Defective CD8 T Cell Responses in Aged Mice Are Due to Quantitative and Qualitative Changes in Virus-Specific Precursors

This information is current as Vilma Decman, Brian J. Laidlaw, Travis A. Doering, Jin of October 1, 2021. Leng, Hildegund C. J. Ertl, Daniel R. Goldstein and E. John Wherry J Immunol 2012; 188:1933-1941; Prepublished online 13 January 2012; doi: 10.4049/jimmunol.1101098 http://www.jimmunol.org/content/188/4/1933 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2012/01/13/jimmunol.110109

Material 8.DC1 http://www.jimmunol.org/ References This article cites 55 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/188/4/1933.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on October 1, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*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 © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Defective CD8 T Cell Responses in Aged Mice Are Due to Quantitative and Qualitative Changes in Virus-Specific Precursors

Vilma Decman,*,† Brian J. Laidlaw,*,† Travis A. Doering,*,† Jin Leng,‡,x Hildegund C. J. Ertl,{ Daniel R. Goldstein,‡,x and E. John Wherry*,†

Aging is associated with suboptimal CD8 T cell responses to viral infections. It is not clear whether these poor responses are due to environmental influences or quantitative and qualitative changes in the pool of responding CD8 T cells. Our studies demonstrated several deleterious age-related changes in the pool of Ag-specific CD8 T cells that respond to infection. The majority of CD8 T cells from uninfected aged mice was CD44Hi and had increased expression of inhibitory receptors including PD1, LAG3, 2B4, and CD160. These aged CD44Hi CD8 T cells were transcriptionally similar to exhausted CD8 T cells found during chronic infections. In addition, the number of virus-specific precursors in aged mice prior to infection was decreased up to 10-fold, and many of these Downloaded from Ag-specific precursors had high expression of CD44 and PD1. Finally, TCR transgenic studies demonstrated that the CD44Hi Ag- specific CD8 T cells from unimmunized aged and young mice were qualitatively inferior compared with CD44Lo CD8 T cells from aged or young donors. Thus, a decrease in precursor frequency as well as qualitative changes of CD8 T cells during aging are directly related to impaired immunity. The Journal of Immunology, 2012, 188: 1933–1941.

ging is associated with a decline in T cell function and crease in homeostatic proliferation of the existing T cell pool. http://www.jimmunol.org/ an increase in susceptibility to infections. Some of the At least in young mice, T cells generated during lymphopenia- immunological changes associated with aging include induced homeostatic proliferation can have increased functional- A Lo reduced thymic output, a decrease in naive T cells, expansion of ity after in vivo Ag stimulation compared with naive CD44 memory phenotype CD8 T cells, cellular senescence, changes in T cells (9), and these homeostatically induced CD44Hi T cells telomerase expression or telomere length, reduced proliferation, provided protective immunity to bacterial infections (10). How- altered signaling, and reduced IL-2 production (1–5). Two prom- ever, two subpopulations of proliferating T cells exist under inent age-associated changes, the involution of the thymus and lymphopenic conditions, a slow proliferative subset, and a subset a global shift of T cell phenotype from naive to memory, lead that rapidly proliferates to CFSE negative (11). Interestingly, the to a decrease in the number of naive T cells (CD44Lo) and an rapidly proliferative subset is thought to be driven by environ- by guest on October 1, 2021 increase in the proportion of T cells with memory phenotype mental and/or self-Ags and upregulates the inhibitory receptor (CD44Hi) (6, 7). Several studies showed that the transfer of programmed death-1 (PD1) (12). It is unclear, however, if changes 2 2 CD44Lo CD8 T cells into RAG-1 / recipients (lymphopenic in the T cell compartment prior to infection, whether driven by environment) results in the transition to a CD44Hi memory phe- lymphopenia or other signals, impact antiviral immunity. notype within 3 wk. Using adoptive transfer of different numbers A diverse repertoire of naive T cells is essential for a vigorous 2 2 of naive (CD44Lo) cells were transferred to syngeneic RAG-1 / response to infection (13, 14). Reduced thymic output can result recipients, the transition to a memory (CD44Hi) phenotype was in a decline in naive Ag-specific precursors, and this decline can demonstrated to be directly proportional to the extent of homeo- impact the generation of an effective immune response (15, 16). static proliferation (8), suggesting that the gradual decline in new Moreover, clonal expansions of memory phenotype T cells have naive T cell production during aging could lead to a gradual in- been observed in mice, primates, and humans (17). These ex- pansions are believed to be driven by chronic infections, altered *Department of Microbiology, University of Pennsylvania Perelman School of Med- cytokine milieu of the host, or both and could further restrict the icine, Philadelphia, PA 19104; †Institute for Immunology, University of Pennsylvania T cell repertoire and/or compete with other T cell responses (18). Perelman School of Medicine, Philadelphia, PA 19104; ‡Department of Internal Medicine, Yale University School of Medicine, New Haven, CT 06510; xDepartment Thus, changes in the global T cell pool and/or changes in T cell of Immunobiology, Yale University School of Medicine, New Haven, CT 06510; and competition might account for some changes in T cell immunity { Immunology Program, Wistar Institute, Philadelphia, PA 19104 observed with age. Received for publication April 15, 2011. Accepted for publication December 13, Although less well documented, cell-intrinsic changes of naive 2011. T cells could also contribute to impaired immunity with aging. This work was supported by the American Heart Association (to V.D.), National Institutes of Health Grants HHSN266200500030C (to E.J.W.) and R01AG028082 Recently, several studies observed phenotypic alterations in the (to D.R.G.), and the Ellison Medical Foundation (to E.J.W.). CD4 T cell pool in aged mice including changes in the expression Address correspondence and reprint requests to Dr. E. John Wherry, University of of several molecules such as PD1, ICOS, CTLA4, and KLRG1, Pennsylvania Perelman School of Medicine, 421 Curie Boulevard, Room 312, Phil- although the consequences of these phenotypic changes remain adelphia, PA 19104. E-mail address: [email protected] poorly understood (19–21). Recent studies also demonstrated that The online version of this article contains supplemental material. decreased thymic output with age leads to an accumulation of Abbreviations used in this article: GSEA, Gene Set Enrichment Analysis; MNV, long-lived naive CD4 T cells that express low amounts of Bim and mouse norovirus; PD1, programmed death-1. have functional defects (22). Less information is available re- Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 garding the impact of aging on the naive CD8 T cell pool. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1101098 1934 DEFECTIVE ANTIVIRAL CD8 T CELL PRECURSORS DURING AGING

In this study, we demonstrated that aging in mice was associated were done as described previously (9, 25). CD8 and CD3 were used as with an increase in the expression of several inhibitory receptors positive gates. B220, CD19, CD4, CD11b, CD11c, and MHC II were used including PD1, LAG3, 2B4, and CD160 on CD8 T cells. These as dump gates. In addition to tetramer staining, samples were stained for CD44 and PD1. changes were especially pronounced on the CD44Hi population of CD8 T cells from uninfected aged mice with changes first be- Flow cytometry, intracellular cytokine staining, and BrdU coming apparent sometime between 8 and 12 mo of age concur- staining Lo Hi rent with a shift from CD44 to predominantly CD44 T cells. Lymphocytes were stained using standard techniques and analyzed by flow We also demonstrated that the precursor frequency of virus- cytometry. Virus-specific CD8 T cells were quantified using MHC class I specific CD8 T cells from uninfected aged mice declines by peptide tetramer staining. MHC class I peptide tetramers were made and $10-fold. Moreover, aging resulted not just in a decreased pre- used as described previously (23). Abs to CD8, CD44, PD1 (RPMI-30), CD3ε (145-2C11), and CD28 (37.51) were purchased from BioLegend cursor frequency but also in changes in the quality of these pre- (San Diego, CA). Abs to LAG3, 2B4, CD160, and IFN-g were purchased cursors. Many Ag-specific CD8 T cell precursors were CD44Hi from eBioscience (San Diego, CA). Abs to CD4, Mip1a, KLRG1, and and PD1 positive. Transcriptional profiling of total CD44Hi CD8 Tbet were purchased from Invitrogen (Carlsbad, CA), R&D Systems T cells from naive aged mice revealed strong similarities to (Minneapolis, MN), Beckman Coulter (Fullerton, CA), or Santa Cruz Biotechnology (Santa Cruz, CA), respectively. All other Abs were pur- exhausted CD8 T cells found during chronic infections. Studies chased from BD Biosciences (San Diego, CA). Staining and analysis were with aged and young TCR transgenic T cells indicated substantial performed as described previously (23). proliferative defects of CD44Hi T cells from uninfected mice. Function was investigated by intracellular cytokine staining following 6 These studies suggest both quantitative and qualitative defects in stimulation. Briefly, 1 3 10 splenocytes were cultured in the absence or the pool of Ag-specific CD8 T cells prior to immunization or presence of the anti-CD3 (2 mg/ml) and anti-CD28 (1 mg/ml) for 5 h at

37˚C. Following staining for surface Ags as described above, cells were Downloaded from infection in the elderly and point to possible opportunities for stained for intracellular cytokines (IFN-g,IL-2,andMip1a)usingthe reversing such age-related defects. Cytofix/Cytoperm kit (BD Biosciences). For BrdU analysis, mice received a single i.p. injection (1 mg/mouse) of Materials and Methods BrdU 12 h before sacrifice. BrdU staining was carried out using a BrdU Mice Flow Kit (BD Pharmingen) in accordance with the manufacturer’s in- structions. BrdU was detected using FITC-conjugated anti-BrdU Ab. Female mice were purchased from Taconic at either 3–5 or 8 mo of age and Samples were collected using an LSR II flow cytometer (BD Biosciences, then aged to 12–14 and 17–22 mo old (aged mice) in the Wistar Institute San Jose, CA). http://www.jimmunol.org/ aging colony. Additional female mice were purchased from Taconic at 2–3 Cell sorting and microarray analysis mo of age. These mice were housed in the same colony but used when 3–5 mo old (young mice). Animals were housed in microisolator cages at the CD8 T cells from uninfected young and aged mice were purified (.90% American Association for the Accreditation of Laboratory Animal Care- purity) using magnetic beads (CD8+ T cell isolation kit, MACS beads; approved Animal Care Facility at the Wistar Institute and screened by an Miltenyi Biotec) and then sorted directly into TRIzol (Life Technologies, independent commercial service (Charles River Laboratories) every 6 mo Rockville, MD) on the basis of CD44 expression. Purity was assessed by for a panel of 17 rodent pathogens including LCMV and mouse norovirus FACS analysis and was .95–98% for all samples. Total RNA was isolated (MNV). All screens for rodent pathogens over the 5 y of existence of this from sorted CD44Hi CD8 T cells from uninfected young and aged mice aging colony were negative. In an independent set of studies, aged mice using TRIzol, according to the manufacturer’s instructions. cDNA was

(17–22 mo old) were obtained from the aging colony at the National fragmented, labeled, and amplified as described previously (26). Sam- by guest on October 1, 2021 Institutes on Aging and examined directly upon receipt. Aged animals with ples were hybridized to the Affymetrix GeneChip Mouse Exon 1.0 ST obvious cancer or lymphoma were excluded from the studies. For all arrays (Affymetrix, Santa Clara, CA) at the University of Pennsylvania adoptive transfer experiments, congenic mice (obtained from National Microarray Core Facility. Transcript levels were summarized from the Cancer Institute) differing in Ly5 (Ly5.1 versus Ly5.2) or Thy (Thy1.1 Affymetrix-CEL files using the Robust Multichip Average algorithm (27), versus Thy1.2) were used. P14 TCR transgenic mice were bred at the as implemented by the Affymetrix Power Tool apt-probe set-summarize. 2 2 Wistar Institute. OT1 Rag2 / female mice were obtained from Taconic at There were four technical replicates for each young and aged CD44Hi 2 mo of age. All mice were used in accordance with Institutional Animal population. Differentially expressed genes were identified using the Class Care and Use Committee guidelines. Neighbors module of GenePattern (28) and the Significance Analysis of Microarrays algorithm (29). Gene Set Enrichment Analysis (GSEA) was Adoptive transfer, virus, and infection performed as described previously (26). CD8 T cells from young and aged P14 Ly5.1 or P14 Thy1.1 mice were Statistical analysis purified (.90% purity) using magnetic beads (CD8+ T cell isolation kit, MACS beads; Miltenyi Biotec) and then sorted on the basis of CD44 Data were analyzed using a two-tailed Student t test, and p # 0.05 was expression. After sorting, 5 3 103 P14 (DbGP33-specific) CD8 T cells considered significant. For microarray data analysis, we used the Class from each group (young CD44Lo, young CD44Hi, aged CD44Lo, and Neighbors and Significance Analysis of Microarrays algorithms as de- aged CD44Hi) were transferred i.v. into wild-type recipient mice (Ly5.2+ scribed previously (28, 29). Thy1.2+), which were then infected with LCMVArmstrong (2 3 105 PFU) i.p. Donor populations were monitored in the peripheral blood by retro- Results orbital blood collection as described previously (23). Similarly, CD8 T cells from young OT1 Rag22/2 mice were purified using magnetic beads Phenotypic changes of T cells from uninfected mice during (CD8+ T cell isolation kit, MACS beads; Miltenyi Biotec) and then sorted aging 3 3 b on the basis of CD44 expression. After sorting, 15 10 OT1 (K OVA To begin to understand age-related changes in antiviral immunity, specific) CD8 T cells from each group (young CD44Lo and young CD44Hi) were transferred i.v. into wild-type recipient mice (Ly5.2+), which were we analyzed the percentages and numbers of T cells in mice at then infected with recombinant vesicular stomatitis virus expressing OVA different times in their lifespan. The analysis of mice at 4, 8, 12, and (2 3 106 PFU) i.v. Donor populations were monitored in the peripheral 18 mo of age demonstrated that the percentages and numbers of blood by retro-orbital blood collection as described previously (23). CD8 T cells were diminished with age. This was true in both spleen Isolation of lymphocytes from tissues and blood (Fig. 1A, Supplemental Fig. 1A), consistent with pre- vious studies (30–33). Moreover, we demonstrated that a previ- Lymphocytes were isolated from peripheral blood and spleens as described previously (24). ously observed global shift from naive to memory phenotype T cells (13, 34, 35) that results in an increase in the frequency and Tetramer staining and magnetic bead sorting number of CD44Hi T cells and a concomitant decrease in the Lo Spleens were isolated from uninfected mice at 4, 7, 12, 14, 17, and 19 mo of number of CD44 naive T cells started as early as 8–12 mo of age age; tetramer staining (DbGP33 or KbOVA) and magnetic bead sorting (Fig. 1A–C). The Journal of Immunology 1935 Downloaded from http://www.jimmunol.org/

FIGURE 1. Age-associated changes in CD8 T cells. Frequencies of CD8 T cells were analyzed in the spleens and blood of uninfected mice at 4, 8, 12, and 18 mo of age (A). The percentages and numbers of CD44Hi and CD44Lo CD8 T cells in the spleens of uninfected mice at age of 4–5, 8, 12, and 18 mo are shown in B and C, respectively. CD8 T cells were isolated from uninfected young and aged spleens at 4, 8, 12, and 18 mo of age and stained for the expression of inhibitory receptors PD1, LAG3, 2B4, and CD160. The representative graphs of individual stains of CD8 T cells from spleen are shown in D. by guest on October 1, 2021 The percentages of different inhibitory receptors on CD44Hi CD8 T cells from spleens are shown in E. A total of 1.48 6 0.09% (mean 6 SEM) of CD44Hi CD8 T cells in the spleens of 4-mo-old mice are PD1 positive compared with 2.58 6 0.33, 15.93 6 6.12, and 43.18 6 2.81% of PD1-positive CD44Hi CD8 T cells from the spleens of 8-, 12-, and 18-mo-old mice, respectively. Data are representative of two independent experiments including 12–20 mice/ experiment. Some data are representative of several pooled experiments. *0.05 . p . 0.01, **0.01 . p . 0.001, ***p , 0.001 by unpaired two-tailed t test.

We next examined the phenotype of T cells from the spleen and stricted to the CD44Hi subset (Fig. 1D; data not shown). Although blood of uninfected aged and young mice and found an increase in many of these inhibitory receptors are involved in self-tolerance the expression of KLRG1 and Tbet for CD8 T cells from uninfected and are upregulated by self-reactive T cells (38), these molecules aged mice (Supplemental Fig. 1B,1C), which could reflect se- also can be induced during chronic infections (39). Mice in our nescence or terminal differentiation as reported previously (36, aging colony were routinely screened for common rodent patho- 37). Expression of other markers such as CD44, CD62L, and gens and were always negative for all agents tested including CD49d (a4 integrin) was also altered (data not shown), but one of LCMV and MNV (data not shown). Moreover, we observed the most striking observations was the difference in expression of a similar upregulation of inhibitory receptors on T cells from an inhibitory receptors between CD8 T cells from uninfected aged independent source of aged mice by examining mice from the and young mice. In particular, CD44Hi CD8 T cells from aged aging colony operated by the National Institutes on Aging (Sup- mice had higher expression of PD1, LAG3, 2B4, and CD160 plemental Fig. 1G,1H). compared with CD8 T cells from young mice (Fig. 1D). The upregulation of inhibitory receptor expression started between 8 The increase in the inhibitory receptors on CD8 T cells from and 12 mo of age and became more prominent between 12 and 18 aged mice is not due exclusively to oligoclonal expansions mo (Fig. 1E, Supplemental Fig. 1D). Mean fluorescence intensity To investigate whether the increase in inhibitory receptors on CD8 was also increased for both PD1 and LAG3 on CD44Hi CD8 T cells from aged mice was due to oligoclonal expansions, which T cells from 12- to 18-mo-old mice (Supplemental Fig. 1E,1F). have been associated with aging (17), we examined TCR Vb ex- Although the mean fluorescence intensity was unchanged for 2B4 pression on CD8 T cells from young (4 mo old) and aged (20–22 or decreased for CD160 (Supplemental Fig. 1E,1F), the percen- mo old) uninfected mice. Aged mice had higher variability in the tages of CD44Hi cells that were 2B4 and CD160+ increased at expression of Vb receptors compared with young animals in 12 and 18 mo of age (Fig. 1D,1E, Supplemental Fig. 1D). The agreement with prior work (Fig. 2A,2B) (40), and in some mice, change in inhibitory receptors expression with age was also evi- the clonal expansions of certain Vb receptors (e.g., Vb4, Vb8.3, dent in the total population of CD8 T cells, though most of the Vb9, Vb13, and Vb14) were detected. Although modest clonal increased expression of PD1, LAG3, 2B4, and CD160 was re- expansions were occasionally present, they represented at most 1936 DEFECTIVE ANTIVIRAL CD8 T CELL PRECURSORS DURING AGING

2C,2D). The expression patterns of these markers on oligoclo- nally expanded populations were variable and ranged from in- distinguishable to slightly or very different from the total CD8 T cell pool, depending on the specific oligoclonal expansion and marker tested. Thus, these data suggested that the increase in the inhibitory receptors on CD8 T cells from aged mice was not due exclusively to oligoclonal expansions. Age-related changes in Ag-specific CD8 T cell precursor numbers and phenotype in uninfected mice In young animals, the majority of Ag-specific but inexperienced CD8 T cells are CD44Lo. Given the loss in CD44Lo CD8 T cells in aged mice, we next investigated the number and phenotype of virus-specific CD8 T cells in aged mice prior to infection. Using MHC class I/peptide tetramer enrichment staining as previously described (9, 25), we compared the numbers of DbGP33-specific and KbOVA-specific CD8 T cell precursors in the spleens of young and aged uninfected mice. Young mice (4 mo old) contained ∼260 DbGP33-specific precursors/spleen, consistent with previous esti- mates using this (41) and other techniques (Fig. 3A,3B) (42). This Downloaded from number declined in mice 12–14 mo old and was further reduced by 17 mo. Aged mice (17 mo) exhibited at least a 10-fold re- duction in the number of DbGP33-specific CD8 T cell precursors per spleen (∼20–30 DbGP33-specific precursors/spleen). Simi- larly, the number of KbOVA-specific CD8 T cell precursors in the spleens of aged mice was significantly reduced compared with the http://www.jimmunol.org/ number of these precursors in the spleens of young mice. Young mice (7 mo old) contained ∼61 KbOVA-specific precursors/spleen, whereas in aged mice (19 mo), the number of KbOVA-specific CD8 T cell precursors per spleen was ∼13 (Fig. 3E). These observations suggest that to reach the same numbers of Ag- specific T cells at the peak of the response, virus-specific CD8 T cells in aged mice would have to undergo at least three extra rounds of division, assuming Ag-specific precursors in aged mice by guest on October 1, 2021 are qualitatively similar to those in young mice. To investigate this second issue, we examined differentiation markers expressed by virus-specific precursors in uninfected aged versus young mice. In aged mice, a higher proportion of DbGP33- or KbOVA-specific precursors were CD44Hi (Fig. 3C,3F), whereas the vast major- ity of these DbGP33- or KbOVA-specific precursors in young mice were CD44Lo. In addition, many of the DbGP33- or KbOVA- specific aged CD8 T cells expressed increased levels of PD1 (Fig. 3D,3G). Thus, the numbers of Ag-specific precursors de- clined with age, and the Ag-specific precursors that remained were phenotypically altered. FIGURE 2. The increase in the inhibitory receptors on CD8 T cells from aged mice is not due exclusively to oligoclonal expansions. CD8 T cells Functional changes of T cell populations from naive mice were isolated from spleens of uninfected young (4 mo old, s) and aged during aging (20–21 mo old, d) mice and were stained for panel of different Vb We next investigated whether the changes in the phenotype and receptors. The percentages of different Vb populations on young and aged precursor frequency of the CD8 T cell pool from aged uninfected naive CD8 T cells are shown in A and B, respectively. A clonal expansion mice were accompanied by changes in the functionality of these was defined as at least 3 SDs above the mean Vb use in young mice. C, cells. T cells from young and aged naive mice were stimulated Histograms of different markers on total or oligoclonally expanded pop- ulation of CD8 T cells from one young and several aged mice. D, The in vitro using anti-CD3 and anti-CD28 Abs, and cytokine pro- percentages of expression of the same markers as shown in C on total and duction was examined. During polyclonal stimulation, there were oligoclonally expanded cells. Data are representative of two independent numerically more IFN-g producers in aged mice (Fig. 4A,4B). experiments including 10 mice/experiment. *0.05 . p . 0.01, **0.01 . Young and aged CD8 T cells produced similar amounts of Mip1a, p . 0.001, ***p , 0.001 by unpaired two-tailed t test. but CD8 T cells from aged mice were poor producers of IL-2 compared with CD8 T cells from young mice in agreement with a quarter and usually ,10% of the total CD8 T cell pool (Fig. 2B). previous data examining CD4 T cells from aged mice (43). We Oligoclonally expanded populations were examined for the ex- failed to detect IL-17 in the in vitro CD8 T cell assay, but we pression of different markers including CD44, CD49d (a4 integ- observed that aged CD8 T cells had slightly higher IL-10 pro- rin), CD122, and PD1. The expression profile of these markers duction than young CD8 T cells upon activation (data not shown). within each Vb fraction was then compared with the expression Although both aged and young CD44Hi CD8 T cells produced profiles on the remaining CD8 T cells within the same mouse (Fig. IFN-g, there were some aged mice in which this function was well The Journal of Immunology 1937

The transcriptional profiles of CD44Hi CD8 T cells from uninfected aged mice indicate a signature similar to that of exhausted CD8 T cells We next examined the global gene-expression profiles of CD44Hi CD8 T cells from young and aged mice. CD44Hi CD8 T cells from young and aged uninfected mice showed differential expression of a number of transcripts known to play important roles in CD8 T cells (Fig 5A, Supplemental Tables I, II). Using both the Class Neighbors and the Significance Analysis of Microarrays algo- rithms, we identified many transcripts that have also been impli- cated in modulating CD8 T cell exhaustion or dysfunction during chronic infections (e.g., PD1, LAG3, IL-10, and Blimp-1; Sup- plemental Tables I, II) (26). To further probe the similarities be- tween aging and exhaustion, we used GSEA to evaluate the global similarity between our samples and a previously defined signature of exhausted CD8 T cells (26). There was a clear enrichment of the transcriptional signature of CD8 T cell exhaustion in the aged CD44Hi CD8 T cells, with a p , 0.001 (Fig. 5B). Moreover, there

is a corresponding bias of a similarly defined memory signature Downloaded from toward the young CD44Hi CD8 T cell samples, p , 0.001. Taken together, these gene expression data indicate that aging is asso- ciated with a transcriptional program in CD8 T cells that has similarities to that found in exhausted T cells. One notable set of differences, however, was the increase in KLRG1 and Tbet ex- pression by CD8 T cells from aged mice. In contrast, expression of http://www.jimmunol.org/ these molecules is reduced in exhausted CD8 T cells during chronic LCMV infection (26, 44). Nonetheless, inhibitory recep- tors, cell death pathways, and transcription factors were identified as specific genes and pathways and may provide important targets for future analysis. Poor responsiveness of Ag-specific CD44Hi CD8 T cells from uninfected aged and young mice

We next examined the impact of qualitative changes in the pool of by guest on October 1, 2021 Ag-specific CD8 T cells from uninfected aged mice in vivo during viral infection. Hence, we isolated DbGP33-specific CD8 T cells from uninfected young and aged P14 TCR transgenic mice that are reactive to the GP33 epitope of LCMV. Note that, although aged P14 mice had a clear population of CD44Hi DbGP33 FIGURE 3. Age-related changes in Ag-specific CD8 T cell precursor tetramer-positive CD8 T cells, all PD1 expressing CD8 T cells in numbers and phenotype in naive mice. Spleens of uninfected mice at 4, 7, these mice were tetramer negative (Supplemental Fig. 2). This 12, 14, 17, and 19 mo of age were stained with DbGP33 or KbOVA tet- ramers. Representative staining for DbGP33-specific precursors are shown observation suggests the use of a nontransgenic TCR and perhaps in A. The absolute numbers of DbGP33-specific precursors at different ages a role for TCR stimulation in age-related increases in PD1 ex- are shown in B. C and D, The percentage of expression of CD44 and PD1 pression. We sorted these P14 CD8 T cells into two groups on Lo Hi on DbGP33-specific precursors, respectively. The absolute numbers of the basis of CD44: CD44 (naive CD8 T cells) and CD44 KbOVA-specific precursors at different ages are shown in E. F and G, The (memory-like CD8 T cells, Fig. 6A). To test the impact of a shift percentage of expression of CD44 and PD1 on KbOVA-specific precursors, from CD44Lo to CD44Hi CD8 T cells with age, equal numbers of respectively. Data are from two independent experiments and are displayed either young or aged CD44Lo and CD44Hi P14 cells were purified . . . . as the pooled results from both. *0.05 p 0.01, **0.01 p 0.001, and adoptively transferred to young congenic recipient mice that , ***p 0.001 by unpaired two-tailed t test. were subsequently challenged with LCMV (Fig. 6C). Although there was clearly a hierarchy in the responsiveness of these dif- preserved and another subset of animals in which the CD8 T cells ferent donor populations (Fig. 6D,6E), the CD44Hi P14 cells from appeared less functional per cell (Fig. 4C,4D). IFN-g production aged mice were 10- to 20-fold less efficient in the initial expansion was also compared with PD1 expression. Very little IFN-g was and accumulated poorly in the blood compared with the other found in the small PD1+ subset of CD8 T cells from young naive populations. Also, a lower percentage of these CD44Hi aged donor mice (Fig. 4E). Although there were some aged mice in which cells incorporated BrdU compared with young P14 cells (17.3 6 IFN-g was produced by PD1+ CD8 T cells, in general, the PD1+ 1.2 versus 26.7 6 1.9%; p = 0.0056; Fig. 6F,6G). Despite perhaps CD8 T cells from aged mice made less IFN-g per cell compared a slight disadvantage of CD44Lo CD8 T cells from aged mice in with the PD1lo cells. Overall, these results indicate that aging expansion (Fig. 6E), we found similar BrdU incorporation be- alters some components of CD8 CD44Hi T cell cytokine respon- tween young CD44Lo and aged CD44Lo cells (∼29.6 6 2.3 and ses, although there was clearly heterogeneity with preserved 34.4 6 2.3%, respectively). We also observed that, in the spleen, functionality in some mice and perhaps moderately reduced IFN-g CD44Hi P14 cells from young donors had moderately diminished production per cell in other mice. responses compared with CD44Lo cells from same donors though 1938 DEFECTIVE ANTIVIRAL CD8 T CELL PRECURSORS DURING AGING

FIGURE 4. Functional changes of naive T cells during aging. CD8 T cells were isolated from the spleens of uninfected young (4–5 mo old) and aged (19–20 mo old) mice. Splenocytes were stimulated in vitro using anti-CD3 and anti-CD28 Abs. Five hours later, cells were stained for the production of different cytokines. The representative graphs for IFN-g, IL-2, and Mip1a staining are shown in A. Graphs in B show absolute numbers of young and aged naive CD8 T cells that produce IFN-g, IL-2, and Mip1a. Representative plots of CD8 T cells show IFN-g production versus CD44 expression (C). The Downloaded from percentages of CD44Hi CD8 T cells that produce IFN-g are shown in D. E is gated on CD8 T cells from young and aged mice and shows IFN-g versus PD1 expression. Data are representative of two independent experiments including eight to nine mice per experiment. *0.05 . p . 0.01, **0.01 . p . 0.001, ***p , 0.001 by unpaired two-tailed t test.

CD44Hi P14 cells from aged mice clearly had the most defective vigorously in response to bacterial challenge compared with Lo proliferative expansion of the populations examined (Fig. 6E,6G). CD44 CD8 T cells (10). We therefore extended our studies of http://www.jimmunol.org/ Previous studies indicated that CD44Hi CD8 T cells generated by naturally arising CD44Hi CD8 T cells from young mice to another homeostatic proliferation in lymphopenic hosts responded more setting. KbOVA-specific CD8 T cells were isolated from unin- fected young OT1 TCR transgenic mice on Rag22/2 background (OT1 Rag22/2 mice). In OT1 Rag22/2 mice, all CD8 T cells are OVA specific. We sorted these CD8 T cells into CD44Lo (naive CD8 T cells) and CD44Hi (memory-like CD8 T cells; Fig. 6H). Equal numbers of purified CD44Lo and CD44Hi OVA-specific CD8 T cells from young mice were adoptively transferred to young congenic recipient mice, and those recipients were infected with by guest on October 1, 2021 vesicular stomatitis virus expressing OVA. Once again, naturally arising CD44Hi cells from young mice were 3- to 4-fold less ef- ficient in the initial expansion and accumulated to lower frequency and absolute number in the blood and spleen compared with the CD44Lo populations (Fig. 6J,6K). Thus, using two different TCR transgenic models and two different types of infections, we demonstrated that although CD44Hi CD8 T cells from young mice can serve as precursors, their response to infection is reduced compared with responses that arise from CD44Lo precursors. These data suggest that at least part of the reason for poor antiviral CD8 T cell responses in old mice is due to the non–lymphopenia- induced natural shift of some Ag-specific precursors to a CD44Hi phenotype.

Discussion Aging-related susceptibility to infectious diseases and cancer is FIGURE 5. The transcriptional profiles of CD44Hi CD8 T cells from associated with diminished immune function including reduced uninfected aged mice indicate a signature similar to that of exhausted CD8 T cell responses. Although some aspects of declining T cell re- T cells. CD8 T cells were isolated from spleens of uninfected young (4 mo sponses with age reflect systemic changes (e.g., an increase in old) and aged (20–21 mo old) mice and sorted, based on CD44 expression. regulatory T cells, etc.), others are due to cell-intrinsic defects (45). RNA was isolated and hybridized to the Affymetrix GeneChip Mouse In the current study, we demonstrated both quantitative and Exon 1.0 ST arrays. Class Neighbors analysis was performed to identify Hi qualitative age-related changes in T cells and found that these genes that differentiate CD44 CD8 T cells from uninfected young and changes were linked to defective responses upon viral infection. aged mice (n = 4 for each set, A). To evaluate the global similarity to The decrease in peripheral T cell numbers likely reflects thymic previously defined signatures of CD8 T cell memory and exhaustion, we performed GSEA of CD44Hi CD8 T cells from uninfected young and aged involution that starts early (perhaps as early as 6 wk in mice and mice (B). There is a clear enrichment of the exhaustion signature in the age 1 in humans) and continues through life (3, 46). Although aged CD44Hi CD8 T cell samples; p , 0.001. There is a corresponding some functional thymic tissue can be found at old age (47), T cell bias of a similarly defined memory signature toward the young CD44Hi numbers in the periphery are thought to be retained increasingly CD8 T cell samples; p , 0.001. by homeostatic proliferation as thymic output declines. Both ho- The Journal of Immunology 1939 Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 6. Poor responsiveness of Ag-specific CD44Hi CD8 T cells from uninfected aged and young mice. P14 (DbGP33 specific) CD8 T cells were purified from uninfected young and aged P14 mice (Ly5.1+) and sorted, based on the expression of CD44 (A). B,DbGP33 tetramer staining on sorted P14 populations. After sorting, equal numbers of young and aged CD44Hi and CD44Lo P14 cells were transferred into recipient mice (Ly5.2+). Recipient mice were then challenged i.p. with LCMV (2 3 105 PFU/mouse, C). The responses of the donor P14 CD8 T cells were monitored in the blood (D) and spleens (day 8 postinfection, E–G) of recipient mice. Lower frequencies of CD44Hi P14 CD8 T cells derived from aged and young donors were observed compared with their CD44Lo counterparts. E and F, The absolute number and frequency of BrdU+ donor P14 CD8 T cells derived from young and aged CD44Hi and CD44Lo cells in the spleens of recipient mice at day 8 postinfection. In these experiments, 12 h preanalysis, mice were injected with BrdU i.p. Data are representative of two independent experiments including four to five mice per experiment. Similarly, OT1 (KbOVA specific) CD8 T cells were purified from uninfected young OT1 Rag22/2 mice (Ly5.1+) and sorted, based on the expression of CD44 (H). I,KbOVA tetramer staining on sorted OT1 populations. After sorting, equal numbers of young CD44Hi and CD44Lo OT1 cells were transferred into recipient mice (Ly5.2+). Recipient mice were then challenged i.v. with VSV-OVA (2 3 106 PFU/mouse). The responses of the donor OT1 CD8 T cells were monitored in the blood (day 7 postinfection, J) and spleens (day 8 postinfection, K) of recipient mice. Lower numbers of CD44Hi OT1 CD8 T cells (s) were observed compared with CD44Lo OT1 CD8 T (N). *0.05 . p . 0.01, **0.01 . p . 0.001, ***p , 0.001 by unpaired two-tailed t test. meostatic proliferation and exposure to environmental Ags are function of LCMV-, HIV-, or HCV-specific CD8 T cells and that believed to be responsible for the phenotypic changes in T cells the blockade of PD1 and/or other inhibitory receptors results in that occur with age. Indeed, T cells that undergo a homeostasis improved CD8 T cell functionality (39). Furthermore, CD4 T cells driven-proliferation under lymphopenic conditions have been from aged mice can also express increased PD1 (19), and PD1- shown to upregulate CD44 (8, 48). Interestingly, CD44Hi T cells positive CD4 T cells from aged mice exhibit proliferative hypo- can be also found in germ-free mice (9, 49), suggesting that these responsiveness in vitro (20) and in vivo (52). Blockade of PD1/ cells can arise during physiological lymphopenia in the process of PDL1 pathway in vitro moderately improved cytokine production, aging in the absence of commensal microbes. but did not restore proliferation of PD1-positive T cells from aged CD44Hi CD8 T cells from uninfected aged mice express high mice (21). The impact of high expression of PD1 on CD8 T cells levels of several inhibitory receptors including PD1, LAG3, 2B4, during aging remains less well defined. In addition, PD1 is clearly and CD160. Upregulated expression of inhibitory receptors, such not the only inhibitory receptor expressed by these cells, and as as PD1, is typically found on T cells responding to persisting Ag occurs during chronic infections, coexpression of inhibitory recep- stimulation and chronic infection in both mice and humans, but tors might influence the function of these aged cells. In combina- PD1 and related pathways also have a central role in regulat- tion with microarray studies, these observations suggest a global ing peripheral self-tolerance and autoimmunity (50, 51). Recent age-associated change in gene expression in CD8 T cells from studies showed that PD1 expression correlates with impaired uninfected mice that has similarities to what occurs for exhausted 1940 DEFECTIVE ANTIVIRAL CD8 T CELL PRECURSORS DURING AGING

CD8 T cells. In addition to inhibitory receptors, these gene ex- that the shift of some naive precursors into the CD44Hi pool in the pression profiles also identified a number of other immunoregula- absence of overt Ag stimulation or severe lymphopenia is associ- tory and/or transcriptional pathways that differed between CD8 ated with reduced proliferative capacity and potentially other T cells from young and aged mice. These differences included functional defects. These observations suggest the possibility that components of AP-1 (e.g., Fos, Fosb, Jun, and JunB), other tran- generating high-quality memory T cells prior to this shift to de- scription factors (Myb, Klf4, Maf, and Prdm1), and IL10 and FasL. fective precursors (i.e., when young) might enhance immunity into It will be interesting to investigate the importance of these tran- old age. Studies to directly test this notion are under way. scriptional changes in the altered responses of T cells during aging. Another key finding of the current study was that the precursor An interesting question that arises is, what is the cause for frequency of Ag-specific CD8 T cells present prior to infection was accumulating inhibitory receptor expression with age. Despite the 10-fold lower in aged compared with young mice. Previous studies similarities to T cells found during chronic infections, it is im- demonstrated that age-associated changes in immunodominance portant to note that our aged mice were consistently negative for all could be due to TCR repertoire changes and/or low precursor common rodent pathogens including viruses such as LCMV and frequency leading to holes in the repertoire (15). We have quan- MNV. This information suggests that, although signatures of T cell tified this loss of virus-specific precursors for two different Ag exhaustion might reflect exposure to unknown infectious agents, specificities and further demonstrated that many Ag-specific pre- other age-related events (such as increased self-reactivity) might cursors from uninfected aged mice are CD44Hi and express PD1. underlie the inhibitory receptor expression and other transcriptional The decrease in both the number and quality of Ag-specific pre- changes. Indeed, T cells undergoing recognition of self-Ag in vivo cursors in aged mice could mean that these cells must undergo express PD1 during the induction of tolerance (38). Although most greater numbers of divisions upon infection to achieve the same of these tolerized T cells are typically physically deleted in young clonal expansion causing a delay in the early antiviral effector Downloaded from mice, we propose that during aging some of these tolerized T cells CD8 T cell response in aged mice consistent with previous escape deletion, persist, and accumulate over time. Oligoclonally observations (24). This delay in clonal expansion during an acute expanded T cells have also been previously shown to express PD1 viral infection could result in a substantial increase in viral rep- (40). Although we observed some oligoclonal expansion of the lication and dissemination. Moreover, the increased number of CD8 T cells in uninfected aged mice, these expansions could not divisions could also result in defects in effector and memory explain all of the PD1 expression, because the percentage of oli- T cell responses, drive terminal differentiation or even skew the http://www.jimmunol.org/ goclonally expanded cells was typically ,10–15%, whereas 20– responding T cells toward functional exhaustion. The expression 60% of CD8 T cells expressed PD1. It is interesting that PD1- of PD1 prior to infection could mean even further difficulties in positive cells were found predominantly in the polyclonal pool of proliferation, expansion and /or functionality of these cells and CD44Hi CD8 T cell from naive mice. Although tetramer-positive enhanced PD1 signaling could alter CD8 T cell differentiation CD44Hi cells from aged P14 transgenic mice were PD1 negative, through induction of transcriptional pathways such as basic leu- these mice did contain a population of tetramer-negative CD8 cine zipper transcription factor, ATF-like (54). T cells that were PD1 positive (Supplemental Fig. 2). It is possible Thus, we demonstrate in this study that the ability of CD8 T cells that tetramer-negative CD8 PD1Hi T cells in these P14 mice cross- from young and aged mice to respond to infection is at least by guest on October 1, 2021 react with self- or environmental Ags. Future studies are necessary partially influenced by intrinsic and population-based changes in to test this idea, but if true, it could suggest that the pool of Ag- CD8 T cells. Precursor frequency plays a significant role in T cell specific CD8 T cells in aged mice prior to infection (i.e., CD44Hi memory quality (55), and our data now demonstrate major PD1+ CD8 T cells) would be more cross-reactive compared with changes in Ag-specific CD8 T cell precursor frequency are asso- the CD8 T cell pool in young animals. ciated with age-dependent defects in antiviral immunity. Under- It is interesting to note, however, that even for the CD44HiPD1Lo standing these aging-associated changes could help define the CD8 T cells from naive aged animals, responses were qualitatively molecular mechanisms underlying defective CD8 T cell respon- inferior compared with CD44Lo CD8 T cells. We also observed ses. Strategies that result in vigorous production of new CD44Lo diminished proliferative responses of naturally arising CD44Hi cells truly naive T cells from the thymus, prevention of T cell senes- from young transgenic animals compared with CD44Lo cells. These cence, and modulation/manipulation of immunoregulatory path- observations were somewhat surprising given the work by others ways are some of the attractive approaches that could enhance demonstrating enhanced responsiveness of homeostatically gener- immunity and responses to vaccines in the elderly. ated CD44Hi CD8 T cells (9, 10). Taken together, these data suggest Hi that naturally arising CD44 CD8 T cells that accumulate with age Acknowledgments might be functionally distinct from those generated by strong ho- Lo We thank Mohammed-Alkhatim A. Ali and Kathleen Mansfield for techni- meostatic emptiness following adoptive transfer of CD44 T cells cal assistance and the members of the Wherry laboratory for helpful dis- 2/2 into lymphopenic hosts (e.g., over ∼3–4 wk in irradiated or Rag cussions. We also thank Dr. Ross M. Kedl at the University of Colorado mice). One possible explanation for some of these differences could Health Sciences Center for help with tetramer staining and bead sorting be that transient bacterial release after sublethal irradiation influ- methodology. ences the functional quality of CD44Hi memory cells; if these CD44Hi cells are generated in antibiotic-treated hosts, they show Disclosures dramatic impairment in protective immunity against bacterial in- The authors have no financial conflicts of interest. fections (53). In addition, the decreased ability of CD44Hi pre- cursors to respond to infection is consistent with the changes in global gene expression in these cells, including the prominent in- References crease in multiple cell intrinsic (e.g., PD1) and potentially cell 1. Effros, R. B. 2007. Telomerase induction in T cells: a cure for aging and disease? Exp. Gerontol. 42: 416–420. extrinsic (e.g., IL-10) negative regulatory pathways. The global 2. Maue, A. C., E. J. Yager, S. L. Swain, D. L. Woodland, M. A. Blackman, and shift to CD44Hi CD8 T cells with age is obviously associated with L. Haynes. 2009. T-cell immunosenescence: lessons learned from mouse models of aging. Trends Immunol. 30: 301–305. the accumulation of functional, Ag-induced memory T cells gen- 3. Linton, P. J., and K. Dorshkind. 2004. Age-related changes in lymphocyte de- erated by infection or vaccination. However, our data now suggest velopment and function. Nat. Immunol. 5: 133–139. The Journal of Immunology 1941

4. Dicarlo, A. L., R. Fuldner, J. Kaminski, and R. Hodes. 2009. Aging in the 29. Tusher, V. G., R. Tibshirani, and G. Chu. 2001. Significance analysis of context of immunological architecture, function and disease outcomes. Trends microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. Immunol. 30: 293–294. USA 98: 5116–5121. 5. Weng, N. P. 2006. Aging of the immune system: how much can the adaptive 30. Lazuardi, L., B. Jenewein, A. M. Wolf, G. Pfister, A. Tzankov, and B. Grubeck- immune system adapt? Immunity 24: 495–499. Loebenstein. 2005. Age-related loss of naı¨ve T cells and dysregulation of T-cell/ 6. Naylor, K., G. Li, A. N. Vallejo, W. W. Lee, K. Koetz, E. Bryl, J. Witkowski, B-cell interactions in human lymph nodes. Immunology 114: 37–43. J. Fulbright, C. M. Weyand, and J. J. Goronzy. 2005. The influence of age on 31. Sansoni, P., A. Cossarizza, V. Brianti, F. Fagnoni, G. Snelli, D. Monti, T cell generation and TCR diversity. J. Immunol. 174: 7446–7452. A. Marcato, G. Passeri, C. Ortolani, E. Forti, et al. 1993. Lymphocyte subsets 7. Woodland, D. L., and M. A. Blackman. 2006. Immunity and age: living in the and natural killer cell activity in healthy old people and centenarians. Blood 82: past? Trends Immunol. 27: 303–307. 2767–2773. 8. Cho, B. K., V. P. Rao, Q. Ge, H. N. Eisen, and J. Chen. 2000. Homeostasis- 32. Fagnoni, F. F., R. Vescovini, G. Passeri, G. Bologna, M. Pedrazzoni, G. Lavagetto, stimulated proliferation drives naive T cells to differentiate directly into memory A. Casti, C. Franceschi, M. Passeri, and P. Sansoni. 2000. Shortage of circu- T cells. J. Exp. Med. 192: 549–556. lating naive CD8+ T cells provides new insights on immunodeficiency in aging. 9. Haluszczak, C., A. D. Akue, S. E. Hamilton, L. D. Johnson, L. Pujanauski, Blood 95: 2860–2868. L. Teodorovic, S. C. Jameson, and R. M. Kedl. 2009. The antigen-specific CD8+ 33. Callahan, J. E., J. W. Kappler, and P. Marrack. 1993. Unexpected expansions of T cell repertoire in unimmunized mice includes memory phenotype cells bearing CD8-bearing cells in old mice. J. Immunol. 151: 6657–6669. markers of homeostatic expansion. J. Exp. Med. 206: 435–448. 34. Nikolich-Zugich, J., and B. D. Rudd. 2010. Immune memory and aging: an 10. Hamilton, S. E., M. C. Wolkers, S. P. Schoenberger, and S. C. Jameson. 2006. infinite or finite resource? Curr. Opin. Immunol. 22: 535–540. The generation of protective memory-like CD8+ T cells during homeostatic 35. Haynes, L., and S. L. Swain. 2006. Why aging T cells fail: implications for proliferation requires CD4+ T cells. Nat. Immunol. 7: 475–481. vaccination. Immunity 24: 663–666. 11. Sprent, J., and C. D. Surh. 2011. Normal T cell homeostasis: the conversion of 36. Weng, N. P., A. N. Akbar, and J. Goronzy. 2009. CD28– T cells: their role in the naive cells into memory-phenotype cells. Nat. Immunol. 12: 478–484. age-associated decline of immune function. Trends Immunol. 30: 306–312. 12. Lin, S. J., C. D. Peacock, K. Bahl, and R. M. Welsh. 2007. Programmed death-1 37. Pawelec, G. 2006. Immunity and ageing in man. Exp. Gerontol. 41: 1239–1242. (PD-1) defines a transient and dysfunctional oligoclonal T cell population in 38. Keir, M. E., M. J. Butte, G. J. Freeman, and A. H. Sharpe. 2008. PD-1 and its acute homeostatic proliferation. J. Exp. Med. 204: 2321–2333. ligands in tolerance and immunity. Annu. Rev. Immunol. 26: 677–704. 13. Goronzy, J. J., W. W. Lee, and C. M. Weyand. 2007. Aging and T-cell diversity. 39. Wherry, E. J. 2011. T cell exhaustion. Nat. Immunol. 12: 492–499. Exp. Gerontol. 42: 400–406. 40. Clambey, E. T., J. White, J. W. Kappler, and P. Marrack. 2008. Identification of Downloaded from 14. Cicin-Sain, L., S. Smyk-Pearson, N. Currier, L. Byrd, C. Koudelka, T. Robinson, two major types of age-associated CD8 clonal expansions with highly divergent G. Swarbrick, S. Tackitt, A. Legasse, M. Fischer, et al. 2010. Loss of naive properties. Proc. Natl. Acad. Sci. USA 105: 12997–13002. T cells and repertoire constriction predict poor response to vaccination in old 41. Obar, J. J., K. M. Khanna, and L. Lefranc¸ois. 2008. Endogenous naive CD8+ primates. J. Immunol. 184: 6739–6745. T cell precursor frequency regulates primary and memory responses to infection. 15. Yager, E. J., M. Ahmed, K. Lanzer, T. D. Randall, D. L. Woodland, and Immunity 28: 859–869. M. A. Blackman. 2008. Age-associated decline in T cell repertoire diversity 42. Blattman, J. N., R. Antia, D. J. Sourdive, X. Wang, S. M. Kaech, K. Murali- leads to holes in the repertoire and impaired immunity to influenza virus. J. Exp. Krishna, J. D. Altman, and R. Ahmed. 2002. Estimating the precursor frequency

Med. 205: 711–723. of naive antigen-specific CD8 T cells. J. Exp. Med. 195: 657–664. http://www.jimmunol.org/ 16. Ahmed, M., K. G. Lanzer, E. J. Yager, P. S. Adams, L. L. Johnson, and 43. Haynes, L., P. J. Linton, S. M. Eaton, S. L. Tonkonogy, and S. L. Swain. 1999. M. A. Blackman. 2009. Clonal expansions and loss of receptor diversity in the Interleukin 2, but not other common gamma chain-binding cytokines, can re- naive CD8 T cell repertoire of aged mice. J. Immunol. 182: 784–792. verse the defect in generation of CD4 effector T cells from naive T cells of aged 17. Clambey, E. T., J. W. Kappler, and P. Marrack. 2007. CD8 T cell clonal mice. J. Exp. Med. 190: 1013–1024. expansions & aging: a heterogeneous phenomenon with a common outcome. 44. Kao, C., K. J. Oestreich, M. A. Paley, A. Crawford, J. M. Angelosanto, Exp. Gerontol. 42: 407–411. M. A. Ali, A. M. Intlekofer, J. M. Boss, S. L. Reiner, A. S. Weinmann, and 18. Nikolich-Zugich, J. 2008. Ageing and life-long maintenance of T-cell subsets in E. J. Wherry. 2011. Transcription factor T-bet represses expression of the in- the face of latent persistent infections. Nat. Rev. Immunol. 8: 512–522. hibitory receptor PD-1 and sustains virus-specific CD8+ T cell responses during 19. Channappanavar, R., B. S. Twardy, P. Krishna, and S. Suvas. 2009. Advancing chronic infection. Nat. Immunol. 12: 663–671. age leads to predominance of inhibitory receptor expressing CD4 T cells. Mech. 45. Lages, C. S., I. Suffia, P. A. Velilla, B. Huang, G. Warshaw, D. A. Hildeman, Ageing Dev. 130: 709–712. Y. Belkaid, and C. Chougnet. 2008. Functional regulatory T cells accumulate in

20. Shimada, Y., M. Hayashi, Y. Nagasaka, Y. Ohno-Iwashita, and M. Inomata. aged hosts and promote chronic infectious disease reactivation. J. Immunol. 181: by guest on October 1, 2021 2009. Age-associated up-regulation of a negative co-stimulatory receptor PD-1 1835–1848. in mouse CD4+ T cells. Exp. Gerontol. 44: 517–522. 46. Taub, D. D., and D. L. Longo. 2005. Insights into thymic aging and regeneration. 21. Lages, C. S., I. Lewkowich, A. Sproles, M. Wills-Karp, and C. Chougnet. 2010. Immunol. Rev. 205: 72–93. Partial restoration of T-cell function in aged mice by in vitro blockade of the PD- 47. Hale, J. S., T. E. Boursalian, G. L. Turk, and P. J. Fink. 2006. Thymic output in 1/PD-L1 pathway. Aging Cell 9: 785–798. aged mice. Proc. Natl. Acad. Sci. USA 103: 8447–8452. 22. Tsukamoto, H., K. Clise-Dwyer, G. E. Huston, D. K. Duso, A. L. Buck, 48. Murali-Krishna, K., and R. Ahmed. 2000. Cutting edge: naive T cells L. L. Johnson, L. Haynes, and S. L. Swain. 2009. Age-associated increase in masquerading as memory cells. J. Immunol. 165: 1733–1737. lifespan of naive CD4 T cells contributes to T-cell homeostasis but facilitates 49. Huang, T., B. Wei, P. Velazquez, J. Borneman, and J. Braun. 2005. Commensal development of functional defects. Proc. Natl. Acad. Sci. USA 106: 18333–18338. microbiota alter the abundance and TCR responsiveness of splenic naive CD4+ 23. Wherry, E. J., J. N. Blattman, K. Murali-Krishna, R. van der Most, and T lymphocytes. Clin. Immunol. 117: 221–230. R. Ahmed. 2003. Viral persistence alters CD8 T-cell immunodominance and 50. Freeman, G. J., E. J. Wherry, R. Ahmed, and A. H. Sharpe. 2006. Reinvigorating tissue distribution and results in distinct stages of functional impairment. J. Virol. exhausted HIV-specific T cells via PD-1–PD-1 ligand blockade. J. Exp. Med. 77: 4911–4927. 203: 2223–2227. 24. Decman, V., B. J. Laidlaw, L. J. Dimenna, S. Abdulla, K. Mozdzanowska, 51. Sharpe, A. H., E. J. Wherry, R. Ahmed, and G. J. Freeman. 2007. The function of J. Erikson, H. C. Ertl, and E. J. Wherry. 2010. Cell-intrinsic defects in the programmed cell death 1 and its ligands in regulating autoimmunity and in- proliferative response of antiviral memory CD8 T cells in aged mice upon fection. Nat. Immunol. 8: 239–245. secondary infection. J. Immunol. 184: 5151–5159. 52. Shimatani, K., Y. Nakashima, M. Hattori, Y. Hamazaki, and N. Minato. 2009. 25. Moon, J. J., H. H. Chu, M. Pepper, S. J. McSorley, S. C. Jameson, R. M. Kedl, PD-1+ memory phenotype CD4+ T cells expressing C/EBPa underlie T cell and M. K. Jenkins. 2007. Naive CD4+ T cell frequency varies for different immunodepression in senescence and leukemia. Proc. Natl. Acad. Sci. USA epitopes and predicts repertoire diversity and response magnitude. Immunity 27: 106: 15807–15812. 203–213. 53. Hamilton, S. E., and S. C. Jameson. 2008. The nature of the lymphopenic en- 26. Wherry, E. J., S. J. Ha, S. M. Kaech, W. N. Haining, S. Sarkar, V. Kalia, vironment dictates protective function of homeostatic-memory CD8+ T cells. S. Subramaniam, J. N. Blattman, D. L. Barber, and R. Ahmed. 2007. Molecular Proc. Natl. Acad. Sci. USA 105: 18484–18489. signature of CD8+ T cell exhaustion during chronic viral infection. Immunity 27: 54. Quigley, M., F. Pereyra, B. Nilsson, F. Porichis, C. Fonseca, Q. Eichbaum, 670–684. B. Julg, J. L. Jesneck, K. Brosnahan, S. Imam, et al. 2010. Transcriptional 27. Irizarry, R. A., S. L. Ooi, Z. Wu, and J. D. Boeke. 2003. Use of mixture models analysis of HIV-specific CD8+ T cells shows that PD-1 inhibits T cell function by in a microarray-based screening procedure for detecting differentially repre- upregulating BATF. Nat. Med. 16: 1147–1151. sented yeast mutants. Stat. Appl. Genet. Mol. Biol. 2: Article 1. 55. Badovinac, V. P., J. S. Haring, and J. T. Harty. 2007. Initial T cell receptor 28. Reich, M., T. Liefeld, J. Gould, J. Lerner, P. Tamayo, and J. P. Mesirov. 2006. transgenic cell precursor frequency dictates critical aspects of the CD8+ Tcell GenePattern 2.0. Nat. Genet. 38: 500–501. response to infection. Immunity 26: 827–841. SUPPLEMENTAL FIGURE 1. Analysis of CD8 T cells from uninfected mice. A

shows numbers of CD8 T cells in the spleens of uninfected mice at age of 4, 8, 12 and 18 months

of age. CD8 T cells were isolated from the spleens of uninfected mice at 4, 8, 12 and 18 months

of age and analyzed for the expression of KLRG1 and Tbet. The percentages of these markers on

CD44Hi or total CD8 T cell population are shown in B and C, respectively. Additionally, CD8 T

cells were isolated from the blood (D and F) and spleens (E) of uninfected mice at 4, 8, 12 and

18 months of age and stained for expression of inhibitory receptors PD1, LAG3, 2B4 and

CD160. The graphs in D show the percentages of PD1, LAG3 and 2B4 on CD44Hi CD8 T cells

from blood. E and F show mean fluorescence values for individual stainings on CD44Hi CD8 T cells in spleen and blood, respectively. Data are representative of 2 independent experiments including 12-20 mice/experiment. * = 0.05 > P > 0.01, ** = 0.01 > P > 0.001, *** = P < 0.001 by unpaired two-tailed t test. G and H show analysis of CD8 T cells from young and aged mice obtained from National Institutes of Aging (NIA). H shows percentages of PD1 positive cells on the CD44Hi population of CD8 T cells from young (Mean ± SEM = 1.74 ± 0.19 %) and aged

(Mean ± SEM = 38.86 ± 5.76%) mice obtained from NIA. I shows PD1 expression on CD8 T

cells from uninfected aged mice obtained from different vendors. CD8 T cells were either isolated from the spleens of uninfected aged mice (18 months of age) obtained from NIA or from

the spleens of mice that were obtained from Taconic and then aged in the Wistar Institute aging

colony.

SUPPLEMENTAL FIGURE 2. P14 (DbGP33-specific) and OT1 (KbOVA-specific)

CD8 T cells were purified from uninfected young and aged P14 mice and uninfected young OT1 Rag2-/- mice, respectively. Graphs show PD1 expression on CD8 T cells from spleens of P14 and

OT1 Rag2-/- mice.

SUPPLEMENTAL TABLE 1. Signatures of young versus aged CD44Hi CD8 T cells by

Class Neighbors analysis. CD8 T cells were isolated from spleens of uninfected young (4 months

old) and aged (20-21 months old) mice and sorted based on CD44 expression. After RNA isolation samples were hybridized to the Affymetrix GeneChip Mouse Exon 1.0 ST arrays. Fold changes for a panel of genes that are differentially expressed in CD44Hi CD8 T cells from uninfected young and aged mice are shown in the table (n=4 for each set). Fold changes in red indicate higher expression in aged CD44Hi CD8 T cells. Fold changes in green indicate higher

expression in young CD44Hi CD8 T cells.

SUPPLEMENTAL TABLE 2. Differentially expressed genes of young versus aged

CD44Hi CD8 T cells with q-value < 0.05. CD8 T cells were isolated from spleens of uninfected

young (4 months old) and aged (20-21 months old) mice and sorted based on CD44 expression.

After RNA isolation samples were hybridized to the Affymetrix GeneChip Mouse Exon 1.0 ST

arrays. Table shows the analysis for a panel of genes that are differentially expressed in CD44Hi

CD8 T cells from uninfected young and aged mice (n=4 for each set). Fold changes in red indicate higher expression in aged CD44Hi CD8 T cells. Fold changes in green indicate higher

expression in young CD44Hi CD8 T cells.

Decman et al, Supplemental Figure 1

A. B. CD44Hi CD8 T cells/Spleen C. Total CD8 T cells/Spleen * * * ** * * * ** * * * 7 * * * * 2.0x10 40 76 40 76

1.6x107 Hi Hi 30 57 Hi 30 57 Hi 1.2x107 20 38 20 38 8.0x106

% Tbet % Tbet 4.0x106 % KLRG1 10 19 10 19

% KLRG1

0 0 0 0 0 CD8T cells/Spleen Absolute number of 4mo 8mo 12mo 18mo 4mo 8mo12mo 18mo 4mo 8mo 12mo 18mo 4mo 8mo12mo 18mo 4mo 8mo 12mo 18mo

D. CD44Hi CD8 T cells/106 PBMC

** ** ***** ***** ** ***** **** *** ** * ** 50 35 35

40 28 28 Hi Hi Hi 30 21 21

20 14 14 % PD1 % 2B41 10 % LAG3 7 7

0 0 0 4mo 8mo12mo18mo 4mo 8mo12mo18mo 4mo 8mo12mo18mo

E. CD44Hi CD8 T cells/Spleen

**** ** *** *** *** * ** *** 700 160 200 180

525 120 150 120 350 80 100 60 MFI 2B4 MFI PD1 175 50

MFI LAG3 40 MFI CD160

0 0 0 0 4mo 8mo12mo18mo 4mo 8mo12mo18mo 4mo 8mo12mo18mo 4mo 8mo12mo18mo

F. CD44Hi CD8 T cells/106 PBMC G. I. Uninfected aged * 20 mice (18 mo) *** *** ** ** 105 *** *** 15 23 51 * * * 4 180 100 100 10 10 Taconic 103

75 75 Spleen 120 2 5 10 0

50 50 cells/ T CD8 % 11 0102 103 104 105 60 0 MFI PD1 25 MFI 2B4 25 104 MFI LAG3 YA 48 18 0 0 0 H. 103 4mo 8mo12mo18mo 4mo 8mo12mo18mo 4mo 8mo12mo18mo 2 80 *** NIA 10 Hi

101 60 11 0 10 100 101 102 103 104

40 4 of CD44 10

Hi 24 42 20 103 CD8 T cells

NIA 102 0 % PD1 YA 101 24 100 CD44 100 101 102 103 104 PD1 Decman et al, Supplemental Figure 2 A. B. 105 92.6 0.5 105 99.3 0.7

104 104 Young Young 3 (4 mo) 10 (7 mo) 103

2 10 102

0 0

0.4 b K OVA K 2 3 4 5 010 10 10 10 0102 103 104 105 PD1 105 26.1 1.1

104 Aged (21 mo) 103

102

0

b 39.2 D GP33 D 0102 103 104 105 PD1 Supplemental Table 1: Young and aged CD8 T cell signatures by ClassNeighbors algorithm. Red fold changes have higher expression in young cells. Entrez ID Gene Symbol Gene Title Fold Change 193740 Hspa1a heat shock protein 1A -3.38 11931 Atp1b1 ATPase, Na+/K+ transporting, beta 1 polypeptide -3.02 16627 Klra1 killer cell lectin-like receptor, subfamily A, member 1 -3.02 109711 Actn1 actinin, alpha 1 -2.99 17863 Myb myeloblastosis oncogene -2.95 76747 Dapl1 death associated protein-like 1 -2.72 381417 RP23-357I14.1 predicted gene 14085 -2.71 93971 Klra19 killer cell lectin-like receptor, subfamily A, member 19 -2.68 16637 Klra6 killer cell lectin-like receptor, subfamily A, member 6 -2.52 17872 Myd116 protein phosphatase 1, regulatory (inhibitor) subunit 15A -2.35 19038 Ppic peptidylprolyl isomerase C -2.28 15903 Id3 inhibitor of DNA binding 3 -2.25 68259 Ift80 intraflagellar transport 80 homolog (Chlamydomonas) -2.25 27424 Klra16 killer cell lectin-like receptor, subfamily A, member 16 -2.21 24064 Spry2 sprouty homolog 2 (Drosophila) -2.11 217316 Slc16a5 solute carrier family 16 (monocarboxylic acid transporters), member-2.09 5 12775 Ccr7 chemokine (C-C motif) receptor 7 -2.05 14281 Fos FBJ osteosarcoma oncogene -2.04 93970 Klra18 killer cell lectin-like receptor, subfamily A, member 18 -1.97 16431 Itm2a integral membrane protein 2A -1.97 19848 Rnu2 U2 small nuclear RNA -1.95 20440 St6gal1 beta galactoside alpha 2,6 sialyltransferase 1 -1.94 258861 Olfr763 olfactory receptor 763 -1.93 11910 Atf3 activating transcription factor 3 -1.92 16476 Jun Jun oncogene -1.86 20620 Plk2 polo-like kinase 2 (Drosophila) -1.85 19252 Dusp1 dual specificity phosphatase 1 -1.77 654450 Klra33 killer cell lectin-like receptor, subfamily A, member 33 -1.76 16477 Junb Jun-B oncogene -1.74 20449 St8sia1 ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 1 -1.73 277146 Gm701 T cell receptor alpha variable 6D-4 -1.71 11852 Rhob ras homolog gene family, member B -1.70 13197 Gadd45a growth arrest and DNA-damage-inducible 45 alpha -1.69 15370 Nr4a1 nuclear receptor subfamily 4, group A, member 1 -1.68 63872 Zfp296 zinc finger protein 296 -1.68 56508 Rapgef4 Rap guanine nucleotide exchange factor (GEF) 4 -1.66 16190 Il4ra interleukin 4 receptor, alpha -1.66 14282 Fosb FBJ osteosarcoma oncogene B -1.62 52377 Rcn3 reticulocalbin 3, EF-hand calcium binding domain -1.62 245381 Ankrd58 ankyrin repeat domain 58 -1.62 11898 Ass1 argininosuccinate synthetase 1 -1.61 11641 Akap2 A kinase (PRKA) anchor protein 2 -1.60 18613 Pecam1 platelet/endothelial cell adhesion molecule 1 -1.60 16600 Klf4 Kruppel-like factor 4 (gut) -1.58 76263 Gstk1 glutathione S-transferase kappa 1 -1.57 109978 Art4 ADP-ribosyltransferase 4 -1.55 68393 Mogat1 monoacylglycerol O-acyltransferase 1 -1.54 15936 Ier2 immediate early response 2 -1.52 105827 Amigo2 adhesion molecule with Ig like domain 2 -1.51 243461 U29423 cDNA sequence U29423 -1.49 16635 Klra4 killer cell lectin-like receptor, subfamily A, member 4 -1.48 331524 Xkrx X Kell blood group precursor related X linked -1.46 258998 Olfr177 olfactory receptor 177 -1.45 259040 Olfr1416 olfactory receptor 1416 -1.43 56390 Sssca1 Sjogren's syndrome/scleroderma autoantigen 1 homolog (human)-1.43 244334 Defb8 defensin beta 8 -1.43 50994 Mtag2 metastasis associated gene 2 -1.42 226098 Hectd2 HECT domain containing 2 -1.41 383243 Olfr128 olfactory receptor 128 -1.40 258304 Olfr498 olfactory receptor 498 -1.39 692166 Igkv15-97 --- -1.39 258344 Olfr390 olfactory receptor 390 -1.38 16701 Krtap6-2 keratin associated protein 6-2 -1.37 258331 Olfr1330 olfactory receptor 1330 -1.35 116913 Tpbpb trophoblast specific protein beta -1.34 258553 Olfr872 olfactory receptor 872 -1.33 64059 Oxct2a 3-oxoacid CoA transferase 2A -1.33 258180 Olfr699 olfactory receptor 699 -1.33 266615 Svet1 subventricular expressed transcript 1 -1.32 258955 Olfr531 olfactory receptor 531 -1.30 258724 Olfr784 olfactory receptor 784 -1.23 24088 Tlr2 toll-like receptor 2 -1.22 11731 Ang2 angiogenin, ribonuclease A family, member 2 -1.20 258898 Olfr1203 olfactory receptor 1205 -1.17 258564 Olfr1015 olfactory receptor 1015 -1.14 259162 Olfr427 olfactory receptor 427 1.25 259143 Olfr503 olfactory receptor 503 1.30 258566 Olfr1002 olfactory receptor 1002 1.40 404321 Olfr776 olfactory receptor 776 1.41 258584 Olfr1101 olfactory receptor 1101 1.45 22236 Ugt1a2 UDP glucuronosyltransferase 1 family, polypeptide A2 1.46 73149 Clec4a3 C-type lectin domain family 4, member a3 1.50 258462 Olfr1392 olfactory receptor 1392 1.53 14714 Gnrh1 gonadotropin releasing hormone 1 1.56 404239 Mrgprb5 MAS-related GPR, member B5 1.56 50928 Klrg1 killer cell lectin-like receptor subfamily G, member 1 1.59 54486 Ptgds2 hematopoietic prostaglandin D synthase 1.60 246256 Fcgr4 Fc receptor, IgG, low affinity IV 1.61 233806 Tmem159 transmembrane protein 159 1.61 319579 Defb20 defensin beta 20 1.62 71907 Serpina9 serine (or cysteine) peptidase inhibitor, clade A (alpha-1 antiproteinase,1.64 antitrypsin), member 9 17996 Neb nebulin 1.64 171230 V1re7 vomeronasal 1 receptor 231 1.65 70809 Clec2g C-type lectin domain family 2, member g 1.67 258177 Olfr1222 olfactory receptor 1222 1.67 227394 Slco4c1 solute carrier organic anion transporter family, member 4C1 1.71 258977 Olfr1273 olfactory receptor 1273, pseudogene 1.72 240873 Tnfsf18 tumor necrosis factor (ligand) superfamily, member 18 1.75 207742 Rnf43 ring finger protein 43 1.76 258987 Olfr1270 olfactory receptor 1270 1.76 72747 Ttc39c tetratricopeptide repeat domain 39C 1.77 56375 B4galt4 UDP-Gal:betaGlcNAc beta 1,4-galactosyltransferase, polypeptide 41.80 81006 Gpr63 G protein-coupled receptor 63 1.81 17952 Birc1f NLR family, apoptosis inhibitory protein 6 1.85 17951 Naip5 NLR family, apoptosis inhibitory protein 5 1.87 12771 Ccr3 chemokine (C-C motif) receptor 3 1.87 435921 Clec2f C-type lectin domain family 2, member g 1.88 170741 Pilrb1 paired immunoglobin-like type 2 receptor beta 1 1.88 240427 Setbp1 SET binding protein 1 1.89 18566 Pdcd1 programmed cell death 1 1.91 433070 Vmn2r96 vomeronasal 2, receptor 96 1.93 435207 Taar7f trace amine-associated receptor 7F 1.94 83925 Trps1 trichorhinophalangeal syndrome I (human) 1.97 620246 Gpr52 G protein-coupled receptor 52 1.98 71720 Osbpl3 oxysterol binding protein-like 3 2.01 238393 Serpina3f serine (or cysteine) peptidase inhibitor, clade A, member 3F 2.01 19249 Ptpn13 protein tyrosine phosphatase, non-receptor type 13 2.01 11871 Art2a ADP-ribosyltransferase 2a, pseudogene 2.02 12494 Cd38 CD38 antigen 2.05 13074 Cyp17a1 cytochrome P450, family 17, subfamily a, polypeptide 1 2.08 12142 Prdm1 PR domain containing 1, with ZNF domain 2.12 385253 Gm1524 predicted gene 1524 2.13 67790 Rab39b RAB39B, member RAS oncogene family 2.18 16017 Ighg1 immunoglobulin heavy constant gamma 1 (G1m marker) 2.23 14103 Fasl Fas ligand (TNF superfamily, member 6) 2.33 56078 Car5b carbonic anhydrase 5b, mitochondrial 2.35 56461 Kcnip3 Kv channel interacting protein 3, calsenilin 2.40 16153 Il10 interleukin 10 2.42 22138 Ttn titin 2.45 26432 Plod2 procollagen lysine, 2-oxoglutarate 5-dioxygenase 2 2.46 12772 Ccr2 chemokine (C-C motif) receptor 2 2.48 433904 Ociad2 OCIA domain containing 2 2.62 14939 Gzmb granzyme B 2.62 12363 Casp4 caspase 4, apoptosis-related cysteine peptidase 2.63 13837 Epha3 Eph receptor A3 2.66 15937 Ier3 immediate early response 3 2.74 20302 Ccl3 chemokine (C-C motif) ligand 3 2.75 12362 Casp1 caspase 1 2.77 232408 Klrb1f killer cell lectin-like receptor subfamily B member 1F 2.88 16768 Lag3 lymphocyte-activation gene 3 2.97 1E+08 Tigit T cell immunoreceptor with Ig and ITIM domains 3.21 20715 Serpina3g serine (or cysteine) peptidase inhibitor, clade A, member 3G 3.35 667838 Naip3 NLR family, apoptosis inhibitory protein 3 3.49 114564 Csprs component of Sp100-rs 3.57 22164 Tnfsf4 tumor necrosis factor (ligand) superfamily, member 4 3.96 620017 Gm1409 immunoglobulin kappa variable 5-39 4.59 14945 Gzmk granzyme K 4.95 71690 Esm1 endothelial cell-specific molecule 1 5.02 17132 Maf avian musculoaponeurotic fibrosarcoma (v-maf) AS42 oncogene homolog5.26 18619 Penk preproenkephalin 6.88 Supplemental Table 2: Genes differentially expressed by a q-value threshold of 0.05. Green fold changes are higher in aged cells. EntrezGeneGene SymbolGene Title Fold Change q-value 11931 Atp1b1 ATPase, Na+/K+ transporting, beta 1 polypeptide -3.02 0.0319 109711 Actn1 actinin, alpha 1 -2.99 0.0312 17863 Myb myeloblastosis oncogene -2.95 0.0286 93971 Klra19 killer cell lectin-like receptor, subfamily A, member 6 /// killer cell -2.68lectin-like0.0497 receptor, subfamily A, member 19 19038 Ppic peptidylprolyl isomerase C -2.28 0.0440 68259 Ift80 oxysterol binding protein-like 9 /// intraflagellar transport 80 homolog-2.25 (Chlamydomonas)0.0359 20449 St8sia1 ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 1 -1.73 0.0347 16195 Il6st interleukin 6 signal transducer -1.67 0.0381 244233 Cd163l1 CD163 molecule-like 1 -1.65 0.0416 228765 Sdcbp2 syndecan binding protein (syntenin) 2 1.50 0.0469 114774 Pawr PRKC, apoptosis, WT1, regulator 1.54 0.0440 338349 Cntln centlein, centrosomal protein 1.55 0.0474 12367 Casp3 caspase 3 1.56 0.0381 12571 Cdk6 cyclin-dependent kinase 6 1.56 0.0498 52589 Ncald neurocalcin delta 1.59 0.0488 16574 Kif5c kinesin family member 5C /// predicted gene 10240 1.60 0.0469 56434 Tspan3 tetraspanin 3 1.62 0.0319 67492 Anubl1 AN1, ubiquitin-like, homolog (Xenopus laevis) 1.69 0.0448 68526 Gpr155 G protein-coupled receptor 155 1.70 0.0381 18671 Abcb1a ATP-binding cassette, sub-family B (MDR/TAP), member 1A /// ATP-binding1.73 0.0356 cassette, sub-family B (MDR/TAP), member 1B 19651 Rbl2 retinoblastoma-like 2 1.73 0.0406 239849 Cd200r4 CD200 receptor 1 /// CD200 receptor 4 1.74 0.0438 11492 Adam19 a disintegrin and metallopeptidase domain 19 (meltrin beta) 1.74 0.0286 74732 Stx11 syntaxin 11 1.75 0.0319 240873 Tnfsf18 tumor necrosis factor (ligand) superfamily, member 18 1.75 0.0356 207742 Rnf43 ring finger protein 43 1.76 0.0455 72747 Ttc39c tetratricopeptide repeat domain 39C 1.77 0.0469 381924 Itgad integrin, alpha D 1.78 0.0369 56375 B4galt4 UDP-Gal:betaGlcNAc beta 1,4-galactosyltransferase, polypeptide 4 1.80 0.0474 229521 Syt11 synaptotagmin XI 1.85 0.0286 258508 Olfr99 olfactory receptor 99 1.86 0.0359 17951 Naip5 NLR family, apoptosis inhibitory protein 5 /// NLR family, apoptosis1.87 inhibitory0.0409 protein 6 /// NLR family, apoptosis inhibitory protein 7 16169 Il15ra interleukin 15 receptor, alpha chain 1.88 0.0337 100340 Smpdl3b sphingomyelin phosphodiesterase, acid-like 3B 1.89 0.0440 12495 Entpd1 tectonic family member 3 /// ectonucleoside triphosphate diphosphohydrolase1.91 0.0316 1 18566 Pdcd1 programmed cell death 1 1.91 0.0319 22038 Plscr1 phospholipid scramblase 1 1.95 0.0312 83925 Trps1 trichorhinophalangeal syndrome I (human) 1.97 0.0381 620246 Gpr52 G protein-coupled receptor 52 1.98 0.0448 12774 Ccr5 chemokine (C-C motif) receptor 5 2.01 0.0319 71720 Osbpl3 oxysterol binding protein-like 3 2.01 0.0286 238393 Serpina3f serine (or cysteine) peptidase inhibitor, clade A, member 3F 2.01 0.0286 19249 Ptpn13 protein tyrosine phosphatase, non-receptor type 13 2.01 0.0286 11871 Art2a ADP-ribosyltransferase 2b /// ADP-ribosyltransferase 2a 2.02 0.0497 12494 Cd38 CD38 antigen 2.05 0.0337 252838 Tox thymocyte selection-associated high mobility group box 2.06 0.0286 70086 Cysltr2 cysteinyl leukotriene receptor 2 2.08 0.0347 13074 Cyp17a1 cytochrome P450, family 17, subfamily a, polypeptide 1 2.08 0.0406 12142 Prdm1 PR domain containing 1, with ZNF domain 2.12 0.0352 67790 Rab39b RAB39B, member RAS oncogene family 2.18 0.0286 14103 Fasl Fas ligand (TNF superfamily, member 6) 2.33 0.0400 56078 Car5b carbonic anhydrase 5b, mitochondrial 2.35 0.0286 56461 Kcnip3 Kv channel interacting protein 3, calsenilin 2.40 0.0316 16153 Il10 interleukin 10 /// SMT3 suppressor of mif two 3 homolog 2 (yeast) 2.42/// microRNA0.0319 684-1 /// microRNA 684-2 26432 Plod2 procollagen lysine, 2-oxoglutarate 5-dioxygenase 2 2.46 0.0286 12772 Ccr2 chemokine (C-C motif) receptor 2 2.48 0.0286 433904 Ociad2 OCIA domain containing 2 2.62 0.0316 14939 Gzmb granzyme B 2.62 0.0337 12363 Casp4 caspase 4, apoptosis-related cysteine peptidase 2.63 0.0316 13837 Epha3 Eph receptor A3 2.66 0.0286 15937 Ier3 immediate early response 3 2.74 0.0316 20302 Ccl3 chemokine (C-C motif) ligand 3 2.75 0.0469 12362 Casp1 caspase 1 2.77 0.0347 232408 Klrb1f killer cell lectin-like receptor subfamily B member 1F 2.88 0.0333 16768 Lag3 lymphocyte-activation gene 3 2.97 0.0316 1E+08 Tigit T cell immunoreceptor with Ig and ITIM domains 3.21 0.0286 20715 Serpina3g serine (or cysteine) peptidase inhibitor, clade A, member 3G 3.35 0.0286 667838 Naip3 NLR family, apoptosis inhibitory protein 3 /// T-cell receptor gamma,3.49 variable0.0286 2 114564 Csprs component of Sp100-rs 3.57 0.0286 22164 Tnfsf4 tumor necrosis factor (ligand) superfamily, member 4 3.96 0.0286 14945 Gzmk granzyme K 4.95 0.0286 71690 Esm1 endothelial cell-specific molecule 1 5.02 0.0333 17132 Maf avian musculoaponeurotic fibrosarcoma (v-maf) AS42 oncogene homolog5.26 0.0286 18619 Penk preproenkephalin 6.88 0.0286