Short-Lived Effector CD8 T Cells Induced by Genetically Attenuated Malaria Parasite Vaccination Express CD11c Laura A. Cooney, Megha Gupta, Sunil Thomas, Sebastian Mikolajczak, Kimberly Y. Choi, Claire Gibson, Ihn K. Jang, Sam Danziger, John Aitchison, Malcolm J. Gardner, Stefan H. I. Kappe and Ruobing Wang Infect. Immun. 2013, 81(11):4171. DOI: 10.1128/IAI.00871-13. Published Ahead of Print 26 August 2013.
These include: REFERENCES
CONTENT ALERTS
This article cites 33 articles, 14 of which can be accessed free at: http://iai.asm.org/content/81/11/4171#ref-list-1 Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»
Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
Updated information and services can be found at: http://iai.asm.org/content/81/11/4171
Short-Lived Effector CD8 T Cells Induced by Genetically Attenuated Malaria Parasite Vaccination Express CD11c Laura A. Cooney,a Megha Gupta,a Sunil Thomas,b Sebastian Mikolajczak,a Kimberly Y. Choi,a Claire Gibson,a Ihn K. Jang,a Sam Danziger,a,c John Aitchison,a,c Malcolm J. Gardner,a Stefan H. I. Kappe,a Ruobing Wanga Seattle Biomedical Research Institute, Seattle, Washington, USAa; Department of Immunology, University of Washington, Seattle, Washington, USAb; Institute for Systems Biology, Seattle, Washington, USAc
M
alaria is a severe public health problem worldwide, and there is a pressing need for an effective malaria vaccine. Immunizations with irradiated and genetically attenuated Plasmodium sporozoites (SPZ) are among the most promising preerythrocytic malaria vaccination strategies, as they provide both complete and long-lasting protection in rodent models of malaria (1–6). Elucidating the basic immune effector mechanisms that mediate protection in these animal models will greatly enhance our efforts to design safe and efficacious vaccines against malaria in humans. Mice infected with Plasmodium yoelii genetically attenuated parasites (PyGAP), such as those lacking the UIS4 gene (PyUIS4⫺), undergo an abortive infection in the liver that can induce protection against wild-type P. yoelii challenge after only one dose (5). Protracted sterile protection after intravenous (i.v.) sporozoite challenge conferred by PyGAP immunization was completely dependent on CD8⫹ T lymphocytes. We previously demonstrated that PyGAP-induced CD8⫹ T cells induce apoptosis of liver-stage (LS)-infected hepatocytes, and considerable evidence indicates that cytotoxic T lymphocytes (CTLs) are the primary mediators of preerythrocytic-stage immunity in attenuated SPZ vaccination models (7–9). More specifically, previous work from our group and others showed that protection correlates with expansion of effector memory rather than central memory CD8⫹ T cells (9, 10). Taken together, these studies demonstrate a key role for multifunctional effector memory CD8⫹ T cells in protection against liver-stage malaria. In addition to functional T cell properties measured by in vitro assays, there are a variety of surface-expressed T cell activation markers that can be used to monitor immune responses that may be more applicable as biomarkers of protection in human vaccine
November 2013 Volume 81 Number 11
studies. The surface markers CD25, CD45RB, CD43glyco, and CD44 have been found to be upregulated on CD8⫹ T cells in malaria protection models (11–14). In addition to these classical markers, beta-2 integrins are emerging as a new class of activation markers in various infection models (14–18). Rai and colleagues highlighted the importance of CD11a in antigen-specific CD8⫹ T cell responses during viral and bacterial infections (19), and they demonstrated that the CD8␣lo CD11ahi subset marks antigen-experienced, malaria-specific T cells in the radiation-attenuated malaria SPZ vaccine model (14). Similarly, CD11c has been shown to be an indicator of antigen-specific T cell activation in viral infections, and CD11c⫹ CD8⫹ T cells were functionally more potent than their CD11c⫺ counterparts (15, 16, 18, 20). Following respiratory syncytial virus (RSV) infection, CD11c⫹ but not CD11c⫺ CD8⫹ T cells showed signs of recent activation, including upregulation of CD11a and expression of CD11b and CD69, and were recruited preferentially to the lung. In addition, CD11c⫹ CD8⫹ T cells were the major subset responsible for gamma interferon (IFN-␥) production, induction of targeted cell apoptosis in vitro, and reduction of viral titers in vivo (15).
Received 5 August 2013 Accepted 17 August 2013 Published ahead of print 26 August 2013 Editor: J. H. Adams Address correspondence to Ruobing Wang,
[email protected]. L.A.C. and M.G. are co-first authors. Copyright © 2013, American Society for Microbiology. All Rights Reserved. doi:10.1128/IAI.00871-13
Infection and Immunity
p. 4171– 4181
iai.asm.org
4171
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
Vaccination with a single dose of genetically attenuated malaria parasites can induce sterile protection against sporozoite challenge in the rodent Plasmodium yoelii model. Protection is dependent on CD8ⴙ T cells, involves perforin and gamma interferon (IFN-␥), and is correlated with the expansion of effector memory CD8ⴙ T cells in the liver. Here, we have further characterized vaccine-induced changes in the CD8ⴙ T cell phenotype and demonstrated significant upregulation of CD11c on CD3ⴙ CD8bⴙ T cells in the liver, spleen, and peripheral blood. CD11cⴙ CD8ⴙ T cells are predominantly CD11ahi CD44hi CD62Lⴚ, indicative of antigen-experienced effector cells. Following in vitro restimulation with malaria-infected hepatocytes, CD11cⴙ CD8ⴙ T cells expressed inflammatory cytokines and cytotoxicity markers, including IFN-␥, tumor necrosis factor alpha (TNF-␣), interleukin-2 (IL-2), perforin, and CD107a. CD11cⴚ CD8ⴙ T cells, on the other hand, expressed negligible amounts of all inflammatory cytokines and cytotoxicity markers tested, indicating that CD11c marks multifunctional effector CD8ⴙ T cells. Coculture of CD11cⴙ, but not CD11cⴚ, CD8ⴙ T cells with sporozoite-infected primary hepatocytes significantly inhibited liver-stage parasite development. Tetramer staining for the immunodominant circumsporozoite protein (CSP)-specific CD8ⴙ T cell epitope demonstrated that approximately two-thirds of CSP-specific cells expressed CD11c at the peak of the CD11cⴙ CD8ⴙ T cell response, but CD11c expression was lost as the CD8ⴙ T cells entered the memory phase. Further analyses showed that CD11cⴙ CD8ⴙ T cells are primarily KLRG1ⴙ CD127ⴚ terminal effectors, whereas all KLRG1ⴚ CD127ⴙ memory precursor effector cells are CD11cⴚ CD8ⴙ T cells. Together, these results suggest that CD11c marks a subset of highly inflammatory, short-lived, antigen-specific effector cells, which may play an important role in eliminating infected hepatocytes.
Cooney et al.
MATERIALS AND METHODS Ethics statement. All animal studies and procedures were approved by the Institutional Animal Care and Use Committee of the Seattle Biomedical Research Institute (Seattle Biomed) according to NIH guidelines for animal housing and care. Mice and parasites. The mice used for immunological studies were 6to 8-week-old female BALB/c mice (Jackson Laboratories). Female 6- to 8-week-old Swiss Webster mice (Charles River Laboratories) were used for maintenance of the malaria parasite life cycle. Wild-type Plasmodium yoelii 17X NL (nonlethal strain) clone 1.1 parasites expressing green fluorescent protein (GFP) and luciferase and P. yoelii UIS4 knockout (KO) parasites (PyUIS4⫺) were cycled between Anopheles stephensi mosquitoes and Swiss Webster mice, as previously described (5). P. yoelii sporozoites (SPZ) were isolated from the salivary glands of infected mosquitoes 14 days after an infective blood meal. The infected mosquitoes were washed with 70% ethanol and extensively with RPMI 1640 medium (Gibco BRL). The salivary glands were removed, ground with a mortar and pestle, collected into microcentrifuge tubes, and centrifuged at 800 rpm for 3 min. SPZ were collected from the supernatant and diluted to appropriate concentrations for immunization. Immunization and challenge. Groups of BALB/c mice (five mice per group) were immunized by i.v. injection with 50,000 PyUIS4⫺ SPZ in 100 l via the tail vein. Booster injections were administered at 2-week intervals. In parallel, control mice were given mock immunizations with uninfected mosquito salivary gland debris. Both immunized and control mice were challenged 14 days after the last immunizations by i.v. injection of 10,000 wild-type P. yoelii SPZ. Sterile protection was achieved if no blood infection was detected by thin blood smear within 18 days postchallenge. Lymphocyte preparation and isolation of CD8ⴙ T cells. After PyUIS4⫺ SPZ or mock immunizations, lymphocytes from blood, lymph nodes (LN), spleen, and liver were prepared for phenotypic analysis, as described previously (9). Briefly, lymph nodes and splenocytes were isolated by passage through a 70-m cell strainer (BD Biosciences). Livers were perfused with a collagenase solution, passed through a 100-m cell strainer, resuspended in 44% Percoll buffer (GE Life Sciences), and underlaid with 67% Percoll buffer. Percoll gradients were centrifuged at 2,000 rpm, and mononuclear cells at the gradient interface were extracted, washed, and resuspended in complete RPMI 1640 medium containing 10% fetal bovine serum (FBS), penicillin-streptomycin, and glutamine. CD8⫹ T cells were isolated from organs and blood lymphocyte fractions
4172
iai.asm.org
by using magnetically activated cell sorting (MACS) negative-selection beads according to the manufacturer’s instructions (Miltenyi Biotech). For both phenotypic and functional analyses, CD8⫹ T cells were further purified from organ lymphocyte preparations by using a CD8⫹ T cell positive-selection kit according to the manufacturer’s instructions (Miltenyi Biotech). The viability and purity of the enriched CD8⫹ T cell preparation were assessed by flow cytometry and trypan blue exclusion and were found to be ⬎95% and ⬎90%, respectively. Antibodies and flow cytometric analysis. For identification and analysis of CD11c⫹ CD8⫹ T cells, cells isolated from spleen, liver, or blood were stained by using a combination of the following monoclonal antibodies (MAbs): anti-CD3 clone 145-2C11, anti-CD8b clone h35-17.2, anti-CD11c clone N418, anti-CD11a clone M17/4, anti-CD44 clone IM7, anti-CD62L clone MEL-14, anti-CD25 clone PC61, anti-CD43gly clone 1B11, anti-KLRG1 clone 2F1, anti-CD127 clone A7R34, and anti-Ki67 clone SolA15 (Biolegend and eBioscience). For estimating cytokine production and cytotoxic potential, cells were restimulated with a wild-type P. yoelii-infected primary hepatocyte culture, followed by surface staining with anti-CD107a antibody clone 1D4B and intracellular staining with antiperforin antibody clone MAK-D. Alternatively, cells were treated with brefeldin A (Sigma) during in vitro stimulation, stained for surface markers, and stained intracellularly with anti-IFN-␥ clone XMG1.2, anti-tumor necrosis factor alpha (TNF-␣) clone MP6-XT22, and anti-interleukin-2 (IL-2) clone JES6-5H4 (eBioscience). Briefly, 0.5 ⫻ 106 to 1 ⫻ 106 cells were incubated with fluorescently labeled antibodies in fluorescenceactivated cell sorter (FACS) buffer (1% bovine serum albumin [BSA] and 0.05% sodium azide in phosphate-buffered saline [PBS]) for 30 min on ice in the dark. The stained cells were subsequently washed with cold FACS buffer and resuspended in PBS with 2% paraformaldehyde. Cells were permeabilized in Cytofix/Cytoperm (BD Pharmingen) for 10 min before intracellular staining. The samples were analyzed on a BD LSR II instrument. For assessment of caspase activation, liver lymphocytes were cultured for 1 h in the presence of FAM-FLICA (Immunochemistry Technologies) according to the manufacturer’s protocol, washed, and stained for surface markers, as described above. Analysis of FACS data was performed by using FlowJo software (TreeStar, Inc.). Cellular inhibition of liver stage development assay (ILSDA). Mouse liver perfusion was done by steady-state manual injection with liver perfusion medium (Invitrogen). Primary hepatocytes were isolated from BALB/c mice as previously described (22). Purified hepatocytes (40,000) were resuspended in 100 l of complete RPMI 1640 medium per well in 96-well flat-bottom tissue culture plates (Nunc, Inc.) and incubated overnight at 37°C in 5% CO2. Hepatocytes were then infected with 30,000 wild-type P. yoelii GFP-luciferase-expressing SPZ in 50 l of medium per well in a 96-well plate. CD11c⫹ CD8⫹ or CD11c⫺ CD8⫹ T cells sorted from liver-infiltrating lymphocytes of PyUIS4⫺ SPZ-immunized mice 5 days after the boost immunization were added to the infected hepatocyte cultures and incubated for 16 h. Parasite liver-stage development was assessed by using the Bright-Glo luciferase assay system and a Centro LB 960 microplate luminometer (Berthold Technologies). Tetramer pulldown. Pooled lymphocytes from spleens and livers of naive or PyUIS4⫺ SPZ-immunized mice were stained with allophycocyanin (APC)-labeled MHC H-2Kd tetramers containing residues 280 to 288 of the immunodominant P. yoelii CSP generated by the NIH tetramer core facility. Following tetramer staining, cells were labeled with anti-APC magnetic beads (Miltenyi Biotech) and applied onto magnetic columns (Miltenyi Biotech). The column flowthrough was collected as the unbound fraction, and the bound fraction was eluted from the column. Bound and unbound cells were stained for CD3, CD8b, CD11c, CD11a, and CD44 as described above. For the bound fractions, the entirety of the sample was analyzed by flow cytometry, typically yielding ⬃200,000 events in the lymphocyte gate by forward and side scatter. RNA purification and microarray analysis. Lymphocytes were isolated from livers of mice 5 days after PyUIS4⫺ SPZ immunization and stained for CD3, CD8b, and CD11c as described above. For each sample,
Infection and Immunity
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
In the present study, we found that PyGAP immunization induced significant upregulation of CD11c on CD3⫹ CD8b⫹ T cells. CD11c⫹ CD8⫹ T cells coexpressed markers of antigen-experienced effector cells and were functionally superior to CD11c⫺ CD8⫹ T cells in terms of both inflammatory cytokine production and cytotoxicity. Using a major histocompatibility complex (MHC) class I tetramer specific for circumsporozoite protein (CSP), the immunodominant LS malaria antigen, we demonstrated that CD11c is expressed by a subset of highly proliferative effector cells at the peak of the CD8⫹ T cell response and is lost from CSP-specific T cells during the memory/contraction phase. Furthermore, CD11c⫹ CD8⫹ T cells are largely KLRG1⫹ CD127⫺, consistent with short-lived effector cells (21), while all KLRG1⫺ CD127⫹ memory precursor cells are CD11c⫺. Most importantly, CD11c⫹ cells were both more proliferative and more apoptotic than CD11c⫺ cells, supporting the hypothesis that these are short-lived effector cells (SLECs). Thus, while CD11a or CD44 is uniformly upregulated by all antigen-experienced cells, CD11c more specifically marks effector CD8⫹ T cells with potent inflammatory functions that may play an important role in controlling SPZ invasion and LS parasite development.
Short-Lived Effector CD8 T Cells Express CD11c
RESULTS
PyGAP immunization induces significant expression of CD11c on CD8ⴙ T cells in liver, spleen, and peripheral blood mononuclear cells. We have demonstrated previously that PyUIS4⫺ GAPinduced protection depends largely on contact-dependent interactions between cytotoxic CD8⫹ T cells and infected hepatocytes. IFN-␥ production did not correlate directly with the protection induced by PyUIS4⫺ SPZ immunization (9). Therefore, in this study, we further delineated the phenotypic and functional associations of CD8⫹ T cell populations with protective immunity in mice induced by PyUIS4⫺ SPZ immunizations. Our goal was to identify biomarkers which may be used in the future to monitor vaccine efficacy or to select vaccine candidate antigens that are associated with protective immunity. Based on our previously reported data (5), we immunized mice with 50,000 PyUIS4⫺ parasites and observed significant increases in numbers of CD3⫹ CD11c⫹ CD8a⫹ T cells in the liver. The peak of the CD11c⫹ CD8⫹ T cell expansion after the priming immunization always occurred on day 5 and declined rapidly after day 7 (Fig. 1A). In contrast, CD11a expression on CD8⫹ T cells was sustainable, and the majority of CD8⫹ T cells were still expressing CD11a on day 12 (Fig. 1B), indicating that these integrins may play different roles during CD8 T cell activation and differentiation. Although PyUIS4⫺ parasites are no longer detectable in the liver at 40 h postvaccination (5), these kinetics are consistent with the antigen-independent expansion and contraction of CD8⫹ T cells seen in many microbial infection models (5, 23, 24). This observation led us to investigate the PyUIS4⫺ SPZ-induced CD11c⫹ CD8⫹ T cell populations in the lymphoid organs on day 5 after each immunization in an attempt to associate the CD8 T cell phenotypes with PyUIS4⫺ SPZ-induced protection. Lymphocytes were isolated from liver, spleen, peripheral blood, and lymph nodes of BALB/c mice immunized three times with 50,000 PyUIS4⫺ SPZ, an immunization regimen that reproducibly induces sterile protection in 100% of immunized mice against a massive challenge with 10,000 wild-type P. yoelii SPZ. Flow cytometric analysis revealed that upon priming with 50,000 PyUIS4⫺ SPZ, there were statistically significant increases in CD11c expression levels in CD8⫹ T cells in the liver, spleen, and peripheral blood (P ⬍ 0.001), but not in lymph nodes, on day 5 after the first PyUIS4⫺ SPZ immunization (Fig. 1C, left). In contrast, CD8⫹ T cells from mock-immunized mice expressed CD11c
November 2013 Volume 81 Number 11
at levels equivalent to those in cells from naive mice. Strikingly, over 60% of the CD8⫹ T cells from livers of PyUIS4⫺ SPZ-immunized mice expressed CD11c, whereas only 14.3% or 12.2% of the splenic or circulating CD8⫹ T cells, respectively, expressed the CD11c marker. CD11c expression on CD8⫹ T cells declined slightly but consistently after the second and third immunizations, demonstrating that the priming immunization induced the peak frequency of the CD11c⫹ CD8⫹ T cell subset (Fig. 1C, right). To understand the population dynamics of CD11c⫹ CD8⫹ and CD11c⫺ CD8⫹ T cells induced by PyUIS4⫺ SPZ immunizations, we studied them within the CD3⫹ CD8blo CD11ahi population of malaria-specific cells in the liver during expansion, contraction, and restimulation following PyGAP immunizations. We demonstrated that the early response to the primary immunization consists almost entirely of CD11c⫹ CD8⫹ T cells, and during the transition from the effector phase to the memory phase, the CD11c⫹ CD8⫹ T cell subset subsides, while the CD11c⫺ CD8⫹ T cell subset begins to preponderate (Fig. 1D). Furthermore, after the secondary and tertiary immunizations, the response is less skewed toward CD11c⫹ CD8⫹ T cells, and the CD11c⫺ CD8⫹ T population predominates. Thus, we postulate that parasite-specific CD11c⫹ CD8⫹ T cells rapidly proliferate to clear an infection but are short lived, whereas long-lived memory cells are CD11c⫺. Subsequent booster immunizations may shift the balance from short-lived effectors toward memory precursors. Quantitative CD11cⴙ CD8ⴙ T cell responses associated with PyGAP-induced protection. To test the above-mentioned hypothesis, we first immunized groups of BALB/c mice with a single dose of either 1,000, 10,000, or 50,000 PyUIS4⫺ SPZ to assess SPZ dose-dependent CD11c expression on CD8⫹ T cells. The failure to protect mice against challenge with the either of the low immunizing doses (1,000 or 10,000 PyUIS4⫺ SPZ) was correlated with the lack of or low-level CD11c expression on CD8⫹ T cells (Fig. 2A). There was no detectible increase in levels of CD11c⫹ CD8⫹ T cells in any of the mice that received 1,000 PyUIS4⫺ SPZ, and all mice became infected after challenge. All five mice that received immunization with 10,000 PyUIS4⫺ SPZ and 2 of 5 mice that received immunization with 50,000 PyUIS4⫺ SPZ exhibited delayed patent parasitemia, which had an increased level of CD11c⫹ CD8⫹ T cells. Only 3 mice were sterilely protected after a single dose of 50,000 PyUIS4⫺ SPZ, which had an average of 60% CD11c⫹ CD8⫹ T cells in the liver (Fig. 2A). We then calculated the absolute numbers of CD11c⫹ CD8⫹ T cells induced by PyUIS4⫺ SPZ immunization in order to determine the number of CD11c⫹ CD8⫹ T cells associated with protective immunization. We found that a minimum of 120,000 liver CD11c⫹ CD8⫹ T cells at day 5 was associated with sterile protection against a challenge with 10,000 wild-type P. yoelii SPZ 2 weeks after immunization (Fig. 2B). Finally, we quantified PyUIS4⫺ SPZ-induced CD11c⫹ CD8⫹ T cells in the liver and spleen that were enhanced by boosting immunizations. We measured the absolute numbers of liver and spleen CD11c⫹ CD8⫹ T cells on days 1, 3, and 5 after the boost immunization, using an immunization regimen that consistently protects 100% of BALB/c mice from infection. As presented in Fig. 2C and D, after the boost immunization, the number of liver CD11c⫹ CD8⫹ T cells was approximately 38,000 on day 1 and increased to over 280,000 on day 5. In contrast, the number of spleen CD11c⫹ CD8⫹ T cells was approximately 2 million on day 1 and decreased to 1.5 million by day 5, indicating that antigen-
iai.asm.org 4173
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
lymphocytes from seven mice were pooled and sorted into CD3⫹ CD8b⫹ CD11c⫹ and CD3⫹ CD8b⫹ CD11c⫺ populations by using a FACSAria instrument (BD Biosciences). Sorted cells were lysed, and RNA was extracted by using Qiagen RNeasy columns according to the manufacturer’s instructions. RNA concentration and quality were measured by using a NanoDrop spectrophotometer (NanoDrop) and an Agilent 2100 Bioanalyser (Agilent). RNA was labeled via a single round of linear amplification and hybridized to Agilent 8⫻60K mouse gene expression arrays, and the data were filtered to show only genes up- or downregulated by at least 2-fold between the two cell populations, with a P value of ⬍0.05. Log2 expression data represent the log2 transformation of the mean mRNA expression level in CD11c⫹ CD8⫹ cells divided by the mean mRNA expression level in CD11c⫺ CD8⫹ cells. Data analysis. Data are presented as the means ⫾ standard errors of the means (SEM). Statistical analysis was performed by using GraphPad Prism 5, and means were compared by two-tailed Student’s t tests or one-way analysis of variance (ANOVA). A P value of ⬍0.05 was considered statistically significant. Analysis of FACS data was performed by using FlowJo software (TreeStar).
20 3
5
7
12
U
IS 4
CD8b
D 80
Py
CD11a+ CD11c+ CD11a+ CD11cCD11a- CD11c+
*
60 40 20 0
5
0
10
15
20
25
30
35
day post immunizations ⫹
FIG 1 Induction and kinetics of CD11c versus CD11c⫺ CD8⫹ T cell populations in mice after PyGAP immunizations. (A and B) Kinetics of CD11c (A) and
CD11a (B) expression on CD8b⫹ T cells after primary immunization with mock control (uninfected mosquito debris) or 50,000 PyUIS4⫺ SPZ. Shown are percentages of CD11c⫹ CD3⫹ CD8b⫹ or CD11a⫹ CD3⫹ CD8b⫹ T cells in relation to the total proportion of CD3⫹ CD8b⫹ T cells in the liver. (C, left) Representative dot plots of CD11c expression on CD3⫹ CD8b⫹ T cells in the liver, spleen, blood, or lymph nodes (pooled inguinal and mesenteric) of mice immunized i.v. with the mock control or 50,000 PyUIS4⫺ SPZ. The percentages of CD11c⫹ T cells in relation to the total proportions of CD3⫹ CD8b⫹ T cells are indicated in each panel. (Right) Average data from several experiments representing the percent CD11c⫹ CD3⫹ CD8b⫹ T cell population in the liver, spleen, lymph node, and blood of naive or mock- or PyUIS4⫺ SPZ-immunized mice on day 5 after each immunization (n ⫽ 5 to 15; mean ⫾ SEM) (ⴱ, P ⬍ 0.05; ⴱⴱ, P ⬍ 0.01; ⴱⴱⴱ, P ⬍ 0.001 [versus mock infection, determined by one-way ANOVA]). Mice were immunized i.v. with uninfected mosquito debris (mock) or 50,000 PyUIS4⫺ SPZ once, twice, or three times at 2-week intervals. (D) Mice were immunized three times with 50,000 PyUIS4⫺ SPZ at 2-week intervals. Liver lymphocytes were collected on day 0, 1, 3, 5, 7, or 12 after each immunization and stained for flow cytometry as described in the text. Lymphocytes were gated on CD3⫹ CD8b⫹ cells and divided into four subsets based on expression of CD11a and CD11c (data from two independent experiments; n ⫽ 5 to 15; mean ⫾ SEM).
specific CD11c⫹ CD8⫹ T cells may migrate from the spleen to the liver and significantly contribute to the elimination of LS parasites after infection. PyGAP-induced CD11cⴙ CD8ⴙ T cells constitute the activated effector T cell population. To gain insight into the phenotype and function of CD11c⫹ CD8⫹ T cells following PyGAP immunization, we assessed the activation/differentiation status of CD11c⫹ CD8⫹ T cells compared to CD11c⫺ CD8⫹ T cells from the liver or spleen. We analyzed their expression levels of markers of activation and memory differentiation at the peak of the CD11c⫹ CD8⫹ T cell response to PyGAP immunization by flow cytometry, including CD44, CD11a, CD62L, CD43gly, and CD25. The results in Fig. 3A and B show that CD11c⫹ CD8⫹ T cells from the liver are almost entirely CD44hi, CD11ahi, and CD62Lneg, indicative of activated effector cells. Furthermore, over 40% of the liver CD11c⫹ CD8⫹ T cells also expressed the activation marker CD43gly or CD25, whereas only 20% or 40% of spleen CD11c⫹
4174
iai.asm.org
CD8⫹ T cells expressed CD43gly or CD25, respectively. Therefore, the PyUIS4⫺ SPZ-specific CD11c⫹ CD8⫹ T cells in the liver appear to be in an activated state. In the spleen, although CD11c⫹ CD8⫹ T cells were largely CD44hi, only 60% were CD11ahi or CD62Lneg (Fig. 3A and B), suggesting that exposure to the PyUIS4⫺ SPZ is required for the full activation of PyUIS4-specific CD11c⫹ CD8⫹ T cells. There was no significant difference in the expression levels of these activation markers on CD11c⫹ CD8⫹ T cells from the liver or spleen after the primary compared to the secondary or tertiary immunizations (data not shown). CD11cⴙ CD8ⴙ T cells are cytotoxic T cells against LS parasite-infected hepatocytes. It was demonstrated previously that perforin-dependent cytotoxicity and inflammatory cytokines (IFN-␥ and TNF-␣) play an important role in sterile protection against liver-stage malaria infection (9, 25). Thus, it was essential to determine the functional characteristics of CD11c⫹ CD8⫹ T cells to assess whether they are potential mediators of protection
Infection and Immunity
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
na
1
0.8
4.1
**
(-) IS , 2x 4( -), 3x
40
**
U
60
***
**
Py
***
k
***
***
x
***
1.1
(-) ,1
80
12.2
***
***
IS 4
12
U
7
Py
5
***
iv e
3
LNs
% CD11ahi (of CD3+ CD8+)
3.2
80 60 40 20 0 80 60 40 20 0 80 60 40 20 0 80 60 40 20 0
oc
1
14.3
***
100
0
% CD3+CD8b+
***
*
20
mock 23.6
m
40
PyUIS4(-),1x 61.0
Liver
60
C
Spleen
80
0
B
naive mock 1x PyUIS4
Blood
% CD11c+ (of CD3+ CD8+)
100
CD11c
A
% CD11c+ (of CD3+ CD8+)
Cooney et al.
Short-Lived Effector CD8 T Cells Express CD11c
150 100
300 200 ***
100
* **
ive
na
**
(-) (-) ck IS4 UIS4 mo yU y P P 1x 2x
5000
Spleen
***
***
1000
ive
na
7.2
) (-) ck 4(IS4 mo UIS y yU P P 1x 2x ⫺
induced protection. Mice were immunized i.v. with different doses of PyUIS4 SPZ, and levels of CD11c⫹ CD8⫹ T cells were measured on day 5. (A and B) Comparison of the percentage (A) and absolute number (B) of CD11c⫹ CD8⫹ T cells in the liver after immunization correlated with infection or protection against wild-type SPZ challenge. (C and D) Quantitative measurement of CD11c⫹ CD8⫹ T cells from liver (C) or spleen (D) on days 1, 3, and 5 after primary and boost immunizations with 50,000 PyUIS4⫺ SPZ at 2-week intervals. This analysis was conducted with the same sets of data shown in Fig. 1. Sample collection, flow cytometry, and statistical analyses are detailed in the Fig. 1 legend.
against liver-stage malaria. CD8⫹ T cells isolated from PyUIS4⫺ SPZ-immunized mice at 5 days after the boost immunization were restimulated in vitro by coculture with infected hepatocytes to induce activation and subsequently analyzed for intracellular perforin expression and for degranulation via CD107a expression, two markers associated with cytotoxicity. The results show a striking difference in the cytotoxic potential of CD11c⫹ CD8⫹ versus CD11c⫺ CD8⫹ T cells (Fig. 4A and B). Approximately 62.9% of CD11c⫹ CD8⫹ T cells were CTLs; 46.6% or 16.3% expressed perforin or CD107a, respectively. In contrast, only 0.2% or 0.1% of CD11c⫺ CD8⫹ T cells expressed perforin or CD107a markers, respectively, indicating that the CD11c⫺ CD8⫹ T cells were not cytotoxic T cells. In addition to cytotoxicity, we also compared the expression levels of inflammatory cytokines in CD11c⫹ CD8⫹ and CD11c⫺ CD8⫹ T cells, including IFN-␥, TNF-␣, and IL-2, which may play an indirect role in protective immunity against malaria challenge and correlates with vaccine efficacy in other models (reviewed in reference 26). Liver lymphocytes were restimulated in vitro with infected hepatocytes for 16 h, with brefeldin A added for the final 2 h, and stained for intracellular cytokines. Similar to the cytotoxicity markers, CD11c⫹ CD8⫹ T cells expressed significantly more of all three cytokines than did CD11c⫺ CD8⫹ T cells (Fig. 4A and B). However, there were no significant differences in expression levels of cytokines or cytotoxicity markers between CD11c⫹ CD8⫹ T cells and CD11c⫺ CD8⫹ T cells after additional booster immunizations (data not shown). To further characterize the differences between CD11c⫹ CD8⫹
100 80 60 40 20 0
100 80 60 40 20 0
***
31.1
48.4
CD11a
***
0
18.9
CD44
B
3000 2000
13.6
71.9
FIG 2 Quantitative CD11c⫹ CD8⫹ T cell responses associated with PyGAP-
November 2013 Volume 81 Number 11
17.4
16.5
CD11c
) 4(-
CD11c
S IS IS UI yU yU K P 0K P 10 5
day 1 day 3 day 5
4000
) 4(-
Spleen
Py
) 4(-
Liver
***
0
1K
ck
% marker+
400
day 1 day 3 day 5
8.6
24.9
mo
D
Liver
500
61.5
50 0
) ) ) ck 4(4(4(mo UIS yUIS yUIS y P P P K K 1K 50 10
64.1
Liver
***
Spleen
20
A
Liver
200
% marker+
40
No CD11c+ CD8+ T cells (x103)
***
B
***
***
CD11c + CD8 + CD11c - CD8 +
***
***
CD44+ CD62L- CD11a+ CD43+ CD25+
***
* *** *** ***
CD44+ CD62L- CD11a+ CD43+ CD25+
FIG 3 PyGAP-induced CD11c⫹ CD8⫹ T cells are activated effector T cells.
Mice were immunized i.v. with 50,000 PyUIS4⫺ SPZ, and lymphocytes were prepared from liver and spleen 5 days later. Cells were stained with CD3, CD8b, CD11c, CD25, CD43glyco, CD44, CD62L, and CD11a antibodies and analyzed by flow cytometry. (A) Representative contour plots showing CD11c versus CD11a and CD44 expression, gated on CD3⫹ CD8b⫹ T cells from the liver or spleen of a PyUIS4⫺ SPZ-immunized mouse. (B) Overall phenotypic analysis of lymphocytes expressing various activation markers within the CD3⫹ CD8b⫹ CD11c⫹ and CD3⫹ CD8b⫹ CD11c⫺ gates. (n ⫽ 10 to 15; mean ⫾ SEM) (ⴱ, P ⬍0.05; ⴱⴱ, P ⬍ 0.001 [versus CD11c⫺, determined by two-way ANOVA]).
and CD11c⫺ CD8⫹ T cells in the livers of PyUIS4⫺ SPZ-immunized mice, we isolated these two cell populations by flow cytometry and compared their gene expression profiles by microarray analyses. A total of 671 genes were found to have a ⬎2-fold difference in expression levels (P ⬍ 0.05). A heat map is shown in Fig. 4C. The patterns of gene expression further revealed by clustering analysis of pathways, including cell cycle, cytotoxicity, and memory response, that are differentially regulated in CD11c⫹ CD8⫹ T cells versus CD11c⫺ CD8 T cells are compared in Table 1. As predicted, CD11c⫹ CD8⫹ cells overexpressed transcripts encoding many markers of effector function, such as IFN-␥, TNF-␣, and granzyme B, compared to CD11c⫺ CD8 T cells. Moreover, CD11c⫹ CD8⫹ T cells were highly enriched in transcripts of genes associated with the cell cycle, suggesting that these cells are more proliferative, while CD11c⫺ CD8⫹ cells were enriched for transcripts known to be expressed in memory T cells, such as CD27, CD28, and CD127 (Table 1). In addition, we observed differential expression of many chemokine receptor genes, which suggests that these two populations may undergo selective migration. Finally, to ascertain whether PyGAP-induced CD11c⫹ CD8⫹ T cells are cytotoxic to infected hepatocytes, we performed an in
iai.asm.org 4175
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
No CD11c+ CD8+ T cells (x103)
***
60
0
C
Liver
80
No CD11c+ CD8+ T cells (x103)
% CD11c+ (of CD3+ CD8+)
A
Cooney et al.
mediators in the CD11c⫹ CD8⫹ T cell subset versus the CD11c⫺ CD8⫹ T cell subset. Lymphocytes were prepared from livers of PyUIS4⫺ SPZ-immunized mice and restimulated in vitro with wild-type P. yoelii SPZ-infected hepatocyte cultures for 5 h, followed by surface staining for CD8b, CD11c, and CD107a and intracellular staining for perforin. Alternatively, cells were restimulated in vitro with wild-type P. yoelii SPZ-infected hepatocyte cultures for 16 h and with brefeldin A for the final 2 h, followed by surface staining for CD8b and CD11c and intracellular staining for IFN-␥, TNF-␣, and IL-2. (B) Bars represent percentages of cells expressing each marker within the CD11c⫹ CD8⫹ T cell subset versus the CD11c⫺ CD8⫹ T cell subset (n ⫽ 6; mean ⫾ SEM) (ⴱⴱⴱ, P ⬍ 0.001 by two-way ANOVA). (C) Transcriptome analysis by microarray. Shown is a heat map representation of genes regulated upon PyUIS4⫺ SPZ immunization in CD11c⫺ CD8⫹ and CD11c⫹ CD8⫹ T cells. Liver lymphocytes from PyUIS4⫺ SPZ-immunized mice were stained for CD3, CD8b, and CD11c and sorted into the CD3⫹ CD8b⫹ CD11c⫹ and CD3⫹ CD8b⫹ CD11c⫺ subsets by using a FACSAria instrument. RNA was extracted, and gene expression was analyzed via microarrays. Each sample in a column contained pooled lymphocytes from seven mice to yield adequate material. Data are from 671 genes with a ⬎2-fold difference in expression (P ⬍ 0.05). Expression levels for each gene were scaled to be between ⫺2 and 2 to emphasize the relative difference between CD11c⫺ CD8⫹ and CD11c⫹ CD8⫹ T cells. (D) CD11c⫹ CD8⫹ or CD11c⫺ CD8⫹ T cells were isolated by cell sorting on day 5 after the second immunization with 50,000 PyUIS4⫺ SPZ and incubated with hepatocytes infected with GFP-luciferase-expressing P. yoelii SPZ for 16 h. Bioluminescence in lysates of infected hepatocytes was measured by using the Bright-Glo substrate kit. Infected hepatocytes that had not been incubated with CD8⫹ T cells were used as a control. Bioluminescence values are shown on the graph (n ⫽ 3; mean ⫾ SEM) (ⴱⴱⴱ, P ⬍ 0.001, determined by unpaired, two-tailed t tests). The result is representative of two independent experiments performed with five mice in each group.
vitro cellular inhibition of liver stage development assay (ILSDA). Primary hepatocytes infected with transgenic P. yoelii GFP-luciferase-expressing SPZ were cocultured with either 100,000 or 500,000 CD11c⫹ CD8⫹ or CD11c⫺ CD8⫹ T cells that had been isolated from the livers of PyUIS4⫺ SPZ-immunized mice. The mock immunization failed to induce a sufficient number of CD11c⫹ CD8⫹ T cells for use as a negative control, so we used
4176
iai.asm.org
primary hepatocytes infected with GFP-luciferase-expressing P. yoelii SPZ instead (Fig. 4D). In the ILSDA, CD11⫺ CD8⫹ T cells from the same PyUIS4⫺ SPZ-immunized mice were assayed simultaneously with the cytotoxicity control for comparison with CD11c⫹ CD8⫹ T cells. Bioluminescence was used to measure parasite load in infected primary hepatocytes after 16 h of incubation with T cells. A concentration-dependent reduction in the number of
Infection and Immunity
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
FIG 4 CD11c⫹ CD8⫹ T cells are LS parasite-specific cytotoxic T cells. (A) Representative flow cytometric contour plots for the levels of cytokines and cytotoxic
Short-Lived Effector CD8 T Cells Express CD11c
TABLE 1 CD11c⫹ CD8⫹ T cells overexpress cell cycle genes and killer T cell genes but underexpress memory genes compared to CD11c⫺ CD8⫹ T cells Log2 expression level 1.372247 1.302702 1.282885 0.200407 1.256356 1.251194 1.264975 1.237407 1.220909 1.220834 1.205737 1.189483 1.155581 1.120607 1.120054 1.317399 1.081535 1.077778 1.072817 1.056288 1.046406 1.00737 1.007047 0.93921
Killer T cell Klrg1 Cd160 Ifng Tnf Gzma Gzmb Gzmc Gzmk
1.132168987 1.485929648 1.116497274 1.222764819 0.34819641 1.232615286 0.636191704 0.601461607
Memory T cell Ccr7 Cd28 Cd127 Cd27 Ctla4 Bcl2 Socs3 Eomes
⫺4.408515888 ⫺1.028787977 ⫺3.398772535 ⫺0.563361512 ⫺0.645008893 ⫺2.456437576 ⫺2.180554636 ⫺1.924305863
infected hepatocytes was observed with CD11c⫹ CD8⫹ T cells compared to CD11c⫺ CD8⫹ T cells. Inclusion of 100,000 and 500,000 CD11c⫹ CD8⫹ T cells resulted in 48% (P ⫽ 0.056) and 85% (P ⬍ 0.001) reductions of liver-stage-infected hepatocytes in the culture, respectively (Fig. 4D), whereas the CD11c⫺ CD8⫹ T cells failed to eliminate infected hepatocytes. These results indicate that infected hepatocytes were targeted by PyUIS4⫺ SPZ-specific CD11c⫹ CD8⫹ T cells and that these are cytotoxic effector CD8⫹ T cells. Coculture of CD8ⴙ T cells with infected hepatocytes induces CD11c expression on splenic CD8 T cells. To understand why immunization with PyUIS4⫺ SPZ resulted in the majority (60 to 70%) of liver CD8⫹ T cells expressing CD11c, compared to ⬍20%
November 2013 Volume 81 Number 11
FIG 5 Coculture of CD11c⫺ CD8⫹ T cells with LS-infected hepatocytes in-
duces CD11c expression on CD8 T cells. Splenocytes (2 ⫻ 106) or purified spleen CD11c⫺ CD8⫹ T cells were isolated on day 5 after the second immunization with PyUIS4⫺ SPZ or the mock control and incubated with wild-type P. yoelii SPZ-infected hepatocytes for 14 h, followed by staining with CD8b and CD11c antibodies for flow cytometric analysis. Cells that were not cocultured with infected hepatocytes served as controls. (Left) Representative dot plots of CD11c expression on CD3⫹ CD8b⫹ T cells; (right) percentages of CD11c⫹ CD8⫹ T cell populations before and after restimulation (n ⫽ 3; mean ⫾ SEM) (ⴱⴱⴱ, P ⬍ 0.001). Statistics were determined by unpaired, two-tailed t tests. The results are indicative of three independent experiments performed with three mice in each group.
iai.asm.org 4177
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
Gene group and gene Cell cycle Bub1 Birc5 Cenpn Nuf2 Ska1 Bub1b Kif2c Cenpe Ndc80 Sgol1 Sgol2 Cdca8 Spc24 Kntc1 Spc25 Birc5 Cenpi Kif18a Cenph Casc5 Mad2l1 Incenp Cenpp Zwilch
of cells from the spleen, blood, or LN, we examined whether CD11c expression on CD8⫹ T cells occurred after exposure to liver-stage (LS)-infected hepatocytes. We postulated that CD11c expression in liver CD8⫹ T cells would result from activation of CD8⫹ T cells that interact with the LS-infected hepatocytes. To test this hypothesis, splenocytes or purified splenic CD11c⫺ CD8⫹ T cells isolated from PyUIS4⫺ SPZ- and mock-immunized mice were cocultured with infected primary hepatocytes. After 14 h, CD11c expression on splenic CD8⫹ T cells was assessed. As demonstrated in Fig. 5, there was a 3-fold increase in CD11c expression levels on CD8⫹ T cells in the pooled splenocytes and a 4-fold increase on purified splenic CD11c⫺ CD8⫹ T cells in PyUIS4⫺ SPZ-immunized mice (P ⱕ 0.001) compared to CD8⫹ T cells from mock-immunized mice. These results suggested that CD11c expression on CD8⫹ T cells might be triggered by the physical contact or proximity between CD8⫹ T cells and LS-infected hepatocytes. It further raises the possibility that splenic CD11c⫺ CD8⫹ T cells could be recruited to the liver during parasite infection and development and that these cells become fully activated effector
Cooney et al.
H-2Kd-CSP
A 1.5
34.9
B
PyUIS4(-) Unbound
PyUIS4(-) Bound 0.2
Day 5
88.7
21.9
41.7
83.7
36.2
0.3
C
34.4
62
36.4
29.8
0.5
0.5
20
10
0 30.4
33.2
91
8.0
**
CD11c+CD8+
CD11c-CD8+
CD11c
FIG 6 PyGAP immunization induces CSP antigen-specific CD11c⫹ CD8⫹ T cells. (A) Liver lymphocytes from naive or PyUIS4⫺ SPZ-immunized mice 5 days after immunization were stained with APC-labeled H-2Kd–CSP(280 –288) tetramer, incubated with magnetic anti-APC beads, and applied onto a magnetic column. Column-bound and unbound flowthrough cells were stained for CD3, CD8b, CD11c, CD11a, and CD44. Representative dot plots are gated on live CD3⫹ CD8b⫹ cells and show H-2Kd–CSP expression versus CD44, CD11a, and CD11c expression. Values in each quadrant show percentages of total CD3⫹ CD8b⫹ cells. Results are representative of three independent experiments. (B) H-2Kd–CSP tetramer enrichment was performed as described above for panel A except that gating was done on CD3⫹ CD8blo CD11ahi cells, and the number of CSP-specific cells 5, 10, or 20 days after primary PyUIS4⫺ SPZ immunization was measured. (C) Liver lymphocytes were isolated from mice 7 days after the first immunization with PyUIS4⫺ SPZ and cultured for 1 h with FAM-FLICA reagent to stain active caspases as a measure of apoptosis, followed by staining for CD3, CD8b, CD11a, and CD11c. Flow data were gated on CD3⫹ CD8blo CD11ahi cells, and the percentage of cells staining positive for FAM-FLICA was determined within the CD11c⫹ CD8⫹ and CD11c⫺ CD8⫹ T cell subsets (n ⫽ 5 mice/group; results are representative of three separate experiments) (ⴱⴱⴱ, P ⬍ 0.001 by paired t tests).
CD8⫹ T cells expressing CD11c in PyGAP-immunized and protected mice. CSP-specific CD11cⴙ CD8ⴙ T cells are major components of PyUIS4ⴚ SPZ-induced protection. CSP is the most abundant SPZ surface protein, and CSP-specific CD8⫹ T cells contribute to protective immunity against wild-type SPZ challenge (27). Therefore, we used CSP as a marker to determine the antigen-specific CD11c⫹ CD8⫹ T cell responses induced by PyUIS4⫺ SPZ. The MHC class I tetramer carrying the immunodominant epitope from P. yoelii CSP was coupled to magnetic beads to enrich parasite-specific cells from the liver. As shown in Fig. 6A, on day 5 after immunization, the CSP-specific CD8⫹ T cells were uniformly CD44hi and CD11ahi, and ⬃80% of these cells expressed CD11c. We then carried out tetramer enrichment of CSP-specific cells from days 10 and 20 after immunization, as the T cell response contracts and transitions from the effector phase to the memory phase. Obviously, the results shown in Fig. 6B demonstrate the loss of CD11c on CSP-specific CD8 T cells during the contraction phase of the immune response. The observation that CD11c expression is not maintained after contraction suggests that CD11c is more likely to be expressed by effector cells but not by memory cells. We demonstrated that CD11c is expressed by parasite-specific T cells at the peak of the response but not in the memory phase, which begs the question of whether CD11c⫹ CD8⫹ T cells were
4178
iai.asm.org
selectively dying or simply downregulating CD11c. To differentiate between these two possibilities, we collected liver lymphocytes from PyGAP-immunized mice on day 7, a time when the numbers of CD11c⫹ CD8⫹ T cells were rapidly declining, and stained them with a fluorescence-labeled polycaspase inhibitor to detect apoptosis (FLICA) (28). Rather than using tetramer enrichment, we gated on the CD3⫹ CD8blo CD11ahi population of malaria-specific effectors, as defined previously by Schmidt et al. (14), and compared caspase activation in CD11c⫹ CD8⫹ T cells to that in CD11c⫺ CD8⫹ T cells. We detected a significant number of CD11c⫹ CD8⫹ T cells undergoing apoptosis at day 7 (Fig. 6C), thus confirming that there is selective contraction of CD11c⫹ CD8⫹ cells over CD11c⫺ CD8⫹ cells. CD11cⴙ CD8ⴙ T cells are short-lived effectors, whereas CD11cⴚ CD8ⴙ T cells are long-lived memory precursor effector cells. To further address the question of whether CD11c⫹ CD8⫹ cells contribute to the development of memory, we compared the expression levels of KLRG1 and CD127 on CD11c⫹ CD8⫹ and CD11c⫺ CD8⫹ T cells. KLRG1 marks cells that have undergone terminal differentiation, while CD127, which is the IL-7 receptor ␣ chain, contributes to memory cell survival. Terminal or shortlived effector cells (SLECs) are KLRG1⫹ CD127⫺, while longlived memory precursor effector cells (MPECs) are KLRG1⫺ CD127⫹ (21, 29). While SLECs may be important for clearing an infection, MPECs possess stem cell-like properties and confer a
Infection and Immunity
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
CD11a 7.7
13.4
40
30 28.1
20.9
CD11c 1.2
% FLICA+
H-2Kd-CSP
1.3
Day 20
15.8
CD44
H-2Kd-CSP
Day 10
0.3
Short-Lived Effector CD8 T Cells Express CD11c
A
CD11c+CD8+ 5.3
11.8
2.6
KLRG1
% marker
30.2
C 80
CD11c-CD8+
16.9
47.7
31
CD11c+CD8+ CD11c-CD8+
60 40
***
20
54.5
0
KLRG1+ CD127-
CD127(IL-7R)
B
CD11c+CD8+
CD11-CD8+
D 100
82
29
***
KLRG1CD127+
CD11c+CD8+ CD11c-CD8+
60 40 20
Ki67 ⫹
0
⫹
FIG 7 CD11c CD8 T cells are highly proliferative short-lived effector cells, whereas CD11c⫺ CD8⫹ T cells are long-lived memory or memory precursor cells.
(A and C) Liver lymphocytes isolated 5 days after PyUIS4⫺ SPZ immunization were stained for flow cytometry as described in the text. Representative dot plots demonstrate expression of KLRG1 and CD127 by antigen-specific CD11c⫹ CD8⫹ versus CD11c⫺ CD8⫹ T cells. Short-lived effector cells are KLRG1⫹ CD127⫺, while memory precursor cells are KLRG1⫺ CD127⫹. Values in each quadrant show percentages of total CD3⫹ CD8blo CD11ahi cells. (B and D) Liver lymphocytes were isolated 5 days after primary immunization with PyUIS4⫺ SPZ and surface stained for CD3, CD8b, CD11a, and CD11c, followed by intracellular staining for Ki67, a marker of cell proliferation. A representative histogram shows Ki67 expression by CD11c⫹ CD8⫹ and CD11c⫺ CD8⫹ T cells after gating on CD3⫹ CD8blo CD11ahi cells. The histogram and bar figures represent percentages of CD11c⫹ CD8⫹ T cells (82%) and CD11c⫺ CD8⫹ T cells (29%) expressing Ki67 (n ⫽ 4; mean ⫾ SEM; results are representative of two independent experiments) (P ⬍ 0.001 by paired t tests).
greater recall potential; thus, the relative development of these two subsets is a key factor in generating long-lasting and effective memory responses following vaccination (29). Five days after the priming PyUIS4⫺ SPZ immunization, CD11c⫹ CD8⫹ T cells are skewed toward KLRG1⫹ CD127⫺ SLECs (Fig. 7A and C). Approximately one-half of the KLRG1⫺ CD127⫹ MPECs, on the other hand, are contained entirely within the CD11c⫺ CD8⫹ T cell population. Thus, CD11c expression may serve as a marker of multifunctional but short-lived effector cells. The results of the microarray and the time course analyses suggest that CD11c⫹ CD8⫹ T cells are more proliferative than CD11c⫺ CD8⫹ T cells. To confirm this hypothesis, we stained liver lymphocytes, gated for antigen-specific CD3⫹ CD8blo CD11ahi cells, and examined the nuclear expression of Ki67, a cell cycle protein and marker of proliferation. The results in Fig. 7B and D show that approximately 80% of CD11c⫹ CD8⫹ T cells were undergoing proliferation on day 5 after immunization, versus ⬃30% of CD11c⫺ CD8⫹ T cells. These results are consistent with previously reported data showing that a higher percentage of SLECs than MPECs enters the cell cycle (21). DISCUSSION
Cell-mediated responses play a crucial role in protective immunity against liver-stage malaria (30, 31). Our group and others previously demonstrated that depletion of CD8⫹ T cells leads to a loss of PyGAP-induced protection in rodents (5, 7, 9, 11). In the present study, we demonstrated a dramatic increase in the expression level of CD11c on CD8 T cells in PyGAP-immunized and protected mice. CD11c expression has been conventionally associated with myeloid antigen-presenting cells and monocytes but has also been described for CD8⫹ T cells during viral infections,
November 2013 Volume 81 Number 11
graft-versus-host disease, and antitumor responses (15–18). Following RSV infection, CD11c⫹ CD8⫹ but not CD11c⫺ CD8⫹ T cells showed signs of recent activation, including upregulation of CD11a and expression of CD11b and CD69, and were recruited preferentially to the lung (15). Similarly, we found that the majority of CD11c⫹ CD8⫹ T cells expressed activation markers, such as CD44 and CD11a, and downregulated CD62L (Fig. 3A and B). These results are also consistent with the idea that CD44 and CD11a mark all previously activated cells irrespective of function or lineage, while CD11c, on the other hand, marks a subset of antigen-specific cells. Together with the data described above, this suggests that CD11c specifically marks the subset of antigen-specific cells with the most potent effector function. Previously, our laboratory and others have demonstrated that CD8⫹ T cells eliminate LS malaria parasites primarily by direct perforin-mediated killing of infected hepatocytes or by indirect release of IFN-␥ (9). In this study, we found that CD11c⫹ CD8⫹ T cells had significantly higher levels of intracellular perforin than did CD11c⫺ CD8⫹ T cells (Fig. 4A and B). Furthermore, CD11c⫹ CD8⫹ T cells expressed CD107a on their surface after restimulation with infected hepatocytes, while CD11c⫺ CD8⫹ T cells did not (Fig. 4A and B). Altogether, this suggests that CD11c⫹ CD8⫹ T cells are potent CTLs that may be instrumental in clearing malaria infection in the liver. In addition, CD11c⫹ CD8⫹ T cells express high levels of IFN-␥, TNF-␣, and IL-2 following restimulation with infected hepatocytes, while CD11c⫺ CD8⫹ T cells expressed very limited quantities (Fig. 4A and B). These results are reinforced by the microarray data, which showed significant enrichment of many transcripts associated with cytokines and cytotoxicity in CD11c⫹ CD8⫹ compared to those in CD11c⫺ CD8⫹ T cells from livers of PyGAP-immunized mice (Fig. 4C and Table 1).
iai.asm.org 4179
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
% KI67+
80
***
Cooney et al.
4180
iai.asm.org
the response toward CD11c⫺ CD8⫹ T cells and away from CD11c⫹ CD8⫹ T cells, suggesting that boosting immunizations significantly increase the memory T cell population, a concept consistent with the long-term protection induced by PyUIS4⫺ SPZ (5). Although sterile protection can be induced by PyGAP immunization with one high dose or several low doses of SPZ, there may be a fine balance between the level of inflammation and the number of immunizations that dictates the relative development of these two competing subsets. Overall, this study defines a population of P. yoelii GAP vaccine-induced CD11c⫹ CD8⫹ T cells with a highly proliferative but short-lived effector phenotype as well as potent cytotoxic and inflammatory functions (33). The kinetics, gene expression analysis, and Ki67 staining all demonstrate that CD11c⫹ CD8⫹ T cells are more proliferative than CD11c⫺ CD8⫹ T cells, which, when combined with their increased effector function, suggests that CD11c⫹ CD8⫹ T cells form the first line of defense against LS malaria infection. CD11c⫺ CD8⫹ T cells, on the other hand, persist longer after immunization and express many markers of memory or memory precursor cells, suggesting that they play an important role in maintaining the pool of memory cells. The trade-off between inducing terminal effector cells to rapidly clear an infection and inducing memory precursor cells for long-lasting protection is an important question currently facing those in the field of vaccine development, and CD11c expression may serve as a useful addition to our current toolbox of biomarkers of CD8⫹ T cell activation for predicting vaccine efficacy. Clearly, it will be important to determine if a similar population of CD11c⫹ CD8⫹ T cells is induced following vaccination with genetically attenuated Plasmodium falciparum in humans and whether it correlates with protection. ACKNOWLEDGMENTS This work was supported by grant 5R01AI76498 from the National Institute of Allergy and Infectious Diseases and by grant OPP1016829 from the Bill & Melinda Gates Foundation (BMGF). We thank the staff of the NIH tetramer core for support, Sampa Pal for her administrative support, and Olivia Finney, Gladys Keitany, Jessica Miller, Ashley Vaughan, and Brandon Sack for their critical review of the manuscript.
REFERENCES 1. Mueller AK, Camargo N, Kaiser K, Andorfer C, Frevert U, Matuschewski K, Kappe SH. 2005. Plasmodium liver stage developmental arrest by depletion of a protein at the parasite-host interface. Proc. Natl. Acad. Sci. U. S. A. 102:3022–3027. 2. Mueller AK, Labaied M, Kappe SH, Matuschewski K. 2005. Genetically modified Plasmodium parasites as a protective experimental malaria vaccine. Nature 433:164 –167. 3. Jobe O, Lumsden J, Mueller AK, Williams J, Silva-Rivera H, Kappe SH, Schwenk RJ, Matuschewski K, Krzych U. 2007. Genetically attenuated Plasmodium berghei liver stages induce sterile protracted protection that is mediated by major histocompatibility complex class I-dependent interferon-gamma-producing CD8⫹ T cells. J. Infect. Dis. 196:599 – 607. 4. Hoffman SL, Goh LM, Luke TC, Schneider I, Le TP, Doolan DL, Sacci J, de la Vega P, Dowler M, Paul C, Gordon DM, Stoute JA, Church LW, Sedegah M, Heppner DG, Ballou WR, Richie TL. 2002. Protection of humans against malaria by immunization with radiation-attenuated Plasmodium falciparum sporozoites. J. Infect. Dis. 185:1155–1164. 5. Tarun AS, Dumpit RF, Camargo N, Labaied M, Liu P, Takagi A, Wang R, Kappe SH. 2007. Protracted sterile protection with Plasmodium yoelii pre-erythrocytic genetically attenuated parasite malaria vaccines is independent of significant liver-stage persistence and is mediated by CD8⫹ T cells. J. Infect. Dis. 196:608 – 616.
Infection and Immunity
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
Similarly, CD11c⫹ CD8⫹ T cells were the major subset responsible for IFN-␥ production, induction of target cell apoptosis in vitro, and reduction of viral titers in vivo (15). Finally, we demonstrated that the CD11c⫹ CD8⫹ T cell subset from PyUIS4⫺ parasite-immunized mice, but not those from mock control mice, significantly inhibited LS parasite development in primary hepatocytes (Fig. 4D), confirming that these cells are cytotoxic T cells against LS malaria. To determine why the majority of CD8 T cells from liver, but not other organs, express CD11c, we isolated CD11⫺ CD8⫹ T cells from spleens of PyUIS4⫺ parasite-immunized mice and cocultured them with LS-infected hepatocytes in vitro. We observed a 3to 4-fold increase in the number of CD11c⫹ CD8⫹ T cells after exposure to the LS-infected hepatocytes (Fig. 5), indicating that the significant increase in the level of CD11c expression on CD8⫹ T cells may require an interaction between CD8 T cells and the infected hepatocytes (9). While CD44 and CD11a are uniformly upregulated on activated T cells, tetramer analysis demonstrated that CD11c is expressed by a subset of CSP-specific T cells during the peak of the effector response, which is lost as cells transition from the effector phase to the memory phase (Fig. 6A and B). Thus, while high levels of CD44 or CD11a can be used to identify cells that have been previously activated, only CD11c can be used to distinguish between recently activated effector cells and circulating memory cells. This distinction can be especially important when examining T cells from the liver, in which activated and memory T cells are greatly overrepresented compared to other organs (32). These results led us to examine the relationship between CD11c expression and memory differentiation. We demonstrated that within the population of antigen-specific T cells in the liver (defined as CD8blo CD11ahi cells), the CD11c⫹ CD8⫹ T subset proliferated more than the CD11c⫺ CD8⫹ T subset at early time points but was also more likely to undergo apoptosis during the contraction phase (Fig. 6C). Thus, over time, the CD11c⫹ CD8⫹ T effector population that initially dominated the response gave way to the CD11c⫺ CD8⫹ T population (Fig. 1D). In addition, CD11c⫹ CD8⫹ T cells were predominantly KLRG1⫹ D127⫺, indicative of short-lived effector cells (Fig. 7A). The population of KLRG1⫺ CD127⫹ memory precursor cells, on the other hand, was exclusively CD11c⫺. In addition, microarray analyses comparing CD11c⫹ and CD11c⫺ CD8⫹ T subsets revealed that CD11c⫺ cells overexpress many genes important for memory and memory precursor cell development, including Eomes, SOCS3, Bcl2, and Tcf7 (Table 1) (21, 29). Taken together, these results demonstrate a negative correlation between CD11c expression and memory cell differentiation. The relative abundance of SLECs versus MPECs following infection has been shown to depend on the cytokine milieu and the degree of inflammation. Specifically, increased inflammation and cytokine expression, but not antigen expression, have been shown to drive increased SLEC differentiation at the cost of MPEC differentiation (21, 29). Interestingly, an increase of the dose of PyGAP SPZ led to increased CD11c expression by CD8⫹ T cells in the liver, but little is known about the activation of the innate immune system and induction of inflammatory cytokines such as IL-12 and IL-27 by PyGAP immunization. Our findings are also supported by the results of Bertolino et al., who reported that hepatocytes have the ability to induce short-lived effector CD8⫹ T cells in numbers comparable to those induced by dendritic cells (33). On the other hand, repeated PyGAP immunization skews
Short-Lived Effector CD8 T Cells Express CD11c
November 2013 Volume 81 Number 11
View publication stats
19.
20. 21.
22. 23. 24. 25. 26. 27.
28.
29.
30. 31. 32. 33.
CD11c⫹CD8⫹ T-cells in melanoma-bearing mice induced by anti-4-1BB monoclonal antibody. Mol. Cells 24:132–138. Rai D, Pham N-LL, Harty JT, Badovinac VP. 2009. Tracking the total CD8 T cell response to infection reveals substantial discordance in magnitude and kinetics between inbred and outbred hosts. J. Immunol. 183: 7672–7681. Springer T, Galfre G, Secher D, Milstein C. 1979. Mac-1: a macrophage differentiation antigen identified by monoclonal antibody. Eur. J. Immunol. 9:301–306. Joshi NS, Cui W, Chandele A, Lee HK, Urso DR, Hagman J, Gapin L, Kaech SM. 2007. Inflammation directs memory precursor and short-lived effector CD8⫹ T cell fates via the graded expression of T-bet transcription factor. Immunity 27:281–295. Sacci J. 2002. Hepatocyte perfusion, isolation, and culture. Methods Mol. Med. 72:503–505. Mercado R, Vijh S, Allen SE, Kerksiek K, Pilip IM, Pamer EG. 2000. Early programming of T cell populations responding to bacterial infection. J. Immunol. 165:6833– 6839. Wong P, Pamer EG. 2003. CD8 T cell responses to infectious pathogens. Annu. Rev. Immunol. 21:29 –70. Butler NS, Schmidt NW, Harty JT. 2010. Differential effector pathways regulate memory CD8 T cell immunity against Plasmodium berghei versus P. yoelii sporozoites. J. Immunol. 184:2528 –2538. Seder RA, Darrah PA, Roederer M. 2008. T-cell quality in memory and protection: implications for vaccine design. Nat. Rev. Immunol. 8:247– 258. Rodrigues MM, Cordey AS, Arreaza G, Corradin G, Romero P, Maryanski JL, Nussenzweig RS, Zavala F. 1991. CD8⫹ cytolytic T cell clones derived against the Plasmodium yoelii circumsporozoite protein protect against malaria. Int. Immunol. 3:579 –585. Smolewski P, Grabarek J, Halicka HD, Darzynkiewicz Z. 2002. Assay of caspase activation in situ combined with probing plasma membrane integrity to detect three distinct stages of apoptosis. J. Immunol. Methods 265:111–121. Obar JJ, Jellison ER, Sheridan BS, Blair DA, Pham Q-M, Zickovich JM, Lefrançois L. 2011. Pathogen-induced inflammatory environment controls effector and memory CD8⫹ T cell differentiation. J. Immunol. 187: 4967– 4978. Tsuji M, Zavala F. 2003. T cells as mediators of protective immunity against liver stages of Plasmodium. Trends Parasitol. 19:88 –93. Overstreet MG, Cockburn IA, Chen YC, Zavala F. 2008. Protective CD8 T cells against Plasmodium liver stages: immunobiology of an ‘unnatural’ immune response. Immunol. Rev. 225:272–283. Crispe I. 2009. The liver as a lymphoid organ. Annu. Rev. Immunol. 27:147–163. Bertolino P, Trescol-Biémont MC, Rabourdin-Combe C. 1998. Hepatocytes induce functional activation of naive CD8⫹ T lymphocytes but fail to promote survival. Eur. J. Immunol. 28:221–236.
iai.asm.org 4181
Downloaded from http://iai.asm.org/ on June 11, 2014 by guest
6. Nussenzweig RS, Vanderberg J, Most H, Orton C. 1967. Protective immunity produced by the injection of X-irradiated sporozoites of Plasmodium berghei. Nature 216:160 –162. 7. Weiss WR, Sedegah M, Beaudoin RL, Miller LH, Good MF. 1988. CD8⫹ T cells (cytotoxic/suppressors) are required for protection in mice immunized with malaria sporozoites. Proc. Natl. Acad. Sci. U. S. A. 85: 573–576. 8. Kumar S, Miller LH, Quakyi IA, Keister DB, Houghten RA, Maloy WL, Moss B, Berzofsky JA, Good MF. 1988. Cytotoxic T cells specific for the circumsporozoite protein of Plasmodium falciparum. Nature 334:258 – 260. 9. Trimnell A, Takagi A, Gupta M, Richie TL, Kappe SHI, Wang R. 2009. Genetically attenuated parasite vaccines induce contact-dependent CD8⫹ T cell killing of Plasmodium yoelii liver stage-infected hepatocytes. J. Immunol. 183:5870 –5878. 10. Hansen SG, Ford JC, Lewis MS, Ventura AB, Hughes CM, CoyneJohnson L, Whizin N, Oswald K, Shoemaker R, Swanson T, Legasse AW, Chiuchiolo MJ, Parks CL, Axthelm MK, Nelson JA, Jarvis MA, Piatak M, Lifson JD, Picker LJ. 2011. Profound early control of highly pathogenic SIV by an effector memory T-cell vaccine. Nature 473:523– 527. 11. White KL, Snyder HL, Krzych U. 1996. MHC class I-dependent presentation of exoerythrocytic antigens to CD8⫹ T lymphocytes is required for protective immunity against Plasmodium berghei. J. Immunol. 156: 3374 –3381. 12. Guebre-Xabier M, Schwenk R, Krzych U. 1999. Memory phenotype CD8(⫹) T cells persist in livers of mice protected against malaria by immunization with attenuated Plasmodium berghei sporozoites. Eur. J. Immunol. 29:3978 –3986. 13. Schmidt NW, Podyminogin RL, Butler NS, Badovinac VP, Tucker BJ, Bahjat KS, Lauer P, Reyes-Sandoval A, Hutchings CL, Moore AC, Gilbert SC, Hill AV, Bartholomay LC, Harty JT. 2008. Memory CD8 T cell responses exceeding a large but definable threshold provide long-term immunity to malaria. Proc. Natl. Acad. Sci. U. S. A. 105:14017–14022. 14. Schmidt NW, Butler NS, Badovinac VP, Harty JT. 2010. Extreme CD8 T cell requirements for anti-malarial liver-stage immunity following immunization with radiation attenuated sporozoites. PLoS Pathog. 6:e1000998. doi:10.1371/journal.ppat.1000998. 15. Beyer M, Wang H, Peters N, Doths S, Koerner-Rettberg C, Openshaw PJ, Schwarze J. 2005. The beta2 integrin CD11c distinguishes a subset of cytotoxic pulmonary T cells with potent antiviral effects in vitro and in vivo. Respir. Res. 6:70. doi:10.1186/1465-9921-6-70. 16. Kim YH, Seo SK, Choi BK, Kang WJ, Kim CH, Lee SK, Kwon BS. 2005. 4-1BB costimulation enhances HSV-1-specific CD8⫹ T cell responses by the induction of CD11c⫹CD8⫹ T cells. Cell. Immunol. 238:76 – 86. 17. Huleatt JW, Lefrançois L. 1995. Antigen-driven induction of CD11c on intestinal intraepithelial lymphocytes and CD8⫹ T cells in vivo. J. Immunol. 154:5684 –5693. 18. Ju S, Park S, Lee S, Kwon B, Kim B. 2007. Marked expansion of
