4-1BBL−/− mice exhibit normal primary CD8 T cell responses to influenza virus, but show decreased CD8 T cell numbers late in the primary response as well as decreased secondary responses. In contrast, CD28−/− mice are defective in initial CD8 T cell expansion. Using agonistic anti-4-1BB Ab to replace the CD28 or 4-1BB signal, we examined the timing of the required signals for CD28 vs 4-1BB costimulation. A single dose of agonistic anti-4-1BB Ab added only during priming restores the secondary CD8 T cell response in CD28−/− mice. Once the T cell numbers in the primary response reach a minimum threshold, a full secondary response is achieved even in the absence of CD28. In contrast, anti-4-1BB added during priming fails to correct the defective secondary response in 4-1BBL−/− mice, whereas addition of anti-4-1BB during challenge fully restores this response. Thus, there is a switch in costimulatory requirement from CD28 to 4-1BB during primary vs recall responses. Adoptive transfer studies show that T cells primed in 4-1BBL−/− or wild-type mice are equally capable of re-expansion when rechallenged in wild-type mice. These studies rule out a model in which signals delivered through 4-1BB during priming program the T cells to give a full recall response and suggest that 4-1BB-4-1BBL interactions take place at later stages in the immune response. The results indicate that anti-4-1BB or 4-1BBL therapy will be most effective during the boost phase of a prime-boost vaccination strategy.
The ability of the immune system to mount faster and more efficient responses to secondary encounter with foreign Ag can be crucial to the host’s survival (1). Simultaneous engagement of TCR and CD28 during the initial stages of T cell activation leads to the production of IL-2 and survival of the T cell, with the net effect being T cell expansion (2). In addition to an initial costimulatory signal provided by CD28, other inducible costimulatory molecules are thought to be important in sustaining or differentiating initial immune responses (3, 4). For example, the inducible TNF receptor family member OX40 plays an important role in secondary CD4 T cell responses and in sustaining CD4 T cell survival (5, 6, 7, 8, 9). For CD8 T cells the inducible TNF receptor family member 4-1BB plays a similar role in sustaining T cell survival and memory (10, 11, 12, 13, 14, 15, 16, 17, 18). Once a CD8 T cell has reached a threshold for activation, it initiates a program of autonomous cell division without a requirement for further antigenic stimulation (19, 20, 21, 22). This is followed by a contraction phase in which the majority of effector cells are destroyed, leaving behind a small pool of memory T cells (23). This destructive process is also preprogrammed, occurring at the same time regardless of Ag dose or rate of clearance (24). The findings that CD8 T cell responses are preprogrammed and that only a 20-h stimulation with Ag is required for CD8 T cells to initiate their expansion program raise the issue of when inducible costimulatory molecules come into play (25).
The immune response to influenza in mice provides an excellent model to study the issue of timing of costimulatory signals for T cell activation, in that the effects of CD28 vs 4-1BB-mediated costimulation are temporally segregated. In the absence of CD28 costimulation, there is only minimal CD8 T cell expansion after influenza A virus infection (13, 14, 26). In contrast, in the absence of 4-1BBL, both the initial expansion and the contraction of CD8 T cell numbers are indistinguishable from those in wild-type (WT)4 mice (14). However, late in the primary response, when T cell numbers are declining very slowly, a significant decrease in CD8+CD62Llow influenza-specific T cells is observed in the spleens of 4-1BBL-deficient compared with WT mice (14). Upon secondary challenge, 4-1BBL−/− mice fail to show an improvement in response over that observed in the primary response. In the present study, using agonistic Abs to replace the 4-1BB signal in 4-1BBL−/− mice, we set out to differentiate between a model of 4-1BB activation in which signals through 4-1BB early in the response impact on long term T cell survival vs a model in which 4-1BB is required again during T cell reactivation.
In this study we show that a single dose of anti-4-1BB added during priming corrects the CD8 T cell defect in CD28−/− mice, resulting in a full secondary response and recall effector function. In contrast, addition of anti-4-1BB during priming fails to correct the defect in 4-1BBL−/− mice; rather, anti-4-1BB added during secondary challenge restores secondary responses in vivo.
Materials and Methods
Anti-4-1BB (M6, rat IgG2a) was produced and purified at Immunex (Seattle, WA) (28). Purified rat IgG was used for controls (Sigma-Aldrich, Oakville, Ontario, Canada).
Influenza virus infection
Seven- to 10-wk-old mice were infected i.p. with 200 hemagglutinin units (HAU) of influenza A HKx31 (H3N2) with either 100 μg of anti-4-1BB Ab (M6) or control rat IgG Ab (Sigma-Aldrich). In a pilot experiment similar effects were seen with administration of 10 or 100 μg of anti-4-1BB during priming of CD28−/− mice, so the higher dose was used throughout to ensure a maximal effect. Influenza virus was produced as previously described (29). At 3 wk postinfection, some mice were challenged with the serologically distinct A/PR8/34 (PR8, H1N1), which shares the NP gene with HKx31, but differs in hemagglutinin and neuraminidase, so that neutralizing Ab does not limit the secondary CTL response. Mice were sacrificed at the indicated time points, and their spleens were harvested for single-cell suspensions.
For adoptive transfer experiments, WT or 4-1BBL−/− mice were immunized with influenza x31. Twenty-one days later, T cells were purified from spleens on mouse T cell enrichment immunocolumns (Cedarlane Laboratories, Hornsby, Ontario, Canada). Cells were resuspended in PBS, and equal numbers of tetramer-positive CD8 T cells were injected in 100 μl through the tail vein of Thy1.1 WT congenic recipients.
Adenovirus-nucleoprotein (NP) infection
The cDNA for PR8 influenza NP (NPw; provided by J. Yewdell, National Institute of Allergy and Infectious Diseases, Atlanta, GA) was subcloned into pDC104 and rescued into an E1/E3-deleted adenovirus using pBHGloxdelE1,3Cre as described by Ng et al. (30) to yield Ad-NPw-004. The expression of NP is controlled by the murine CMV intermediate-early promoter. Ad-NPw-004 was expanded in 293 cells and purified by CsCl gradient. Mice were infected i.p. with 5 × 108 PFU and were sacrificed at the times indicated. Spleen cells were prepared and analyzed as described for influenza infections.
Spleen cell suspensions were prepared in PBS/2% FCS/0.01% sodium azide on ice. Cells were surface-stained with FITC-conjugated anti-mouse CD62L, PE-conjugated anti-mouse CD44, and either allophycocyanin-conjugated anti-mouse CD8 or CD4 Ab (eBioscience, San Diego, CA) to measure activated/memory cells. Cells were stained with PE-conjugated anti-mouse CD8, FITC-conjugated anti-mouse CD62L (eBioscience), and allophycocyanin-labeled tetramers consisting of murine class I MHC molecule H-2Db, β2-microglobulin, and influenza NP peptide, NP366–374 (National Institute of Allergy and Infectious Diseases MHC Tetramer Core Facility) to measure influenza-specific CD8 T cells. For intracellular IFN-γ staining, cell suspensions were restimulated in culture medium (RPMI 1640/10% FCS with antibiotics and 2-ME) for 6 h at 37°C with 1 μM NP366–374 peptide and GolgiStop (BD PharMingen, San Diego, CA). Cells were then harvested, resuspended in PBS/2% FCS/azide, and surface-stained with PE-anti-CD8 and FITC-anti-CD62L as described above. After surface staining, cells were fixed in Cytofix/Cytoperm solution (BD PharMingen) and then stained with allophycocyanin-conjugated anti-mouse IFN-γ diluted in 1× perm/wash solution (BD PharMingen). Samples were analyzed using a FACSCalibur and CellQuest software (BD Biosciences, Mountain View, CA). Statistical analysis between groups was performed using a two-tailed t test.
Spleen cells (5 × 106 cells) were incubated with 250 HAU/ml heat-killed (56°C, 30 min) influenza A HKx31. Supernatants were removed, and the levels of IL-2 and IFN-γ were measured as described previously (14).
Splenocytes from influenza-infected mice were incubated for 2 h at 37°C to remove adherent cells. Serial 3-fold dilutions of effectors were assayed for anti-influenza NP-specific CTL activity against 51Cr-labeled EL4 cells pulsed with 50 μM NP366–374 peptide for 6 h as described previously (14).
Administration of agonistic anti-4-1BB Ab during priming corrects the secondary CD8 T cell defect in CD28−/− mice
The finding that CD28 is important for the initial expansion of T cells has made it difficult to evaluate the role of CD28 in the generation or survival of the memory T cell pool. Previous studies have shown that 4-1BBL can replace CD28 for activation of resting T cells from CD28−/− mice in vitro (31, 32). More recently, Halstead et al. (33) showed that repetitive systemic administration of an agonistic anti-4-1BB Ab restores primary CD8 T cell expansion in CD28−/− mice. The finding that a signal through 4-1BB can replace CD28 for early T cell activation in vivo allows one to ask whether defects in CD8 T cell memory in CD28−/− mice are due to defects in initial expansion alone or whether CD28 is required for long term survival of the T cells or for the secondary response. 4-1BB is induced early and transiently after initial T cell activation in vivo using superantigen (15). After immunization with OVA/LPS, 4-1BB is also expressed early and transiently, concomitantly with CD69 expression on adoptively transferred transgenic T cells.5 Therefore, we reasoned that a single dose of anti-4-1BB Ab administered at the time of infection should be sufficient to replace CD28 function in CD28−/− mice. Previous analysis of the immune response to influenza in C57BL/6 mice using staining with tetramers for the major CD8 T cell epitope, NP366–374, indicated that the primary CD8 T cell response to influenza delivered i.p. peaks in the spleen by day 7 after infection, and at this time point the maximal difference between WT and CD28−/− mice was observed (14). This response declined after day 7, and by day 21 only ∼1.5% of the CD8 T cells in the spleen bound the Db/NP366–374 tetramer. Upon secondary challenge, the response occurred ∼2 days earlier, but was maintained for longer, so that on day 7 after secondary infection there was a major difference in CD8 tetramer-positive T cells in CD28−/− vs WT mice.
Based on these previously established kinetics, we administered anti-4-1BB together with influenza A virus strain HKx31 i.p. on day 0 and then assessed CD8 T cell numbers on day 7 and 21 of the primary response and again on day 7 after secondary challenge (Fig. 1). Intraperitoneal infection of mice with influenza did not result in extensive viral replication, and yields of splenocytes were the same within each group. Therefore, conversion of the percentage of tetramer-positive cells to total T cell numbers gave the same results, and data are reported as the percentage of tetramer-positive cells throughout. On day 7 after influenza HKx31 infection, CD28−/− mice that had received a single dose of anti-4-1BB Ab showed a substantial increase in influenza Db/NP366–374-specific CD62LlowCD8 T cells (Fig. 1,a; 3.5% tetramer+ compared with 0.8% in control Ab-treated mice). This partial restoration of T cell expansion was also reflected by an increase in the number of IFN-γ-producing cells detected after a 6-h restimulation with NP366–374 peptide. In addition, anti-4-1BB Ab increased the number of influenza-specific CD8 T cells in WT mice (11% tetramer+ compared with 7% in control Ab-treated mice). On day 21 of the response, CD28−/− mice that had received anti-4-1BB Ab showed increased levels of tetramer-positive cells compared with CD28−/− mice that received the control Ab (Fig. 1 a). Similar results were seen with the number of IFN-γ-producing NP366–374-pulsed CD8 T cells.
We next asked whether this enhancement of the primary response in CD28−/− mice led to improved secondary responses to influenza. CD28−/− mice were infected i.p. 3 wk after primary exposure with a serologically distinct influenza PR8 strain. The PR8 recombinant strain of influenza virus A retains the same NP gene as the HKx31 strain, but differs in serological epitopes, thereby avoiding a neutralizing Ab response that would limit the secondary CD8 T cell response (34). On day 7 after a second dose of influenza PR8, CD28−/− mice that had received a single dose of anti-4-1BB Ab at the time of the primary infection were able to mount influenza-specific recall CD8 T cell responses, similar to the levels observed in WT mice (Fig. 1,b). Wild-type mice that had received the anti-4-1BB Ab during the primary response did not show any further increase in the number of influenza-specific CD8 T cells in the secondary response compared with control WT mice, even though primary responses were enhanced in these mice. As shown in Fig. 1 a, staining for intracellular IFN-γ gave qualitatively similar results as the tetramer staining.
Treatment of CD28−/− mice with agonistic Ab during priming also restored the direct ex vivo killing activity of secondary CTL to the levels observed in WT mice (Fig. 1 c). The influenza-specific cytotoxic activity observed in the secondary T cells immediately ex vivo correlated with the number of influenza-specific CD8 T cells detected in each mouse type. Thus, anti-4-1BB treatment of CD28−/− mice during the primary response (day 0) restored both CD8 T cell numbers as well as functional activity (cytotoxicity and IFN-γ production by CD8 T cells), measured 28 days later. These data show that beyond its requirements for primary expansion/survival of cells, signaling through CD28 is dispensable for the effector memory CD8 T cell response to influenza virus.
Effects of anti-4-1BB Ab on CD4 T cells and Ab responses in CD28−/− mice
We also asked whether anti-4-1BB Ab could affect CD4 T cell responses in CD28−/− mice. As MHC class II-restricted influenza responses during the primary response to influenza are relatively weak and difficult to follow without restimulation, we measured the number of activated/memory cells by their CD44highCD62Llow surface phenotype. Administration of a single dose of anti-4-1BB Ab at the time of influenza HKx31 infection dramatically increased the numbers of activated/effector memory CD8 T cells (CD44highCD62Llow) in CD28−/− mice to levels comparable to those seen in WT mice at all time points after influenza infection (Fig. 2). These results support those seen with tetramer staining (Fig. 1). In addition, anti-4-1BB Ab increased the numbers of activated/memory CD4 T cells in CD28−/− mice, although the effect was substantially less than that for CD8 T cells, with the percentage of memory CD4 T cells remaining low compared with that in WT mice (Fig. 2). There was also a small enhancement of activated/memory CD8 and CD4 T cells in WT mice that had received anti-4-1BB Ab compared with those that received the control rat Ig Ab.
The effect of anti-4-1BB Ab on virus-specific CD4 T cell production of IL-2 and IFN-γ was also assessed. Anti-4-1BB Ab given at the time of infection had at best a marginal effect on CD4 T cell cytokine responses in CD28−/− or WT mice for either primary or secondary responses (Fig. 3). CD28−/− mice also showed greatly reduced anti-influenza IgG1 and IgG2a responses (14). However, administration of anti-4-1BB had no detectable effect on isotype switching to IgG1 and IgG2a anti-influenza Abs (data not shown). Wild-type mice that had received the agonistic anti-4-1BB Ab also showed no increase in influenza-specific Abs (data not shown), consistent with previous reports that agonistic anti-4-1BB Abs show a preferential effect on CD8 T cells (35). It should be noted that in another study a different agonistic Ab added early in the response was shown to block humoral immunity (36). However, we observed no effect, positive or negative, of anti-4-1BB on the Ab response to influenza in WT or CD28−/− mice (data not shown). The failure of anti-4-1BB to restore CD4 cytokine production or help for B cells in CD28−/− mice could be due to the relatively weaker effects of agonistic anti-4-1BB Abs on CD4 T cells (Fig. 2) (35). Additionally, anti-4-1BB may only replace a subset of CD28 signals, despite their common ability to induce NF-κB, c-Jun N-terminal kinase, and p38 mitogen-activated protein kinase activation (12, 37, 38, 39, 40, 41). Although CD4 T cell responses were only partially restored, this CD4 T cell response is clearly sufficient to provide help for the secondary CD8 T cell response to influenza, the maintenance of which is CD4 T cell dependent (42, 43).
Effects of anti-4-1BB Ab on CD8 T cell responses in 4-1BBL−/− mice
Previously, 4-1BB/4-1BBL interactions were shown to be important in maintaining CD8 T cell numbers late in the primary response to influenza and to influence the size of the secondary response to viruses (13, 14, 17). It was not possible to deduce from previous studies whether the effect of 4-1BBL on T cell recall responses was due to an early transient signal through 4-1BB during initial activation that gave rise to long-lived memory cells or was due to a requirement for 4-1BB/4-1BBL at later times in the response. We therefore tested whether a single dose of anti-4-1BB during priming could influence CD8 T cell numbers late in the primary response or the secondary CD8 T cell response in 4-1BBL−/− mice (Fig. 4).
Both WT and 4-1BBL−/− mice that received a single dose of the anti-4-1BB Ab with influenza HKx31 on day 0 showed a substantial increase in Db/NP366–374-specific CD8 T cells on day 7 (from 7 to 12% tetramer positive; Fig. 4,a). However, in contrast to the results seen in CD28−/− mice, addition of anti-4-1BB Ab at the time of primary infection did not lead to sustained effects and failed to restore secondary CD8 T cell numbers in 4-1BBL−/− mice (Fig. 4,a). However, provision of a single dose of anti-4-1BB Ab in combination with the challenge dose of influenza A PR8 restored the secondary CD8 T cell response in 4-1BBL−/− mice (Fig. 4 b). In fact, the agonistic anti-4-1BB Ab, when given at the time of challenge, significantly enhanced both WT and 4-1BBL−/− CD8 T cell responses above responses seen in WT mice that had received influenza PR8 with control Ab (from 17% tetramer+ CD8 T cells to 29.7% in WT and from 7 to 25.9% in 4-1BBL−/− mice).
Previous results have shown that the secondary CD8 T cell response to influenza is more sustained than the primary response (44). Therefore, it was important to assess whether the anti-4-1BB treatment of 4-1BBL−/− mice resulted in sustained responses after secondary challenge. As shown in Fig. 4 b, by day 28 after secondary challenge, mice that had been treated with anti-4-1BB at the time of secondary challenge retained responses that were substantially higher (p < 0.05) than those of control Ab-treated 4-1BBL−/− mice. These data suggest that 4-1BB signals are required during the secondary response to influenza virus.
Although anti-4-1BB treatment has a weaker effect on CD4 T cells than on CD8 T cells, we do not think the failure of anti-4-1BB provided during priming to correct the defect in secondary response in 4-1BBL−/− mice is due to a defect in CD4 help. Previous studies showed that 4-1BBL−/− mice have no detectable defect in the CD4 T cell response to influenza virus under conditions where a CD8 T cell defect is readily detected (14). Furthermore, the amount of help provided by anti-4-1BB during priming was clearly sufficient to stimulate recovery of the full secondary response in CD28−/− mice.
Once a threshold is reached in the primary response, the magnitude of the secondary CD8 T cell response to influenza virus is independent of the size of the primary response
Comparison of influenza-specific T cell numbers obtained in the primary response with those obtained during the secondary response showed a lack of correlation between the size of the primary CD8 T cell response to Db/NP366–374 epitope and the magnitude of the secondary response (Fig. 5,a). These data are in agreement with previous studies in which manipulation of pathogen dose (24) or adoptive transfer of different numbers of memory cells (45) was used to show that T cell memory can be independent of the size of the primary response. Collectively these data suggest that once the CD8 T cells reach a threshold number in the primary response they are able to generate a complement of memory cells that can give rise to the full secondary response to influenza. Any enhancement of the primary response above 3.5% of CD8 T cells (7 or 11%) does not influence the number of Db/NP366–374-specific T cells generated during the secondary response (Fig. 5 a) even in the absence of CD28.
To further explore the relationship between the numbers of T cells remaining at the end of the primary response and the response following challenge, we compared the effects of priming mice with influenza HKx31 vs immunization with a much stronger immunogen, a recombinant replication-defective adenovirus expressing the influenza NP gene (Ad-NP). On day 21 after infection, WT mice and 4-1BBL−/− mice infected with Ad-NP contained greater numbers of Db/NP366–374-specific CD8 T cells than mice that received influenza A HKx31 (Fig. 5 b). 4-1BBL−/− mice infected with Ad-NP had one-third less Ag-specific CD8 T cells than Ad-NP infected WT mice, consistent with earlier findings that 4-1BBL is required for CD8 T cell survival late in the primary response (14). Even though Ad-NP-immunized mice had 8 times more NP-specific CD8 T cells on day 21 compared with influenza x31-infected mice, Ag-specific CD8 T cells from WT mice primed with either Ad-NP or influenza HKx31 expanded to the same level after challenge with influenza A PR8. Similarly, Ad-NP immunization of 4-1BBL−/− mice resulted in a 4-fold increase in NP-specific CD8 T cells on day 21 compared with influenza-infected mice, yet secondary responses were similar in Ad-NP- vs influenza virus-primed 4-1BBL−/− mice after secondary challenge. These data show that increasing cell numbers late in the primary response does not compensate for the secondary response defect in 4-1BBL−/− mice.
CD8 T cells primed in the absence of CD4 help show a defect in their ability to respond to secondary stimulation in vivo (46, 47, 48). To test whether 4-1BB signals have a similar effect on programming CD8 T cells, we infected WT or 4-1BBL−/− mice with influenza virus x31, and 21 days later transferred purified T cells into Thy1.1 congenic WT recipients, such that each mouse received equal numbers of tetramer-positive T cells. Upon rechallenge with influenza PR8, the T cells primed in WT or 4-1BBL−/− mice expanded to a similar extent (Fig. 5 c), arguing against a programming defect in the T cells accounting for the secondary response defect.
We have shown that a single dose of agonistic anti-4-1BB given only during the primary response to influenza virus can restore secondary CD8 T cell responses to influenza virus in CD28−/− mice. Given that the same treatment does not correct the defect in 4-1BBL−/− mice, the data suggest that, beyond its role in primary T cell activation and expansion, CD28 is dispensable for the secondary CD8 T cell response to influenza virus. In contrast, administration of agonistic Ab only during the secondary response in 4-1BBL−/− mice corrected the secondary response defect and increased secondary CD8 T cell responses in WT and 4-1BBL−/− mice to a level twice the normal ceiling.
Previous results have shown that 4-1BBL−/− mice show normal primary expansion and initial contraction of CD8 T cell responses to influenza virus, but there is a decrease in the number of CD8 T cells late in the primary response, and this decrease correlates with the decreased secondary response (14). Despite the ability of anti-4-1BB to increase primary responses in 4-1BBL−/− mice, we found that anti-4-1BB delivered on day 0 of the primary response did not lead to sustained enhancement of the response and failed to correct the defect in the secondary response in 4-1BBL−/− mice. Furthermore, administration of adenovirus-NP to WT or 4-1BBL−/− mice, although greatly increasing the number of CD8 T cells late in the primary response, did not correct the defect in secondary T cell responses in 4-1BBL−/− mice.
Fig. 6 summarizes the effects of costimulation on the primary vs the secondary response to influenza virus. CD28 is required for a 10-fold increase in the primary expansion of influenza-specific CD8 T cells (Fig. 6,a), but is not required for secondary expansion. In contrast, the primary CD8 T cell responses in WT and 4-1BBL−/− mice are similar; however, the presence of 4-1BBL results in a 2-fold increase in cell numbers in the secondary response (Fig. 6,b) in addition to its effects on T cell numbers late in the primary response. The secondary response to viruses consists of a mixture of secondary T cells as well as additionally recruited naive T cells (45, 49). Therefore, anti-4-1BB could be recruiting additional naive T cells during the secondary response. Even assuming that the secondary response is a mixture of primary and secondary T cells, it is clear that 4-1BBL−/− mice have no defect in initial T cell expansion during priming and yet show a secondary defect. Furthermore, anti-4-1BB during priming induces a 70% increase in CD8 T cell numbers in WT or 4-1BBL−/− mice, whereas the same dose of anti-4-1BB added during challenge increases NP366–374-specific T cell numbers in 4-1BBL−/− mice by almost 4-fold during the secondary response. Addition of anti-4-1BB during priming corrects the secondary response in CD28−/− mice, but fails to correct the defect in 4-1BBL−/− mice. Taken together, these data point to a model in which the effects of 4-1BB signaling early in primary responses are transient and play little role in the presence of an intact CD28 signaling pathway (Fig. 6 c). Rather, the data suggest a more important role for 4-1BB signaling occurring late in the primary response to sustain T cell numbers and again during secondary responses. Adoptive transfer experiments rule out a programming model in which 4-1BB/4-1BBL is required during priming to allow T cells to re-expand during the secondary response. Thus, the effects of 4-1BB costimulation on CD8 T cells are conceptually different from the effects of CD4 T cell help on CD8 recall responses (46, 47, 48). The finding that anti-4-1BB treatment has a more substantial effect when administered during the secondary response suggests that anti-4-1BB or 4-1BBL will be more useful when administered in the boost phase of a prime-boost vaccination strategy.
Costimulation with CD28 involves the provision of proliferation as well as survival signals (50). Like CD28, 4-1BB can give rise to proliferative and survival signals (12, 37, 39, 40). It appears that CD28 is the major provider of costimulatory signals to primary CD8 T cells, but 4-1BB takes over at least part of this role during secondary expansion. Although there have been some studies that show a role for 4-1BB predominantly in T cell survival (15), there is also clear evidence that when given with a signal through the TCR, 4-1BB induces IL-2 production and proliferation (32). Thus, it is likely that during secondary expansion both survival and proliferative signals through 4-1BB are important.
The mechanism by which the CD8 T cell costimulatory molecule dependence changes during primary vs secondary responses is not known. Analysis of secondary tetramer-positive T cells ex vivo failed to detect a defect in response to anti-CD3 plus anti-CD28 in these cells (data not shown). It is possible that distinct APC are involved in priming vs those involved in sustaining effector cells or those involved in secondary expansion. For example, Crowe et al. (51) recently showed that changes in immunodominance during the primary vs the secondary response to influenza virus in vivo reflects differences in Ag presentation by dendritic cells vs nondendritic cells. Interestingly, a unique accessory cell (CD3−CD4+OX40LhighCD30LhighB7llow 4-1BBLlow) was recently shown to be important for providing OX40L signals to primed or memory CD4 T cells (52). Whether an equivalent cell will be found for 4-1BBL-CD8 T cell interactions remains to be determined.
We thank the National Institute of Allergy and Infectious Diseases tetramer facility for provision of Db/NP366–374 tetramers.
This work was supported by a grant from the Canadian Institutes for Health Research (to T.H.W.) and by the Canadian Network for Vaccines and Immunotherapeutics.
Abbreviations used in this paper: WT, wild type; Ad-NP, adenovirus expressing the influenza NP gene; HAU, hemagglutinin unit; NP, nucleoprotein.
W. Dawicki and T. H. Watts. Adoptive transfer and immunization of OT-I and OT-II transgenic T cells in 4-1BBL−/− versus WT mice reveals similar effects of the 4-1BB costimulatory pathway on CD4 and CD8 T cell secondary response. Submitted for publication.