Single-cell -expression profiling reveals qualitatively distinct CD8 T cells elicited by different gene-based vaccines

Lukas Flatza,b,1, Rahul Roychoudhuria,1, Mitsuo Hondaa,1, Abdelali Filali-Mouhimc, Jean-Philippe Gouletc, Nadia Kettafc, Min Lind, Mario Roederera, Elias K. Haddadc,e,Rafick P. Sékalyc,e,2, and Gary J. Nabela,2,3

aVaccine Research Center, National Institute for Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892-3005; bInstitute for Infectious Diseases, University of Bern, CH-3010 Bern, Switzerland; cLaboratoire d’immunologie, Centre de Recherche du Centre Hospitalier de l’Université de Montréal, Saint-Luc, Montréal, QC, Canada H2X 1P1; dFluidigm Corporation, San Francisco, CA 94080; and eVaccine and Gene Therapy Institute, Port St. Lucie, FL 34987

Edited* by Ralph M. Steinman, The Rockefeller University, New York, NY, and approved February 4, 2011 (received for review September 3, 2010)

CD8 T cells play a key role in mediating protective immunity against liferative and cytotoxic potential (14). Similar T-cell subsets have selected pathogens after vaccination. Understanding the mecha- also been defined in humans (15), indicating that these subsets are nism of this protection is dependent upon definition of the hetero- evolutionarily conserved. However, the presence of such cell types fi geneity and complexity of cellular immune responses generated by does not predict vaccine ef cacy with consistency. Some studies different vaccines. Here, we identify previously unrecognized subsets have implicated a protective role of central memory (CM) T cells, of CD8 T cells based upon analysis of gene-expression patterns within although others have suggested a protective role for effector single cells and show that they are differentially induced by different memory (EM) cells (6, 16). These studies highlight the need to understand and discriminate among differences in CD8 T-cell vaccines. Three prime-boost vector combinations encoding HIV Env fi responses generated by different vaccines; we have therefore stimulated antigen-speci c CD8 T-cell populations of similar magni- compared vaccine-elicited cellular immune responses elicited by tude, phenotype, and functionality. Remarkably, however, analysis three different gene-based vaccine regimens through transcrip- fi of single-cell gene-expression pro les enabled discrimination of tional profiling and the application of a technique that has allowed a majority of central memory (CM) and effector memory (EM) CD8 us to measure 96 gene-expression signals from single immune T cells elicited by the three vaccines. Subsets of T cells could be de- cells. We have identified qualitative differences in CD8 T cells fined based on their expression of Eomes, Cxcr3,andCcr7,orKlrk1, elicited by the three vaccine regimens that cannot be detected by Klrg1,andCcr5 in CM and EM cells, respectively. Of CM cells elicited conventional techniques, and that allow individual CD8 T cells by DNA prime-recombinant adenoviral (rAd) boost vectors, 67% elicited by the three immunizations to be readily distinguished. were Eomes− Ccr7+ Cxcr3−, in contrast to only 7% and 2% stimulated Furthermore, we have identified unique subsets of CD8 T cells by rAd5-rAd5 or rAd-LCMV, respectively. Of EM cells elicited by DNA- based upon analysis of of Eomes, Cxcr3, and Ccr7, − − − rAd, 74% were Klrk1 Klrg1 Ccr5 compared with only 26% and or Klrk1, Klrg1, and Ccr5 in CM and EM cells, respectively, that 20% for rAd5-rAd5 or rAd5-LCMV. Definition by single-cell gene pro- were differentially induced by the three vaccines. filing of specific CM and EM CD8 T-cell subsets that are differentially induced by different gene-based vaccines will facilitate the design Results and evaluation of vaccines, as well as enable our understanding Three Different Prime-Boost Vaccine Combinations Stimulate CD8 of mechanisms of protective immunity. T-Cell Responses of Similar Magnitude and Cytokine Functionality. Plasmid DNA, recombinant replication-defective adenovirus and lymphocyte subsets | microarray | immune differentiation lymphocytic choriomeningitis virus (rAd5 and rLCMV, respec- tively), encompassing the same insert derived from a recombinant HIV-1 Env gene (17), were used to immunize BALB/c mice in major aim of vaccines designed to elicit immunity against different prime-boost combinations. The three different prime- A chronic viral pathogens is the generation of pathogen- fi fi boost immunization regimens (DNA-rAd5, rAd5-rAd5, and rAd5- speci c cytotoxic T cells of suf cient magnitude and quality to rLCMV) elicited T-cell responses with mean frequencies ranging mediate protection against disease (1, 2). Cytotoxic T cells can from 20.1to 32.6% (Fig. 1A), as measured by binding to an antigen- control viral load in various models of simian immunodeficiency fi fi fi – speci c tetramer H2-Dd/PA9 (18). To con rm the speci city of virus (SIV) infection in rhesus macaques (3 7). However, the tetramer binding and therefore the accuracy of subsequent tetra- quantitative and qualitative features of vaccine-elicited T-cell fi mer-sorting experiments, we compared the frequency of tetramer- responses that mediate protection are yet to be de ned (8, 9). binding CD8+ T cells in rAd5-rAd5 immunized animals with that in Recently, analysis of T cells by intracellular cytokine staining has fi unimmunized animals (Fig. S1 A and B). The mean percentage of allowed greater precision in measuring antigen-speci c responses, CD8 T cells binding the tetramer was only 0.027% [95% CI (con- facilitating the functional profiling of antigen-specificcellsin fi – γ α dence interval) 0.007 0.047] in naive mice versus 38.3% (95% CI vitro. Simultaneous measurement of IFN- ,TNF- , and IL-2 in 29.29–47.31) in vaccinated mice, predicting a tetramer-sort purity vaccine-elicited CD4 T cells has shown that the proportion of cells secreting multiple cytokines, polyfunctional T cells, correlated with protection in a mouse model of leishmania infection (10). However, T-cell polyfunctionality was not a major predictor of Author contributions: L.F., R.R., M.H., E.K.H., R.P.S., and G.J.N. designed research; L.F., R.R., M.H., N.K., M.L., and E.K.H. performed research; L.F., R.R., M.H., A.F.-M., J.-P.G., outcome in a number of SIV protection studies in rhesus mac- M.R.,E.K.H.,R.P.S.,andG.J.N.analyzeddata;andL.F.,R.R.,M.R.,E.K.H.,R.P.S.,and aques (6, 11) and in other vaccine-induced protection models (12). G.J.N. wrote the paper. The generation of synthetic class I MHC:peptide tetramers Conflict of interest statement: M.L. is a research technician for Fluidigm Corporation. has allowed the identification of vaccine elicited antigen-specific T cells responsive to vaccine epitopes (13). Analysis of surface *This Direct Submission article had a prearranged editor. 1 , such as the interleukin-7 receptor (IL7R) and L- L.F., R.R., and M.H. contributed equally to this work. selectin (CD62L), in mice has allowed the definition of three 2R.P.S. and G.J.N. contributed equally to this work. differentiation states, referred to as effector (IL7R low, CD62L 3To whom correspondence should be addressed. E-mail: [email protected]. + − low), effector memory (IL7R , CD62L ), and central memory This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. + + (IL7R , CD62L ) states characterized by a gradient of pro- 1073/pnas.1013084108/-/DCSupplemental.

5724–5729 | PNAS | April 5, 2011 | vol. 108 | no. 14 www.pnas.org/cgi/doi/10.1073/pnas.1013084108 Downloaded by guest on September 30, 2021 n.s. some parallel genetic features between these phenomena and A B n.s. n.s. T-cell differentiation. These results demonstrate that different DNA-rAd5 rAd5-rAd5 rAd5-rLCMV vaccine vectors elicit CD8 T cells in which distinct gene networks that are associated with multiple immune functions have been induced. However, whether these transcriptional differences were PA9 Tetramer n.s. uniformly induced or caused by the differential induction of dis- 50 n.s. n.s. tinct subsets within these compartments remained unclear. T cells

+ 40 Multiparameter Analysis of Gene Expression in Individual Antigen- 30 60 fi + fi in CD8 Speci c CD8 T Cells. Our ndings suggested that distinct gene sets + 20 40 were induced by alternative modes of immunization. However, /PA9 d these analyses were performed using mRNA isolated from pooled 10 20 populations of vaccine-induced cells; we therefore sought to in- quire whether the gene-expression differences observed were the Percent D Percent 0 0 IFN-γ + + + + - - - - result of differences that were uniformly present in all T cells in- IL-2 + + - - + + - - duced by the immunizations, or whether they were the result of DNA-rAd5 rAd5-rAd5 TNF-α + - + - + - + - rAd5-rLCMV differential induction of subsets of T cells bearing distinct tran- scriptional patterns. We measured mRNA expression of 91 dif- Fig. 1. Gene-based vaccine regimens and assessment of resultant CD8+ T-cell ferent transcripts, selected for their relevance to T-cell phenotype, responses. Animals were immunized intramuscularly with DNA-rAd, rAd-rAd, regulation, survival, and function (Table S2) within individual or rAd-LCMV encoding HIV-1 Env, titrated to elicit similar frequencies of CD8+ resting CD8+ T cells within CM or EM subsets, and identified by T cells specific for the PA9 epitope in HIV-1 Env 3 wk following boost im- H2-Dd/PA9 tetramer-binding 3 wk following boost immunization. munization. (A) pMHC tetramer staining. H2-Dd/PA9 tetramer staining of To confirm clonality of sorted cells and the specificity of the splenocytes isolated at 3 wk postimmunization. (B) Intracellular cytokine qPCR signals, we sorted individual surface-stained CD4 or CD8 analysis following PA9 peptide stimulation. Splenocytes were pulsed with single-positive lymphocytes and analyzed the expression of Cd4 γ α PA9 peptide and stained intracellularly for IFN- , TNF- , and IL2 after 5 h. and Cd8a mRNA within these cells, in addition to the other − Similar results were obtained in three independent experiments. Data show transcripts in the assay. Only 0.59% of CD8+ CD4 cells were Cd4 fi − the mean of ve mice per group of one representative experiment. mRNA-positive, and only 4.34% of CD4+ CD8 sorted cells were CD8a mRNA-positive (Fig. 3A). To evaluate the sensitivity of the assay at levels of expression found within single cells, we titrated of 99.9% (Fig. S1C). We also assessed the proportion of CM and the number of sorted CD8 surface-positive cells from 128 cells EM CD8 T cells within the tetramer-binding populations elicited to a single cell per well and measured the abundance of Cd8a by the three vaccines, and found this to be similar for all three > mRNA. This analysis showed a strongly linear association be- vaccines (P 0.05) (Fig. S2 A and B). In vaccine studies, the tween number of cells and the abundance of CD8a mRNA quality of T-cell memory is often assessed by measuring the ability (Fig. 3B). We also validated the single-cell gene-expression data of cells to produce one or more cytokines in response to antigen + against microarray data of pooled populations of similarly sorted (1). To assess the cytokine functionality of CD8 T cells elicited by cells. We enumerated the proportion of single cells positive for these vectors, splenocytes were isolated from mice 3 wk after the fi each gene transcript among CM and EM subsets separately, and nal boost. Cells were pulsed with PA9 peptide, and in- then compared these ratios with mRNA measurements from tracellularly stained for the cytokines IFN-γ, TNF-α, and IL-2. + corresponding pooled populations of cells assessed by conven- The cytokine functionality of CD8 T cells was similar for all three tional cDNA microarrays. The ratio of positive cells for a given vaccines (Fig. 1B)(P > 0.05), highlighting an inability to detect < fi gene correlated (r = 0.86; P 0.0001) with its relative expression qualitative differences between antigen-speci c T cells elicited by by cDNA microarray hybridization when comparing CM and EM different vaccines using standard cytokine detection assays. cells (Fig. 3C).

+ Distinct Sets of Are Induced in CD8 T Cells by Alternative Efficient Discrimination of Individual CD8+ T Cells Elicited by Different Immunizations. We asked whether the patterns of gene expres- Vector Combinations. To explore the single-cell gene-expression IMMUNOLOGY fi sion in antigen-speci c CD8 T cells differed depending on the profiles of CM and EM cells elicited by the three vaccines, we vaccine vectors used. Global gene-expression patterns were de- performed a two-way hierarchical cluster analysis (Fig. 4A). In termined using cDNA microarray analysis of either CM or + both CM and EM, most of the cells elicited by DNA-rAd were EM H2-Dd/PA9 tetramer-binding CD8 T cells 3 wk after the found to cluster separately from the other groups, confirming final immunization. Each vaccine regimen induced a unique set of results obtained by conventional gene arrays (Fig. 2A). To test genes in either CM or EM subsets (Fig. 2 A and B and Table S1). whether single cells elicited by different immunizations could be In addition, we found that common genes were induced in CM distinguished based upon their single-cell gene-expression pro- and EM subsets when any two vectors were compared (Fig. 2 and files, we performed a Fisher’s linear discriminant analysis. Using Table S1). To gain insight into the nature of the genes that were this supervised learning algorithm, we were able to readily dis- uniquely induced by our vector regimens, we performed an en- tinguish CD8 T cells elicited by the three immunization protocols richment analysis of the data using Ingenuity Pathway Analysis based upon their gene-expression profiles despite separate (Ingenuity Systems), which is similar to Geneset Enrichment analysis of CM and EM cells, with a low rate of misclassification Analysis (19). Fig. 2 C and D depict the most significantly regu- (Fig. 4B) (12.4% of CM cells and 21.2% of EM cells). These lated pathways induced in EM and CM cells with representative analyses therefore demonstrate that the majority of T cells eli- genes from these pathways. Immunization with rAd5-rAd5 se- cited by different gene-based vaccines are qualitatively distinct lectively induced pathways involved in IL-2, IL-4, IL-8, and from one another, and can be readily discriminated by analysis of CTLA-4 signaling (Fig. 2 C and D, pathways in red), whereas single-cell gene-expression profiles. rAd5-rLCMV activated pathways involved in IL-15, Granzyme A To identify the smallest set of genes that could allow clas- and IFN signaling (Fig. 2 C and D, pathways in green) and DNA- sification of T cells elicited by the three different immunization rAd5 induced pathways involved in IL-10, 41BB, CD40 (Fig. 2 C regimens, we performed a supervised decision-tree analysis using and D, pathways in black). We also observed the differential binary expression values of a smaller number of genes pre- induction of pathways involved in metabolic processes (insulin selected using Fisher’s exact test (Tables S3 and S4). We found receptor signaling, oxidative phosphorylation, mitochondrial that a majority of surface IL7R+ CD62L+ CM cells elicited by dysfunction) in T-cells elicited by the different vaccines. In addi- the three immunization groups could be distinguished based tion, some pathways involved in carcinogenesis and neuronal upon the differential expression of a small number of genes, in differentiation were also induced, suggesting the presence of particular, Eomes, Cxcr3, and Ccr7 (Fig. 5A). Notably, 77% of

Flatz et al. PNAS | April 5, 2011 | vol. 108 | no. 14 | 5725 Downloaded by guest on September 30, 2021 Downloaded by guest on September 30, 2021 5726 2. Fig. et irarycmaiosaebsdo aaotie ntoidpnetexperiments. independent two in obtained data on based are comparisons Microarray test. endarm niaetenme fuiul p rdw-euae ee nue ytetreimnzto eies( regimens immunization three the by induced ( genes expression). down-regulated relative or up- ( uniquely list). of complete number the indicate diagrams Venn A5 re:rd-LM;rd A5rd)uigtecnnclptwygn esfo neut aha nlssSfwr.Tega n leclr fthe of colors -Log blue and and side gray left The the Software. on Analysis displayed Pathway Ingenuity are from pathways sets 0.05) gene pathway canonical the using fi rAd5-rAd5) red: rAd5-rLCMV; green: rAd5; 2D/A-idn CD8 H2-Dd/PA9-binding ue hwterpeettv aha ebrhpfrec ee(le rsn nteptwy ry bet.Teidenti The absent). gray: pathway; the in present (blue: gene each for membership pathway representative the show gures | iraryaayi fgoa eeepeso nvcieeiie CD8 vaccine-elicited in expression gene global of analysis Microarray www.pnas.org/cgi/doi/10.1073/pnas.1013084108 D C Cell Cycle:G2/MDNADamageCheckpointRegulation A B Neuropathic Pain Neurons SignalingIn DorsalHorn eeepeso etaso h nqeyu-euae ee nec muiaingop(e niae ihadgenidctslow indicates green and high indicates (red group immunization each in genes up-regulated uniquely the of heatmaps expression Gene ) downregulated upregulated DNA-rAd5 DNA-rAd5 Retinoic acidMediatedApoptosisSignaling NRF2−mediated Oxidative StressResponse C −mediated EndocytosisSignaling and G− CoupledReceptorSignaling CTLA4 Signal.inCyto.TLymphocytes CTLA4 Signal.inCyto. TLymphocytes + Pantothenate andCoABiosynthesis rAd5-rLCMV rAd5-rLCMV Virus EntryviaEndocyticPathways Virus Entry viaEndocyticPathways Entry Virus 4−1BB SignalinginTLymphocytes el rmtesleso muie iewr sltdadgn rncito a nlzdb DAmcora hybridization. microarray cDNA by analyzed was transcription gene and isolated were mice immunized of spleens the from cells T B CellActivatingFactorSignaling Mitotic RolesofPolo−Like Kinase Molecular MechanismsofCancer TypeI DiabetesMellitusSignaling Synaptic LongTerm Potentiation CD28 SignalinginTHelperCells CD28 SignalinginTHelperCells CCR5 SignalinginMacrophages D RANK SignalinginOsteoclasts nihetaayi fteuiul prgltdgnswti MadC oprmnsfrec muiain(lc:DNA- (black: immunization each for compartments CM and EM within genes up-regulated uniquely the of analysis Enrichment ) CM NF−kB Activation by Viruses 14−3−3−mediated Signaling Oxidative Phosphorylation Insulin ReceptorSignaling Mitochondrial Dysfunction Atherosclerosis Signaling Ubiquinone Biosynthesis April MediatedSignaling Erythropoietin Signaling rAd5-rAd5 rAd5-rAd5 N−Glycan Biosynthesis Pyrimidine Metabolism Angiopoietin Signaling Granzyme ASignaling Chemokine Signaling Coagulation System VDR/RXR Activation Interferon Signaling B CellDevelopment Thrombin Signaling Inositol Metabolism Ceramide Signaling P70S6K Signaling Integrin Signaling TREM1 Signaling IL−15 Production GNRH Signaling IL−10 Signaling CD40 Signaling HGF Signaling IL−2 Signaling IL−8 Signaling IL−4 Signaling DNA-rAd5 DNA-rAd5 FYN SRC RAC2 PPP1R10 F2R PTPN11 rAd5-rLCMV EIF2B5 rAd5-rLCMV YWHAB RAC1 INPPL1 PRKCB PIK3CD EM 10 VAMP2 RAPGEF1 p ntergtsd feach of side right the on (p) CXCR4 ITGB2 rAd5-rAd5 AP2B1 rAd5-rAd5 SERPINB1 AP2A2 AP2S1 PSMC5 ROCK1 MYL12B AP2B1 ARPC3 ZYX ARPC1A SERPINA5 RAP2A GNB1 GRINA ICAM1 PLD3 TOP2B CCND3 CSK HMGA1 ILDR1 JAK3

ELF2 B ALOX12 MAP3K3 ARRB2 DNA-rAd5 ARHGEF1 NDUFC1

+ NDUFB3 MAP3K11 COX17 el.( cells. T ATP6V1D HCG 25371 IRF1 NDUFS7 rAd5-rLCMV NDUFV3 IFNGR2 ATP5L NDUFA2

fi PARP9 ATP5G1 NDUFA6 ue h vrersnaints a efre sn Fisher using performed was test overrepresentation The gure. A PPP1R10 ATP5E COX5A h he acn eiesuiul nuedsic eso genes. of sets distinct induce uniquely regimens vaccine three The ) NDUFB6 ITPR3 NDUFA7 METTL1 CAMK2B PSENEN DHRS1 BCL2 DHDH MC IGFBP7 rAd5-rAd5 PPP2CA TPI1 APEX1 DEGS1 TRAT1 HLA−DQB1 CD19 RAC1 HLA−DOA NUDT5 CD86 APOBEC3B NME1 NT5E CTLA4 POLR2J NME7 FOS DAD1 DPM1 S1PR2 MOGS PANK1 DNA-rAd5 SOD2 BCAT2 TRAF6 MAP3K14 CAT JUN NFKBIA DNAJB1 NFKBIE BLVRB MGST3 STAT3 NFKBIB rAd5-rLCMV TLR6 TNFAIP3 SRC STAT5A PRC1 PRKCD

fi PRKCH STAG2 PTK2 r ftems signi most the of ers ME

P TNIP1 CCR5 CRK

< RASA1 GNG2 CALM3 rAd5-rAd5

.5 (see 0.05) MAP3K11 RAP2B GRK4 E2F4 SUV39H1 DUSP6 ADCY6 MDM2 RGS16 AURKA SMAD5 TGFBR2 TBP PRKAR1A 3.81 3.71 3.41 2.07 1.93 1.8 1.71 1.6 1.6 1.56 1.5 1.4 1.43 1.87 1.78 1.59 1.52 1.51 1.33 3.47 2.71 2.21 2.17 2.1 2.13 1.96 1.91 1.89 3.7 3.66 3.56 3.5 2.91 2.87 2.72 2.62 2.56 2.5 8.79 6.18 5.02 4.68 1.88 1.7 1.58 1.43 1.43 4.32 4.32 4.06 4.06 3.87 3.8 al S1 Table 4 4 5 4 4 5 5 5 fi lt tal. et Flatz at( cant ’ o the for exact s P < AB majority of CD8 T cells elicited by the different vaccines could 25 30 be readily distinguished based upon analysis of their single-cell ) 2 gene-expression profiles, despite the fact that we analyzed cells )

2 fi 20 25 within phenotypically de ned subsets. These results support the view that quantitative parameters, such as response magnitude,

15 or qualitative paramaters, such as cytokine functionality and 20 phenotype, may not identify correlates of protective T-cell im- munity in vaccine studies, because they are insensitive to the extent 10 10 of heterogeneity that can be measured in CD8 T-cell responses. In our analysis, cells elicited by DNA-rAd5 and rAd5-rAd5 5 15

CD8 mRNA abundance (log CD8 mRNA could be readily distinguished from each other, indicating the

CD8 mRNA relative abundance (log CD8 mRNA substantial role of the priming immunization in shaping the quality 0 5 024681012141618 1 10 100 of the T-cell response. Beneficial protective outcomes have been CD4 mRNA relative abundance (log2) Number of sorted cells observed since the description of heterologous prime boosting using gene-based viral vectors (20). Recently, Liu et al. found that r=0.86 C 6 p=7.43x10-8 rhesus macaques immunized with rAd26/rAd5 encoding SIV Gag

4 in prime-boost combination exhibited greater reductions of peak and setpoint viraemia, as well as decreased AIDS-related mor- 2 tality compared with macaques that received rAd5/rAd5 or rAd35/ 0 rAd5 following challenge with SIV(MAC251) (4). Although some

Single Cell -2 modest differences in the magnitude and cytokine profile of eli- -4 cited T cells were reported, no qualitative parameter could be fi -6 de ned that substantially predicted outcome. Our results provide a strategy for the evaluation of the independent roles of prime and -6 -4 -2 0 2 4 6 Microarray boost immunization in determining the quality of the subsequent T-cell response and for identifying correlates of immune pro- Fig. 3. Validation of single-cell gene-expression analysis. (A) Evaluation of tection in such studies and others (3–7). false-positive signals. Individual CD4 or CD8 single-positive lymphocytes Based on single-cell gene-expression measurements, we were were sorted by FACS and analyzed for the expression of CD4 and CD8a able to define unique subsets of CD8 T cells that were differ- − mRNA in addition to all other transcripts evaluated. Red dots show CD4 entially induced by the three vaccines. We observed that Klrk1 − − single-positive cells and blue dots show CD8 single-positive cells. (B) Evalu- Klrg1 Ccr5 EM cells were induced with greater frequency as ation of signal linearity. A titration of CD8+ cells through a range of cell a result of priming with plasmid DNA rather than a viral vector. counts from 128 cells down to single cells was sorted by FACS and analyzed Indeed, each of these genes was seen to be progressively up- for the cd8a mRNA abundance in addition to all other transcripts evaluated. regulated in a recently reported experiment where the global gene- (C) Validation of single-cell gene expression measurements. Single-cell gene- expression profiles of CD8 T cells repetitively restimulated in vivo expression data were validated by comparison with microarray analysis of was assessed (21). Additionally, KLRG1 has been previously pooled populations of cells. Single H2-Dd/PA9 binding CM and EM cells were characterized as a marker of terminal differentiation in CD8 T either individually sorted and the proportion of positive cells for each cells, the expression of which is induced with increased antigen- transcript evaluated, or cells were pooled and analyzed by conventional receptor signaling (22). These data suggest that plasmid-primed cDNA hybridization microarrays. The proportion of positive cells for each EM responses may exist in an earlier maturational state than viral transcript was then compared, where possible, with microarray fold-change vector-primed responses as a result of weaker stimulation, con- values. Each point reflects one gene transcript evaluated. Data analysis of sistent with the weaker T-cell immunogenicity of plasmid DNA single-sorted T cells is based on two independent experiments. − compared with viral vectors (23, 24). However, although Eomes − Ccr7+ Cxcr3 CM cells were also induced with greater frequency as a result of priming with plasmid DNA rather than a viral vector, rAd-LCMV–elicited CM CD8 T cells were Eomes+, while only all of these genes were progressively downregulated with in- 14% of DNA-rAd–elicited CD8 T cells and 45% of rAd-rAd– creasing stimulation, indicating the presence of other factors, in IMMUNOLOGY elicited T cells were Eomes+ (Fig. 5A, Right). Additionally, we addition to antigen strength, that may also determine the quality of found that 67% of plasmid-primed (DNA-rAd–elicited) CM T- T-cell memory. − − cells were Eomes Ccr7+ Cxcr3 , but only 7% and 2% of vector- Both microarray analysis of pooled cell populations and single- primed cells (elicited by rAd5-rAd5 and rAd-LCMV, re- cell gene-expression analyses were able to resolve differences in spectively) possessed a similar transcriptional pattern (Fig. 5A, T-cell responses generated by the three immunization regimens. Left), indicating the presence of a subset of CM cells, the fre- Although microarrays allowed an unbiased evaluation of tran- quency of which was dependent upon the vector used to prime scriptional differences between cellular immune responses eli- the immune response. Similarly, this analysis allowed subsets of cited by the three vaccines, single-cell measurements allowed the EM cells to be defined based upon the expression of Klrk1, Klrg1, analysis of coordinate gene expression, and thus the discrimina- and Ccr5, the proportion of which varied with the immunization tion of subsets of cells that would not have been discriminated regimen used (Fig. 5B). We found that 73.8% of plasmid-primed − − − based upon a unidimensional analysis of gene transcription (Fig. (DNA-rAd5–elicited) EM cells were Klrk1 Klrg1 Ccr5 , 4). Furthermore, multidimensional analysis of single-cell gene- whereas only 25.6% and 19.6% of vector-primed cells (rAd5- expression profiles allowed the resolution of vast heterogeneity at rAd5 and rAd5-LCMV, respectively) exhibited this transcrip- a transcriptional level within both CM and EM compartments. tional signature. Thus, analysis of single-cell gene-expression Just as the breadth of T-cell epitope specificity is evaluated in profiles allowed the majority of cells elicited by the three dif- vaccine-induced responses (25, 26), measurement of the diversity ferent vaccines to be distinguished, despite analysis within EM of the T-cell memory response induced by vaccination is possible and CM compartments independently. through analysis of single-cell gene-expression profiles, a param- eter that cannot be measured by microarray analysis of pooled Discussion cell populations, and this may be of substantial utility in identi- In this study, we analyzed gene expression in pooled antigen- fying correlates of protective outcomes in future studies. specific cells and within single cells comprising these pop- Our attempts to design and evaluate CD8 T-cell vaccines are ulations to assess differences in cellular immunity resulting from confounded by our inability to resolve major qualitative differ- immunization with three different prime-boost vaccine regi- ences between the T cells elicited by different vaccines, even when mens. Although the immune responses were indistinguishable by divergent protective outcomes are observed upon infectious chal- assays conventionally used in vaccine studies, we found that the lenge (3–7, 11, 16). Being able to resolve differences between T

Flatz et al. PNAS | April 5, 2011 | vol. 108 | no. 14 | 5727 Downloaded by guest on September 30, 2021 A Color Key and Histogram Color Key and Histogram DNA-rAd5 EM CM rAd5-rLCMV Count Count rAd5-rAd5 0 1000 2000 3000 4000 5000 6000 0 5 10 15 20 00 1000 2000 3000 5 4000 5000 10 15 20 Value Value

Rela Rela Was Was Cd44 pim1 Tcf7 Il2rb pim1 Cd44 Cd28 Prf1 foxo3a Cxcr3 Il2rg Tcf7 Il2rb Il2rg Gzma Cd28 Zfhx1b cmyc Prf1 pdcd1lg1 klrc1 Zap70 Cd11a Bcl2l11 pdcd1lg1 Vav1 Tnf Lef1 Rora Ccr7 Vav1 foxo3a Bcl2l11 Gzmk notch2 Eomes Klrg1 Gata3 Tgfbi klrc1 Gzmk Tgfbi Ccr5 Cd11a Edg8 Mapk8 Zap70 Cd69 Cxcr3 Stat5a Eomes Ccr5 Gata3 Cd160 Tnfsf6 Lamp1 lcos Cd69 Bcl2 Ccr7 Tnf Mapk8 Btla Stat5a Edg8 Btla Klrg1 lcos Gzma cmyc Zfhx1b Lef1 Rora Lamp1 Irf4 Bcl2 Nfkb1 Tbx21 Klrk1 Nfkb1 Bid stam Relc Ifng pd1 gzmb gzmb Cd160 Cd80 mki67 Ifng Klrk1 notch2 pd1 cd154 Irf4 Il15 Relb il2 Foxp3 Lyzs klra3 Tnfsf6 Cd4 tnfsf10 Relc klra3 Ctla4 Relb Cd80 stam Ccl6 Cdh1 tnfsf10 Ccl6 Bid Tbx21 cd154 mki67 B ●

● 4 4 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● 2 ● ● ● ● ●● 2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● 0 ● ● ● 0 ● ● ●● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● LD2 ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● -2 ● ● ● -2 ● ● ● ●● ● ● ● ● ●

-4 ● -4

-6 -4 -2 0 2 4 -6 -4 -2 0 2 4 LD1

DNA-rAd5 rAd5-rLCMV rAd5-rAd5 DNA-rAd5 rAd5-rLCMV rAd5-rAd5 DNA-rAd5 34 44 DNA-rAd5 28 00 2VMCLr-5dAr 39 01 0VMCLr-5dAr 48 5 rAd5-rAd5 2 6 31 rAd5-rAd5 3 7 30

Fig. 4. Analysis of single-cell gene expression in vaccine-elicited CD8+ T cells. Resting H2-Dd/PA9 tetramer binding cells were isolated from the spleens of immunized mice. Expression of the indicated mRNA transcripts was assessed by single cell quantitative RT-PCR. and individually sorted into 96-well plates. Following reverse transcription of cellular mRNA using specific primers, each single-cell cDNA mixture was subjected to 18 cycles of PCR preamplification before microfluidic separation into a further 96 separate reactions for the specific quantification of single-gene transcripts by quantitative RT-PCR. (A) Gene- expression heatmaps of single-cell gene-expression profiles obtained after two-way hierarchical clustering using Euclidean distance and Ward agglomeration methods. The coexpression of gene transcripts within single CD8+ T cells elicited by the three immunizations was analyzed. (B) Linear discriminant analysis. To assess whether single-cell gene-expression profiles could be used to discriminate individual EM (Left)orCM(Right) CD8+ T cells elicited by the three different immunization regimens, we performed a Fisher Linear Discriminant Analysis on the dataset. Each dot represents a single cell (black: DNA-rAd, green: rAd- rLCMV, red: rAd-rAd). The truth tables below each heatmap represent a crossvalidation of the linear discriminant analysis and compare predicted classi- fications (columns) with the actual immunization regimen used (rows). Both EM and CM cells elicited by the three immunizations could be readily dis- criminated, as indicated by a low rate of misclassification (12.4% for CM T cells and 21.2% for EM T cells). Data analysis of single cell sorted T cells is based on two independent experiments.

5728 | www.pnas.org/cgi/doi/10.1073/pnas.1013084108 Flatz et al. Downloaded by guest on September 30, 2021 A CM fectious pathogens as HIV, tuberculosis, and malaria. In this study, No Eomes Yes the analysis of gene transcription within single immune cells has Yes Ccr7 No + No Cxcr3 Yes allowed the qualitative discrimination of CD8 T-cell responses

1.0 from three vaccines that could not be resolved using conventional assays. Such analyses of vaccine-elicited T cells in clinical efficacy 0.8 trials and nonhuman primate studies will likely allow the identifi- 0.6 cation of correlates of protective immunity in these studies. 0.4 Materials and Methods Cell Proportion 0.2 Animals and Immunization Protocols. Mice at 6 to 10 wk of age were im- 0.0 munized with DNA-rAd5, rAd5-rAd5, or rAd5-rLCMV prime-boost vaccine regimens, and splenocytes were analyzed by flow cytometry at 3 wk fol- DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 rAd5-rLCMV rAd5-rLCMV rAd5-rLCMV rAd5-rLCMV lowing the final immunization by surface or intracellular staining, or by microarray or nanofluidic single cell quantitative RT-PCR following FACS as EM described below and in SI Materials and Methods. B No Klrk1 Yes No Klrg1 Yes No Ccr5 Yes No Ccr5 Yes Flow Cytometry and Intracellular Staining. The phenotype and function of 1.0 antigen-specific immune responses were measured by H2Dd/PA9 tetramer

0.8 staining and intracellular cytokine staining following PA9 peptide stimula- tion, as described in SI Materials and Methods. 0.6

0.4 Measurement of Gene Expression Within Single Cells and Microarray Analyses.

Cell Proportion 0.2 Microarray analyses were performed on pooled populations of antigen- specific cells, as described in SI Materials and Methods. 0.0

Measurement of Gene Expression Within Single Cells. Single tetramer-binding DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 DNA-rAd5rAd5-rAd5 + rAd5-rLCMV rAd5-rLCMV rAd5-rLCMV rAd5-rLCMV rAd5-rLCMV CD8 T cells were stained and individually sorted using a BD FACS Aria into 96-well plates. Following cellular lysis and specific generation of single cell Fig. 5. Decision-tree analysis of single-cell gene expression profiles. Gene- cDNA libraries (27, 28), each sample was split nanofluidically into 96 separate expression data were analyzed to determine whether assessment of single- fi cell coexpression in a smaller subset of genes could allow the classification of chambers for speci c qRT-PCR for the individual transcripts indicated in cells elicited by the three immunization groups. Bar graphs showing leaf Table S2, as described in SI Materials and Methods. statistics represent the proportion of cells elicited by each immunization group which satisfy the given transcriptional pattern as indicated in the re- ACKNOWLEDGMENTS. We thank Steve Perfetto, Richard Nguyen, David gression tree. (A) CM, (B) EM CD8 T cells. The genes used for decision tree Ambrozak, Pratip Chattopadhyay, Enrico Lugli, Yolanda Mahnke, and Carl Hogerkorp for help and advice with cell sorting, microarray, and real time PCR; analysis were selected following Fisher’s exact test to identify the top seven fi Jeff Skinner for his advice regarding hierarchical clustering of gene expression and top four most signi cantly discriminating genes for CM (Table S3) and EM data; Derek Macallan and George Griffin, Martin Pieprzyk, Tomer Kalisky, Daniel (Table S4) respectively (Bonferroni threshold 0.0008). Danila, and Nicholas Restifo for their help and advice; and Brenda Hartman for help with the figure preparation and Ati Tislerics for support in manuscript editing. This research was supported in part by the Intramural Research Program of the cells elicited by different vaccines is therefore critically important if Vaccine Research Center, National Institute of Allergy and Infectious Diseases, effective T-cell vaccines are to be developed against such in- National Institutes of Health. L.F. was supported by the Schwyzer Stiftung.

1. Seder RA, Darrah PA, Roederer M (2008) T-cell quality in memory and protection: 16. Acierno PM, et al. (2006) Preservation of functional virus-specific memory CD8+ T Implications for vaccine design. Nat Rev Immunol 8:247–258. lymphocytes in vaccinated, simian human immunodeficiency virus-infected rhesus 2. Sekaly RP (2008) The failed HIV Merck vaccine study: A step back or a launching point –

monkeys. J Immunol 176:5338 5345. IMMUNOLOGY for future vaccine development? J Exp Med 205(1):7–12. 17. Chakrabarti BK, et al. (2002) Modifications of the human immunodeficiency virus 3. Wilson NA, et al. (2009) Vaccine-induced cellular responses control simian im- envelope glycoprotein enhance immunogenicity for genetic immunization. J Virol 76: munodeficiency virus replication after heterologous challenge. J Virol 83:6508–6521. 5357–5368. 4. Liu J, et al. (2009) Immune control of an SIV challenge by a T-cell-based vaccine in 18. Honda M, et al. (2009) Different vaccine vectors delivering the same antigen elicit CD8+ T – rhesus monkeys. Nature 457(7225):87 91. cell responses with distinct clonotype and epitope specificity. JImmunol183:2425–2434. 5. Reynolds MR, et al. (2008) Macaques vaccinated with live-attenuated SIV control 19. Subramanian A, et al. (2005) Gene set enrichment analysis: A knowledge-based – replication of heterologous virus. J Exp Med 205:2537 2550. approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 6. Hansen SG, et al. (2009) Effector memory T cell responses are associated with 102:15545–15550. fi protection of rhesus monkeys from mucosal simian immunode ciency virus challenge. 20. Irvine KR, et al. (1997) Enhancing efficacy of recombinant anticancer vaccines with Nat Med 15:293–299. prime/boost regimens that use two different vectors. J Natl Cancer Inst 89:1595–1601. 7. Tsukamoto T, et al. (2009) Impact of cytotoxic-T-lymphocyte memory induction 21. Wirth TC, et al. (2010) Repetitive antigen stimulation induces stepwise transcriptome without virus-specific CD4+ T-Cell help on control of a simian immunodeficiency virus diversification but preserves a core signature of memory CD8(+) T cell differentiation. challenge in rhesus macaques. J Virol 83:9339–9346. Immunity 33(1):128–140. 8. Letvin NL (2007) Correlates of immune protection and the development of a human 22. Sarkar S, et al. (2008) Functional and genomic profiling of effector CD8 T cell subsets immunodeficiency virus vaccine. Immunity 27:366–369. with distinct memory fates. J Exp Med 205:625–640. 9. Friedrich TC, Watkins DI (2008) Wanted: Correlates of vaccine-induced protection 23. Rao SS, et al. (2006) Comparative evaluation of three different intramuscular delivery against simian immunodeficiency virus. Curr Opin HIV AIDS 3:393–398. methods for DNA immunization in a nonhuman primate animal model. Vaccine 24: 10. Darrah PA, et al. (2007) Multifunctional TH1 cells define a correlate of vaccine- – mediated protection against Leishmania major. Nat Med 13:843–850. 367 373. 11. Engram JC, et al. (2009) Vaccine-induced, simian immunodeficiency virus-specific CD8+ 24. Kong WP, et al. (2003) Immunogenicity of multiple gene and clade human fi – T cells reduce virus replication but do not protect from simian immunodeficiency virus immunode ciency virus type 1 DNA vaccines. J Virol 77:12764 12772. fi disease progression. J Immunol 183:706–717. 25. Seaman MS, et al. (2005) Multiclade human immunode ciency virus type 1 envelope 12. Reyes-Sandoval A, et al. (2010) Prime-boost immunization with adenoviral and immunogens elicit broad cellular and humoral immunity in rhesus monkeys. J Virol modified vaccinia virus Ankara vectors enhances the durability and polyfunctionality 79:2956–2963. of protective malaria CD8+ T-cell responses. Infect Immun 78(1):145–153. 26. Wu L, Kong WP, Nabel GJ (2005) Enhanced breadth of CD4 T-cell immunity by DNA 13. Altman JD, et al. (1996) Phenotypic analysis of antigen-specific T lymphocytes. Science prime and adenovirus boost immunization to human immunodeficiency virus Env and 274(5284):94–96. Gag immunogens. J Virol 79:8024–8031. 14. Kaech SM, et al. (2003) Selective expression of the interleukin 7 receptor identifies 27. Foygel K, et al. (2008) A novel and critical role for Oct4 as a regulator of the maternal- effector CD8 T cells that give rise to long-lived memory cells. Nat Immunol 4:1191–1198. embryonic transition. PLoS ONE 3:e4109. 15. Sallusto F, Lenig D, Förster R, Lipp M, Lanzavecchia A (1999) Two subsets of memory T 28. Diehn M, et al. (2009) Association of reactive oxygen species levels and radioresistance lymphocytes with distinct homing potentials and effector functions. Nature 401:708–712. in cancer stem cells. Nature 458:780–783.

Flatz et al. PNAS | April 5, 2011 | vol. 108 | no. 14 | 5729 Downloaded by guest on September 30, 2021