses to the melanocyte differentiation Ag TRP-2 (5,17). At least a proportion of these TRP-2–specific cells are primedearly during vitiligo initiation as a result of melanoma growth andTreg depletion (5, 17). Therefore it remained possible that naiveCD8 T cells specific for TRP-2, and potentially other melanocyteAgs, are continually primed during vitiligo progression. Becauseof this, it was necessary to formally address the contribution of allnewly primed effectors to vitiligo pathogenesis.To eliminate the ongoing generation of naive T cells over the

course of autoimmunity, we used adult thymectomy. Mice un-derwent thymectomy surgery and were then treated to initiatevitiligo as in Fig. 1A. Over the next 2 mo, vitiligo progression wasfollowed. We found that the course of autoimmune vitiligo wasunaltered in thymectomized mice, as compared with thymus-intactmice, with regard to kinetics (Fig. 3A) and intensity (Fig. 3B).Thus, a continual supply of naive T cells was not required for theprogression of autoimmune vitiligo. This suggests that vitiligo ismaintained by a population of long-lived memory T cells that areprimed early during disease initiation, rather than continually duringdisease progression.

We also tested the capacity of thymectomized, vitiligo-affectedmice to reject B16 challenge tumors inoculated 45 d after sur-gery. Indeed, thymectomized mice with vitiligo were significantlyprotected from melanoma rechallenge, with no significant reduc-tion in tumor protection as compared with thymus-intact mice withvitiligo (Fig. 3C). Thus, long-lived antitumor immunity did notrequire a continual source of naive T cells. This is consistent withthe present findings that naive melanocyte-specific T cells are notefficiently primed during vitiligo progression, and underscores theimportance of long-lived T cell responses, rather than short-livedeffectors, for both autoimmunity and antitumor immunity.

Priming of functional CD8 T cells during vitiligo progressionis rescued by the depletion of CD4 T cells

Our previous work has shown that anti-CD4 depleting Ab elimi-nates Tregs in B16 melanoma tumor-bearing mice, thereby in-ducing the priming of naive pmel cells (19). These pmel cellsattain full effector function and develop into long-lived memory(5, 17). Vitiligo subsequently develops in these mice, althoughTreg cells repopulate within the 2 wk after anti-CD4 treatment(16). Based on this, we speculated that repopulated Tregs exertdominant suppression in mice with vitiligo, in which case anothercourse of anti-CD4 treatment could restore the priming of naivepmel cells.To test this, we adoptively transferred vitiligo-affected mice

with naive pmel cells and then depleted them of CD4 T cells be-fore assessing priming 10 d later (Fig. 4A). Indeed, CD4 T celldepletion significantly increased accumulation of pmel cells inlymph nodes of vitiligo-affected mice, but not naive mice (Fig.4B). In contrast with pmel cells in CD4-intact mice, pmel cells inCD4-depleted mice were also capable of producing significantamounts of IFN-g (Fig. 4C). Surprisingly, significant proportionsof pmel cells were detected in lymph nodes of CD4-depleted micewith vitiligo as long as 30 d after adoptive transfer (Fig. 4D).These T cells were capable of producing IFN-g even at this latetime point (Fig. 4E). Thus, in hosts with vitiligo, depletion of CD4T cells rescues the priming of naive pmel cells that develop intofunctional memory.

DiscussionAlthough autoimmune disease has been extensively studied, muchof the emphasis has been on understanding how autoreactive T cellresponses are initiated (20–24). The effects of ongoing tissuedestruction, and self-Ag liberation, on the naive CD8 T cell rep-ertoire have remained largely unknown. In these studies, we useda model of melanoma-initiated, CD8 T cell–mediated vitiligo todefine the effects of ongoing melanocyte destruction on naive Ag-specific CD8 T cells. We report that melanocyte destruction drivesthe proliferation of Ag-specific T cells in draining lymph nodes;however, vitiligo is insufficient for full functional priming of thesecells. We also demonstrate that newly primed T cells are not re-quired for optimal disease pathology or tumor rejection in micewith vitiligo. Thus, autoimmune melanocyte destruction is itselfa poorly immunogenic process, which does not recruit new ef-fector T cells to the ongoing response.To our knowledge, this work is the first to address whether

naive, self Ag-specific CD8 T cells become functional effectorsduring self-perpetuating CD8 T cell–mediated autoimmune dis-ease. In our studies, transgenic pmel T cells were used to probethe immunogenicity of vitiligo but not to initiate disease. Similarquestions have been addressed in mice expressing OVA as a self-Ag in the pancreas, using adoptively transferred pathogenic OT-1T cells to initiate tissue destruction. In this setting, naive OT-Icells proliferated and eventually underwent deletional tolerance

FIGURE 3. Recent thymic emigrants are not required for vitiligo or for

protection against secondary melanomas. Thymectomized or wild-type

mice were treated as described in Fig. 1A. The kinetics of development (A)

and overall extent (B) of vitiligo was monitored in each group of mice,

with representative vitiligo-affected mice from each group shown in (B).

Data are representative of two experiments, with 8–16 mice per group. No

statistical differences were found between the groups by (A) log-rank

analysis or (B) Student t test. (C) Wild-type vitiligo-affected or thymec-

tomized vitiligo-affected mice were challenged with B16 melanoma cells

45 d after surgery, and tumor growth was monitored; thymectomized,

vitiligo-unaffected mice from the same cohort were used as negative

controls. Data are combined from two identical experiments each involv-

ing 6–16 mice per group. Significance was determined by log-rank anal-

ysis. *p , 0.05.

The Journal of Immunology 1437

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

(15), which is consistent with our findings. However, in contrastwith our studies, OT-1 cells acquired the ability to produce IFN-g(15). Our observed lack of IFN-g production by pmel cells couldbe because of the low-avidity nature of the pmel TCR (18), or thefact that pmel cells were transferred into vitiligo-affected miceduring long-term disease progression, compared with OT-1 cellsthat were transferred during disease initiation (15). Despite this,our studies together support the broad conclusion that CD8 T cell–mediated self-tissue destruction is insufficient for the initiation ofCD8 T cell priming.The incompletely activated phenotype and functional state ac-

quired by naive pmel cells in vitiligo-affected mice is similar towhat has been reported for CD8 T cells recognizing self-Ag in thesteady-state (i.e., in the absence of overt tissue destruction). Instudies where HA-specific CD8 T cells were adoptively transferredinto mice expressing HA as a self-Ag in the pancreas, HA-specificT cells proliferated in draining lymph nodes and upregulated CD44(24). However, these T cells only partially downregulated CD62L,did not produce IFN-g, and eventually disappeared after severalcycles of division (24). Several other groups have made similarfindings using model self-Ags (20, 21, 25). Although we observedno such proliferation of naive pmel cells in response to normalmelanocytes in the steady-state (in hosts lacking vitiligo), this mayagain reflect the low-avidity nature of the pmel TCR, which wasoriginally generated in wild-type mice that express gp100 in theperiphery (18). Our studies in vitiligo-affected mice show that,

even in the presence of overt, ongoing autoimmunity, low-avidityself-reactive CD8 T cells still cannot overcome the thresholdnecessary for priming. Although our studies investigate CD8T cell–mediated vitiligo induced by melanoma, in the future itwould be interesting to determine whether these findings extend toother models of melanocyte destruction (e.g., vitiligo initiated bymelanocytoxic chemicals [26, 27], mAbs [28], or pathogenic CD4T cells [29]).Despite our finding that naive pmel cells were not primed in

hosts with vitiligo, the possibility remained that endogenous CD8T cells with other melanocyte Ag specificities could becomeprimed and contribute to vitiligo pathogenesis. Recruitment ofnaive CD8 T cells has been demonstrated during the course ofcertain chronic viral infections (12–14, 30), and adult thymec-tomized mice have been used to demonstrate a critical role forthese newly primed effectors in immunosurveillance (12). How-ever, the present studies in thymectomized mice demonstrate noapparent role for newly primed effectors in autoimmune pathologyor associated melanoma tumor protection, supporting the idea thatthe autoreactive cytotoxic CD8 T cell response is “self-limiting”(15). This finding also underscores that CD8 T cells primed duringthe initial phase of melanoma therapy (i.e., during B16 melanomagrowth and anti-CD4 treatment; see Fig. 1A), are responsible forthe long-term melanocyte destruction and antitumor immunitythat we observe after tumor excision (5). Given that vitiligo ismediated by these long-lived T cells, it can be speculated that

FIGURE 4. CD4 T cells prevent the functional priming of naive pmel cells that develop into memory in vitiligo-affected hosts. (A) Schematic diagram;

mice were treated as depicted in Fig. 1A; however, they received additional weekly anti-CD4 treatments after adoptive transfer of Ly5.2+ pmel cells. (B and

C) Mice received 106 naive pmel cells 45 d after surgery, with anti-CD4 administered on days 4 and 7, and pmel cell responses were analyzed in lymph

nodes (LNs) on day 10 with regard to proportion of Ly5.2+ cells among total CD8+ cells (B), and the proportion of IFN-g+ cells among Ly5.2+CD44hi cells

(C). (D and E) Mice were treated as in (A), except that vitiligo-affected hosts received only 105 naive pmel cells, and CD4 depletion continued once weekly

until 30 d after T cell transfer. Thirty days posttransfer, the pmel responses in the LNs were analyzed with regard to the proportion of Ly5.2+ pmel cells

among CD8+ cells (D), and production of IFN-g; gated on CD8+Ly5.2+CD44hi cells (E). Data are representative of two experiments, each with three to

seven mice per group. Representative dot plots are shown. Symbols represent individual mice, and horizontal lines depict averages. Statistical significance

determined by t test as indicated by brackets. (C and E) Statistical significance was also determined by paired t test comparing irrelevant (Irrel) or gp100

peptide restimulation within the same group. *p , 0.05, **p , 0.01.

1438 VITILIGO DOES NOT INDUCE NAIVE CD8 T CELL PRIMING

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from

specifically depleting memory T cells would halt disease pro-gression. Thus, these data in thymectomized mice support ourprevious finding that melanoma/melanocyte-specific memory CD8T cells do not become functionally exhausted, and our prior as-sumption that these cells are responsible for tumor protection invitiligo-affected mice (5).Although deficiencies in Treg responses have been documented in

humans with vitiligo (31), our finding that CD4 T cell depletionenables the priming of new Ag-specific T cells suggests that Tregsmaintain some suppressive activity during vitiligo progression.Upon depletion of CD4 T cells, pmel cells acquired both theability to produce IFN-g and persist as functional memory. De-spite this, our past studies to investigate whether CD4 Th cellspromote memory T cell responses in this model showed no neteffect of ongoing anti-CD4 treatment on postsurgical vitiligo ormelanoma tumor protection (16). Taken in conjunction with thepresent findings, this may suggest that CD4+ Treg and Th cells playopposing roles during the course of vitiligo, with Th cells pro-moting the optimal function of memory T cells and Tregs sup-pressing the ongoing priming of naive T cells. Furthermore, the factthat vitiligo progresses despite the presence of Tregs also suggeststhat pathogenic memory T cells may be less susceptible to Tregsuppression than naive T cells. In future studies, the targeted abla-tion of Foxp3+ Tregs without impairing Th cells could help to fur-ther dissect the distinct contributions of these CD4+ T cell subsets.In conclusion, these studies demonstrate that ongoing CD8

T cell–mediated autoimmune vitiligo is a weakly immunogenicevent that is perpetuated by long-lived T cells, as opposed tonewly primed effectors. In addition, our finding that CD4 T cellssuppress the priming of new effector T cells in hosts with vitiligosuggests a dominant role for Treg suppression even in the face ofovert autoimmune disease. These studies reveal autoimmune vit-iligo to be a complex disease setting with multiple layers of T cellactivation and regulation.

AcknowledgmentsWe thank Ed Usherwood and David Mullins for helpful discussions.

DisclosuresThe authors have no financial conflicts of interest.

References1. Matsuzawa, T., M. Watanabe, and T. Kondo. 1953. Case of leukoderma in X-ray

portion of patient with melanosarcoma. Shinshu Med. J. 2: 254–258.2. Nordlund, J. J., J. M. Kirkwood, B. M. Forget, G. Milton, D. M. Albert, and

A. B. Lerner. 1983. Vitiligo in patients with metastatic melanoma: a goodprognostic sign. J. Am. Acad. Dermatol. 9: 689–696.

3. Bystryn, J. C., D. Rigel, R. J. Friedman, and A. Kopf. 1987. Prognostic significance ofhypopigmentation in malignant melanoma. Arch. Dermatol. 123: 1053–1055.

4. Quaglino, P., F. Marenco, S. Osella-Abate, N. Cappello, M. Ortoncelli,B. Salomone, M. T. Fierro, P. Savoia, and M. G. Bernengo. 2010. Vitiligo is anindependent favourable prognostic factor in stage III and IV metastatic mela-noma patients: results from a single-institution hospital-based observationalcohort study. Ann. Oncol. 21: 409–414.

5. Byrne, K. T., A. L. Cote, P. Zhang, S. M. Steinberg, Y. Guo, R. Allie, W. Zhang,M. S. Ernstoff, E. J. Usherwood, and M. J. Turk. 2011. Autoimmune melanocytedestruction is required for robust CD8+ memory T cell responses to mousemelanoma. J. Clin. Invest. 121: 1797–1809.

6. Lehmann, P. V., T. Forsthuber, A. Miller, and E. E. Sercarz. 1992. Spreadingof T-cell autoimmunity to cryptic determinants of an autoantigen. Nature 358:155–157.

7. Vanderlugt, C. L., and S. D. Miller. 2002. Epitope spreading in immune-mediateddiseases: implications for immunotherapy. Nat. Rev. Immunol. 2: 85–95.

8. el-Shami, K., B. Tirosh, E. Bar-Haim, L. Carmon, E. Vadai, M. Fridkin,M. Feldman, and L. Eisenbach. 1999. MHC class I-restricted epitope spreadingin the context of tumor rejection following vaccination with a single immuno-dominant CTL epitope. Eur. J. Immunol. 29: 3295–3301.

9. Brossart, P., S. Wirths, G. Stuhler, V. L. Reichardt, L. Kanz, and W. Brugger.2000. Induction of cytotoxic T-lymphocyte responses in vivo after vaccinationswith peptide-pulsed dendritic cells. Blood 96: 3102–3108.

10. Markiewicz, M. A., F. Fallarino, A. Ashikari, and T. F. Gajewski. 2001. Epitopespreading upon P815 tumor rejection triggered by vaccination with the singleclass I MHC-restricted peptide P1A. Int. Immunol. 13: 625–632.

11. Jackaman, C., D. Majewski, S. A. Fox, A. K. Nowak, and D. J. Nelson. 2012.Chemotherapy broadens the range of tumor antigens seen by cytotoxic CD8(+)T cells in vivo. Cancer Immunol. Immunother. 61: 2343–2356.

12. Vezys, V., D. Masopust, C. C. Kemball, D. L. Barber, L. A. O’Mara,C. P. Larsen, T. C. Pearson, R. Ahmed, and A. E. Lukacher. 2006. Continuousrecruitment of naive T cells contributes to heterogeneity of antiviral CD8 T cellsduring persistent infection. J. Exp. Med. 203: 2263–2269.

13. Kemball, C. C., E. D. Lee, V. Vezys, T. C. Pearson, C. P. Larsen, andA. E. Lukacher. 2005. Late priming and variability of epitope-specific CD8+T cell responses during a persistent virus infection. J. Immunol. 174: 7950–7960.

14. Snyder, C. M., K. S. Cho, E. L. Bonnett, S. van Dommelen, G. R. Shellam,and A. B. Hill. 2008. Memory inflation during chronic viral infection ismaintained by continuous production of short-lived, functional T cells.Immunity 29: 650–659.

15. Parish, I. A., J. Waithman, G. M. Davey, G. T. Belz, J. D. Mintern, C. Kurts,R. M. Sutherland, F. R. Carbone, and W. R. Heath. 2009. Tissue destructioncaused by cytotoxic T lymphocytes induces deletional tolerance. Proc. Natl.Acad. Sci. USA 106: 3901–3906.

16. Cote, A. L., K. T. Byrne, S. M. Steinberg, P. Zhang, and M. J. Turk. 2011.Protective CD8 memory T cell responses to mouse melanoma are generated inthe absence of CD4 T cell help. PLoS ONE 6: e26491.

17. Zhang, P., A. L. Cote, V. C. de Vries, E. J. Usherwood, and M. J. Turk. 2007.Induction of postsurgical tumor immunity and T-cell memory by a poorly im-munogenic tumor. Cancer Res. 67: 6468–6476.

18. Overwijk, W. W., M. R. Theoret, S. E. Finkelstein, D. R. Surman, L. A. de Jong,F. A. Vyth-Dreese, T. A. Dellemijn, P. A. Antony, P. J. Spiess, D. C. Palmer, et al.2003. Tumor regression and autoimmunity after reversal of a functionally tol-erant state of self-reactive CD8+ T cells. J. Exp. Med. 198: 569–580.

19. Turk, M. J., J. A. Guevara-Patino, G. A. Rizzuto, M. E. Engelhorn,S. Sakaguchi, and A. N. Houghton. 2004. Concomitant tumor immunity toa poorly immunogenic melanoma is prevented by regulatory T cells. J. Exp.Med. 200: 771–782.

20. Franck, E., C. Bonneau, L. Jean, J. P. Henry, Y. Lacoume, A. Salvetti,O. Boyer, and S. Adriouch. 2012. Immunological tolerance to muscle auto-antigens involves peripheral deletion of autoreactive CD8+ T cells. PLoSONE 7: e36444.

21. Kurts, C., H. Kosaka, F. R. Carbone, J. F. Miller, and W. R. Heath. 1997. Class I-restricted cross-presentation of exogenous self-antigens leads to deletion ofautoreactive CD8(+) T cells. J. Exp. Med. 186: 239–245.

22. Ludewig, B., K. McCoy, M. Pericin, A. F. Ochsenbein, T. Dumrese,B. Odermatt, R. E. Toes, C. J. Melief, H. Hengartner, and R. M. Zinkernagel.2001. Rapid peptide turnover and inefficient presentation of exogenous anti-gen critically limit the activation of self-reactive CTL by dendritic cells.J. Immunol. 166: 3678–3687.

23. Lang, K. S., M. Recher, T. Junt, A. A. Navarini, N. L. Harris, S. Freigang,B. Odermatt, C. Conrad, L. M. Ittner, S. Bauer, et al. 2005. Toll-like receptorengagement converts T-cell autoreactivity into overt autoimmune disease. Nat.Med. 11: 138–145.

24. Hernandez, J., S. Aung, W. L. Redmond, and L. A. Sherman. 2001. Phe-notypic and functional analysis of CD8(+) T cells undergoing peripheraldeletion in response to cross-presentation of self-antigen. J. Exp. Med. 194:707–717.

25. Mukhopadhaya, A., T. Hanafusa, I. Jarchum, Y. G. Chen, Y. Iwai,D. V. Serreze, R. M. Steinman, K. V. Tarbell, and T. P. DiLorenzo. 2008.Selective delivery of beta cell antigen to dendritic cells in vivo leads to de-letion and tolerance of autoreactive CD8+ T cells in NOD mice. Proc. Natl.Acad. Sci. USA 105: 6374–6379.

26. van den Boorn, J. G., D. I. Picavet, P. F. van Swieten, H. A. van Veen,D. Konijnenberg, P. A. van Veelen, T. van Capel, E. C. Jong, E. A. Reits,J. W. Drijfhout, et al. 2011. Skin-depigmenting agent monobenzone inducespotent T-cell autoimmunity toward pigmented cells by tyrosinase haptenationand melanosome autophagy. J. Invest. Dermatol. 131: 1240–1251.

27. van den Boorn, J. G., D. Konijnenberg, E. P. Tjin, D. I. Picavet,N. J. Meeuwenoord, D. V. Filippov, J. P. van der Veen, J. D. Bos, C. J. Melief,and R. M. Luiten. 2010. Effective melanoma immunotherapy in mice by theskin-depigmenting agent monobenzone and the adjuvants imiquimod and CpG.PLoS ONE 5: e10626.

28. Takechi, Y., I. Hara, C. Naftzger, Y. Xu, and A. N. Houghton. 1996. A mela-nosomal membrane protein is a cell surface target for melanoma therapy. Clin.Cancer Res. 2: 1837–1842.

29. Muranski, P., A. Boni, P. A. Antony, L. Cassard, K. R. Irvine, A. Kaiser,C. M. Paulos, D. C. Palmer, C. E. Touloukian, K. Ptak, et al. 2008. Tumor-specific Th17-polarized cells eradicate large established melanoma. Blood 112:362–373.

30. Wilson, J. J., C. D. Pack, E. Lin, E. L. Frost, J. A. Albrecht, A. Hadley,A. R. Hofstetter, S. S. Tevethia, T. D. Schell, and A. E. Lukacher. 2012. CD8T cells recruited early in mouse polyomavirus infection undergo exhaustion. J.Immunol. 188: 4340–4348.

31. Ben Ahmed, M., I. Zaraa, R. Rekik, A. Elbeldi-Ferchiou, N. Kourda,N. Belhadj Hmida, M. Abdeladhim, O. Karoui, A. Ben Osman, M. Mokni,and H. Louzir. 2012. Functional defects of peripheral regulatory T lym-phocytes in patients with progressive vitiligo. Pigment Cell Melanoma Res.25: 99–109.

The Journal of Immunology 1439

at Dartm

outh College on M

ay 2, 2014http://w

ww

.jimm

unol.org/D

ownloaded from