LCMV immunodominant peptide GP33–41
LCMV immunodominant peptide NP396–414
Peritoneal exudate cell
Vesicular stomatitis virus
CD11b, CD11c, and F4/80 are normally used to define dendritic cell and/or macrophage populations. In this study, the expression of all three markers was observed on CD8+ T cells following infection of mice with several distinct viruses. Using lymphocytic choriomeningitis virus as a model virus, it was found that relatively more CD11b+CD8+ and CD11c+CD8+ T cells were present in the periphery than in primary lymphoid organs; in contrast, the F4/80+CD8+ T cell population was more prevalent in the spleen. All three myeloid markers were detected on virus-specific CTL. The expression of CD11b and CD11c on CD8+ T cells correlated with their level of CTL activity, whereas the F4/80+CD8+ T cell population increased after the peak of the CTL response but did not have higher CTL activity. These data suggest that there is a differential induction of CD11b, CD11c, and F4/80 on virus-specific CD8+ T cells following an acute virus infection.
Various cell surface antigens have been utilized in attempts to identify specific subpopulations of leukocytes. Mac-1 (CD11b/CD18) is a member of the β2-integrin family of adhesion molecules, and can be found on monocytes, neutrophils, peritoneal B-1 cells, CD8+ dendritic cells, NK cells, and a subset of CD8+ T cells 1–3. CD11b binds a diverse group of ligands, which include inactivated complement component C3b (iC3b), fibrinogen, coagulant factor X, and the intercellular adhesion molecule (ICAM)-1. Functional roles of the molecule include cellular adhesion, phagocytosis, extravasation, and chemotaxis. CD11b was originally identified in the mouse as a marker for monocytes/macrophages and granulocytes 2, 3. Subsequently, CD11b was demonstrated as an activation marker for CD8+ CTL during a virus infection 1. Some reports have shown that CD11b expression becomes down-regulated when CD8+ T cells enter the memory phase, which could be used to phenotypically distinguish recently activated effectors cells from memory cells 4–6.
CD11c/CD18 (p150,95) is another member of the β2-integrin family of adhesion molecules. CD11c is expressed on monocytes, granulocytes, macrophages, NK cells, dendritic cells, and subsets of T and B lymphocytes 2, 3. CD11c has functions similar to CD11b. Kim et al. also demonstrated the expression of CD11c on mucosal CD8+ T cells following systemic vesicular stomatitis virus (VSV) infection 7. CD11c seems to be potentially useful as an activation marker for CD8+ T cells. However, there have been no further studies to define the expression of CD11c on antigen-specific CD8+ T effectors in secondary lymphoid organs during a virus infection.
F4/80, a 160-kD glycoprotein, comprises seven epidermal growth factor (EGF)-like domains in the NH2-terminal region 8. As a result of its macrophage-restricted expression, F4/80 has long been used as a marker for murine macrophages 8, 9. F4/80 is down-regulated upon stimulation of Langerhans cells as they migrate to draininglymph nodes, and is absent in T cell areas of the spleen and lymph nodes, implicating a possible role for this molecule in the retention or adhesion of macrophages and other cells in specific tissue areas 10. Although F4/80 knockout mice have been generated, there was no distinct phenotype detected in these mice 11.
In the current report, we have analyzed the expression of CD11b, CD11c, and F4/80 on CD8+ T cells following an acute viral infection. We found that all of these cell surface markers were expressed on the majority of virus-specific CTL.
2.1 CD8+ T cells express CD11b, CD11c, and F4/80 following an LCMV infection
Expression of the myeloid marker CD11b has been found on activated and memory CD8+ T cells from mice infected with LCMV, but neither CD11c nor F4/80 have been investigated in this system. To determine the level of CD11b, CD11c, and F4/80 expression on CD8+ T cells, splenocytes from uninfected or LCMV-infected C57BL/6 mice (3, 6, 8, and 10 days post-infection) were stained with mAb specific for these markers and analyzed by FACS.
As shown in Fig. 1, splenocytes from control mice consisted of a very low percentage of CD11b+ (0.2%), CD11c+ (0.4%), and F4/80+ (0.2%) CD8+ T cells. Upon infection with LCMV, the expression of CD11b, CD11c, and F4/80 on CD8+ T cells could be detected as early as day 6 post-infection. The percentage of CD11b+CD8+ and CD11c+CD8+ T cells in splenocytes was dramatically increased on day 8 post-LCMV infection (24% for CD11b+CD8+ and 43% for CD11c+CD8+, respectively). CD11b+ and CD11c+CD8+ T cells were reduced after the peak of the CTL response (i.e. day 10), whereas the percentage of F4/80+CD8+ T cells actually increased on day 10 (i.e. from 4.2% on day 8 to 7.6% on day 10), but decreased by day 14 (data not shown). Among these subpopulations, CD11c+CD8+ T cells were the most abundant, whereas F4/80+CD8+ T cells were the least. By three-color staining, we found that the majority of either CD11b+ or F4/80+ CD8+ T cells were also CD11c+ on day 8 post-LCMV infection (data not shown).
Naive CD8+ T cells express low levels of the activation/ memory marker CD44, whereas effector and memory CD8+ T cells are mostly CD44+1. We also found that all of the CD11b+, CD11c+, and F4/80+ CD8+ T cells were CD44+ (data not shown), indicating that the expression of these myeloid markers was restricted to activated CD8+ T cells.
2.2 Induction of CD11b, CD11c, and F4/80 on CD8+ T cells post-infection is not unique to an LCMV infection or to C57BL/6 mice
To determine whether other viruses could induce the expression of CD11b, CD11c, and F4/80 on CD8+ T cells, C57BL/6 mice were infected with LCMV, VSV, and vaccinia virus (VV) and the expression of these cell surface markers on CD8+ T cells was assessed by FACS. At the peak of the CTL response (day 8 for LCMV; day 6 for VSV and VV), CD11b+, CD11c+, and F4/80+ CD8+ T cells were all increased (Fig. 2). The pattern of this increase was similar between LCMV, VSV and VV infections. Thus, the CD11c+ population in CD8+ T cells was the largest, whereas the F4/80+ population was the smallest.
Because the Th1/Th2 profile in C57BL/6 is biased towards Th1 cytokines, we also investigated the ability of these viruses to induce these same CD8+ T cell subpopulations in BALB/c mice, which have a Th2 cytokine bias 12, 13. The induction of CD11b, CD11c, and F4/80 expression on CD8+ T cells following VSV and VV infections, as with LCMV, was also observed in BALB/c mice (Fig. 2). Thus, two different mouse strains showed a similar expression pattern of CD11b, CD11c, and F4/80 on CD8+ T cells during an acute infection with three different virus infections, demonstrating that CD11b, CD11c, and F4/80 are commonly induced on CD8+ T cells following a viral infection.
2.3 Tissue distribution of CD11b+, CD11c+, and F4/80+ CD8+ T cells
To determine the tissue distribution of the CD11b+CD8+, CD11c+CD8+, and F4/80+CD8+ populations induced post-infection, we analyzed different tissues from acutely infected (day 8) mice and compared them to naive mice. The tissues harvested for this part of the study included primary lymphoid organs (thymus and bone marrow), secondary lymphoid organs (spleen and lymph node), and peripheral tissues [PBMC, peritoneal exudate cells (PEC), and liver].
Interestingly, there was no expression of CD11b, CD11c, and F4/80 on CD8+ T cells in the thymus before or after infection. With the exception of the thymus, CD11b and CD11c expression could be detected on CD8+ T cells in all of the other tissues on day 8 post-LCMV infection (Table 1). The CD11c+CD8+ T cell population was greater in peripheral tissues (liver, PBMC, and PEC) than in secondary lymphoid organs (spleen and lymph node), whereas the percentage of CD11b+CD8+ T cells was higher in the liver than in other organs. The F4/80+CD8+ T cell population was much greater in the spleen (5.8%) than in the other organs. Although CD11b, CD11c, and F4/80 expression on CD8+ T cells could be detected in the bone marrow on day 8 post-infection, the fraction of CD8+ T cells in this tissue was low (control: 5–6%, LCMV: 7–9%). We also analyzed these markers on CD8+ T cells in these different organs on days 3, 6, and 10 post-LCMV infection. It was found that CD11b and CD11c could be expressed on CD8+ T cells as early as day 6 post-LCMV infection in the spleen, lymph node, PBMC, liver and PEC (the highest level of expression was observed on day 8), whereas F4/80 was not detected in the other organs with the exception of the spleen (data not shown).
2.4 Expression of CD11b, CD11c, and F4/80 on virus-specific CTL
Although we showed above that CD8+ T cells expressed CD11b, CD11c, and F4/80 post-infection, it was important to determine whether these markers were present on virus-specific CTL. To identify the expression of CD11b, CD11c, and F4/80 on virus-specific CD8+ T cells, we used MHC class I (H-2Db) tetramers loaded with the LCMV Db-restricted immunodominant peptides GP33–41 (GP33) and NP396–414 (NP396) 14, respectively, to stain the CD8+ T cell subpopulations. As expected, CD8+ T cells from day 8 LCMV-infected mice could be stained with these tetramers (Fig. 3A). The tetramer+CD8β+ population was then gated to assess CD11b, CD11c, and F4/80 expression on those cells. The majority of the tetramer+ cells expressed CD11b, CD11c, and F4/80, regardless of whether the Db tetramers were loaded with GP33 or NP396 (GP33: 84% CD11b+, 95% CD11c+, and 79% F4/80+; NP396: 69% CD11b+, 90% CD11c+, and 68% F4/80+). As seen in the bulk CD8+ T cell population (e.g. Fig. 1), tetramer+CD8+ T cells had higher levels of CD11c expression than either CD11b or F4/80. Therefore, these data suggest that virus-specific CD8+ T cells have CD11b, CD11c, and F4/80 on their cell surface.
The CD8+ splenocytes were further analyzed for the co-expression of CD11b, CD11c or F4/80 and intracellular IFN-γ following in vitro stimulation with GP33 and NP396 14 by three-color FACS analysis. C57BL/6 mice were infected with 1×105 PFU of LCMV and splenocytes were harvested on day 8 post-infection. As shown in Fig. 3B, CD11b, CD11c, and F4/80 expression was found on 74%, 96%, and 50% of the virus-specific (IFN-γ+) CD8+ T cells for GP33 stimulation, and 89%, 95%, and 50% for NP396 stimulation, respectively, on day 8 post-LCMV infection. These data are consistent with the tetramer staining observed on the CD8+ T cell subpopulation (Fig. 3A).
To determine whether there was a correlation between CD11b, CD11c, or F4/80 expression and CTL activity, we analyzed virus-specific killing by these CD8+ T cell subpopulations following an acute LCMV infection. CD8+ T cells were magnetically sorted into positive and negative populations based on CD11b, CD11c, or F4/80 expression. The sorted cells were then analyzed for virus-specific killing activity in a standard 6-h 51Cr-release assay. As shown in Fig. 3C, the CD11b+ and CD11c+ CD8+ T cell populations mediated higher levels of virus-specific killing activity than CD8+ T cells without these markers. There was no significant difference in killing activity between the F4/80+ and F4/80– CD8+ T cell populations.
In this study, we have shown that CD11b, CD11c and F4/80 are specifically expressed on LCMV-specific CD8+ T cells. F4/80 is generally used as a macrophage marker 8. However, this is the first report demonstrating F4/80 expression on activated CD8+ T cells. Therefore, F4/80 can be used as a marker for the subpopulations of activated CD8+ T cells, rather than simply used as a macrophage marker. In addition, we found that CD11c+CD8+ T cells had high levels of CTL activity following an acute infection. CD11c expression has been detected on activated CD8+ T cells in humans and mice under different stimulation conditions 7, 15–17; however, its relevance to virus specificity and functional antiviral CTL activity was not studied in those reports. We found that CD11b+CD8+, CD11c+CD8+, and F4/80+CD8+ T cells were mainly present in the CD44hi CD8+ T cell population, which is consistent with the expression of these markers on activated CD8+ T cells.
Because a high level of CD44 expression may not be an ideal marker to identify the effector and memory CD8+ T cells in some mouse strains 18, CD11c, which was shown in the current study to be specifically expressed on effector as well as memory CTL (data not shown), may thus be a better marker for these populations. CD11a is also a member of the β2-integrin family of adhesion molecules and its expression can be induced on CD8+ T cells following activation 15, 19. However, unlike CD11b and CD11c, its expression is already very high in naive CD8+ T cells, and is thus not ideal for identifying activated CD8+ cells (data not shown). We also analyzed the expression of CD11b, CD11c, and F4/80 on CD4+ T cells following LCMV infection. The percentage of CD4+ T cells expressing CD11b, CD11c, and F4/80 was very low (Y.L. and R.R.B., unpublished observations). This is consistent with a previous report indicating that CD11b+CD4+ T cells are barely detectable following a virus infection 1.
Several cytokines have been used to study the induction of CD11b and CD11c expression in vitro, including IL-2, IL-4, IL-7, IL-12, IL-15, GM-CSF, IFN-γ, and TNF-α 20–22. However, none of these cytokines could stimulate the expression of CD11b, CD11c, or F4/80 on naive CD8+ T cells or maintain their expression on effector CD8+ T cells (data not shown). We found that PMA + ionomycin treatment could maintain CD11b and CD11c expression on pre-activated CD8+ T cells (e.g. LCMV-specific CTL) in vitro (data not shown), suggesting that a continuous stimulation of these cells is necessary for expression of these markers. Restricted expression of CD11b and CD11c on activated (rather than naive) CD8+ T cells may be important for controlling the distribution of CD8+ T cells, which may prevent nonspecific (or bystander) tissue damage in sites other than where there is active virus replication. Unlike CD11b and CD11c, F4/80 expression was increased after culture (data not shown), indicating that such treatment is a better means to induce F4/80 on CD8+ T cells. In addition, F4/80 expression on CD8+ T cells from either naiveor immune splenocytes increased dramatically after PMA + ionomycin treatment (data not shown). Thus, the induction of F4/80 expression on CD8+ T cells may not be specific for a viral infection, and occurs as a consequence of activation.
Antigen-specific CD8+ T cells have been shown to migrate to nonlymphoid tissues and can be long-lived memory cells 23, 24. CCR5 expression on humanCD8+ T cells was suggested to be critical for the migration of antigen-specific effector and differentiated memory CD8+ T cells to inflamed tissues and secondary lymphoid organs 25. Nielsen et al. have suggested that CD11b can facilitate the homing of CD8+ T cells to inflammatory sites, as an anti-CD11b antibody prevented local injury following the adoptive transfer of antigen-primed donor cells 26. Consistent with that report, our data showed a higher percentage of CD11c+CD8+ and CD11b+CD8+ T cells in the periphery than in lymphoid tissues, suggesting that the expression of CD11c and CD11b on CD8+ T cells might confer to these cells the ability to adhere to endothelial cells and may be necessary for chemotaxis and homing.
The ligand for (and function of) F4/80 molecules has not been clearly defined. Warschkau et al. have shown that an anti-F4/80 mAb could inhibit IL-12, TNF-α and IFN-γ production in macrophages and NK cell cultures, suggesting that F4/80 may be involved in the direct cell-to-cell signaling during macrophage/NK cell interactions 9. In the current study, F4/80 expression on activated CD8+ T cells was mainly found in the spleen. The percentage of F4/80+CD8+ T cells in the spleen was even higher after the peak of an acute CTL response (i.e. day 10 post-infection; data not shown). Therefore, the expression of F4/80 on virus-specific CD8+ T cells may not be important in migration. Instead, it may help CD8+ T cells interact with other cells in the spleen and possibly play a role in the functional maturation of CD8+ T cells. Other markers, such as CCR7, are known to define subsets of memoryand effector T lymphocytes with different homing potentials in the human system 27, suggesting that F4/80 might be useful in identifying the murine equivalent of "central memory" and "effector memory" T cells, as defined by Lanzavecchia and colleagues 27.
Our report demonstrating the expression of myeloid cell surface markers on virus-specific CD8+ T cells is the beginning of studies that will help us to understand the functional relevance of these markers. In addition, the differences in CD11b, CD11c and F4/80 expression on acute antiviral CD8+ T cells will enable monitoring of the CD8+ T cell compartment during the progression and generation of CTL in response to an acute virus infection.
4 Materials and methods
Male and female C57BL/6 and BALB/c mice were obtained from the Jackson Laboratory (Bar Harbor, ME). All mice were age- and sex-matched and were used between 6 and 12 weeks of age. All animal procedures were approved by the Indiana University School of Medicine Institutional Animal Care and Use Committee.
4.2 Virus and infection of mice
The Armstrong strain of LCMV was kindly provided by Dr. R. M. Welsh (University of Massachusetts Medical Center, Worcester, MA). Viral stocks were propagated in baby hamster kidney (BHK) cellsand titrated on Vero cells. The Indiana strain of VSV, kindly provided by Drs. J. Yewdell and J. Bennink (Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD) was prepared and titrated in BHK cells. The WR strain of VV, also provided by Drs. Yewdell and Bennink, was prepared and titrated in human osteosarcoma 143B cells. Mice were infected i.p. with 1×105–2×105 PFU of LCMV, or 1×106 PFU of VSV and VV.
4.3 Organ isolation
Organs including thymi, bone marrow, lymph node, spleen, liver mononuclear cells (MNC), and PEC, and peripheral blood were harvested from uninfected and LCMV-infected mice at the indicated time points and processed into single-cell suspensions. Splenocyte, liver MNC, and PEC isolation was described previously 28. Peripheral blood was collected from euthanized mice via cardiac puncture and treated with 0.5 mM EDTA and stained with the appropriate mAb. Erythrocytes were lysed in 0.84% NH4Cl and washed twice with Iscove's modified Dulbecco's medium (BioWhittaker, Walkersville, MD) containing 10% FBS, 2 mM L-glutamine and antibiotics (complete medium). Aliquots were used for FACS analysis.
4.4 Antibodies and FACS analysis
Single-cell suspensions were stained for cytofluorography with the following anti-mouse mAb, all purchased from PharMingen (San Diego, CA): biotinylated anti-mouse CD8β, FITC-conjugated mAb against mouse TCRβ, CD11b, and CD11c, and PE-conjugated mAb against CD8α. FITC- or PE-conjugated rat IgG2b mAb was used as an isotype control. The biotinylated mAb were subsequently stained with streptavidin-allophycocyanin (APC). FITC-conjugated anti-F4/80 mAb was from Serotec, Inc. (Raleigh, NC). Mouse H-2Db tetramers loaded with GP33 and NP366 14 and conjugated with PE were kindly provided by Dr. Liisa Selin (University of Massachusetts Medical School, Worcester, MA). The labeling procedure and FACS analyses were performed as described previously 28, 29. Samples were analyzed on a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA), with 10,000–50,000 events acquired per sample as indicated. For tetramer staining, cells were stained with tetramers at 4°C for 1 h, followed by staining with the mAb. For these experiments, 100,000 events were acquired per sample.
4.5 Cell sorting
Different subpopulations of cells were immunomagnetically separated using the MACS system (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions.
4.6 Cytotoxicity assay
To analyze LCMV-specific CTL activity, MC57G (H-2b) cells were infected with LCMV at a multiplicity of infection of 0.1. Two days later, the cells were labeled with Na251CrO4 and used as targets in a 6-h 51Cr-release assay as previously described 29.
4.7 Cytofluorimetric analysis of intracellular cytokines
Splenocytes from uninfected or LCMV-infected mice were cultured at 37°C in 5% CO2 for 6 h in the presence or absence of synthetic GP33 or NP396 14, purchased from Sigma-Genosys (Pittsburgh, PA), at a concentration of 1 μg/ml in complete medium. Brefeldin A (Sigma) was added at a final concentration of 2 μg/ml for the entire incubation period. After incubation, the cells were washed and pretreated with 2.4G2 supernatant. The cells were next stained for CD8β, CD11b, CD11c, and F4/80, and then permeabilized using the Cytofix/Cytoperm kit (PharMingen) according to the manufacturer's instructions. Intracellular antigens were detected using a PE-conjugated anti-IFN-γ mAb (PharMingen). For these experiments, 50,000 events were acquired per sample.
The authors would like to thank Drs. Ray Welsh, Liisa Selin, Jon Yewdell and Jack Bennink for kindly providing reagents used in this study. The technical assistance of Phil Spence and David Jay is also gratefully acknowledged. This work was supported in part by the National Institutes of Health, Grant RO1 AI46455 to R.R.B. T.J.R. was supported by a Predoctoral Minority Fellowship (AI46455–01S1) from the National Institutes of Health.