Signatures of Pathogen-Specific T Cells II: Memory T Cells in Acute and Chronic Infection

This information is current as Bennett Davenport, Jens Eberlein, Tom T. Nguyen, of September 24, 2021. Francisco Victorino, Verena van der Heide, Maxim Kuleshov, Avi Ma'ayan, Ross Kedl and Dirk Homann J Immunol published online 18 September 2020 http://www.jimmunol.org/content/early/2020/09/17/jimmun

ol.2000254 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2020/09/17/jimmunol.200025 Material 4.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 24, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2020 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published September 18, 2020, doi:10.4049/jimmunol.2000254 The Journal of Immunology

Chemokine Signatures of Pathogen-Specific T Cells II: Memory T Cells in Acute and Chronic Infection

Bennett Davenport,*,†,‡,x,{ Jens Eberlein,*,† Tom T. Nguyen,*,‡ Francisco Victorino,*,†,‡ Verena van der Heide,x,{ Maxim Kuleshov,‖,# Avi Ma’ayan,‖,# Ross Kedl,† and Dirk Homann*,†,‡,x,{

Pathogen-specific memory T cells (TM) contribute to enhanced immune protection under conditions of reinfection, and their effective recruitment into a recall response relies, in part, on cues imparted by that coordinate their spatiotemporal

positioning. An integrated perspective, however, needs to consider TM as a potentially relevant chemokine source themselves. In this study, we employed a comprehensive transcriptional/translational profiling strategy to delineate the identities, expression + patterns, and dynamic regulation of chemokines produced by murine pathogen-specific TM. CD8 TM, and to a lesser extent + Downloaded from CD4 TM, are a prodigious source for six select chemokines (CCL1/3/4/5, CCL9/10, and XCL1) that collectively constitute a + prominent and largely invariant signature across acute and chronic infections. Notably, constitutive CCL5 expression by CD8 TM serves as a unique functional imprint of prior antigenic experience; induced CCL1 production identifies highly polyfunctional + + + CD8 and CD4 TM subsets; long-term CD8 TM maintenance is associated with a pronounced increase of XCL1 production + capacity; chemokines dominate the earliest stages of the CD8 TM recall response because of expeditious synthesis/secretion kinetics (CCL3/4/5) and low activation thresholds (CCL1/3/4/5/XCL1); and TM chemokine profiles modulated by persisting viral Ags exhibit both discrete functional deficits and a notable surplus. Nevertheless, recall responses and partial virus control in http://www.jimmunol.org/

chronic infection appear little affected by the absence of major TM chemokines. Although specific contributions of TM-derived chemokines to enhanced immune protection therefore remain to be elucidated in other experimental scenarios, the ready visu-

alization of TM chemokine-expression patterns permits a detailed stratification of TM functionalities that may be correlated with differentiation status, protective capacities, and potential fates. The Journal of Immunology, 2020, 205: 000–000.

athogen-specific memory T cells (TM) are an integral The choreography of these events is, in part, governed by che- component of the anamnestic immune response and can mokines, a large family of mostly secreted small molecules that P provide immune protection by curtailing secondary (II˚) regulates the spatiotemporal positioning of motile cells (6–8). by guest on September 24, 2021 infections, limiting morbidity, and forestalling potential host death Pathogen-specific TM, by virtue of their distinct chemokine (1–5). These clinical outcomes are the net result of highly com- expression patterns, are acutely attuned to varied che- plex and coordinated interactions between multiple organ systems, mokine cues, as demonstrated in numerous in vitro and in vivo tissues, cell types, and extracellular factors that are marshaled into studies (6–11); however, as has been known for over two decades action following pathogen detection, and the relevant contribu- (12), T cells are also a relevant source for certain chemokines tions of specific TM to these processes are grounded in three themselves, notably for CCL3, CCL4, and CCL5, which, beyond fundamental determinants: their numbers, their location, and their their chemotactic functions, can also act as competitive inhibitors differentiation status (i.e., the particular phenotypic, molecular, of HIV binding to its coreceptor CCR5 (13–15); at micromolar and epigenetic makeup that permits TM populations to respond concentrations, CCL5 may exert receptor-independent cellular with the elaboration of rapid effector activities as well as coop- activation, , and even antimicrobial activity, although erative cellular interactions, local and systemic mobilization, II˚ some of the evidence is contradictory and the in vivo relevance effector [TE] differentiation, and proliferative expansion). unclear (16–21). Several other chemokines, including CCL1,

*Barbara Davis Center for Childhood Diabetes, University of Colorado Anschutz JDRF Grant CDA 2-2007-240 (to D.H.). The funders had no role in study design, Medical Campus, Aurora, CO 80045; †Department of Immunology and Microbiol- data collection and analysis, decision to publish, or preparation of the manuscript. ogy, University of Colorado Anschutz Medical Campus, Aurora, CO 80045; The microarray data presented in this article have been submitted to Expression ‡Department of Anesthesiology, University of Colorado Anschutz Medical Campus, x Omnibus (https://www.ncbi.nlm.nih.gov/geo/) under accession number GSE143632. Aurora, CO 80045; Diabetes, Obesity and Metabolism Institute, Icahn School of Medicine at Mount Sinai, New York, NY 10029; {Precision Immunology Institute, Address correspondence and reprint requests to Prof. Dirk Homann, Diabetes, Icahn School of Medicine at Mount Sinai, New York, NY; ‖Department of Pharma- Obesity and Metabolism Institute, Icahn School of Medicine at Mount Sinai, cological Sciences, Icahn School of Medicine at Mount Sinai, New York, NY 10029; One Gustave L. Levy Place, Box 1152, New York, NY 10029. E-mail address: and #Mount Sinai Center for Bioinformatics, Icahn School of Medicine at Mount [email protected] Sinai, New York, NY 10029 The online version of this article contains supplemental material. ORCIDs: 0000-0002-6268-1767 (J.E.); 0000-0003-4222-4099 (T.T.N.); 0000-0002- + Abbreviations used in this article: Arm, Armstrong; BFA, brefeldin A; CD8 TMP, 3602-4405 (V.v.d.H.); 0000-0002-7812-7752 (M.K.); 0000-0002-7622-5754 (D.H.). + memory-phenotype CD8 T cell; CHX, cycloheximide; cl13, clone 13; ET50, 50% Received for publication March 9, 2020. Accepted for publication August 7, 2020. effective time; FC, flow cytometry; FRC, fibroblastic reticular cell; Gzm, granzyme; I˚, primary; II˚, secondary; LCMV, lymphocytic choriomeningitis virus; MHC-I, This work was supported by National Institutes of Health (NIH) Grants AG026518 MHC class I; rLM-OVA, recombinant Listeria monocytogenes expressing full- and AI093637, a Barbara Davis Center Pilot and Feasibility grant, NIH/Diabetes length OVA; T , central T ;T, effector T cell; T , effector T ;T , memory Endocrinology Research Center Grant P30-DK057516 (to D.H.), NIH Grants U54- CM M E EM M M T cell. HL127624 and U24-CA224260 (to A.M.), and NIH Training Grants T32 AI07405, T32 AI052066, and T32 DK007792 (to B.D.). This work was also supported by Copyright Ó 2020 by The American Association of Immunologists, Inc. 0022-1767/20/$37.50

www.jimmunol.org/cgi/doi/10.4049/jimmunol.2000254 2 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS

CCL9/10, and XCL1, have further been reported as products of (IACUC-2014-0170), and all efforts were made to minimize suffering of animals. pathogen-specific TM (22–26), but to date, experimental evidence in support of pathogen-specific, TM-derived chemokines as non- Mice, pathogens, and challenge/vaccination protocols redundant contributors to effective immune protection at the level C57BL6/J (B6), congenic B6.CD90.1 (B6.PL-Thy1a/CyJ), congenic B6.CD45.1 of II˚ T expansion, pathogen control, and/or host survival re- a b 2/2 tm1Unc/ E (B6.SJL-Ptprc Pepc /BoyJ), and B6.CCL3 (B6.129P2-Ccl3 J) mains limited to CCL3 and possibly XCL1 in some, but not other, mice on a B6 background as well as BALB/c mice were purchased murine model systems (27, 28). from The Jackson Laboratory; p14 TCR transgenic mice were obtained + Chemokine synthesis and secretion by TM, similar to other on a B6.CD90.1 background from Dr. M. Oldstone (CD8 T cells from effector functions, such as and TNFSF ligand pro- these mice [p14 cells] are specific for the dominant lymphocytic choriomeningitis virus [LCMV]-GP33–41 determinant restricted by duction, typically require a brief period of TCR activation, a Db). B6.CCL52/2 mice (36) were obtained from Dr. M. von Herrath prerequisite that may provide a safeguard against inappro- (these mice are identical to the commercially available B6.129P2- tm1Hso 2/2 priate TM activation and immunopathology (29). It is there- Ccl5 /J strain), and B6.CCL1 mice were a gift from Dr. S. fore of interest that CCL5 is expressed in a constitutive Manes (37). LCMV Armstrong (Arm; clone 53b) and clone 13 (cl13) fashionbyhumanCD8+T cell subsets (22, 30, 31), in which it were obtained from Dr. M. Oldstone, stocks were prepared by a single passage on BHK-21 cells, and plaque assays for determination of is confined to a unique subcellular compartment and released virus titers were performed as described (38). Recombinant Listeria with near-instantaneous kinetics upon TCR engagement (30). monocytogenes expressing full-length OVA (rLM-OVA) (39) was grown Other studies, however, have demonstrated a preferential as- and titrated as described (34, 40). For acute infections, 8–10-wk-old 3 5 sociation of constitutively expressed CCL5 with cytolytic mice were inoculated with a single i.p. dose of 2 10 PFUs LCMV Arm or 2 3 103 CFU rLM-OVA i.v.; for persistent infections, mice were

+ Downloaded from granules in HIV-specific CD8 T cell clones or primary (I˚) challenged with 2 3 106 PFU LCMV cl13 i.v. (control cohorts were + human CD8 T cells (13, 31), but murine memory-phenotype acutely infected with 2 3 106 PFU LCMV Arm i.v.). Combined TLR/CD40 + + CD8 T cells (CD8 TMP), which contain abundant Ccl5 in vaccinations were performed as described (34, 41) by immunizing addition to Ccl3 and Ccl4 transcripts, apparently do not ex- mice i.p. with 500 mg OVA (Sigma-Aldrich) in combination with 50 mg press the corresponding , the synthesis of which re- anti-CD40 (FGK4.5; Bio X Cell) and 50 mg polyinosinic/polycytidylic acid [poly(I:C); Amersham/GE Healthcare]; all vaccinations were quires TCR stimulation (32, 33). Our recent work on chemokine performedbymixingeachcomponentinPBSandinjectioninavolume + signatures of pathogen-specific CD8 TE may provide clues for of 200 ml. http://www.jimmunol.org/ a reconciliation of these discrepancies (34): at the peak of the isolation, T cell purification, and effector response, CD8+T express constitutive CCL5 in a E stimulation cultures subcellular compartment distinct from granzyme (Gzm)–containing cytolytic granules, but upon TCR stimulation, these structures Our procedures for isolation of from lymphatic and non- partly coalesce just prior to secretion; in fact, release of pre- lymphoid tissues, including total body perfusion with PBS, are detailed elsewhere (42, 43). To generate p14 chimeras, naive CD90.1-congenic p14 stored CCL5 proceeds so rapidly that temporarily, and within T cells were enriched by negative selection, and ∼5 3 104 cells were the confines of the immunological synapse, CCL5 concentra- transferred i.v. into sex-matched B6 recipients that were challenged 24 h tions well in excess of 1 mM may be achieved (34), a potential later with LCMV (25). For microarray analyses, splenic p14 TM were foundation for in vivo CCL5 action at supraphysiological levels positively selected using anti-CD90.1-PE Ab and PE-specific magnetic by guest on September 24, 2021 . (16, 17). beads (STEMCELL Technologies) and further enriched to 99% purity by FACS sorting (BD Biosciences FACSAria) as described (25). Adoptive Importantly, in the same study, we provide detailed evidence that transfer/rechallenge experiments (Supplemental Fig. 3) were conducted the induced production of all of the above (CCL1/3/4/5/9/10 and with spleen cells obtained from LCMV-immune B6 mice and depleted of XCL1), but no other chemokines, constitutes a largely invariant CD19+B cells or CD19+B and CD4+T cells using magnetic beads prior to functional signature of specific CD8+ and CD4+T generated in i.v. transfer into B6 congenic recipients and LCMV challenge (25). For E specific T cell stimulation, spleen cells were cultured for 5 h with MHC response to acute viral and bacterial infections, as well as pro- class I (MHC-I) (1 mg/ml)– or MHC-II (5 mg/ml)–restricted peptides in the tective immunization (34). Building on this work, we have now presence (for flow cytometry [FC] analyses) or absence (for ELISA assays) extended our investigations to an interrogation of the identities, of 1 mg/ml brefeldin A (BFA; Sigma-Aldrich) (43). In some cases, highly patterns, regulation, and relevance of chemokines expressed by purified p14 TM were stimulated for 3 h with plate-bound anti-CD3 pathogen-specific T populations maintained after resolution of (10 mg/ml) and soluble anti-CD28 (2 mg/ml) prior to processing for M microarray hybridization. acute and under conditions of chronic infections. To be sure, because of the fundamentally different disease courses in these In vivo T cell activation and chemokine blockade experimental scenarios, such T populations present with both + M In vivo activation of CD8 TM (Fig. 5D) was performed as described (44) distinctive and shared properties (35). If those differences war- (i.e., vaccine-immune mice were injected with 100 mgOVA257 peptide rant a terminological distinction may remain a matter of debate, i.v., followed by 250 mg BFA/PBS i.p. 30 min later); spleens were har- vested 2 h after peptide injection, processed, and analyzed by FC. For but we favor an emphasis on relevant commonalities, as argued + in vivo chemokine neutralization in the context of CD8 TM recall re- by Jameson and Masopust (5). Hence, regardless of the acute or sponses, we employed experimental designs, Ab dosages, and treatment chronic course of an infection, we, in this study, refer to pop- schedules as detailed in the legend to Supplemental Fig. 3E and 3F using ulations generated in the immediate wake of a pathogen chal- the following Abs for CCL5 blockade: anti-CCL5 clone R6G9 [mIgG1 (45)] or mIgG1 isotype clone MOPC-21 (Sigma-Aldrich); combined lenge (∼1wk)asTE, and to populations present at later stages (.1mo)asT . CCL3/4/5 blockade: anti-CCL3 clone 756605 (rIgG1), anti-CCL4 clone M 46907 (rIgG2a), anti-CCL5 clone 53405 (rIgG2a) (R&D Systems), or rIgG control; and XCL1 blockade: polyclonal anti-XCL1 AF486 or goat IgG Materials and Methods control AB-108-C (R&D Systems). Ethics statement Microarray analyses

All procedures involving laboratory animals were conducted in accordance Microarray experiments were conducted with highly purified p14 TM with recommendations in the Guide for the Care and Use of Laboratory populations (directly ex vivo or after 3 h in vitro stimulation with Animals of the National Institutes of Health; the protocols were approved anti-CD3/anti-CD28) and Affymetrix M430.2 arrays as detailed (25, 34). by the Institutional Animal Care and Use Committees of the University Selected data are shown in Figs. 1A and 2C, Supplemental Figs. 1A of Colorado (permit numbers 70205604[05]1F, 70205607[05]4F, and and 2A and 2C, and the entire datasets, including those for ex vivo and B-70210[05]1E) and the Icahn School of Medicine at Mount Sinai anti-CD3/anti-CD28–stimulated p14 TE, can be retrieved from the Gene The Journal of Immunology 3

Expression Omnibus repository accession number GSE143632 (https:// independent of the fact that not all effector functions are induced in all TM www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE143632); MAS5, RMA, of a given specificity. and GC-RMA normalization were performed and yielded essentially similar results (data not shown). Results + Peptides and MHC tetramers Chemokine signatures of virus-specific CD8 TM Peptides corresponding to the indicated pathogen epitopes were obtained To define the range of chemokines produced by virus-specific from Peptidogenic, the National Jewish Health Molecular Core Facil- CD8+T , we employed a comprehensive transcriptional/ . M ity, or GenScript at purities of 95%; their MHC restriction and translational profiling strategy previously used for the charac- amino acid sequences are indicated. LCMV epitopes are the following: + b b terization of CD8 TE generated in response to an acute infection GP33–41 (D /KAVYNFATC), GP276–286 (D /SGVENPGGYCL), NP396–404 b b (D /FQPQNGQFI), GP64–80 (IA /GPDIYKGVYQFKSVEFD), and NP118–126 with LCMV (34). In brief, so-called p14 chimeras were generated d b (L /RPQASGVYM); and rLM-OVA epitopes: OVA257–264 (K /SIINFEKL) by transfusing congenic B6 mice with a trace population of naive b + and LLO190–201 (IA /NEKYAQAYPNVS). Note that the LCMV-GP33 b + + b TCR transgenic CD8 T cells specific for the dominant LCMV- peptide also stimulates K GP34 CD8 TE/M not captured by D GP33 tet- GP33–41 determinant; following infection with LCMV Arm (2 3 ramer stains; we therefore refer to the T cells activated in GP33 stimula- 5 + b b b 10 PFU i.p.), p14 T rapidly differentiate, expand, and contribute tion cultures as GP33/4-specific CD8 TE/M.DNP396,DGP33,DGP276, E d b b L NP118,KOVA257, and IA GP66 complexes were obtained from the to efficient virus control before contracting and developing into National Institutes of Health Tetramer Core Facility as APC or PE con- p14 TM populations ∼6 wk later (25, 49, 50); highly purified p14 jugates and/or biotinylated monomers, and CD1/aGalCer tetramers were a b TM were then subjected to gene array analyses conducted directly gift from Dr. L. Gapin. Note that the IA GP66 tetramer identifies the same population of CD4+T responsive to stimulation with the longer GP ex vivo or after a 3-h in vitro TCR stimulation, as detailed in

E/M 64–80 Downloaded from peptide (46). Tetramer staining was performed as described (i.e., for Materials and Methods and previous studies (25, 34). 45 min at 4˚C or room temperature for MHC-I tetramers and for 90 min at In these analyses, 11 distinct chemokine mRNA species scored 37˚C in the presence of sodium azide for MHC-II tetramers) (47). as present in p14 TM and, according to their expression patterns, Abs, staining protocols, and FC could be clustered into three groups (Fig. 1A): 1) absence of ex vivo detectable mRNA but robust transcription after TCR stimu- All Abs and FC staining protocols used in the current study are described lation (Ccl1 and Ccl17); 2) constitutive mRNA expression that elsewhere. Specifically, this includes Abs/protocols for standard cell surface and intracellular staining as well as Ab conjugation (25), our detailed Ab significantly increased upon TCR engagement (Ccl3, Ccl4, Ccl9/10, http://www.jimmunol.org/ characterization and staining strategies for detection practically all murine Ccl27, Cxcl10, and Xcl1); and 3) chemokine mRNA species that chemokines (22, 34), and the procedure for visualization of LCMV-NP remained unaffected or were downregulated by TCR activation o expression in infected I cells (42, 48); perforin stains were performed with [Ccl5, Ccl6,andCcl25; this group also contains Cklf and Cklfsf3/6/7 Ab clone S16009B conjugated to PE (BioLegend). All samples were ac- (Supplemental Fig. 1A), members of the related chemokine-like quired on FACSCalibur, LSR II, or LSRFortessa (BD Biosciences) or Attune NxT (Thermo Fisher Scientific) flow cytometers and analyzed with factor superfamily, about which relatively little remains known to DIVA (BD Biosciences) and/or FlowJo (Tree Star) software. To estimate date (51)]. Altogether, the chemokine mRNA expression patterns + the relative magnitude of the CD8 TM chemokine response in the context of p14 T are notably similar to the corresponding transcriptional + M of other CD8 TM activities (Fig. 3D), we determined the percentage of + profiles of p14 TE (34). individual NP396-specific CD8 TM (∼d42) subsets expressing constitutive by guest on September 24, 2021 (GzmA/B and perforin) or inducible (CCL1/3/4/5/9/10 and XCL1; IFN-g, Complementary expression analyses were conducted IL-2, IL-3, and GM-CSF; TNF-a, FasL, and CD40L; and degranulation) with chemokine FC using a portfolio of extensively characterized effector functionalities as shown/described in this study (Figs. 3A, 3B, 4A, Abs that permit detection of practically all murine chemokines at 4B) or in Ref. 25; details about our calculations to estimate the relative the single-cell level (22, 34). Upon TCR stimulation with GP33 magnitude and distribution of individual functional activities as displayed peptide, nearly all p14 T produced CCL3, CCL4, and CCL5; a in Fig. 3D are provided (34). M large subset made XCL1; and smaller fractions expressed CCL1 ELISA and CCL9/10. Of note, neither Ccl6, Ccl17, Ccl25, Ccl27, Cxcl10, Quantitation of CCL3, CCL4, and CCL5 in sera and/or peptide-stimulated nor any other chemokine mRNA was translated by p14 TM tissue culture supernatants (Figs. 2F, 5D) was performed using respective (Fig. 1B and data not shown). Remarkably, most p14 TM Quantikine ELISA Kits and protocols provided by the manufacturer expressed abundant CCL5 even in the absence of TCR activation (R&D Systems). To determine secretion kinetics of prestored CCL5 by + (Fig. 1B), establishing an expression pattern that is unique not specific CD8 TM (Fig. 5E), spleen cells from LCMV-immune mice were + preincubated for 30 min at 37˚C with 10 mg/ml cycloheximide (CHX; only for CD8 TM chemokines but for at large (25). With only six chemokine mRNA species serving as templates for in- Sigma-Aldrich) to prevent translation and were stimulated with NP396 peptide, and CCL5 in the supernatant was quantified by ELISA (the duced, and in the case of Ccl5, also constitutive translation, the amount of CCL5 secreted in the absence of peptide stimulation was chemokine signatures of p14 TM are therefore equivalent to those subtracted from stimulated samples at all time points). To calculate of p14 T (34). CCL5 release on a per-cell basis, FC analyses were performed in parallel E b + + to calculate the exact numbers of D NP396 CD8 TM in the stimulation Constitutive CCL5 expression by endogenously generated culture. + pathogen-specific CD8 TM Statistical analyses We next extended our investigations to endogenously generated + Data handling, analysis, and graphic representation was performed using LCMV-specific CD8 TM. Just like p14 TM, these populations, Prism 4.0 and 6.0c (GraphPad Software, San Diego, CA). All data sum- regardless of epitope specificity, constitutively expressed CCL5 marized in bar and line diagrams are expressed as mean 6 1 SE. Asterisks but no other chemokine, cytokine, or TNFSF ligand (Fig. 2A, Ref. 25, indicate statistical differences calculated by unpaired or paired Student and data not shown). Moreover, this expression pattern was largely t test and adopt the following convention: *p , 0.05, **p , 0.01, and identical in CD8+T recovered from lymphatic and nonlymphoid ***p , 0.001. The 50% effective time (ET50) values (response kinetics; M Fig. 5A–C) and EC50 values (activation thresholds; Fig. 5F–H) were cal- tissues, was independent of mouse strain, and pertained to specific + culated by plotting the fraction of specific TM demonstrating detectable CD8 TM generated in response to infection with the bacterium chemokine/cytokine staining (FC) or the amount of secreted chemokines 26 L. monocytogenes (Fig. 2A, 2B). We note, however, that consti- (ELISA) as a function of stimulation time (10 M peptide for 0–5 h) or + peptide concentration (1026–10211 M peptide for 5 h), followed by non- tutive CCL5 expression by CD8 TM in peripheral lymph nodes linear regression analysis using appropriate data format and analysis was somewhat reduced in comparison with all other tissues (Fig. 2A, functions in the Prism software. Note that ET50 and EC50 values are 2B). Because lymph nodes are enriched for CD62L-expressing 4 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

+ FIGURE 1. Chemokine mRNA and protein expression by virus-specific CD8 TM.(A) p14 TM were obtained from spleens of LCMV-immune p14 chimeras (d46), enriched to .99% purity and processed for RNA extraction (either immediately or after 3 h anti-CD3/anti-CD28 stimulation) and gene array analysis (n = 4 individual mice). The bar diagrams display MAS5-normalized values of chemokine mRNA expression of p14 TM–analyzed ex vivo (gray bars) or after TCR stimulation (black bars). Statistically significant differences are indicated by asterisks, and the broken line indicates the detection threshold set at an MAS5 value of 40; coverage: 39/40 chemokines (Ccl26 not on chip). (B) Left, p14 TM (d68) were analyzed for chemokine protein after 5 h culture in the absence (gray histograms) or presence of GP33 peptide (black tracings); all histograms are gated on p14 TM (the broken line histogram in the CCL5 panel indicates a negative control stain). Right, Summary of constitutive (no peptide) and induced (GP33 peptide stimulation) chemokine ex- pression by p14 TM (SEM, n = 3 individual mice, data from one of three similar experiments, asterisks indicate significant expression differences between unstimulated and stimulated p14 TM). *p , 0.05, **p , 0.01, ***p , 0.001.

central TM (TCM), we evaluated the possibility that TCM in the pe- (32, 33). To reconcile these discrepancies, we evaluated constitutive ripheryexpresslowerCCL5levelsthaneffectorTM (TEM). Although CCL5 expression in major hematopoietic cell subsets isolated this was indeed the case (Fig. 2A), the differences were small, from lymphatic and nonlymphoid organs of naive mice (Fig. 2E, and the progressive TEM→TCM conversion in aging pathogen- Supplemental Fig. 2). In this study, ex vivo detectable CCL5 was immune mice (25, 52) was associated with only a slight re- restricted in all tissues to two hematopoietic lineages, NK and duction of Ccl5 message and overall CCL5 content in splenic CD3ε+T cells. Moreover, substantial CCL5 expression by NK cells + CD8 TM (Fig. 2C, 2D). Thus, constitutive CCL5 expression, as reported previously (22) contrasted with the weak CCL5 ex- + in contrast to the rapid (GzmB) or gradual (GzmA) decline pression by CD8 TMP that was slightly more pronounced in the of constitutively expressed components within the cytolytic CD122+ subset (Supplemental Fig. 1B–E); additional minor CCL5+ + + effector pathway, is a distinctive hallmark of most periph- T cell subsets included gdTCR T cells as well as CD4 TMP and + eral CD8 TM populations maintained in long-term memory NKT cells that harbored constitutive CCL5 at marginal levels (Fig. 2D and Ref. 25). (Supplemental Fig. 1C, 1F, 1G). In no case did we find ex vivo Our findings are in apparent contrast to the reported presence of CCL5 expression by B cells, , , dendritic hi + Ccl5 mRNA but absence of protein in murine CD44 CD8 TMP cell (DC) subsets, or granulocytes (Supplemental Fig. 2B). The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

+ FIGURE 2. Constitutive CCL5 expression by pathogen-specific TM.(A) Constitutive chemokine expression by LCMV-specific CD8 TM was analyzed in different tissues on d175 (top row, spleen) or d55 (middle and bottom rows) after acute LCMV infection. Bottom row, middle, PBMCs from rLM-OVA– + b + immune mice were analyzed on d83. Two additional contour plots (right) gated on epitope-specific CD8 TM recovered from LCMV-immune B6 (D NP396 , d + hi d175, spleen) or BALB/c (L NP118 , d174, PBMC) mice demonstrate a slight, but significant, reduction of constitutive CCL5 expression levels by CD62L + lo + CD8 TCM versus CD62L TEM subsets. (B) Summary of constitutive CCL5 expression by LCMV-specific CD8 TM recovered from lymphatic and nonlymphoid organs. (C) Temporal regulation of ex vivo detectable Ccl5 mRNA and CCL5 protein expression by aging p14 TM (the gray shaded area demarcates the transitional period from effector [d8] to early memory [d42] stage, and asterisks identify significant differences comparing young [∼d40–50] b + + to older p14 TM). (D) Complementary ex vivo CCL5 expression data by endogenously generated D NP396 CD8 TM; the dot plot insert shows constitutive b + + + CCL5 versus GzmA expression in blood-borne d53 D NP396 CD8 TM.(E) Ex vivo CCL5 expression by CD8 T cells from (Figure legend continues) 6 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS

+ + The low-level CCL5 content of CD8 TMP in naive mice pro- of aging CD8 TM, as did the CCL1 production potential by a vides an explanation for the seeming absence of CCL5 in these major subset (Fig. 3C, Supplemental Fig. 2B). In contrast, in- cells (32, 33) and stands in stark contrast to the robust CCL5 ducible CCL9/10 and XCL1 expression tracked with the modu- + expression pattern of CD8 TMP in age-matched, LCMV-immune lation of the respective mRNA patterns in the memory phase + mice (Fig. 2E). Thus, a viral challenge can leave an imprint of (i.e., aging CD8 TM populations featured a decreasing potential prior infection by shifting the relative distribution of constitutive for CCL9/10, but increasing proficiency for XCL1 production) + CCL5 expressors from both NK cells and CD8 TMP (each con- (Supplemental Figs. 2A, 2B, 3C). In fact, within the temporal ∼ + + tributing 50% to the CCL5 population in naive mice) to context of increasingly diversified CD8 TM activities (25), the + + CD8 TMP as the predominant constituent of hematopoietic CCL5 rising XCL1-production competence constitutes the most pro- . + cells ( 90%) (Supplemental Fig. 2H, 2I). These observations, nounced functional gain for aging CD8 TM, prompting further together with the presence of CCL5 in the sera of naive mice at investigations, as discussed below. Finally, the relative proportions ∼ + levels of 1 ng/ml (Fig. 2F), may warrant a reclassification of of chemokine-producing CD4 TM subsets, despite the gradual + CCL5 as a dual function rather than inflammatory chemokine. decline of total specific CD4 TM numbers (47), did not signifi- + Although CD8 TM would be an obvious source for the serum cantly change over time (Fig. 3C). CCL5, levels were not increased in LCMV-immune mice (Fig. + + CD8 TM-produced chemokines in context 2F), and the absence of spontaneous CCL5 release by CD8 TMP confirms earlier reports about the need for TCR stimulation to The full spectrum of rapid effector functions elaborated by + induce CCL5 secretion (32, 33). pathogen-specific CD8 TM remains, at present, unknown. Al-

though the induction of cytokines (IFN-g,IL-2,IL-3,and Downloaded from Induced chemokine synthesis by pathogen-specific CD8+ GM-CSF), TNFSF ligands (TNF-a, CD40L, and FasL) and de- and CD4+T M granulation are readily observed in short-term restimulation cul- + To delineate the principal chemokine production capacity of en- tures (25, 43, 53); virus-specific, CD8 TM-producing IL-4, IL-5, + dogenously generated pathogen-specific CD8 TM, we visual- IL-10, IL-13, or IL-17 are rare, if present at all, in our experi- ized induced chemokine expression by peptide restimulation of mental systems. To provide a provisional estimate for the relative +

CD8 TM obtained from LCMV- and rLM-OVA–immune mice. contribution of chemokine production to the gamut of established http://www.jimmunol.org/ + + Similar to p14 TM, nearly all specific CD8 TM readily made CD8 TM functionalities, we quantified the fractions of NP396- + CCL3/4/5, large subsets produced XCL1, and 15–25% expressed specific CD8 TM capable of individual chemokine, cytokine, and CCL1 and CCL9/10; no other chemokines were synthesized in TNFSF ligand synthesis, constitutive GzmA/B and perforin ex- response to TCR activation (Fig. 3A, 3B and data not shown). pression, and degranulation (Fig. 3D); although at best a rough Because these expression patterns were further consistent across approximation, our calculations indicate that chemokine synthesis + different epitope specificities [and therefore also independent of may account for .40% of CD8 TM activities and thus suggest a + immunodominant determinants and functional avidities (25, 34)], potential importance of CD8 TM-derived chemokines in the they apparently constitute a largely invariant functional property context of recall responses. + of pathogen-specific CD8 T . Similar analyses conducted with by guest on September 24, 2021 M Induced CCL1 expression as a determinant for polyfunctional LCMV- and rLM-OVA–specific CD4+T also revealed subsets M CD8+ and CD4+T subsets producing CCL3/4/5 (40–50%), CCL1 (20–30%), XCL1 (5–15%), M or CCL9/10 (,5%) (Fig. 3A, 3B); these expression patterns es- The extent to which individual effector molecules contribute to I˚ sentially recapitulate the chemokine response of the respective T cell–mediated pathogen control has been delineated in multiple + pathogen-specific CD4 TE (34). experimental systems, yet comparatively little information is + The long-term maintenance of CD8 TM populations is associ- available about their precise role in immune protection afforded ated with a gradual remodeling process that promotes their by recall responses. Rather, the concept of T cell poly- functional diversification, as reflected in an increasing capacity functionality, typically measured in IFN-g/TNF-a/IL-2 coex- for IFN-g, IL-2, CD40L, and FasL synthesis (25). We therefore pression analyses, argues that the presence of TM with a monitored the temporal modulation of chemokine mRNA species diversified and robust functional repertoire correlates with better in aging p14 TM and observed four distinct patterns of longitudinal pathogen control irrespective of the precise mechanisms by which mRNA expression: enduring absence (Ccl1), a continuous de- the analyzed T cell functions may in fact contribute to this task crease (Ccl5 and Ccl9), a progressive increase (Xcl1), and an (53, 54). In this study, we demonstrate that induced CCL1 ex- + + initial ∼5 mo decline followed by a subsequent rise (Ccl3 and pression by virus-specific CD8 and CD4 TM identifies subsets Ccl4) (Supplemental Fig. 2A), a somewhat unusual dynamic characterized by highly diversified effector functions and in- comparable to that previously reported for Ifng, Tnf, Prf1, and creased production of multiple effector molecules (because we Gzmb/k/m (25). The relevance of the latter kinetics, however, re- were unable to concurrently visualize more than two to three mains unclear because inducible CCL3/4 production was not al- chemokines, we relied on a series of successive and comple- + tered in aging NP396-orGP33/4-specific CD8 TM (Fig. 3C, mentary stains to define complex functional TM profiles according Supplemental Fig. 2B). Similarly, and notwithstanding the slight to 11 parameters) (Fig. 4). + decline of constitutive CCL5 expression (Fig. 2C, 2D), the ca- In addition to IFN-g, LCMV-specific CD8 TM subsets rapidly pacity for stimulated CCL5 synthesis remained a stable property synthesize TNF-a, IL-2, GM-CSF, and CD40L (25, 43), and the

age-matched naive versus LCMV-immune (day 45 [d45]) mice; the corresponding bar diagram quantifies the fraction (percentage) and expression levels + + (geometric mean fluorescence intensity [GMFI]) of CCL5 CD8 TMP in naive versus young and aged LCMV-immune mice (statistical significance in- dicated by asterisks, no difference between d45 and d206 LCMV-immune mice). (F) Serum CCL5 levels as a function of time after LCMV infection of B6 mice (top) and age-matched uninfected B6 mice (bottom); the broken line indicates the detection threshold as determined in LCMV-immune B6.CCL52/2 mice. All summary data represent SEM with three to four individual mice analyzed per group or time point in multiple independent experiments [(A, B, and E) two to four experiments; (C, D, and F) one to two experiments]. *p , 0.05, **p , 0.01. The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

+ FIGURE 3. Induced chemokine expression by pathogen-specific TM.(A) Induced chemokine expression by splenic GP33/4-specific CD8 (top row) and + GP64-specific CD4 TM (bottom row) evaluated on d55 after acute LCMV challenge of B6 mice. (B) Top, Summary of induced chemokine expression by + + + LCMVepitope–specific (IFN-g ) CD8 and CD4 TM (SEM, three mice per group, d42). Bottom, Induced chemokine profiles of specific TM recovered from the spleens of rLM-OVA–immune mice (SEM, three mice per group, d83). Data are representative for two to three independent experiments conducted at various + + time points after LCMV or rLM-OVA infection, and asterisks indicate significant differences between respective pathogen-specific CD8 and CD4 TM.(C) Induced + + chemokine production by NP396-specific CD8 TM (black) and GP64-specific CD4 TM (gray) as a function of time after LCMV challenge; details for data display and statistics as in legend to Fig. 2C (SEM of three to six mice per time point; all data were generated with splenic TM, with the exception of XCL1 production, for + which blood-borne NP396-specific CD8 TM were analyzed, and were combined from four independent experiments). (D)CompositionoftheNP396-specific + CD8 TM response (∼d42) as reflected by the relative magnitude of individual subsets expressing constitutive (GzmA/B and perforin) or inducible (all other in- cluding CCL5) effector activities (data for the pie chart represent averages of $3 mice analyzed in up to five separate experiments). *p , 0.05, **p , 0.01, ***p , 0.001. 8 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS level of IFN-g expression in TNF-a+, IL-2+, GM-CSF+, and 10 min for induced CCL3/4 and IFN-g production, whereas XCL1 CD40L+ subsets was consistently and significantly higher than in synthesis was significantly delayed (∼2.0 h, p , 0.0002), yet + CD8 TM subsets, producing only IFN-g (p , 0.05, analysis of faster than the eventual induction of CCL1 (∼3.5 h, p , 0.0001) + NP396-, GP33/4-, and GP276-specific CD8 TM; data not shown). (Fig. 5A, 5B). In fact, a direct comparison with other effector Combinatorial analysis of these functions demonstrated that IL-2+ functions demonstrated that induced chemokine production + + + and CD40L , but not GM-CSF , CD8 TM belong to subsets that dominates the earliest (CCL3/4) and later (CCL1 and CCL9/10) + also produce IFN-g and TNF-a at higher levels. Among the IL-2 stages of the CD8 TM response (Fig. 5B) and also illustrates that and CD40L producers, however, only ∼25% exhibited coex- the presence/absence of pre-existing mRNA species is not the sole pression, whereas another ∼25% were IL-2+/CD40L2, and ∼50% predictor for the temporal elaboration of effector functions: al- expressed CD40L in the absence of IL-2 (Fig. 4A). Similar to though absent at the mRNA level in resting p14 TM, Il2, Csf2/ + + CD8 TE, CD8 TM coproduced CCL3/4 (Fig. 4B, upper left Gmcsf, and Cd40lg mRNA species were rapidly transcribed fol- panel), the particularly tight association of which may, in part, be lowing TCR stimulation (Supplemental Fig. 2C), but translation of due to the formation of intracellular CCL3/4 hetero-oligomers CD40L, IL-2, and GM-CSF proteins occurred with significantly (34). Further comparing the extent of induced CCL3 production different kinetics (Fig. 5B). Conversely, despite the presence of in subsets expressing additional chemokines with those that do Ccl9 mRNA (Supplemental Fig. 2A), synthesis of the corre- not, CCL3 was significantly increased in CCL1+ and XCL1+ and, sponding protein was a relatively late event (Fig. 5B). Because to a lesser extent, CCL9/10+ populations (Fig. 4B, upper panels). differential mRNA stability may shape the temporal order in In fact, the CCL1+ population is a subset of XCL1+ cells that which encoding inflammatory mediators are expressed produced significantly more CCL3 and IFN-g compared with the (56), it is conceivable that this phenomenon also contributes to Downloaded from + 2 + XCL1 /CCL1 CD8 TM, which in turn, exhibited higher CCL3/ the observed differences in translation kinetics. Nevertheless, 2 2 IFN-g levels than the XCL1 /CCL1 population. In addition, the significant increase of Xcl1 message in aging p14 TM although IL-2 producers were found in both CCL12 and CCL1+ (Supplemental Fig. 2A) allowed for an analysis of the kinetics populations, the latter subset was significantly enriched for IL-2+ with which an individual chemokine is synthesized as a function cells (∼45 versus 23%), which also produced IL-2 at higher levels of changing mRNA levels. As shown in Fig. 5C, not only did more 2 + (p = 0.0206) (Fig. 4B, middle panels). CCL1 and CCL1 pop- aged p14 TM (∼21 mo) produce XCL1, but they did so ∼20 min http://www.jimmunol.org/ ulations were also stratified according to CD40L expression: in faster than younger p14 TM (∼9 wk). In contrast, the kinetics of this study, the CCL1+ subset was clearly enriched for both IL-2+/ CCL1/3/4 and IFN-g production were not significantly different in + 2 + + CD40L and IL-2 /CD40L subsets (Fig. 4B, lower panels). Fi- young and old CD8 TM (data not shown). + nally, the majority (∼3/4) of CCL9/10 expressors were found To relate the preceding in vitro data to a specific in vivo CD8 TM among CCL12 cells and, expectedly, expressed significantly lower response, B6 mice previously immunized with combined TLR/ amounts of CCL3 (Fig. 4B, lower panels). Therefore, induced CD40 vaccination were subjected to a peptide challenge and co- CCL1 expression defines a highly polyfunctional, virus-specific administration of BFA to allow for the intracellular accumulation + CD8 TM subset comprising ∼25% of epitope-specific cells that of chemokines and other effector molecules in vivo (34, 44). hi hi + + hi + by guest on September 24, 2021 are characterized by a CCL3 /CCL4 /CCL5 /XCL1 /IFN-g / Within 2 h after challenge, ∼2/3 of the vaccine-specific CD8 TM TNF-ahi phenotype, are enriched for IL-2hi and/or CD40L+ cells, became activated, as determined by CD69 expression, and the and exhibit variable GM-CSF and/or CCL9/10 expression (CCL5 majority of these cells readily synthesized CCL3/4 and XCL1 + + analyses are not displayed in Fig. 4B because all specific CD8 TM (Fig. 5D), demonstrating that the in vivo CD8 TM response fea- are capable of induced CCL5 production, shown in Fig. 3A–C). tures rapid and abundant chemokine production. Finally, we Similar functional stratifications were also performed for determined the secretion kinetics and quantities of CCL5, con- + + LCMV-specific CD4 TM, taking into account that their relative stitutively expressed by LCMV-specific CD8 TM (day 45 [d45]) in IL-2+ and CD40L+ subsets are significantly larger than the cor- an ELISA format. Interestingly, both the rapid speed of CCL5 + responding CD8 TM subpopulations (43) (Fig. 4C, 4D). Collec- release and the amount of prestored CCL5 secreted by individual + + tively, these analyses demonstrate that induced CCL1 production CD8 TM (Fig. 5E) were equivalent to CD8 TE (34). Thus, the + + is also useful to delineate a polyfunctional CD4 TM subset that CD8 TM chemokine response proceeds through an ordered se- + accounts for ∼25% of the specific CD4 TM compartment and is quence of overlapping events that are, in part, dictated by the defined by a CCL3hi/CCL4hi/IFN-ghi/TNF-ahi/IL-2+/CD40Lhi extent of pre-existing protein and/or mRNA species: CCL5 → expression pattern with variable induction of XCL1, GM-CSF, CCL3/4 and IFN-g → TNF-a → CD40L → XCL1 → IL-2 → GM- and/or CCL9/10. CSF → CCL1 → CCL9/10. Given the particularly fast release of CCL3/4/5, stimulated CD8+T rather than DCs or other APCs Orchestration of the CD8+T chemokine response (I): kinetics M M (57) likely constitute the principal source for these chemokines in of chemokine production a recall response. The presence of chemokine transcripts in TM has been suggested to + confer a kinetic advantage that permits more rapid protein syn- Orchestration of the CD8 TM chemokine response (II): thesis (32, 55), but this hypothesis has, to our knowledge, not yet activation thresholds for chemokine production and secretion been experimentally tested. The distinct expression patterns of ex In addition to the temporal regulation of the T cell chemokine vivo detectable chemokine mRNA species in LCMV-specific response, production of individual chemokines may be controlled + CD8 TM (absence of Ccl1 mRNA, abundant Ccl3/4/5 mRNA, by distinct activation thresholds. The functional avidities of specific and progressively increasing Xcl1 mRNA; Supplemental Fig. 2A) T cells are often determined by measuring induced IFN-g synthesis would therefore suggest that early synthesis of CCL3/4 is fol- in response to graded concentrations of antigenic peptide (34, 58). lowed by that of XCL1 and, finally, CCL1 protein. Experiments Although the interpretation of such data provides insights into performed with LCMV-immune p14 chimeras indeed confirmed properties intrinsic to the interaction between TCRs and peptide/ this prediction: by calculating the time point at which p14 TM MHC complexes, it should be noted that the experimental read- demonstrate detectable chemokine expression in half of the re- out is modulated by complex signaling events that precede the sponder population (ET50), we found ET50 values of ∼1 h and induction of a given T cell function. For example, functional The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

+ + FIGURE 4. Induced CCL1 expression as a determinant for highly polyfunctional CD8 and CD4 TM subsets. (A) Patterns of induced cytokine and + + TNFSF ligand expression by GP33/4–specific CD8 TM (d67, spleen). Middle and bottom rows, Red dots designate specific CD8 T cells producing a third cytokine (identified in red font) in addition to IFN-g/TNF-a (or CD40L/IL-2) expression displayed in conventional dot plots. (B) Top row, left, CCL3/4 + + coexpression by all GP33/4–specific CD8 TM; top row, right, specific CD8 TM subsets producing CCL1, CCL9/10, or XCL1 also express higher levels of CCL3 (the color-coded CCL3 geometric mean fluorescence intensity [GMFI] values refer to cells identified with the same color; statistical significanceis indicted by asterisks). Middle row, Induced CCL1/XCL1 expression and corresponding CCL3 and IFN-g expression levels in subpopulations identified by color-coded histograms and corresponding GMFI values. In addition, IL-2 production by CCL12 (black tracing) versus CCL1+ (red tracing) subsets is shown. Bottom row, left, Approximately eighty percent of CCL1+ cells fail to induce CCL9/10 expression, suggesting that XCL1 and CCL9/10 are expressed in a reciprocal fashion in CCL1+ subsets (compare with above CCL1/XCL1 plot). Lack of CCL9/10 production is (Figure legend continues) 10 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS avidities for induced IFN-g production were reported to be lower time, constitutive CCL5 and induced CCL3/4/5/9/10 expression than those required for IL-2 production by the same cells, sug- were undistinguishable (Supplemental Fig. 3A, 3B). In contrast, + gesting that limiting Ag concentrations may permit the elaboration II˚ CD4 TE presented with a comparatively enhanced functional of direct T cell effector functions (IFN-g), whereas induction of diversity, as reflected in substantially larger fractions of TNF-a+, proliferative T cell responses, mediated by T cell–produced IL-2, CCL1+,CCL3+,CCL4+,CCL5+,andXCL1+ subsets, a chemo- + preferentially occurs at higher Ag loads (53). kine profile that in fact resembles the response of I˚ CD8 TE We therefore compared the functional avidities/activation (Supplemental Fig. 3B). We also extended these analyses to the + thresholds for induced chemokine and cytokine/TNFSF expres- formation of II˚ memory, a stage at which II˚ CD8 TM typically + + sion by LCMV-specific CD8 TM (Fig. 5F–H). Although the acti- display a less-mature phenotype than I˚ CD8 TM (Ref. 25 and + vation thresholds for 10 distinct CD8 TM effector functions Supplemental Fig. 3C). In agreement with this notion, II˚ + varied, at most, by a factor of only ∼3.0 (Fig. 5G), several findings CD8 TM produced more CCL9/10 and less XCL1 (Supplemental + are noteworthy: 1) the activation thresholds for CCL3/4 produc- Fig. 3D), the only two chemokines subject to CD8 TM maturation- tion were only slightly lower than for IFN-g induction, yet we associated modulations (cf. Fig. 3C); induced CCL1/3/4/5 expres- + + consistently observed the emergence of a small CCL3 or sion, however, was largely similar between I˚ and II˚ CD8 TM.II˚ + + CCL4 T cell subset in the absence of detectable IFN-g at limiting CD4 TM, in contrast, mostly preserved the functional gains accrued peptide concentrations. In fact, at a peptide concentration of 10210 in the II˚ effector phase and exhibited enhanced CCL1/3/4/5 + M, the very-small subset of responding p14 TM consisted to and XCL1 production capacity in comparison with I˚ CD4 TM .50% of CCL3/4+ cells that were IFN-g2. In contrast, no IFN-g (Supplemental Fig. 3D). induction was observed in the absence of CCL3/4 expression To determine if the major TM-derived chemokines contribute to Downloaded from + (Fig. 5F). Parallel analyses of aged p14 TM (∼21 mo) yielded the regulation of II˚ CD8 TE responses, we conducted adoptive identical results and indicate that subtle differences pertaining to transfer/rechallenge experiments with CD8+T cell–enriched donor the thresholds for defined effector functions are preserved in long- populations under conditions of systemic chemokine neutraliza- term memory (data not shown). 2) Although production of CCL1 tion. We first focused on the potential role of CCL5 and observed and XCL1 was initiated with delayed kinetics (Fig. 5B), the ac- that its blockade resulted only in a very minor impairment of II˚ + tivation thresholds for these chemokines were relatively low and CD8 TE responses (Supplemental Fig. 3E). A similarly negligible http://www.jimmunol.org/ + comparable to CCL3/4 (Fig. 5G). And 3) the highest activation impact on the concurrent I˚ CD8 TE expansions (Supplemental + thresholds were recorded for CD40L and, in agreement with Fig. 3E) is in general agreement with the unperturbed I˚ CD8 TE earlier observations (53), IL-2 (Fig. 5G). To directly compare the responses generated by B6.CCL52/2 mice in the wake of an acute thresholds for CCL3/4 synthesis with CCL5 release and produc- LCMV challenge (34, 61). Our conclusion that CCL5 is dis- + tion, we determined the respective functional avidities by ELISA pensable for CD8 TM recall responses is further supported by + (Fig. 5H). The activation thresholds for CCL3 and CCL4 secretion experiments that demonstrated commensurate II˚ CD8 TE reac- + were expectedly comparable but, importantly, significantly higher tivities of wild-type and CCL5-deficient CD8 TM (data not shown than that for induced CCL5 release. In summary, at limiting Ag and Ref. 61). Somewhat surprisingly, even the combined neu-

+ by guest on September 24, 2021 concentration, the IFN-g response is preceded by secretion of the tralization of CCL3/4/5 or XCL1 failed to alter II˚ CD8 TE re- chemokines CCL5 and, to a certain extent, CCL1/3/4 and XCL1. sponses (Supplemental Fig. 3F). Thus, at least in the context of an + These chemokines may stabilize interactions between T cells and acute LCMV rechallenge, none of the major CD8 TM chemokines + target cells or APCs (59), may provide costimulatory signals (60), appear to contribute to the regulation of II˚ CD8 TE responses. exert direct effector functions (27), or mediate additional functions Chemokine profiles of specific T cells in chronic LCMV currently under investigation. infection (I): effector stage + + II˚ CD8 and CD4 TE/M immunity: expression profiles and Challenge of naive mice with the LCMV variant cl13 (2 3 106 PFU functional roles of T cell chemokines i.v.) results in a chronic infection characterized by prolonged virus Given the prominence of the TM chemokine response, we next persistence and T cell exhaustion, as reflected in a progressive assessed the potential modulation of chemokine-expression pro- functional deterioration of their IL-2, TNF-a, and IFN-g pro- files in the context of a recall response. To this end, we employed duction capacity (48, 62, 63). We previously noted that an esca- an adoptive transfer system in which T cell–enriched populations lation of acute pathogen infection dosage depressed CCL1/3/4/5 + + from LCMV-immune mice are transferred into naive congenic and XCL1 synthesis potential of specific CD8 , but not CD4 TE recipients that are subsequently challenged with LCMV; in this (34) and therefore speculated that a similar CD8+T cell incapac- study, II˚ TE responses derived from the transferred population itation would also occur under conditions of chronic viral infec- can be concurrently visualized with the I˚ TE response generated tion. Accordingly, we inoculated mice with LCMV cl13 to initiate + by the host (25). At the level of II˚ CD8 TE responses, we ob- a persistent infection and performed a first set of analyses at the served a distinct reduction of functional diversities (i.e., in height of the TE stage (d8; control cohorts were infected with 2 3 + 6 comparison with I˚ CD8 TE, the former cells produced less IFN- 10 PFU LCMVArm i.v., which does not result in viral persistence g,TNF-a, and CD40L as well as CCL1 and XCL1); at the same or T cell exhaustion). In agreement with its previously described

associated with higher CCL3 (as well as IFN-g and IL-2) expression. Bottom row, right, CCL4 GMFI values and IL-2/CD40L expression patterns by CCL12 (blue) versus CCL1+ (red) subsets. All values listed in (A) and (B) are the average of three mice analyzed. (C) Complementary analyses of cytokine and + + + hi hi TNFSF ligand production by GP64-specific CD4 TM [data display as in (A); note that IL-2 cells are CD40L , IFN-g , and TNF-a ]. (D) Top row, + + Coproduction of CCL3/4 by a major CD4 TM subset and enrichment of CCL1, CCL9/10, and XCL1 expression in the CCL3/4 subpopulation. Middle row, Progressive enrichment of CD40L+ IL-2 producers in CCL12/42 → CCL12/4+ → CCL1+/4+ subsets (color-coded gating strategy). Bottom row, No clear + + association of induced CCL9/10 and XCL1 production with CCL1 expression. However, CCL1 CD4 TM not only contain more IL-2 producers, the latter cells are also enriched for a pronounced IFN-ghi phenotype. Representative data from two independent experiments are shown, and the text boxes below + + (A)–(D) summarize the functional characteristics of LCMV-specific CD8 TM and CD4 TM subsets identified by induced CCL1 expression. *p , 0.05, **p , 0.01, ***p , 0.001. The Journal of Immunology 11 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

+ FIGURE 5. Kinetics and activation thresholds for induced chemokine production by specific CD8 TM.(A and B) Spleen cells from LCMV-immune p14 chimeras (∼10 wk) were stimulated for 0–5 h with GP33 peptide, and expression of indicated chemokines, cytokines, and TNFSF ligands was determined as a function of stimulation time. The dot plots (gated on p14 TM) show representative data for the temporal regulation of induced CCL1/3/4, XCL1, and IFN- g synthesis, and the bar diagram ranks inducible chemokine (black) and other effector molecule (gray) production according to ET50 values (time required to induce effector functions in 50% of the population capable of producing a given effector molecule); significant differences are indicated by asterisks. Overall, induced CCL3/4 and IFN-g synthesis proceeds with significantly faster kinetics as compared with other effector functions (p , 0.003), with the exception of TNF-a.(C)ET50 values for induced XCL1 expression by young (d64) versus aged (d637) p14 TM were determined as above. (D) Ap- proximately two months after OVA/anti-CD40/poly(I:C) vaccination, B6 mice were injected with BFA only or OVA257 peptide and BFA as indicated, and b + + activation status (CD69) and in vivo IFN-g, TNF-a, and chemokine production by K OVA257 CD8 TM was determined as described in Materials and b + + Methods; values are the percentage of K OVA257 CD8 TM in respective quadrants synthesizing indicated cytokines/chemokines. (E) Kinetics of prestored + CCL5 release by individual NP396-specific CD8 TM (d45) were determined in stimulation cultures supplemented with CHX to prevent protein neosyn- thesis; for comparison, CCL5 secretion in the absence of CHX is featured for the 5-h time point (gray symbol). (F) Intracellular IFN-g and CCL3/4 content of p14 TM (∼9 wk) determined after 5 h stimulation with graded doses of GP33 peptide (plots gated on p14 TM). The adjacent (Figure legend continues) 12 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS tropism (64, 65), LCMV cl13 preferentially infected fibroblastic re- complicate this picture because in both LCMV cl13 and Arm ticular cells (FRCs) and APC subsets but largely spared lymphocytes infection, ∼3.5-fold fewer IFN-g–producing than tetramer+ + + (Fig. 6A and data not shown); furthermore, CD8 TE expansions CD4 TM were recovered from the spleen (Fig. 7D). Thus, in the after cl13 infection were diminished in the expected epitope- dynamic interplay between specific TM and persisting virus that + + dependent fashion (i.e., NP396 . GP33/34 . GP276), and the may constrain (CD8 TM) or expand (CD4 TM) reactive T cell + functionalimpairmentofCD4TE appeared particularly pro- pools, concurrent functional impairments, at least at the level b + nounced (compare numbers of IA GP66 and corresponding IFN- of IFN-g production, may become uncoupled such that lesser + g–producing, GP64-specific CD4 TE) (Fig. 6B, 6C; note that numerical reduction can be associated with greater functional b D GP33 tetramer stains and GP33 peptide stimulation cannot be handicaps. directly compared because the latter method also activates We therefore sought to broaden the functional characterization of b + + K GP34 CD8 TE). In regard to constitutive chemokine expres- specific TM populations in cl13 infection. At the level of inducible + + sion by specific CD8 TE, CCL5 was the only ex vivo detectable cytokines and TNFSF ligands, GP33/4-specific CD8 TM, and to a + chemokine after high-dose Arm i.v. infection (Fig. 6D), pre- similar, although lesser, extent also GP64-specific CD4 TM, pro- senting with a chemokine profile practically identical to that duced less TNF-a and CD40L, whereas no significant differences observed after lower dose i.p. Arm infection (34). In contrast, were noted for residual IL-2, GM-CSF, IL-3, or IL-10 expression + CD8 TE generated in response to cl13 infection expressed, in ad- (Fig. 7E and data not shown). Importantly, although constitutive + dition to CCL5, low levels of all other T cell chemokines indicative chemokine expression by specific CD8 TM in both infection of recent TCR activation in the presence of enhanced viral repli- models was restricted to CCL5, induced chemokine production in cation (Fig. 6D). The latter pattern (i.e., low-level ex vivo detectable cl13-infected mice revealed the emergence of CCL3+, CCL4+, and Downloaded from + + chemokine expression) was also observed for specific CD4 TE,yet XCL1 subsets that lacked IFN-g expression; at the same time, + its presence after both cl13 and Arm infection indicates that the very few CD8 TM synthesized CCL1 or CCL9/10 (Fig. 7F). We + CD4 TE chemokine response is more impervious to chronic viral therefore extended and summarized these observations in two challenge. Indeed, this contention is supported by a comparison of complementary ways: 1) traditional data display quantifying + + induced chemokine production profiles in which CD8 TE in cl13- chemokine-producing cells as a fraction of IFN-g TM subsets infected mice displayed impaired CCL1/3/4 and XCL1 (and ele- demonstrated impaired chemokine responses (CCL1/3/4/5 and http://www.jimmunol.org/ vated CCL9/10) expression in an epitope-dependent manner (NP396 XCL1) in cl13 infection that were more pronounced in NP396-as + + . GP33/34 and GP276), whereas the CD4 TE chemokine response compared with GP33/4- and GP276-specific CD8 TM; differences + was not compromised and, if anything, presented with greater were less evident in the CD4 TM compartment, although similar CCL9/10 and XCL1 expression (Fig. 6E). to the TE stage, GP64-specific tended to make more CCL9/10 and XCL1, whereas their CCL3/4 expression was somewhat Chemokine profiles of specific T cells in chronic LCMV diminished (Fig. 8A). And 2) by including IFN-g–negative infection (II): memory stage CCL3+, CCL4+,andXCL1+ subsets in our analyses and relat- Extending our investigations to the d30 time point, at which chronic ing the functional distributions (IFN-g2/chemokine+,IFN- LCMV infection is fully established, we found preferential in- g+/chemokine+,andIFN-g+/chemokine2) to the number of by guest on September 24, 2021 fection of APC subsets and, in particular, lymphoid reticular cells tetramer+ cells quantified in parallel assays, we found that the 2 + identified as CD45 /podoplanin cells (66) in cl13-challenged, but seeming numerical reduction of functional GP276-specific + not Arm-challenged, mice (Fig. 7A), a numerical reduction of CD8 TM subsets as determined solely by IFN-g production + specific CD8 TM compartments according to epitope specificity largely disappears (Figs. 7D, 8B, left). Thus, in the context of . . + (NP396 GP33 GP276) but also a trend toward increased spe- these complementary analyses, CD8 TM in chronic LCMV in- + cific CD4 TM numbers in comparison with the corresponding fection are found to incur both functional deficits (reduced + + CD4 TM pool in Arm-immune mice (Fig. 7B, 7C). Although the chemokine production by the IFN-g fraction) and functional + effects of cl13 infection on compromised CD8 TE expansions are gains (chemokine production in the absence of IFN-g expres- well known, a direct enumeration of specific TM numbers by ex sion). We also note that the limited expression of CCL1, in- vivo tetramer staining versus induced IFN-g production can help cluding the lasting absence of a CCL1+/IFN-g2 subset (Fig. 8B), to clarify certain peculiarities pertaining to impaired functional- may make this chemokine a particularly useful marker to as- + + ities (Fig. 7D): reduction of the splenic NP396-specific CD8 TM certain the overall functional capacities of CD8 TM in chronic pool by a factor of ∼6 is captured in absolute numbers by both infection. Finally, analyses conducted on d107 after cl13 infec- analysis modalities, implying that despite the considerable pres- tion (i.e., at a time point when infectious virus is eliminated from + sure of the chronic infection (67), the IFN-g production capacity many tissues, and a phenotypic/functional recovery of CD8 TM + of surviving NP396-specific CD8 TM is reasonably well preserved. is underway) (63, 68) revealed a slightly smaller functional b + + + In contrast, a minor numerical decrease of D GP276 CD8 TM is GP276-specific CD8 TM pool with a relative reduction of the associated with 2.5-fold fewer IFN-g–producing, GP276-specific compartments expressing only CCL3/4 or XCL1 in favor of + CD8 TM, demonstrating a considerable functional incapacitation subsets coproducing IFN-g (Fig.8B,right).Atthesametime, + + (Fig. 7D). Similar comparisons of GP64-specific CD4 TM further the NP396-specific CD8 TM pool, likely as a consequence of its

diagrams summarize the emergence of p14 TM expressing effector functions as a function of GP33 peptide concentration. (G) Summary of activation thresholds (EC50 values) determined for individual chemokines (black bars) and other effector functions (gray bars) elaborated by splenic p14 TM stim- 211 26 ulated with graded doses of GP33 peptide (1 3 10 –1 3 10 M; data are combined results from three similar experiments evaluating three p14 chimeras each; to account for interexperimental variability, all samples were costained for IFN-g, the average IFN-g EC50 value was set at 1.0 [dotted line], and the EC50 values of other effector functions were calculated accordingly). Statistically significant differences comparing the IFN-g activation threshold to CCL1, + CCL4, CD40L, and IL-2 are indicated by asterisks. (H) CCL3/4/5 secreted by p14 CD8 TM as a function of GP33 peptide concentration was determined by ELISA (5 h stimulation), and EC50 values are summarized in the adjacent bar diagram. All data in (B), (C), and (E)–(H) are representative for individual experiments conducted two to four times (SEM, n = 3–4 mice per experiment). *p , 0.05, **p , 0.01. The Journal of Immunology 13 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 6. Chemokine profiles of specific T cells in chronic LCMV infection (I): effector stage. B6 mice were infected with 2 3 106 PFU LCMVArm or cl13 i.v. as indicated and subsequently analyzed on d8. (A) LCMV–nucleoprotein (NP) expression in myeloid and lymphoid cell subsets (the APC dot plots are gated on CD3ε2CD192NK1.12 cells and feature LCMV-NP+ subsets in red; the histograms are gated on indicated lymphoid cell subsets and compare

LCMV-NP expression in Arm [gray] versus cl13 [black tracing] infection). (B and C) Enumeration of specific TE frequencies and numbers by tetramer and IFN-g stains; the arrows/values indicate the factor by which respective TE frequencies or numbers differ between Arm- and cl13-infected mice. (D)Exvivo + + chemokine expression by specific CD8 TE (top) and CD4 TE (bottom) in Arm and cl13 infection. (E) Quantification of induced chemokine production by + + epitope-specific CD8 and CD4 TE; asterisks indicate significant differences between Arm- and cl13-infected mice (all summary data are SEM with n =4 mice per group and are representative for results obtained in two to three independent experiments). *p , 0.05, **p , 0.01, ***p , 0.001. 14 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS

FIGURE 7. Chemokine profiles of specific T cells in chronic LCMV infection (II): memory stage. Analyses of

LCMV Arm- or cl13-infected B6 mice (2 3 106 PFU i.v.) Downloaded from as conducted on d30. (A) Left, LCMV–nucleoprotein (NP) expression (red) by APC and other cell subsets in cl13, but not Arm, infection (plots gated on CD3ε2CD192NK1.12 cells). Right, LCMV-NP detection in fibroblastic reticular (CD452podoplanin+) in cl13 but not Arm infection. (B and C) Enumeration of specific TM frequencies and numbers by tetramer and IFN-g stains; http://www.jimmunol.org/ the arrows/values indicate the factor by which respective

TE frequencies or numbers differ between Arm- and cl13- infected mice (black font, significant differences; gray font, NS). (D) Comparison of epitope-specific TM num- bers in the spleen as based on tetramer (ex vivo) versus IFN-g (stimulated) stains; the arrows/values indicate significant differences emerging as a consequence of the two distinct analysis modalities used for TM enumeration. E ( ) Cytokine and TNFSF ligand production by GP33/4- by guest on September 24, 2021 + specific CD8 TM.(F) Ex vivo (top) and induced (bottom) + chemokine expression by GP276-specific CD8 TM in Arm and cl13 infection (all plots gated on CD8+T cells). All summary data are SEM, with n = 4 mice per group, and represent results from two independent experiments. *p , 0.05, **p , 0.01, ***p , 0.001.

protracted recovery, exhibited particularly prominent fractions of B6, B6.CCL12/2, B6.CCL32/2, and B6.CCL52/2 mice with cl13 IFN-g2/chemokine+ subsets (Fig. 8B, right). and quantified viral titers in the serum in biweekly intervals over a period of ∼10 wk. In B6 mice, high titers of infectious virus in the Chronic LCMV infection under condition of systemic serum are maintained for ∼1 mo before onset of a slow decline chemokine deficiency and eventual virus clearance from this compartment by ∼3mo Finally, to explore potential roles for major T cell–derived che- (48, 63). In the present experiments, the kinetics of virus con- mokines in control of a chronic LCMV infection, we challenged trol were identical in B6, B6.CCL12/2, and B6.CCL32/2 mice, The Journal of Immunology 15 whereas B6.CCL52/2 mice, as reported in an earlier study (61), Accordingly, although we cannot categorically rule out the pos- presented with a modest impairment of late LCMV control sibility that minor quantities of other chemokines are produced by (Fig. 8C). Furthermore, virus titers in various tissues as well as pathogen-specific TM under some experimental or naturally oc- specific TM frequencies, numbers, phenotypes, and functional- curring conditions, we are confident that the visualization of in- ities were not significantly different in B6, B6.CCL12/2, and duced CCL1/3/4/5/9/10 and XCL1 production accurately captures B6.CCL32/2 mice (Supplemental Fig. 4 and data not shown). their distinctive chemokine production potential. Related caveats Although these experiments cannot directly address the role of also pertain to our attempt to position the prima facie already + TM-derived CCL1 or CCL3, the absence of a phenotype under considerable magnitude of the CD8 TM chemokine response in a conditions of systemic chemokine deficiency strongly suggests larger frame of reference: to estimate the relative contribution of + that TM-produced CCL1 or CCL3 are unlikely to contribute to the chemokine production to the overall size of the inducible CD8 TM partial control of a chronic LCMV infection. functionome, we quantified a broad although likely incomplete + spectrum of specific CD8 TM activities and estimate that .40% + Discussion of CD8 TM effector functions are dedicated to chemokine pro- We previously profiled the chemokine response of pathogen- and duction. Although clearly only a provisional approximate, it vaccine-specific TE and defined a series of distinctive properties nonetheless suggests a remarkable prominence of chemokine + pertaining to its constituents (CCL1/3/4/5/9/10 and XCL1, but no synthesis as part of the immediate pathogen-specific CD8 TM other chemokines), apparent magnitude, quantitative differences response. + + between CD8 and CD4 TE, consistency of expression patterns By extending our investigations to the visualization of complex across different challenge models, and to unique aspects of indi- chemokine coexpression patterns, we further expand the notion of Downloaded from vidual chemokine synthesis, coexpression, cooperative regulation, T cell polyfunctionality and, at the same time, propose a potential and/or secretion (34). We now demonstrate that most of these analytical simplification as based on inducible CCL1 expression. properties are carried forward into the memory stage, identify T cell polyfunctionality as a diagnostically relevant concept was discrete traits subject to further temporal modulation, and delin- first introduced by R. Seder’s group (69), who identified a simple + + + eate notable adjustments of the TM chemokine response under metric (composed of the number of IFN-g /TNF-a /IL-2 -specific + conditions of pathogen persistence. Collectively, these character- CD4 TM [triple producers] as well as the amount of cytokines http://www.jimmunol.org/ istics establish TM-derived chemokines as particularly prominent, produced by these cells [measured by the mean fluorescence in- robust, and rapidly mobilized components of the recall response; tensity of cytokine stains]) that can predict the extent of immune in addition, we demonstrate that the analytical visualization of TM protection in a model for Leishmania major infection. Although it chemokine-expression patterns permits a more detailed stratifi- is important to note that T cell polyfunctionality is a correlate for cation of TM functionalities that may be correlated with differ- immune protection (i.e., it does not make claims about the precise entiation status, protective capacities, and potential fates. mechanisms by which specific T cell activities contribute to im- + + Similar to pathogen-specific CD8 TE (34), specific CD8 TM mune protection), the concept has been successfully applied to evaluated 6–12 wk after acute challenge with LCMV or L. mon- other pathophysiological scenarios and CD8+T cells and may ocytogenes present with a spectrum of inducible chemokines that further incorporate cytotoxic capacities (54, 70). Our stratification by guest on September 24, 2021 + is both circumscribed and distributed across larger and smaller of LCMV-specific CD8 TM responses according to five cytokines/ subsets (CCL3/4/5, 85–95%; XCL1, 60–70%; and CCL1 and TNFSF ligands and six chemokines now reveals that visualization + CCL9/10, 15–30%). The same select chemokines are also pro- of induced CCL1 marks a highly polyfunctional CD8 TM subset + duced by pathogen-specific CD4 TM, although in a resemblance that produces larger quantities of IFN-g, TNF-a, CCL3, and to the corresponding effector populations (34), induced chemokine CCL4, readily expresses CCL5 and XCL1, and is enriched for synthesis is restricted to somewhat smaller subsets (CCL3/4/5, IL-2hi and/or CD40L+ cells; a corresponding dissection of LCMV- + 40–50%; XCL1, 5–15%; CCL1, 20–30%; and CCL9/10, ,5%). specific CD4 TM populations demonstrates a similar utility of + + These CD8 and CD4 TM chemokine signatures are largely pre- CCL1 expression analyses because this subset presents with a served throughout long-term memory but with two significant distinct IFN-ghi/TNF-ahi/CCL3hi/CCL4hi/IL-2+/CD40Lhi pheno- + exceptions: a precipitous decline of Ccl9 mRNA in aging CD8 TM type. Furthermore, and in contrast to all other chemokines, CCL1 diminishes their corresponding protein production capacity, and a synthesis is strictly dependent on de novo mRNA transcription, + + progressive increase of Xcl1 message permits larger CD8 TM and its induction in specific CD8 TM is compromised under subsets to synthesize XCL1 with faster kinetics. Escalating XCL1 conditions of persistent viral infection. Although the simple vi- production potential, part of an intriguing functional diversifica- sualization of CCL1-expressing TM subsets does not allow for an + tion associated with long-term CD8 TM maintenance (25), is integration of T cell activities into a polyfunctionality index (54), + especially pronounced in blood-borne CD8 TM and arguably it may nevertheless serve as a singular metric for the quantification constitutes one of the largest functional gains bestowed on aging of highly polyfunctional T cells (acute infection) or T cells with + CD8 TM populations. At the same time, however, the repertoire of residual functional reserves (chronic infection). Mechanistically, + CD8 TM-produced chemokines is not broadened with age and however, CCL1 appears to be dispensable, as shown by the normal remains strictly confined to the same six chemokines distinctive virus control kinetics achieved by CCL1-deficient mice in acute for effector and early memory stage. (34) and chronic LCMV infection. + Our conclusion about the defined, restricted, and mostly in- Perhaps the most striking aspect of the CD8 TM chemokine + + variant nature of CD8 and CD4 TM chemokine responses shared response is the constitutive expression of CCL5 by specific + across different infectious disease models is contingent on the CD8 TM in lymphatic and nonlymphoid organs alike throughout sensitivity, specificity, and breadth of our assay systems. As we long-term memory and even in chronic LCMV infection. Unique + have detailed in our earlier reports (22, 34), the analytical power among CD8 TM chemokines, cytokines, and TNFSF ligands, the of gene array and FC technology is delimited by a variety of CCL5 expression pattern is reminiscent of that reported for con- factors that, together with the specifics of the experimental con- stituents of the perforin/Gzm pathway. However, following acute text, inform the degree of certainty with which statements about infection, GzmB expression is quickly downregulated to low or the presence/absence of individual gene products can be made. undetectable levels in LCMV-, L. monocytogenes–, and other 16 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS

FIGURE 8. Total TM chemokine responses and viral persistence under conditions of systemic chemokine deficiency. (A) Quantification of chemokine production Downloaded from + + by specific CD8 and CD4 TM on d30 after Arm or cl13 infection (significant differences are indicated by asterisks). As before, these analyses depict chemokine- producing TM subsets as a fraction of the respective + IFN-g epitope–specific TM populations. (B) Quanti- + fication of chemokine-producing CD8 TM subsets on d30 (left; Arm and cl13) and d107 (right; cl13 only). In http://www.jimmunol.org/ b + b + this study, the percentage of D GP276 and D NP396 + CD8 TM as revealed by tetramer staining in parallel experiments (data not shown) was set at 100%; the relative fraction of IFN-g2 and chemokine-producing + CD8 TM populations was calculated accordingly and was further stratified into IFN-g+/chemokine2 (light gray), IFN-g+/chemokine+ (medium gray), and IFN-g2/chemokine+ (dark gray) subsets [the summary data in (A) and (B) are SEM with n = 4 mice per group by guest on September 24, 2021 analyzed in two independent experiments]. (C) Quan- tification of LCMV serum titers in cl13-infected B6, B6.CCL12/2, CCL32/2, and CCL52/2 mice (n =5 mice per group and time point; data summarized from individual time course experiments). *p , 0.05, **p , 0.01, ***p , 0.001.

+ + pathogen-specific CD8 TM recovered from lymphatic tissues (25, expression by pathogen-specific CD8 TM in the absence of in- 52, 71–73), and constitutive GzmA expression by LCMV-specific flammatory stimuli as a distinctive functional signature. + CD8 TM is subject to a protracted downmodulation (25) (although Several additional aspects pertaining to the constitutive CCL5 + + perforin is expressed by practically all CD8 TM in the early expression by CD8 TM are noteworthy: 1) acute pathogen infec- memory stage, its subsequent temporal regulation remains, tions leave a conspicuous and lasting imprint that is readily dis- + at present, unknown). Together, these dynamics provide further cernible even without an analytical focus on specific CD8 TM:in + evidence for the differential regulation of cytotoxic effector naive mice, ∼25% of CD8 TMP express low levels of constitutive + mechanisms (74–76) and establish enduring constitutive CCL5 CCL5; in contrast, ∼70% of CD8 TMP in LCMV-immune mice The Journal of Immunology 17 present with a robust CCL5+ phenotype. In fact, in such mice, (22, 23, 34), raising the possibility that in some model systems, + + CD8 TMP rather than innate immune cells constitute the pre- II˚ CD8 TE-derived XCL1 is the defining factor for productive + + + dominant hematopoietic source of prestored CCL5 (CD8 TMP, engagement of XCR1 DCs and the regulation of II˚ CD8 TE + ∼90%; NK cells, and ∼10%; other [mostly CD4 TMP], ,5%). 2) proliferative responses. + As shown earlier, constitutive CCL5 expression is a dynamic In regard to the functional properties of II˚ CD8 TE and, in process that continuously replenishes intracellular CCL5 storage particular, TM generated in the wake of a recall response, earlier pools (31, 34); given the associated bioenergetics demands, the work has described discrete alterations in comparison with the + + + long-term preservation of a CCL5 CD8 TM phenotype may respective I˚ CD8 TE/M populations, such as reduced IL-2 pro- therefore provide crucial kinetic advantages for the recall re- duction or enhanced GzmB expression, killing capacities, and sponse. 3) Under steady-state conditions, however, prestored TNF-a induction (72, 73). Together with corresponding pheno- + CCL5 is not released from CD8 TM, as evidenced by comparable typic alterations, these observations are consistent with a greater + + CCL5 serum levels in LCMV-immune and naive mice; rather, degree of II˚ CD8 TE differentiation and delayed II˚ CD8 TM CCL5 secretion requires TCR stimulation, as reported previously maturation (25). In agreement with this interpretation, we dem- + (32, 33). And 4) yet very similar to CD8 TE (34), the initial burst onstrate reduced IL-2, TNF-a, CD40L, CCL1, and XCL1 pro- + + of TCR-triggered CCL5 secretion by CD8 TM occurs indepen- duction by II˚ CD8 TE, as well as a somewhat greater induction of dently of transcriptional/translational regulation and proceeds with TNF-a, CCL3/4, and CCL9/10 but decreased XCL1 expression by + + extraordinary speed; we estimate that within 30 min after TCR II˚ CD8 TM. Interestingly, II˚ CD4 TE avoid a corresponding at- + activation, individual CD8 TM release as much as 1.0 fg of pre- trition of functional diversity and display a substantial increase of stored CCL5. CCL1/3/4/5 and XCL1 production that also extends into the II˚ Downloaded from + The latter kinetics also impinge on the precise coordination of CD4 TM stage. Given the distinct CCL3/4 coexpression pattern of + + the CD8 TM chemokine response; because of a very low activa- a major I˚ CD4 TM subset, it might be tempting to postulate the tion threshold and its rapid release, TCR-induced CCL5 secretion existence of yet another Th cell subset, but we consider this option + constitutes the first step in a cascade of CD8 TM activities that a distraction (5) and have focused ongoing investigations on the + otherwise require efficient protein synthesis. In this study, a question if the functional skewing observed in II˚ CD4 TE re- + combination of pre-existing mRNA levels, functional avidities sponses involves a conversion of I˚ CD4 TM previously incapable http://www.jimmunol.org/ (EC50 values), and protein production/secretion kinetics (ET50 of CCL1/3/4 production or the preferential recruitment of CCL1/ values) particular for individual effector functions shapes the co- 3/4 expressors into the recall response (data not shown). + ordinated elaboration of the CD8 TM response. Accordingly, and Chronic viral infections are associated with molecular, pheno- in direct comparison with cytokine (IFN-g, IL-2, and GM-CSF) typic, and functional alterations of the specific T cell compartments and TNFSF ligand (TNF-a and CD40L) synthesis, chemokines that are aptly described by the term T cell exhaustion because play a preeminent role because the production of CCL3/4 and cardinal TE activities are compromised to varying degrees (77). + IFN-g dominates the early stages of the CD8 TM response and, For our interrogation of the TE/M chemokine signatures in chronic together with CCL1 and XCL1, is preferentially induced under LCMV infection, not previously undertaken in a systematic by guest on September 24, 2021 conditions of limited Ag availability. These conclusions are fur- fashion (61, 78), we therefore had to account for the fact that TE/M ther reinforced by in vivo experiments in which just 2 h after dysfunctions also affect inducible IFN-g production, the extent of + peptide injection, .55% of specific CD8 TM produce CCL3/4 and which is furthermore regulated in an epitope-specific fashion for + XCL1, whereas ∼45% express IFN-g and only ∼25% TNF-a. CD8 TM (63, 67). Because concurrent analyses of MHC tetramer Collectively, this work identifies a previously unappreciated and binding and peptide-induced functionalities are technically not striking prominence of chemokine production as part of the feasible for murine T cells, we conducted careful complementary + CD8 TM recall response. Nevertheless, the evidence for a decisive analyses to assess the extent of TE/M abundance and dysfunction. role of TM-produced chemokines in the regulation of II˚ responses Although these experiments expectedly confirmed a hierarchical + remains limited. Our own experiments, in agreement with a pre- numerical deficit of CD8 TE/M as a function of epitope specificity vious report (61), demonstrate that CCL5 is largely dispensable in LCMV cl13– versus high-dose Arm–infected control mice, + + + for the proliferative expansion of II˚ CD8 TE and even combined we also noted a modest increase of both tetramer CD4 T cell CCL3/4/5 or XCL1 blockade does not compromise the LCMV- frequencies/numbers and IFN-g producers in the memory, but not specific recall response. To our knowledge, there are two reports effector, stage. This unexpected expansion is reminiscent of the + documenting a pertinent role for CD8 TM-derived chemokines phenomenon of memory inflation described for some epitope- + under conditions of an II˚ pathogen challenge. In the first, II˚ L. specific CD8 TM populations in persistent infections with CMV, + monocytogenes–specific CD8 TE were shown to provide immune HSV-1, polyomavirus, or parvovirus (79–81) and may be related + protection through CCL3 secretion and induction of TNF-a and to a more restricted exposure of CD4 TM to persisting LCMV reactive oxygen species in phagocytes (27). The evidence in the because of its preferential localization to FRCs (64). Although second report is more indirect. In their elegant study, Alexandre FRCs may express MHC-II under certain inflammatory condi- et al. (28) conditionally depleted DCs bearing XCR1 (the sole tions (82), LCMV-infected FRCs do not (64), thus limiting the + XCL1 receptor) in pathogen-immune mice and found that II˚ productive engagement of specific CD4 TE/M populations to + CD8 TE expansions generated in response to L. monocytogenes, MHC-II–expressing, LCMV-infected APCs (42). Accordingly, + vesicular stomatitis virus, and vaccinia virus, but not CMV, were and despite the fact that IFN-g production by CD4 TEM in both compromised. Complementary experiments revealed that early LCMV cl13 and high-dose Arm infection is partially reduced, we + IFN-g production by II˚ CD8 TE was contingent on an NK cell/ predicted and indeed observed an overall lesser functional im- + + IFN-g–dependent induction of IL-12 and CXCL9 in XCR1 pairment of CD4 TE/M at the chemokine production level. + DCs, but although IL-12, CXCR3, or NK cell neutralization/ In contrast, although specific CD8 TE generated immediately depletion all depressed the emergence of the early IFN-g+ II˚ after a cl13 infection preserve a constitutive CCL5+ phenotype, + CD8 TE phenotype, neither IL-12, CXCL9, nor NK cells were their chemokine production capacity is modestly altered (reduced + required for efficient II˚ CD8 TE expansions (28). Other than NK CCL1/3/4 and XCL1 and increased CCL9/10), and the extent of + + cells, the only major hematopoietic XCL1 source are CD8 TE/M impaired CCL1 induction tracks with CD8 TE epitope specificity 18 CHEMOKINE SIGNATURES OF PATHOGEN-SPECIFIC MEMORY T CELLS

(i.e., NP396 . GP33/4 . GP276). The compromised chemokine 3. Chang, J. T., E. J. Wherry, and A. W. Goldrath. 2014. Molecular regulation of effector and memory T cell differentiation. Nat. Immunol. 15: 1104–1115. profiles, as a property of IFN-g–producing subsets, are main- 4. Laidlaw, B. J., J. E. Craft, and S. M. Kaech. 2016. The multifaceted role of tained, if not somewhat exacerbated, by the time chronic LCMV CD4(+) T cells in CD8(+) T cell memory. Nat. Rev. Immunol. 16: 102–111. infection is fully established but, importantly, are complemented 5. Jameson, S. C., and D. Masopust. 2018. Understanding subset diversity in T cell + + + memory. Immunity 48: 214–226. by the emergence of CCL3 , CCL4 , and XCL1 subsets that do 6. Rot, A., and U. H. von Andrian. 2004. Chemokines in innate and adaptive host not make IFN-g. In fact, inclusion of these subsets into a calcu- defense: basic chemokinese grammar for immune cells. Annu. Rev. Immunol. 22: lation of GP -specific CD8+T frequencies confirms an abun- 891–928. 276 M 7. Zlotnik, A., and O. Yoshie. 2012. The chemokine superfamily revisited. Im- dance that is roughly equivalent to the corresponding functional munity 36: 705–716. + CD8 TM subset in acute Arm infection. At later stages of cl13 8. Bachelerie, F., A. Ben-Baruch, A. M. Burkhardt, C. Combadiere, J. M. Farber, . + G. J. Graham, R. Horuk, A. H. Sparre-Ulrich, M. Locati, A. D. Luster, et al. infection ( 3 mo), a gradual functional resurgence of CD8 TM 2013. International union of basic and clinical pharmacology. [corrected]. occurs (68) and is reflected at the chemokine level by a contraction LXXXIX. Update on the extended family of chemokine receptors and intro- of the CCL3/4/XCL1-only compartment in favor of their coex- ducing a new nomenclature for atypical chemokine receptors. [Published erra- tum appears in 2014 Pharmacol. Rev. 66: 467.] Pharmacol. Rev. 66: 1–79. pression with IFN-g. Of note and in further support of CCL1 as a 9. Griffith, J. W., C. L. Sokol, and A. D. Luster. 2014. Chemokines and chemokine + marker for CD8 TM with residual functionalities, CCL1-production receptors: positioning cells for host defense and immunity. Annu. Rev. Immunol. is not observed in the absence of IFN-g coexpression in either early 32: 659–702. 10. Schulz, O., S. I. Hammerschmidt, G. L. Moschovakis, and R. Fo¨rster. 2016. or late phases of cl13 infection. We therefore suggest that future Chemokines and chemokine receptors in lymphoid tissue dynamics. Annu. Rev. functional profiling of T cells in chronic viral infection and related Immunol. 34: 203–242. 11. Lauvau, G., M. Boutet, T. M. Williams, S. S. Chin, and L. Chorro. 2016. scenarios of prolonged Ag persistence such as cancer and autoim- Memory CD8(+) T cells: innate-like sensors and orchestrators of protection. munity should include, ideally in a highly multiplexed fashion, a Trends Immunol. 37: 375–385. Downloaded from visualization of the T cell chemokine response to account for the 12. Price, D. A., P. Klenerman, B. L. Booth, R. E. Phillips, and A. K. Sewell. 1999. Cytotoxic T lymphocytes, chemokines and antiviral immunity. Immunol. Today full presence and diversity of reactive T cell populations. 20: 212–216. Finally, and in an echo of a seeming disconnect between the 13. Wagner, L., O. O. Yang, E. A. Garcia-Zepeda, Y. Ge, S. A. Kalams, prominence of T and T chemokine responses and their limited B. D. Walker, M. S. Pasternack, and A. D. Luster. 1998. Beta-chemokines are E M released from HIV-1-specific cytolytic T-cell granules complexed to proteogly- relevance for the regulation of I˚ (34) and II˚ TE expansions, cans. Nature 391: 908–911. systemic CCL1 or CCL3 deficiency also has no bearing on virus 14. DeVico, A. L., and R. C. Gallo. 2004. Control of HIV-1 infection by soluble http://www.jimmunol.org/ factors of the immune response. Nat. Rev. Microbiol. 2: 401–413. control or T cell exhaustion in chronic LCMV infection; a modest 15. Demers, K. R., M. A. Reuter, and M. R. Betts. 2013. CD8(+) T-cell effector impact observed under conditions of CCL5 deficiency corrobo- function and transcriptional regulation during HIV pathogenesis. Immunol. Rev. rates an earlier report (61) but leaves unclear the specific contri- 254: 190–206. 16. Appay, V., and S. L. Rowland-Jones. 2001. RANTES: a versatile and contro- bution of T cell–produced CCL5. As argued previously (34), an versial chemokine. Trends Immunol. 22: 83–87. apparent absence of pronounced phenotypes in mice deficient for 17. Wong, M. M., and E. N. Fish. 2003. Chemokines: attractive mediators of the major T cell–produced chemokines may be grounded in redun- immune response. Semin. Immunol. 15: 5–14. 18. Levy, J. A. 2009. The unexpected pleiotropic activities of RANTES. J. Immunol. dancies of the chemokine system and thus emerge only in the 182: 3945–3946. context of compound chemokine mutations or may be observed in 19. Domingo-Gonzalez, R., O. Prince, A. Cooper, and S. A. Khader. 2016. Cyto-

kines and chemokines in Mycobacterium tuberculosis infection. Microbiol. by guest on September 24, 2021 other experimental or natural scenarios characterized by a distinc- Spectr. DOI: 10.1128/microbiolspec.TBTB2-0018-2016. tive dynamic interplay between specific T cells and persisting Ags. 20. Roychoudhury, K., and S. Roy. 2004. Role of chemokines in Leishmania in- In summary, the present work defines six chemokines (CCL1/3/ fection. Curr. Mol. Med. 4: 691–696. 21. Nguyen, L. T., and H. J. Vogel. 2012. Structural perspectives on antimicrobial 4/5/9/10 and XCL1) as a major component integral to the func- chemokines. Front. Immunol. 3: 384. tional repertoire of pathogen-specific T cells in long-term memory 22. Eberlein, J., T. T. Nguyen, F. Victorino, L. Golden-Mason, H. R. Rosen, and and chronic viral infection. Unique expression patterns, such as D. Homann. 2010. Comprehensive assessment of chemokine expression profiles by flow cytometry. J. Clin. Invest. 120: 907–923. constitutive CCL5 and inducible CCL1 expression, may serve 23. Crozat, K., R. Guiton, V. Contreras, V. Feuillet, C. A. Dutertre, E. Ventre, targeted diagnostic purposes, and the relative dominance exerted by T. P. Vu Manh, T. Baranek, A. K. Storset, J. Marvel, et al. 2010. The XC che- mokine receptor 1 is a conserved selective marker of mammalian cells homol- T cell–derived chemokines in the early recall response establishes ogous to mouse CD8alpha+ dendritic cells. J. Exp. Med. 207: 1283–1292. TM populations as a principal source for CCL1/3/4/5 and XCL1 in 24. Bolinger, B., S. Sims, L. Swadling, G. O’Hara, C. de Lara, D. Baban, N. Saghal, particular. How such prodigious chemokine production, which L. N. Lee, E. Marchi, M. Davis, et al. 2015. Adenoviral vector vaccination in- duces a conserved program of CD8(+) T cell memory differentiation in mouse necessitates the mobilization of not inconsiderable bioenergetic and man. Cell Rep. 13: 1578–1588. resources, exactly contributes to enhanced immune protection, 25. Eberlein, J., B. Davenport, T. Nguyen, F. Victorino, K. Haist, K. Jhun, however, remains to be investigated in greater detail. A. Karimpour-Fard, L. Hunter, R. Kedl, E. T. Clambey, and D. Homann. 2016. Aging promotes acquisition of naive-like CD8+ memory T cell traits and en- hanced functionalities. J. Clin. Invest. 126: 3942–3960. Acknowledgments 26. Brinza, L., S. Djebali, M. Tomkowiak, J. Mafille, C. Loiseau, P. E. Jouve, S. de 2/2 Bernard, L. Buffat, B. Lina, M. Ottmann, et al. 2016. Immune signatures of We thank Dr. S. Manes for the gift of B6.CCL1 mice, Dr. L. Lenz for protective spleen memory CD8 T cells. Sci. Rep. 6: 37651. rLM-OVA, Dr. T. Lane for the CCL5 Ab clone R6G9, Dr. L. Gapin for 27. Narni-Mancinelli, E., L. Campisi, D. Bassand, J. Cazareth, P. Gounon, CD1/aGalCer tetramers, Drs. P. Marrack and J. Kappler and the entire N. Glaichenhaus, and G. Lauvau. 2007. Memory CD8+ T cells mediate anti- Kappler/Marrack laboratory for critical discussion, and Dr. F. Mortari bacterial immunity via CCL3 activation of TNF/ROI+ phagocytes. J. Exp. Med. 204: 2075–2087. (R&D Systems) for the generous gift of most of the chemokine Abs used 28. Alexandre, Y. O., S. Ghilas, C. Sanchez, A. Le Bon, K. Crozat, and M. Dalod. in this study. 2016. XCR1+ dendritic cells promote memory CD8+ T cell recall upon sec- ondary infections with Listeria monocytogenes or certain viruses. J. Exp. Med. 213: 75–92. Disclosures 29. Slifka, M. K., and J. L. Whitton. 2000. Antigen-specific regulation of T cell- The authors have no financial conflicts of interest. mediated cytokine production. Immunity 12: 451–457. 30. Catalfamo, M., T. Karpova, J. McNally, S. V. Costes, S. J. Lockett, E. Bos, P. J. Peters, and P. A. Henkart. 2004. Human CD8+ T cells store RANTES in a unique secretory compartment and release it rapidly after TcR stimulation. Im- References munity 20: 219–230. 1. Whitton, J. L., and M. B. Oldstone. 2001. The immune response to viruses. In 31. Stegelmann, F., M. Bastian, K. Swoboda, R. Bhat, V. Kiessler, A. M. Krensky, Field’s Virology, 4th Ed. D. Knipe, P. Howley, D. Griffin, R. Lamb, M. Martin, M. Roellinghoff, R. L. Modlin, and S. Stenger. 2005. Coordinate expression of and S. Strauss, eds. Lippincott Williams & Wilkins, Philadelphia, p. 285–320. CC chemokine ligand 5, granulysin, and perforin in CD8+ T cells provides a host 2. Sallusto, F., A. Lanzavecchia, K. Araki, and R. Ahmed. 2010. From vaccines to defense mechanism against Mycobacterium tuberculosis. J. Immunol. 175: memory and back. Immunity 33: 451–463. 7474–7483. The Journal of Immunology 19

32. Swanson, B. J., M. Murakami, T. C. Mitchell, J. Kappler, and P. Marrack. 2002. 56. Hao, S., and D. Baltimore. 2009. The stability of mRNA influences the temporal RANTES production by memory phenotype T cells is controlled by a post- order of the induction of genes encoding inflammatory molecules. Nat. Immunol. transcriptional, TCR-dependent process. Immunity 17: 605–615. 10: 281–288. 33. Walzer, T., A. Marc¸ais, F. Saltel, C. Bella, P. Jurdic, and J. Marvel. 2003. Cutting 57. Castellino, F., A. Y. Huang, G. Altan-Bonnet, S. Stoll, C. Scheinecker, and edge: immediate RANTES secretion by resting memory CD8 T cells following R. N. Germain. 2006. Chemokines enhance immunity by guiding naive CD8+ antigenic stimulation. J. Immunol. 170: 1615–1619. T cells to sites of CD4+ T cell- interaction. Nature 440: 890– 34. Eberlein, J., B. Davenport, T. T. Nguyen, F. Victorino, K. Jhun, V. van der Heide, 895. M. Kuleshov, A. Ma’ayan, R. Kedl, and D. Homann. 2020. Chemokine signa- 58. Slifka, M. K., and J. L. Whitton. 2001. Functional avidity maturation of CD8(+) tures of pathogen-specific T cells I: effector T cells. J. Immunol. 205: 2169– T cells without selection of higher affinity TCR. Nat. Immunol. 2: 711–717. 2187. 59. Henry, C. J., D. A. Ornelles, L. M. Mitchell, K. L. Brzoza-Lewis, and 35. Kahan, S. M., E. J. Wherry, and A. J. Zajac. 2015. T cell exhaustion during E. M. Hiltbold. 2008. IL-12 produced by dendritic cells augments CD8+ T cell persistent viral infections. Virology 479–480: 180–193. activation through the production of the chemokines CCL1 and CCL17. 36. Makino, Y., D. N. Cook, O. Smithies, O. Y. Hwang, E. G. Neilson, L. A. Turka, J. Immunol. 181: 8576–8584. H. Sato, A. D. Wells, and T. M. Danoff. 2002. Impaired T cell function in 60. Molon, B., G. Gri, M. Bettella, C. Go´mez-Mouto´n, A. Lanzavecchia, RANTES-deficient mice. Clin. Immunol. 102: 302–309. C. Martı´nez-A, S. Man˜es, and A. Viola. 2005. T cell costimulation by chemokine 37. Mira, E., B. Leo´n, D. F. Barber, S. Jime´nez-Baranda, I. Goya, L. Almonacid, receptors. Nat. Immunol. 6: 465–471. G. Ma´rquez, A. Zaballos, C. Martı´nez-A, J. V. Stein, et al. 2008. Statins induce 61. Crawford, A., J. M. Angelosanto, K. L. Nadwodny, S. D. Blackburn, and recruitment via a CCL1 dependent pathway. J. Immunol. 181: E. J. Wherry. 2011. A role for the chemokine RANTES in regulating CD8 T cell 3524–3534. responses during chronic viral infection. PLoS Pathog. 7: e1002098. 38. Homann, D., A. Tishon, D. P. Berger, W. O. Weigle, M. G. von Herrath, and 62. Wherry, E. J. 2011. T cell exhaustion. Nat. Immunol. 12: 492–499. M. B. Oldstone. 1998. Evidence for an underlying CD4 helper and CD8 T-cell 63. Wherry, E. J., J. N. Blattman, K. Murali-Krishna, R. van der Most, and defect in B-cell-deficient mice: failure to clear persistent virus infection after R. Ahmed. 2003. Viral persistence alters CD8 T-cell immunodominance and adoptive immunotherapy with virus-specific memory cells from muMT/muMT tissue distribution and results in distinct stages of functional impairment. J. Virol. mice. J. Virol. 72: 9208–9216. 77: 4911–4927. 39. Pope, C., S. K. Kim, A. Marzo, D. Masopust, K. Williams, J. Jiang, H. Shen, and 64. Mueller, S. N., M. Matloubian, D. M. Clemens, A. H. Sharpe, G. J. Freeman, L. Lefranc¸ois. 2001. Organ-specific regulation of the CD8 T cell response to S. Gangappa, C. P. Larsen, and R. Ahmed. 2007. Viral targeting of fibroblastic Downloaded from Listeria monocytogenes infection. J. Immunol. 166: 3402–3409. reticular cells contributes to immunosuppression and persistence during chronic 40. Humann, J., R. Bjordahl, K. Andreasen, and L. L. Lenz. 2007. Expression of the infection. Proc. Natl. Acad. Sci. USA 104: 15430–15435. p60 autolysin enhances NK cell activation and is required for listeria mono- 65. Sevilla, N., S. Kunz, A. Holz, H. Lewicki, D. Homann, H. Yamada, K. P. Campbell, cytogenes expansion in IFN-gamma-responsive mice. J. Immunol. 178: 2407– J. C. de La Torre, and M. B. Oldstone. 2000. Immunosuppression and resultant 2414. viral persistence by specific viral targeting of dendritic cells. J. Exp. Med. 192: 41. Kurche, J. S., M. A. Burchill, P. J. Sanchez, C. Haluszczak, and R. M. Kedl. 1249–1260. 2010. Comparison of OX40 ligand and CD70 in the promotion of CD4+ T cell 66. Farr, A. G., M. L. Berry, A. Kim, A. J. Nelson, M. P. Welch, and A. Aruffo. 1992.

responses. J. Immunol. 185: 2106–2115. Characterization and cloning of a novel expressed by stromal cells http://www.jimmunol.org/ 42. Homann, D., D. B. McGavern, and M. B. Oldstone. 2004. Visualizing the viral in T-dependent areas of peripheral lymphoid tissues. J. Exp. Med. 176: 1477– burden: phenotypic and functional alterations of T cells and APCs during per- 1482. sistent infection. J. Immunol. 172: 6239–6250. 67. Zajac, A. J., J. N. Blattman, K. Murali-Krishna, D. J. Sourdive, M. Suresh, 43. Lenz, D. C., S. K. Kurz, E. Lemmens, S. P. Schoenberger, J. Sprent, J. D. Altman, and R. Ahmed. 1998. Viral immune evasion due to persistence of M. B. Oldstone, and D. Homann. 2004. IL-7 regulates basal homeostatic pro- activated T cells without effector function. J. Exp. Med. 188: 2205–2213. liferation of antiviral CD4+T cell memory. Proc. Natl. Acad. Sci. USA 101: 68. Fuller, M. J., A. Khanolkar, A. E. Tebo, and A. J. Zajac. 2004. Maintenance, loss, 9357–9362. and resurgence of T cell responses during acute, protracted, and chronic viral 44. Haluszczak, C., A. D. Akue, S. E. Hamilton, L. D. S. Johnson, L. Pujanauski, infections. J. Immunol. 172: 4204–4214. L. Teodorovic, S. C. Jameson, and R. M. Kedl. 2009. The antigen-specific CD8+ 69. Darrah, P. A., D. T. Patel, P. M. De Luca, R. W. B. Lindsay, D. F. Davey, T cell repertoire in unimmunized mice includes memory phenotype cells bearing B. J. Flynn, S. T. Hoff, P. Andersen, S. G. Reed, S. L. Morris, et al. 2007. markers of homeostatic expansion. J. Exp. Med. 206: 435–448. Multifunctional TH1 cells define a correlate of vaccine-mediated protection

45. Glass, W. G., M. J. Hickey, J. L. Hardison, M. T. Liu, J. E. Manning, and against Leishmania major. Nat. Med. 13: 843–850. by guest on September 24, 2021 T. E. Lane. 2004. targeting of the CC chemokine ligand 5 results in 70. Panagioti, E., P. Klenerman, L. N. Lee, S. H. van der Burg, and R. Arens. 2018. diminished leukocyte infiltration into the central nervous system and reduced Features of effective T cell-inducing vaccines against chronic viral infections. neurologic disease in a viral model of multiple sclerosis. J. Immunol. 172: 4018– Front. Immunol. 9: 276. 4025. 71. Wolint, P., M. R. Betts, R. A. Koup, and A. Oxenius. 2004. Immediate cyto- 46. Homann, D., H. Lewicki, D. Brooks, J. Eberlein, V. Mallet-Designe´, L. Teyton, toxicity but not degranulation distinguishes effector and memory subsets of and M. B. Oldstone. 2007. Mapping and restriction of a dominant viral CD4+ CD8+ T cells. J. Exp. Med. 199: 925–936. T cell core epitope by both MHC class I and MHC class II. Virology 363: 113– 72. Jabbari, A., and J. T. Harty. 2006. Secondary memory CD8+ T cells are more 123. protective but slower to acquire a central-memory phenotype. J. Exp. Med. 203: 47. Homann, D., L. Teyton, and M. B. Oldstone. 2001. Differential regulation of 919–932. antiviral T-cell immunity results in stable CD8+ but declining CD4+ T-cell 73. Masopust, D., S. J. Ha, V. Vezys, and R. Ahmed. 2006. Stimulation history memory. Nat. Med. 7: 913–919. dictates memory CD8 T cell phenotype: implications for prime-boost vaccina- 48. Korns Johnson, D., and D. Homann. 2012. Accelerated and improved quantifi- tion. J. Immunol. 177: 831–839. cation of lymphocytic choriomeningitis virus (LCMV) titers by flow cytometry. 74. Jenkins, M. R., K. Kedzierska, P. C. Doherty, and S. J. Turner. 2007. Hetero- PLoS One 7: e37337. geneity of effector phenotype for acute phase and memory influenza A virus- 49. Pircher, H., K. Bu¨rki, R. Lang, H. Hengartner, and R. M. Zinkernagel. 1989. specific CTL. J. Immunol. 179: 64–70. Tolerance induction in double specific T-cell receptor transgenic mice varies 75. Moffat, J. M., T. Gebhardt, P. C. Doherty, S. J. Turner, and J. D. Mintern. 2009. with antigen. Nature 342: 559–561. Granzyme A expression reveals distinct cytolytic CTL subsets following influ- 50. Eberlein, J., B. Davenport, T. T. Nguyen, F. Victorino, T. Sparwasser, and enza A virus infection. Eur. J. Immunol. 39: 1203–1210. D. Homann. 2012. Multiple layers of CD80/86-dependent costimulatory activity 76. Metkar, S. S., C. Menaa, J. Pardo, B. Wang, R. Wallich, M. Freudenberg, S. Kim, regulate primary, memory, and secondary lymphocytic choriomeningitis virus- S. M. Raja, L. Shi, M. M. Simon, and C. J. Froelich. 2008. Human and mouse specific T cell immunity. J. Virol. 86: 1955–1970. granzyme A induce a proinflammatory cytokine response. Immunity 29: 720– 51. Lu, J., Q. Q. Wu, Y. B. Zhou, K. H. Zhang, B. X. Pang, L. Li, N. Sun, 733. H. S. Wang, S. Zhang, W. J. Li, et al. 2016. Cancer research advance in CKLF- 77. Wherry, E. J., and M. Kurachi. 2015. Molecular and cellular insights into T cell like MARVEL transmembrane domain containing member family (review). exhaustion. Nat. Rev. Immunol. 15: 486–499. Asian Pac. J. Cancer Prev. 17: 2741–2744. 78. Kao, C., K. J. Oestreich, M. A. Paley, A. Crawford, J. M. Angelosanto, M.- 52. Wherry, E. J., V. Teichgra¨ber, T. C. Becker, D. Masopust, S. M. Kaech, R. Antia, A. A. Ali, A. M. Intlekofer, J. M. Boss, S. L. Reiner, A. S. Weinmann, and U. H. von Andrian, and R. Ahmed. 2003. Lineage relationship and protective E. J. Wherry. 2011. Transcription factor T-bet represses expression of the in- immunity of memory CD8 T cell subsets. Nat. Immunol. 4: 225–234. hibitory receptor PD-1 and sustains virus-specific CD8+ T cell responses during 53. Seder, R. A., P. A. Darrah, and M. Roederer. 2008. T-cell quality in memory and chronic infection. Nat. Immunol. 12: 663–671. protection: implications for vaccine design. [Published erratum appears in 2008 79. O’Hara, G. A., S. P. Welten, P. Klenerman, and R. Arens. 2012. Memory T cell Nat. Rev. Immunol. 8: 486.] Nat. Rev. Immunol. 8: 247–258. inflation: understanding cause and effect. Trends Immunol. 33: 84–90. 54. Boyd, A., J. R. Almeida, P. A. Darrah, D. Sauce, R. A. Seder, V. Appay, 80. Welten, S. P. M., N. S. Baumann, and A. Oxenius. 2019. Fuel and brake of G. Gorochov, and M. Larsen. 2015. Pathogen-specific T cell polyfunctionality is memory T cell inflation. Med. Microbiol. Immunol. (Berl.) 208: 329–338. a correlate of T cell efficacy and immune protection. [Published erratum appears 81. Klenerman, P. 2018. The (gradual) rise of memory inflation. Immunol. Rev. 283: in 2015 PLoS One 10: e0138395.] PLoS One 10: e0128714. 99–112. 55. Comerford, I., and R. J. Nibbs. 2005. Post-translational control of chemokines: a 82. Brown, F. D., and S. J. Turley. 2015. Fibroblastic reticular cells: organization and role for decoy receptors? Immunol. Lett. 96: 163–174. regulation of the T lymphocyte life cycle. J. Immunol. 194: 1389–1394. Supplemental Figure 1

Supplemental Figure 1. CKLFSF mRNA expression by p14 T (d46): CKLFSF family A M p14 TM & hematopoietic sources of constitutively Cklf Cklfsf1 & 2a expressed CCL5. A., gene expression analyses of p14 TM Cklfsf2b (d46) were performed as detailed in the legend to Fig.1A Cklfsf3 Cklfsf4 not on chip and Methods. The bar diagrams display MAS5-normalized Cklfsf5 Cklfsf6 values of chemokine-like factor superfamily (CKLFSF) Cklfsf7 * ex vivo mRNA expression by p14 TM analyzed ex vivo (gray bars) or Cklfsf8 αCD3/αCD28 after a 3h TCR stimulation with aCD3/aC28 (black bars). 100 101 102 103 104 105 mRNA (MAS5 normalized values) Statistically significant ex vivo or induced Cklfsf gene expression above the threshold of the MAS5 value of 40 CD11b+ myeloid (broken line) was demonstrated for Cklf, Cklfsf3, Cklfsf6 and B CD3ε+T cells NK cells cell subsets CD19+ B cells Cklfsf7, and the asterisks indicates significant 3 1 3

1 2

ε downregulation of Cklfsf7 after p14 TM activation (n=4 32.6 0.46 CD3 CD11b mice/group). B.-I., with the exception of panels H & I, all 2 3.16 data were generated with naïve B6 mice (n≥3 mice/group NK1.1 CCL5 CCL5 NK1.1 CCL5 CCL5 and experiment) analyzed directly ex vivo for expression of

CD3ε+CD8+T cells CD3ε+CD4+T cells various surface markers and intracellular CCL5. B., dot plots C CD8+T + demonstrating gating strategies for identification of blood-

CD4 T + CD8 TMP borne T cells, NK cells and CD11b-expressing myeloid cells CD44 + CD4 TMP (numbers in plots refer to gates employed in adjacent control CCL5 control CCL5 0 5 10 15 % CCL5+ histograms; black tracing: CCL5 stains, gray-shaded histograms: control stains). CD11b-expressing cells, + + + + excluding mature CD11b NK cells, comprise monocytes, D CD3ε CD8 T cells E CD8 T cells subsets 2 3 1 macrophages, and DC subsets. Lack of CCL5 Spleen 2 PC + 3 expression by B cells is demonstrated after gating on CD19 CD44 CD44 1 cells. C., dot plots are gated on the indicated blood-borne T CD122 CCL5 cell subsets; the adjacent bar diagram displays the Liver MLN + CD3ε+ cells frequencies of CCL5 T cells in indicated T cell subsets (TMP: CD44 CD44 F 31.7 CD44hi memory-phenotype subsets). D., mice were perfused ±3.1%

TCR TCR with PBS and spleen, mesenteric lymph nodes (MLN), bone γδ Lung BM marrow (BM), peritoneal cavity (PC), liver and lung were CCL5 processed as detailed in Methods to yield single cell suspension. Dot plots display constitutive CCL5 expression

CCL5 CCL5 G Spleen 0.6% + restricted to the CD8 TMP compartment regardless of tissue spleen origin; the corresponding bar diagram summarizes these tetramer tetramer MLN * BM ** data, and statistical differences were calculated in Liver PC GalCer 21% + α comparison to spleen. E., dot plot gated on splenic CD8 T liver +

lung CD1/ cells; histograms show CCL5 expression of individual CD8 T 0 20 40 60 80 100 lo hi lo + + cell subsets: CD44 (gray filled), CD44 CD122 (broken % CCL5 CD8 TMP CD3ε CCL5 tracing) and CD44hiCD122hi (black tracing); the difference between the CCL5 GMFI of CD44hiCD122lo vs. CCL5+ cells 100 *** naive H LCMV d45 hi hi 80 CD44 CD122 cells is small but significant (*p=0.0132, naive LCMV LCMV LCMV d206

+ ε d45 206 60 *** paired t-test). F., gated on splenic CD3e T cells, the dot plot cells (%) cells(%)

+ 40 CD3 identifies CCL5+ gdTCR+ T cells; values indicates the 20 ns distribution of of distribution CCL5 NK1.1 NK1.1 NK1.1 0 frequencies of gdT cells expressing ex vivo detectable CCL5 T cells NK cells other

(SEM). G., splenic and hepatic immune cells were prepared + 100 CCL5+ CD3ε+T cells *** from PBS-perfused mice, and NKT cells were identified by I 80 60 combined CD3e and CD1/aGalCer tetramer stains; + + T cells (%) (%) cells T 40 + histograms are gated on CD3e CD1/aGalCer tetramer CD44 CD44 LCMV LCMV ε *** naive d45 206 20 NKT cells (black tracing: CCL5 stains, gray-shaded CD3 0 hi + hi CD8α CD8α CD8α CCL5 of distribution CD44 CD8 other CD44 histograms: control stains). H., PBMCs obtained from naïve as well as young and aged LCMV-immune B6 mice were analyzed for their relative contribution to the CCL5+ compartment. Note that a sizable fraction of CCL5+T cells in the LCMV d45 group expresses low levels of NK1.1; therefore, no clear distinction between conventional T cells expressing NK1.1 and NKT cells can be made in the LCMV d45 group, and the corresponding bar diagram distinguishes T cells (CD3e+), NK cells (NK1.1+CD3e-) and other cells (NK1.1-CD3e-); asterisks indicate significant differences among indicated populations in naïve vs. LCMV d45 or d206 groups (ns: not significant; no significant differences between LCMV d45 and d206 groups). Note that a history of LCMV infection permanently alters the distribution of CCL5+ cells by making T cells its primary source. I., representative data and + hi + summary of CCL5 distribution among T cell subsets; CCL5 CD44 T cells not expressing CD8a belong mostly to the CD4 TMP compartment (not shown and panel C); asterisks indicate significant differences among indicated populations in naïve vs. LCMV d45 or d206 groups (no significant differences between LCMV d45 and d206 groups). Supplemental Figure 2

A p14 TE → TM mRNA (ex vivo) C p14 TM (d46)

4 Ifng 10 104 104 *** Tnf *** 103 mRNA mRNA 103 103 Il2 *** 2 10 ** ** Csf2/Gmcsf *** ** *** 102 ** 102 Cd40lg 101 *** ***

chemokine 0 Ccl1 1 Ccl3 1 Ccl4 10 10 10 Gzma 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400 * Gzmb ex vivo days after LCMV days after LCMV days after LCMV αCD3/αCD28 100 101 102 103 104 105 * * * 104 104 Ccl9 104 mRNA (MAS5 normalized values) *** *** mRNA mRNA 103 103 103 Supplemental Figure 2. Temporal regulation of chemokine + 102 102 *** *** 102 mRNA and protein expression by LCMV-specific CD8 TE/M *** & expression of selected effector function-related genes Ccl5 Xcl1 chemokine 101 101 101 A., temporal regulation of chemokine gene expression by p14

0 100 200 300 400 0 100 200 300 400 0 100 200 300 400 TE/M as a function of time after LCMV challenge. Data were days after LCMV days after LCMV days after LCMV generated with highly purified p14 TE/M as detailed in ref.(25), the gray shaded area demarcates the transitional period from effector (d8) to early memory (d42) stage, the broken line GP -specific CD8+T T B 33 E → M designates the detection threshold set at a MAS5 value of 40, chemokine protein (induced) black asterisks indicate significant expression differences 100 100 100 comparing young (d46) and older p14 TM, and the red 80 80 80

asterisks refer to significantly elevated Ccl4 levels as + + + 60 60 60 compared to the d156 time point. Note that the Ccl5 panel is

40 40 40 identical to one of the panels shown in Fig.2C, and the Xcl1 % CCL3 % CCL1 % CCL4 20 20 20 panel displays MAS5-normalized data previously featured in RMA-normalized fashion in ref.(25). B., induced chemokine 0 0 0 expression by endogenously generated GP33/4-specific 0 50 100 150 0 50 100 150 0 50 100 150 + days after LCMV days after LCMV days after LCMV CD8 TE/M as a function of time after LCMV challenge. Asterisks indicate a significant decrease (CCL9/10) or increase (XCL1) 100 100 100 + ** of chemokine protein expression by d164 CD8 TM in

80 + 80 80 +

+ + comparison to younger CD8 TM (CCL9/10: d55; XCL1: d42). 60 60 60 All values in panels A and B are the mean ± 1 SEM with n ≥ 3 40 40 40 individual mice per time point, and statistical differences in %XCL1 % CCL5 ** 20 % CCL9/10 20 20 panels A and B were calculated by one-way ANOVA. C.,

0 0 0 selected cytokine, TNFS ligand and granzyme gene 0 50 100 150 0 50 100 150 0 50 100 150 expression by p14 TM (d46) analyzed directly ex vivo (gray) or days after LCMV days after LCMV days after LCMV after 3h aCD3/aCD28 in vitro stimulation (n=4 mice/group). Supplemental Figure 3

A C E mIgG1 αCCL5 host donor host donor } } } }

396 NP b D

CD90.1 CD90.1 o CD27 I memory * o II memory o o + CD62L I & II CD8 TE expansions (d7)

CD122 *** PBMC * 2.1x CD127 Spleen 1.1x KLRG1 ** MLN 1.8x 0 10 20 30 40 50 60 70 80 90 100 Liver 1.9x % surface markerhi * Lung 2.0x B Io & IIo NP -specific CD8+T (d6) D Io & IIo NP -specific CD8+T (d70) 396 E 396 M PC 3.1x IL-2 *** IL-2 BM mIgG1 * 2.2x TNFα * TNFα ** αCCL5 CD40L *** CD40L 101 102 103 104 105 106 b + o + 6 D NP396 I CD8 TE per 10 cells 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 % cytokine or CD40L+ % cytokine or CD40L+ PBMC ** 1.4x CCL1 * CCL1 Spleen 0.8x CCL3 CCL3 * MLN ** 1.9x CCL4 CCL4 ** Liver 1.1x CCL5 CCL5 Lung 1.1x CCL9/10 CCL9/10 * o PC I response Io memory 1.1x XCL1 ** o XCL1 * II response IIo memory BM 1.3x 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 + + % chemokine % chemokine 101 102 103 104 105 106 b + o + 6 D NP396 II CD8 TE per 10 cells o o + o o + I & II GP64-specific CD4 TE (d6) I & II GP64-specific CD4 TM (d70)

IL-2 IL-2 ** o + I CD8 TE 1.5x TNFα *** TNFα o + 1.2x II CD8 TE CD40L CD40L 102 103 104 105 106 107 108 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 b + o o + D NP396 I & II CD8 TE per spleen % cytokine or CD40L+ % cytokine or CD40L+

CCL1 *** CCL1 *** IIo DbNP + CD8+T expansions (d8) CCL3 CCL3 *** F 396 E

CCL4 *** CCL4 ** rat IgG αCCL3/4/5 CCL5 *** CCL5 **

CCL9/10 CCL9/10 goat IgG *** o o XCL1 ** I response XCL1 I memory αXCL1 *** IIo response IIo memory 2 3 4 5 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 10 10 10 10 b + o + 6 % chemokine+ % chemokine+ D NP396 II CD8 TE per 10 PBMC

o + + Supplemental Figure 3. II CD8 and CD4 TE/M immunity: expression profiles and functional roles of T cell chemokines. A. & B., AT/RC experiments were conducted with 5x106 spleen cells prepared from CD45.1-congenic LCMV-immune donors (d201), 6 o o infection of B6 recipients with 2x10 pfu LCMV cl13 i.v., and specific I and II TE populations were analyzed six days later (dot plots + o o + are gated on all CD8 T cells or indicated host/donor subsets). Functional analyses of I and II NP396-specific CD8 and GP64- + + specific CD4 TE were conducted after 5h peptide stimulation and the bar diagrams display the fraction of IFNg subsets expressing additional cytokines, TNFSF members or chemokines (n=4 mice/group; asterisks indicate significant differences between Io and IIo o o TE populations). C. & D., similar AT/RC experiments as in panels A & B followed by I and II TM analyses performed on d70 (n=4 o o + mice/group; asterisks indicate significant differences between I and II TM populations). E., CD8 T cell-enriched spleen cells from 4 b + + LCMV-immune B6.CD90.1 donors containing ~2.5x10 D NP396 CD8 TM were transferred i.v. into B6 recipients treated with 250µg 5 o o b + mIgG1 isotype or aCCL5 i.p. on days -1, 1, 3 and 5 in relation to LCMV infection with 2x10 pfu LCMV Arm i.p. I and II D NP396 + + CD8 TE expansions were quantified 7 days later in lymphatic and non-lymphoid tissues; the extent of differential CD8 TE expansions in various tissues of control and aCCL5-treated mice is indicated (n=3 mice/group). F., similar AT/RC experiments were 3 b + + conducted with 5.0x10 D NP396 donor CD8 TM (d286), combined CCL3/4/5 (3x80µg aCCL3/4/5 each or 3x240µg rIgG i.p.) or XCL1 blockade (3x70µg polyclonal aXCL1 or 3x70µg goat IgG i.p.) administered on days 0, 2 and 4 in relation to LCMV infection, o + and quantification of II CD8 TE expansions on d8 (n=3 mice/group). Supplemental Figure 4

+ A specific CD8 TM functionalities (d107) B6 IL-2 IL-2 IL-2 γ IFN GM-CSF GM-CSF GM-CSF B6.CCL1

TNFα TNFα TNFα γ IFN CD40L CD40L CD40L -/- CCL1 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 % cytokine+ or TNFSF+ of % cytokine+ or TNFSF+ of % cytokine+ or TNFSF+ of + + + + + + IFNγ NP396-specific CD8 TM IFNγ GP33/4-specific CD8 TM IFNγ GP276specific CD8 TM

CCL1 ND CCL1 ND CCL1 ND CCL3 CCL3 CCL3

CCL4 CCL4 CCL4

CCL5 CCL5 CCL5

CCL9/10 CCL9/10 CCL9/10 - XCL1 B6 XCL1 XCL1 B6.CCL1-/- 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 % chemokine+ of IFNγ+ % chemokine+ of IFNγ+ % chemokine+ of IFNγ+ + + + NP396-specific CD8 TM GP33/4-specific CD8 TM GP276-specific CD8 TM

B virus titers (d78) 107

106

105

104

LCMV/g tissue LCMV/g 103

pfu B6.WT 2 10 B6.CCL3-/- spleen liver lung kidney

+ + C specific CD8 & CD4 TM frequencies, numbers & PD-1 phenotype (d78 blood) 15 105 100 B6.WT B6.WT B6.CCL3-/- B6.CCL3-/- 80 104

10 + T cells

+ 60 T cells/ 3 + 10

PBMC 40 5 6 2 %PD-1 10 10 20 B6.WT

tetramer -/- %tetramer 0 101 0 B6.CCL3 b + b + b + b + b + b + b + b + b + b + b + b + D NP396 D GP33 D GP276 I-A GP66 D NP396 D GP33 D GP276 I-A GP66 D NP396 D GP33 D GP276 I-A GP66

+ + + + + + specific CD8 or CD4 TM specific CD8 or CD4 TM specific CD8 or CD4 TM

+ + D specific CD8 & CD4 TM frequencies, numbers & PD-1 phenotype (d78 spleen) 107 100 15 B6.WT B6.WT -/- B6.CCL3-/- B6.CCL3 80 6 10 + T cells 10

+ 60 T cells/ + 40 5 spleen 105 %PD-1 20 B6.WT tetramer -/- %tetramer 0 104 0 B6.CCL3 b + b + b + b + b + b + b + b + b + b + b + b + D NP396 D GP33 D GP276 I-A GP66 D NP396 D GP33 D GP276 I-A GP66 D NP396 D GP33 D GP276 I-A GP66

+ + + + + + specific CD8 or CD4 TM specific CD8 or CD4 TM specific CD8 or CD4 TM

+ + E specific CD8 & CD4 TM functionalities (d78 spleen) 10 107 100 B6.WT B6.WT B6.WT -/- B6.CCL3-/- B6.CCL3 B6.CCL3-/- 80 T cells

6 + 10 γ 60 T cells + 5 IFN T cells/ γ

+ 40 γ 5 of spleen 10 + IFN

α 20 IFN %

4 0 10 TNF 0

NP GP GP NP GP GP % NP GP GP 396 33/4 276 GP64 396 33/4 276 GP64 396 33/4 276 GP64

+ + + + + + specific CD8 or CD4 TM specific CD8 or CD4 TM specific CD8 or CD4 TM

Supplemental Figure 4. Chronic LCMV infection under conditions of CCL1- or CCL3-deficiency. A., B6 and B6.CCL1-/- mice 6 + were infected with 2x10 pfu LCMV cl13 i.v. and functional profiles of splenic NP396-, GP33/4- and GP276-specific CD8 TM were + -/- determined on d107. No significant differences between any CD8 TM functionalities in B6 vs. B6.CCL1 mice other than the lack of + CCL1 production by the mutant CD8 TM (see insert in upper right corner) were noted (ND: not detected; n=3 mice/group). B.-E., B6 and B6.CCL3-/- mice were infected with 2x106 pfu LCMV cl13 i.v. and analyzed 78 days later. B., virus titers in tissues. C. & D., + + specific CD8 and CD4 TM frequencies, numbers and PD-1 expression status in peripheral blood and spleen. E., functional profiles + + of CD8 and CD4 TM recovered from the spleen. No significant differences were noted for any of the above readouts (n=5 mice/group).