INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Major histocompatibility complex (MHC) class I-restricted cytotoxic T lymphocytes (CTL) can play a critical role in antiviral immune responses (40). These T cells facilitate viral clearance by directly lysing target cells harboring productive virus infection. In the case of the human immune response to herpes simplex virus type 1 (HSV-1), CD8⁺ T cells can be isolated from infected individuals (32). In addition, the recent identification of the HSV-encoded ICP47 protein, which blocks class I-restricted peptide presentation, suggests that CTL lysis plays an important role during the antiviral response (11, 15, 39). In mouse model systems, CD8⁺ T cells can be isolated from lymph nodes draining the site of cutaneous primary infection, and these T cells effectively protect animals against subsequent infection (2, 3, 13, 22, 28). However, unlike many primary antiviral responses, T cells isolated from lymph nodes draining sites of cutaneous HSV-1 infection did not appear to kill infected target cells without being subjected to a period of in vitro culture, usually in the absence of exogenous antigen (6, 27, 30). This culture period was thought to reflect a need for extralymphoid differentiation into armed cytotoxic effectors capable of dealing with the peripheral infection. Such a proposal is consistent with the notion that while CTL precursors are activated within the draining lymph nodes, the cytotoxic effectors are required only within the actual sites harboring infective virus, in this case the infected footpads.

We have examined the CD8⁺ CTL response to footpad infection with HSV-1. This response is dominated by T cells specific for a single K^b-restricted determinant derived from the glycoprotein B (gB) (28, 34). We have shown that these CD8⁺ T cells are restricted with respect to their T-cell receptor (TCR) expression with approximately 60% bearing a V10⁺ -chain containing a highly conserved junctional sequence (9). This pattern of TCR expression permitted visualization of HSV-specific T cells directly ex vivo and revealed that this specificity dominated the activated CD8⁺ T-cell subset (8). The results suggested that these T cells may be fully activated within the lymph node environment and therefore would potentially not require any period of peripheral arming. We show here that cytotoxicity can indeed be isolated directly from the lymph nodes draining sites of cutaneous infection and further show that the in vitro culture period involves a rapid expansion of gB-specific CTL.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mice, virus, and cell lines. C57BL/6 mice were purchased from the Central Animal Facility at Monash University, Clayton, Victoria, Australia. The KOS strain of HSV-1 was propagated, and the titers were determined by using Vero cells. The gB peptide with the sequence SSIEFARL (single letter amino acid code) corresponding to residues 498 to 505 in HSV-1 gB was synthesized by using an Applied Biosystems model 431A synthesizer (ABI, Foster City, Calif.) and kindly provided by J. Fecondo, Swinburne University of Technology, Hawthorn, Victoria, Australia. The H-2^b thymoma cell line EL4 was grown in Dulbecco's modified Eagle medium (DMEM) containing 10% fetal calf serum (FCS), 4 mM glutamine, 5 × 10⁵ M 2-mercaptoethanol, and antibiotics (complete DMEM).

Injection with HSV-1. Mice were injected in each hind footpad with 4 × 10⁵ PFU HSV-1 in phosphate-buffered saline (PBS), and the draining popliteal lymph nodes were removed 5 days later. Viable cell counts were determined for all lymph node samples by trypan blue exclusion prior to analysis. Lymph node cells were analyzed directly ex vivo or after 3 days in culture in wells of a 96-well flat-bottom plate at a density of 10⁶ cells/well in 250 µl RPMI containing 10% FCS, 2 mM glutamine, 5 × 10⁵ M 2-mercaptoethanol, and antibiotics (complete RPMI) without exogenous antigen.

Assessment of gB-specific CTL activity. CTL lysis was assessed by a 4-h chromium release assay with ⁵¹Cr (150 µCi)-labeled EL4 cells that had or had not been pulsed with 1 µg of gB peptide per ml. Effector/target ratios ranged from 100:1 to 0.1:1 in the assay. Results are expressed as a percentage of specific lysis with spontaneous lysis less than 20% for target cells.

H-2K^b/gB tetramer. Tetrameric H-2K^b/gB peptide complexes were prepared essentially as described by Altman and coworkers (1). Briefly, recombinant H-2K^b and human ₂-microglobulin, produced in Escherichia coli, were dissolved in urea and injected together with the gB peptide into a refolding buffer consisting of 100 mM Tris pH 8.0, 400 mM arginine, 2 mM EDTA, 5 mM reduced glutathione, and 0.5 mM oxidized glutathione. Refolded complexes were purified by anion-exchange chromatography by using DE52 resin (Whatman International, Ltd., Maidstone, England) followed by gel filtration through a Superdex 75 column (Amersham Pharmacia Biotech, Uppsala, Sweden). The refolded H-2K^b/gB peptide complexes were biotinylated by incubation for 16 h at 30°C with the BirA enzyme (Avidity, Boulder, Colo.). Tetrameric MHC-peptide complexes were produced by the stepwise addition of extravidin-conjugated phycoerythrin (Sigma, St. Louis, Mo.) to achieve a 1:4 molar ratio (extravidin-phycoerythrin-biotinylated H-2K^b/gB peptide complexes).

Flow cytometry. Lymph node cells analysed directly ex vivo or after 3 days in culture without exogenous antigen were triple stained with allophycocyanin-labeled anti-CD8 (53-6.7; Pharmingen, San Diego, Calif.), phycoerythrin-labeled anti-CD25 (PC61; Pharmingen), and biotin-labeled anti-V10 (B21.5; Pharmingen), followed by streptavidin-fluorescein isothiocyanate (FITC) (Molecular Probes, Inc., Eugene, Oreg.). Dead cells were excluded by using propidium iodide, and the cells were visualized on a Becton Dickinson FACScalibur. Live gates were set on lymphocytes, using forward and side scatter profiles, and 100,000 live cells were collected for analysis.

CFSE labeling. Popliteal lymph nodes from three mice injected in the hind footpads 5 days earlier with HSV-1 were removed, the two lymph nodes for each mouse were pooled, and viable cell counts were determined for each lymph node sample. Cells were resuspended in PBS containing 0.1% BSA at a density of 10⁷ cells/ml. For fluorescence labeling, 2 µl of a 5,6-carboxy-succinimidyl-fluorescein-ester (CFSE; Molecular Probes, Inc.) stock solution (5 mM in dimethyl sulfoxide) was incubated with 10⁷ cells for 10 min at 37°C as previously described (16, 21). Viable cell counts were again determined for each sample, and the cells were cultured for 3 days in wells of a 96-well flat-bottom plate at a density of 10⁶ cells/well in 250 µl of complete RPMI without exogenous antigen. Cells were then stained with allophycocyanin-labeled anti-CD8 (53-6.7; Pharmingen) and either biotin-labeled anti-V10 (B21.5, Pharmingen) or biotin-labeled anti-V3 (KJ25; Pharmingen), followed by phycoerythrin-streptavidin (Southern Biotechnology Associates, Birmingham, Ala.). Dead cells were excluded by using propidium iodide, and the cells were visualized with a Becton Dickinson FACScalibur apparatus. Live gates were set on lymphocytes, using forward and side scatter profiles, and 50,000 live cells were collected for analysis.

PCR analysis. The hind feet and draining popliteal lymph nodes from HSV-1-footpad injected mice were removed 12, 24, and 72 h postinoculation and digested overnight at 55°C in 500 µl of a solution consisting of 50 mM Tris-HCl (pH 8.0), 100 mM EDTA, 0.5% sodium dodecyl sulfate and 0.52 mg of proteinase K per ml. Genomic DNA was isolated by phenol-chloroform extraction, followed by precipitation with ethanol. The pellets were washed with 70% ethanol and resuspended in 100 µl of TE buffer (10 mM Tris-HCl, pH 8.0; 1 mM EDTA). HSV-1 DNA was amplified by PCR by using 25 ng of genomic DNA, 50 pmol of primers, 0.2 mM deoxynucleoside triphosphate and 1.5 U of Taq polymerase (Gibco BRL, Rockville, Md.). The primers used for PCR amplification were HSV-1a (5'-CCCTGTCTCGCGCGACGGAC-3') and HSV-1b (5'-TCACCGACCCATACGCGTAA-3') (20). Amplification involved a single cycle of 95°C for 5 min, 55°C for 1 min, and 72°C for 2 min, followed by 30 cycles of 93°C for 30 s, 55°C for 1 min, and 72°C for 1 min and one final cycle of 93°C for 30 s, 55°C for 1 min, and 72°C for 7 min with a DNA Thermal Cycler (Perkin-Elmer Cetus, Norwalk, Conn.). For each specimen, 25 ng of genomic DNA was evaluated for amplification competence by PCR by using insulin primers (5'-CGAGCTCGAGCCTGCCTATCTTTCAGGTC-3' and 5'-CGGGATCCTAGTTGCAGTAGTTCTCCAG-3'). A 20-µl aliquot of each PCR product was analyzed by electrophoresis on a 2% agarose gel stained with ethidium bromide.

Determination of virus titer. The hind feet of mice infected with HSV-1 were removed at the ankle, and individual feet were ground in a 5-ml homogenizer (Laboratory Supply) with 1.5 ml of complete RPMI to make a 20% (wt/vol) suspension. The draining popliteal lymph nodes were similarly ground in a homogenizer to produce a 2% (wt/vol) suspension. These suspensions were frozen at 70°C and then thawed rapidly and centrifuged at 12,000 × g for 10 min at 4°C. The supernatant fluid was used immediately in a PFU assay on Vero cells to determine the virus titer. Briefly, serial 10-fold dilutions of the supernatant fluids were made in serum-free minimal essential medium (MEM) and added to confluent monolayers of Vero cells in six-well multiwell dishes (0.9 ml of supernatant/well). After 1 h of incubation at room temperature with occasional rocking, 3 ml of 1% agarose in MEM-2 (MEM containing 2% FCS, 4 mM glutamine, 5 × 10⁵ M 2-mercaptoethanol, and antibiotics) was added to each well. The plates were incubated at 37°C for 4 days before being fixed with 10% formalin in phosphate buffer (4 ml/well). After 1 h of incubation at room temperature, the formalin and agarose were removed, and the cell monolayers were stained with 0.01% crystal violet to visualize the plaques. Results are expressed as the log of the number of PFU of virus per tissue (i.e., footpad or lymph node). The minimum level of virus that could be reliably detected was log₁₀ 0.48.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

HSV-1 can be detected in footpads after immunization but not in the draining popliteal lymph nodes. Cutaneous infection with HSV-1 has in the past been associated with an absence of cytotoxicity in lymph nodes draining the site of infection (27, 30). This lack of lytic activity has been attributed to an absence of virus replication in this site since it has been assumed that cytotoxicity would only be necessary in the presence of virus-infected cells (17). We wanted to confirm the absence of virus within these draining lymph nodes. Figure 1 shows a comparison of virus isolation from footpads after infection with HSV-1 and from popliteal lymph nodes draining the infected tissues. It is obvious that replicating virus is found within the site of cutaneous infection but not in the draining lymph nodes. Thus, it would seem that any lymph node-derived cytotoxicity would not be required for actual clearance of infectious virus from this site.

View larger version (19K): [in this window] [in a new window]   FIG. 1.   Virus isolation from footpads and draining lymph nodes following HSV-1 infection. Mice were infected in the hind footpads with 4 × 105 PFU of HSV-1 and 1 to 8 days later the hind feet and popliteal lymph nodes were removed and homogenized. The resulting supernatant fluids were used in a PFU assay on Vero cells to determine the virus titer. The results are expressed as the log of the number of PFU of virus per tissue (i.e., footpad or lymph node). Each bar represents the mean and standard deviation of the footpads or lymph nodes from three mice. The minimum level of detection is log10 0.48.

View larger version (51K): [in this window] [in a new window]   FIG. 2.   Presence of virus DNA in footpads and draining lymph nodes following HSV-1 infection. Mice were infected in the hind footpads with 4 × 105 PFU of HSV-1, and 12, 24, and 72 h later the hind feet and popliteal lymph nodes were removed and digested overnight with proteinase K. The genomic DNA was isolated by phenol-chloroform extraction, and 25 ng of genomic DNA was amplified by PCR with primers specific for HSV-1 or for insulin. Genomic DNA, extracted from popliteal lymph node cells which had been isolated from mice 5 days after HSV-1 infection and cultured for 3 days without exogenous antigen, was also amplified by PCR with primers specific for HSV-1 or for insulin (in vitro).

In vitro culture of popliteal lymph node cells results in an increase in HSV-specific CTL activity. Early studies had shown that while CTL from lymph nodes draining sites of HSV-1 infection failed to lyse infected targets directly ex vivo, these same cells could be converted to good effectors by culturing them for 3 days in vitro in the absence of exogenous antigen (30, 31). We had previously shown that HSV-1 gB-specific CD8⁺ T cells could be detected within the activated subset of popliteal lymph nodes draining infected tissues with the V10 marker in combination with TCR -chain sequences, which are both signatures for this specificity (8). We reasoned that by using relatively high effector/target ratios we might detect some cytotoxicity associated with these gB-specific T cells without the need for in vitro culture. Figure 3 shows this to be the case. However, these effectors are still about 40-fold less efficient, on the basis of the effector/target ratio that gives half maximal lysis, than those derived after 3 days in culture.

View larger version (17K): [in this window] [in a new window]   FIG. 3.   An increase in gB-specific CTL is evident after 3 days of culture without exogenous antigen. Three mice were infected in the hind footpads with 4 × 105 PFU of HSV-1, and 5 days later the cells from the individual draining popliteal lymph nodes were collected, viable cell counts were performed, and 1 × 106 to 2 × 106 cells were placed into culture without exogenous antigen. The remainder of the cells were analysed directly ex vivo in a 4-h chromium release assay with 51Cr-labeled EL4 cells that had or had not been pulsed with 1 µg of gB peptide per ml at the effector/target ratios indicated. After 3 days the cultured cells were harvested, viable cell counts were performed, and the in vitro cells were also analyzed in a 4-h chromium release assay. Specific lysis of 51Cr-labeled EL4 cells alone was <2% by both ex vivo and in vitro effector cells.

The increase in cytotoxicity evident after in vitro culture correlates with a corresponding expansion of CD8⁺ T cells expressing biased V10⁺ TCR indicative of gB specificity. The increase in cytotoxicity after in vitro culture could be linked to maturation to full effector function or "arming" of the effector cells (23, 24). The alternative explanation is that this increase reflects the in vitro expansion of the specific T cells. We had previously shown that the gB-specific CD8⁺ T cells preferentially express a V10⁺ TCR with a highly restricted pattern of CDR3 sequence conservation (9). This V10⁺ T-cell bias can be used to detect the presence of gB-specific CD8⁺ T cells within freshly isolated lymph node cells where they preferentially reside in the activated T-cell subset (8). Figure 4 shows the preferential expression of V10⁺ TCR in the CD25⁺ CD8⁺ lymph node population, where CD25 is used as a marker for activation. Approximately 45% of the activated CD8⁺ cells isolated directly ex vivo are V10⁺. However, this represents only 0.4% of all lymph node cells. In contrast, 14% of cells recovered after the 3-day in vitro culture of these popliteal cells are V10⁺ CD8⁺ CD25⁺, indicative of a 35-fold increase in the proportion of cells specific for gB. Indeed, the CD25⁺ CD8⁺ cultured lymph node cells are clearly gB specific, as determined by direct staining with a tetrameric class I reagent consisting of H-2K^b and the gB peptide (Fig. 5). This 35-fold increase in the proportion of gB-specific CD8⁺ T cells is comparable to the 40-fold increase in CTL activity evident after the in vitro culture (Fig. 3).

View larger version (36K): [in this window] [in a new window]   FIG. 4.   V10 expression by CD8+ CD25+ activated T cells from the draining lymph nodes of HSV-1-infected mice determined directly ex vivo or after 3 days of culture without exogenous antigen. Three mice were infected in the hind footpads with 4 × 105 PFU of HSV-1, and 5 days later the cells from the draining popliteal lymph nodes were collected, viable cell counts were performed, and 1 × 106 to 2 × 106 cells were placed into culture without exogenous antigen. Another 106 cells were triple stained with V10, CD25, and CD8 antibodies prior to analysis by flow cytometry. After 3 days the cultured cells were similarly analyzed by flow cytometry. Dead cells were excluded by using propidium iodide staining. The dot plots represent CD8 versus CD25 staining for the popliteal lymph node cells ex vivo, and after 3 days in culture without exogenous antigen (in vitro). The histogram shows the V10 receptor expression of the CD8+ CD25+ T cell subsets. The percentage of CD8+ CD25+ cells in the lymph nodes, and the percentage of V10+ T cells among the CD8+ CD25+ T cell subsets are shown as a mean and standard deviation for the three mice analyzed.

View larger version (33K): [in this window] [in a new window]   FIG. 5.   The CD8+ CD25+ T cells from the draining lymph nodes of HSV-1-infected mice analyzed after 3 days of culture without exogenous antigen are predominantly gB specific. A mouse was infected in the hind footpads with 4 × 105 PFU of HSV-1, and 5 days later the cells from the draining popliteal lymph nodes were collected and placed into culture without exogenous antigen. After 3 days, the cultured cells and lymph node cells isolated from a naive mouse were triple stained with CD25 and CD8 antibodies, and the H-2Kb/gB tetramer and analyzed by flow cytometry. Dead cells were excluded by using propidium iodide staining. The dot plot represents CD8 versus CD25 staining for the cultured popliteal lymph node cells. The histograms show the level of H-2Kb/gB tetramer staining of the CD8+ CD25hi and CD8+ CD25lo T cell subsets from the cultured (shaded) and naive (unshaded) lymph node cell populations.

View larger version (10K): [in this window] [in a new window]   FIG. 6.   Proliferation of gB-specific CD8+ T cells following 3 days of culture. Three mice were infected in the hind footpads with 4 × 105 PFU of HSV-1, and 5 days later the cells from the draining popliteal lymph nodes were collected and labeled with the fluorescent dye CFSE prior to culture without exogenous antigen for 3 days. The cells were then stained with CD8 and V10 or V3 antibodies prior to analysis by flow cytometry. Dead cells were excluded by using propidium iodide staining. The histograms show the proliferation (CFSE fluorescence) of the CD8+ V10+ (gB-specific population) or CD8+ V3+ (control population) T cells. The percentage of proliferating cells in each subset is shown as a mean and standard deviation for the three mice analyzed.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Cutaneous inoculation with HSV-1 results in the initial replication of virus in epithelial tissues, followed by infection of neuronal cells and the subsequent movement of virus to the sensory ganglia (36). We show here that while infectious virus does not make its way into the lymph nodes draining the site of infection, the HSV-specific CTL response appears to be nonetheless initiated in these lymphoid areas. This initiation is apparent from the actual development of cytolytic activity (Fig. 3) and previously, from the accumulation of activated CD8⁺ T cells bearing gB-specific TCR -chain sequences with the draining lymph nodes (8).

Given the absence of infection in the draining lymph nodes, CTL priming requires that antigen, but not virus, is transported to this site, most likely by professional antigen-presenting cells (APCs) such as dendritic cells (5, 14). Dendritic cell migration is thought to occur as a consequence of various forms of tissue insult, such as microbial infection (38). The initial activation of CTL directed to antigens found in peripheral tissues can occur within draining lymph nodes in the total absence of any lymphoid-based antigen expression (19). This process of class I-restricted indirect presentation of antigen is thought to play a central role in immune surveillance of antigens whose expression is confined to parenchymal tissues (5, 14). In the case under study here, the peripheral antigen is virus protein arising from cutaneous HSV-1 infection. This is in contrast to certain other localized infections, such as those involving influenza and Sendai viruses, where infected APCs can be isolated from draining lymph nodes where they are presumably directly involved in CTL priming (12, 37).

Given the demonstration that virus is not found within the lymph node compartment, it would have seemed reasonable to suggest that while the specific CTL are activated within this site, full maturation to cytotoxic effectors might occur after the T cell have emerged and commenced their migration to infected tissues. Clearly, this was the sentiment emerging from the previous studies showing an absence of cytotoxicity from lymph node cells isolated directly ex vivo which increased to relatively high levels after a period of in vitro culture (17, 27, 30). However, we have shown that cytotoxic effector function can be detected within these draining lymph nodes, albeit only at modest levels that were probably missed by previous investigators using relatively low effector/target ratios combined with HSV-infected target cells, which are probably less sensitive than the peptide-pulsed targets used here. We took advantage of this CTL detection in combination with the TCR V10 marker, which is a signature of gB-specific T cells (9), to quantify the gB-specific CD8⁺ T-cell numbers. These calculations show that the level of cytotoxicity within the lymph nodes appears to be equivalent on a specific-cell basis to that found after the in vitro culture period.

We cannot exclude some level of extralymphoid effector cell maturation. Recent studies by McNally and coworkers also showed that the appearance of cytotoxicity corresponded to the increased numbers of CD8⁺ CD25⁺ cells within the cultures (23, 24). However, this was attributed to a "conversion" or acquisition of cytotoxic effector function by the CTL precursors. In contrast, our results suggest that the in vitro-mediated increased cytotoxicity is largely a direct result of activation, and probably proliferation, of the gB-specific CTL. In other words, the gB-specific CD8⁺ T cells found within the draining lymph nodes appear to be fully mature cytotoxic effectors. This rapid CTL arming is in line with results from Oehen and Brduscha-Riem (29) showing that CD8⁺ T cells are fully cytotoxic even after just one round of division. Moreover, the appearance of cytotoxic effector function in that study preceded the downregulation of the lymph node homing receptor, CD62L. Combined, these results suggest that effector function appears very shortly after T-cell activation and probably before egress from the lymph nodes. The fact that this occurs in the absence of virus infection within the lymph nodes suggests that the arming may require other elements found in this location, such as a cytokine milieu compatible with the maturation to the fully cytotoxic form of the CD8⁺ T cell (10, 33, 35).

Finally, the apparent paucity of direct ex vivo cytotoxicity in the HSV-1 response is in sharp contrast to other antiviral responses, such as that to lymphocytic choriomeningitis virus (LCMV), where there is a very strong level of CD8⁺ T cell-mediated killing (7, 25). However, LCMV involves a much more widespread, disseminating viral infection that includes all lymphoid tissues compared to the much more localized cutaneous HSV-1 infection examined here. Consequently, while up to 70% of splenic CD8⁺ T cells are specific for LCMV (4, 26), only about 0.4% of cells within the draining lymph nodes are directed to the major HSV-1-derived target determinant (see Fig. 4), although these cells dominate the activated lymph node T-cell subset (8). Therefore, while the lack of lymphoid-based virus infection might have little consequence on CTL arming it could have considerable impact on the level of effector expansion. This proliferation could well require secondary antigen recognition by the armed effectors, which in this case would occur within the infected tissues. The issue of tissue-based T-cell proliferation has not been examined in detail although recent experiments examining the fate of autoreactive T cells has suggested that this can be an important mechanism for CTL expansion (18).

Overall, our results reinforce the notion that cutaneous infection with HSV-1 involves CTL priming within lymph nodes draining the site of infection by antigen transfer to this site. This priming results in the full activation of CTL effectors despite the lack of active lymph node infection. While these cells ultimately leave this site to deal with cells harboring replicating virus, their effector status appears to be determined early in the activation process and does not require any level of extralymphoid maturation.