Materials and methods

Animals and cell lines

Female C57BL/6 (B6, H-2^b), BALB/c (H-2^d), DBA/2 (H-2^b) and (BALB/c C57BL/6) F1 (H-2^b/d) mice were purchased from the National Cancer Institute (Frederick, MD, USA). C3H.SWH2b/SnJ (SW, H-2^d) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). DO11.10 OVA-TCR Tg mice were kindly provided by M Kripke (MD Anderson Cancer Center, Houston, TX, USA). Mice were housed in pathogen-free conventional and biohazard quarters. From the day of irradiation until 2 weeks after BMT, the mice were provided with acidified water (pH 2.5) supplemented with 2 g/l neomycin sulfate (Sigma, St Louis, MO, USA). EL4 cells, a C57BL/6-derived lymphoma cell line (American Type Culture Collection, ATCC, Rockville, MD, USA), were used as targets for cytolytic T cell (CTL) assays. The cells were grown in RPMI-1640 supplemented with 5% heat-inactivated fetal bovine serum (Biowhittaker, Walkersville, MD, USA) and 2 mmol/L L-glutamine.

Influenza vaccination

Murine-passaged human influenza virus A (strain A/PR/8/34, American Type Tissue Culture, Rockville, MD, USA) was grown in embryonated chicken eggs. Naïve SW BMT-donor spleens were mechanically dissociated. Following ACK-mediated erythrocyte lysis, 100 10⁶ spleen cells were mixed with 1 ml of murine-passaged influenza virus at 37°C for 1 h. Following infection, the cells were washed once in fresh RPMI and resuspended at 5 10⁶ cells in 0.2 ml of Hank's balanced salt solution (HBSS). SW donors were immunized at 2-week intervals and were used as bone marrow donors 1–2 weeks after the last immunization. Mice were immunized at least twice before use as donors; in some experiments, the donor group received a third vaccination if more than 3 weeks had elapsed from the second vaccination. The number is specified in the figure legends. BMT recipients had a single exposure to 5 10⁶ influenza-infected spleen cells the day prior to transplant (D-1), 7 days after BMT (D+7) or 10 days after BMT (D+10).

Bone marrow transplantation

BMT recipients received 850 cGy total body irradiation using a ⁶⁰Co source 1 day prior to transplantation (D-1). On the day of transplantation (D-0), recipient C57BL/6, BALB/c, DBA/2 or BALB/cxC57BL/6 F1 mice received 4 10⁶ bone marrow and 10x10⁶ spleen cells intravenously in a total volume of 0.2 ml of HBSS.

Delayed lymphocyte infusion (DLI)

Lymph node T cells for use in post-BMT lymphocyte infusions were purified by passage through two nylon wool columns Depletion was confirmed by flow cytometry using monoclonal antibody to F4/80 (Caltag, San Francisco, CA, USA). Less than 5% of the injected material stained positive for F4/80.

Cytotoxicity assays

Spleen cells were cultured in six-well plates at 1 10⁶ cells/ml in complete media consisting of 10% FBS (Summit Biotech, Collins, CO, USA), 100 U/ml penicillin, 100 g/ml streptomycin, 2 mM L-glutamine, 100 mM sodium pyruvate, 0.1 mM nonessential amino acids and 50 M 2-mercaptoethanol. As stimulators, 0.125 10⁶ naïve SW spleen cells were pulsed with 35 g of NP366-peptide in a total volume of 10 ml. Irradiated stimulators were washed in R10S and resuspended at a final concentration of 1.25 10⁴ cells/ml and added to effectors. After 5 days in culture, the cells were harvested and plated in triplicate with 5 10^{3 51}Cr-labeled target per well at effector to target (E:T) ratios ranging from 200:1 to 12.5:1. Target cells were labeled by combining 5 10⁶ cells in 60 l of R10S and 40 g of peptide 1 g/l and 0.1 ml (100 Ci) sterile Na₂⁵¹CrO₄ (Amersham, Arlington Heights, IL, USA) for 1 h at 37°C. Labeled targets were washed three times before plating with effectors in a total volume of 0.2 ml/well in 96-well, round-bottomed plates. The cultures were incubated at 37°C for 4 h and 0.1 ml of the supernatant was counted in a gamma counter (Wallac, San Francisco, CA, USA). The percentage of lysis was calculated as 100 ((experimental c.p.m.-spontaneous c.p.m.)/(maximum c.p.m.-spontaneous c.p.m.)).

Limiting dilution analysis (LDA)

To determine the precursor frequency of NP366-specific cytolytic lymphocytes, serial dilutions of spleen cells were cultured in replicates of 24–40 in 96-well, round-bottomed plates together with peptide-pulsed 20 Gy-irradiated naïve SW spleen cells (50 000/well) as stimulators in 0.2 ml R10S. Fresh complete media were added on days 4 and 7 by replacing 0.1 ml of supernatant without disturbing the cell pellet. On days 0, 3 and 6, IL-2 was added to each well at a final concentration of 2 IU/ml (Chiron, Emoryville, CA, USA). On day 11, 5 10^{3 51}Cr-labeled EL4 target cells pulsed with or without NP366-peptide were added to each well and cytotoxicity was measured as described above. CTL precursor frequency was estimated based on the Poisson distribution and probability theory, which predicts that an average of one CTL precursor per well is present when 37% of the wells exhibit no NP lytic.^{10, 11}

Peptides

The sequence of the NP366–374 is ASNENMETM. The sequence of OVA 323–339 is ISQAVHAAHAEINEAGR.

Monoclonal antibodies, carboxyfluorsulfoethanimide (CFDA-SE) labeling and flow cytometry

Spleen cells were analyzed by flow cytometry on a FACScan (Becton Dickson, Mountain View, CA, USA). To block nonspecific FcR binding of labeled antibodies, 10 l of mouse IgG (Caltag, San Francisco, CA, USA) was added for 10 min at 4°C. To determine the percentage of donor-type T cells in the spleens, cells were stained with FITC-conjugated CD5.1⁺ and PE-conjugated CD3⁺ rat anti-mouse monoclonal antibodies (Pharmingen, San Diego, CA, USA) at 4°C for 1 h, respectively. The cells were then washed three times in PBS, fixed in 2% paraformaldehyde and analyzed by FACS. Physical gates on cells with low forward and side scatters were used to select the lymphocyte population. For the identification of Tg OVA-TCR CD4⁺ T cells, spleen cells were stained by FITC-labeled KJ1–26 (a gift from P Marrack, Denver, CO, USA) antibody and PE-labeled anti-mouse CD4 (Pharmingen) with or without anti-mouse Tricolor-CD44 antibody. For tetramer staining, 1 10⁶ spleen cells from BMT recipients were initially incubated with FITC-labeled anti-CD8 monoclonal antibody for 1 h at 4°C. The cells were then washed once in PBS and stained with 1:100 diluted NP-PE-tetramer (MHC Tetramer Core Facility, NIAID, Atlanta, GA, USA) for 15 min at room temperature. After washing in PBS, the cells were analyzed immediately by flow cytometry. CFDA-SE staining was carried out according to the manufacturer's protocol (Molecular Probes, Eugene, OR, USA) in 0.1 mM CFSE. CFDA-SE-labeled cells were infused via tail vein. On day 4, the spleen, lymph node, lung, liver, kidney and the intestine were isolated, snap frozen in liquid nitrogen and tissue sections were examined for CFDA-SE-labeled cells using fluorescence microscopy.

Cytokine flow cytometry (CFC)

Splenocytes were harvested and resuspended at 2 10⁶ cells in 0.1 ml. Irradiated (20 Gy) naïve C3H.SW splenocytes were added as antigen-presenting cells. NP-peptide (ASNENMETM) was added to each well at either 2 or 100 g/well in a volume of 100 l along with stimulators. After incubating the plates at 37°C and 5% CO₂ for 1 h, brefeldin A was added to each well at 10 g/ml. The cells were incubated for an additional 5 h. The plates were then washed with cold PBS and incubated in 100 l of PBS with 2 mM EDTA and incubated in 5% CO₂ for 10 min. The wells were washed in PBS and resuspended in 100 l of FACS Lyse solution (Caltag) for 15 min at room temperature. The cells were spun down and resuspended in 100 l of FACS Perm solution (Caltag) for 15 min. The cells were then spun down and resuspended in PBS+1% BSA containing 1:50 dilution of anti-IFN- (Pharmingen) and 1:50 dilution of anti-CD8 (Caltag) antibodies for 30 min on ice.

Statistics

For normally distributed data, Student's t-test was used. For experiments in which data was not normally distributed, the nonparametric median test was used.

Results

Model 1: Antigen exposure enhances the recovery of antigen-specific CD4⁺ T cells

One possible explanation for relatively poor T-cell function after transplantation is that antigen-specific donor T cells do not engraft efficiently. We tested this hypothesis in murine MHC-matched, minor histocompatibility antigen (mHA)-mismatched transplantation. Recipient C57BL/6 (CD5.1^-) mice were lethally irradiated and transplanted with 20 10⁶ spleen cells from SW donor mice (CD5.1⁺). Engraftment was determined using donor-specific B- and T-cell antibodies in flow cytometry 96 h following transplantation. Donor CD5.1⁺ cells were identified in all recipients, but the total number of cells recovered from an exhaustive harvest of lymph nodes and spleens was very low (5–10% of total T cells infused). These observations were confirmed in an independent experiment using CFDA-SE-labeled lymphocytes in which we again failed to see significant number of labeled donor cells in marrow, lymphoid or other visceral organs using flow cytometry and fluorescent microscopy (data not shown). Based on these observations, we chose to study the fate of an antigen-specific clonotypic CD4⁺ subpopulation using Tg BALB/c (H-2^d) DO11.10 mice. The TCR of these Tg animals is specific for the class II restricted OVA peptide 323–339. Tg T cells were transplanted along with normal BALB/c bone marrow into irradiated BALB/c recipients. This approach allowed discrimination between lymphocytes derived from non-Tg hematopoietic stem cells and peripherally expanded mature Tg donor T cells.

We hypothesized that peritransplant antigen exposure would increase the efficiency of transfer of antigen-specific T cells. To test this hypothesis, 2 10⁶ OVA-Tg CD4⁺ T cells and 4 10⁶ normal BALB/c bone marrow cells were transplanted into lethally irradiated BALB/c animals. On the day before transplantation (D-1) cohorts of animals were vaccinated with OVA peptide (300 g) in complete Freund's adjuvant (CFA) subcutaneously or CFA alone. Tg T cells were measured using a TCR-specific monoclonal antibody, KJ1–26.

The number of antigen-specific cells in chimeric animals with early antigen exposure was five- to 10-fold greater than the animals without OVA exposure in the first 48 h of following transplantation (Figure 1a). We studied the persistence of the effect by examining BMT recipients nearly 2 months post-BMT. At 7 weeks following transplantation, the OVA-specific T cells could still be found at four-fold higher numbers in early antigen exposed groups than in antigen-deprived cohorts (Figure 1 b and c). In a third independent experiment, a major HA-matched, minor histocompatibility mismatched allogeneic transplant model was used. BALB/c donors were used for DBA/2 recipients. Results were similar to the syngeneic transplant. DBA/2 transplant recipients injected with OVA on day -1 had an average of 5 10⁵ CD4⁺KJ126⁺ cells (s.d.=1.8 10⁵, n=3), while the animals not exposed to OVA had an average of 1.2 10⁵ CD⁺KJ126⁺ cells recovered from lymph nodes (s.d.=2.6 10⁴, n=3) (P=0.02 by t-test).

Figure 1.

Early exposure to OVA antigen enhances clonal expansion of OVA-specific Tg T cells following syngeneic BMT. Lethally irradiated BALB/c mice were injected with 300 g of OVA peptide in CFA or CFA alone subcutaneously 1 day before transplantation (D-1). The transplanted mice received 2 10⁶ Tg OVA-specific T cells and 6 10⁶ non-Tg naïve BALB/c bone marrow cells. At 2 days (a) and 7 weeks (b) following BMT, draining lymph nodes were isolated and examined for the presence of OVA-specific T cells using FITC-labeled KJ1–26 antibody and PE-labeled anti-mouse CD4 antibody. Naïve BALB/c animals were used as negative controls (c). The absolute number of antigen-specific T cells was determined by multiplying the total cells recovered with the percentage of OVA-specific CD4⁺ T cells identified by flow. T-test was used to compare OVA and CFA groups. In (a), the number in the OVA group was greater than in the CFA group (P<0.05) In (b), the number of cells in the OVA group was greater than in the CFA group (P<0.05). In (c), the difference was not significant.

Full figure and legend (20K)

In these experiments, the T cells that were transplanted were from Tg animals not previously primed with OVA, and thus were phenotypically naïve. We hypothesized that early antigen exposure in the BMT recipient would induce a clonal burst and drive cells to a memory phenotype. To test this hypothesis, irradiated BALB/c mice received 20 10⁶ Tg CD4⁺ T cells and 4 10⁶ normal BALB/c marrow cells together with a single injection of OVA peptide on the day before transplantation (D-1). At 3 months following transplantation, the splenocytes were stained with KJ1–26 and CD44 antibodies in vitro. Naïve BALB/c animals generally express intermediate levels of CD44 on their T-cell surface. However, in animals exposed to antigen, a subpopulation of T cells expressing CD44_high expression could be identified. The CD44_high population accounted for 30% of KJ1–26 CD4⁺ T cells in the OVA-exposed animals, while in BMT recipients not exposed to antigen CD44_high, Tg T cells represented less than 10% of antigen-specific T cells (data not shown).

Acute GVHD prevents clonal expansion of antigen-specific CD4⁺ T cells

In this CD4 system, the effect of early antigen exposure was seen in both the syngeneic BALB/c>BALB/c and the allogeneic minor antigen mismatch BALB/c>DBA/2 transplants. In the latter model, we have observed no clinically significant GVHD. However, active graft-versus-host (GVH) reactions often complicate transplantation. Since transplanted T cells are of donor origin, they will not be the target of alloreactive donor T cells. However, we hypothesized that the environment in which alloreactivity is occurring may influence the efficiency of donor T-cell engraftment and expansion. To test this hypothesis, we compared the size of the Tg clonotypic T-cell population in BMT settings with and without ongoing GVH reactivity. Under syngeneic conditions, (BALB/c C57BL/6) F1 into F1 recipients exposed to antigen, the number of OVA-Tg CD4⁺ T cells was approximately four-fold higher than Tg T cells in (P into F1) acute GVH conditions (Figure 2). The presence of antigen early at transplantation did not enhance the subpopulation of Tg T cells recovered in the active GVH environment.

Figure 2.

Strong GVH reactions reduce the clonal expansion of antigen-specific donor T cells induced by early antigen exposure. BALB/c or BALBc/B6 F1 recipient mice were lethally irradiated and transplanted with either syngeneic (F1) or allogeneic (BALB/c) bone marrow cells (4 10⁶) and spleen cells (15 10⁶) along with OVA-Tg T cells (4 10⁶). Animals were killed on day 14 and examined for the presence of OVA-specific T cells in the lymph nodes cells using FITC-labeled KJ1–26 antibody and PE-labeled anti-mouse CD4 antibody. The absolute number of antigen-specific T cells was determined by multiplying the total cells recovered with the percentage of OVA-specific CD4⁺ T cells identified by flow. The nonparametric median test was used to calculate statistical differences. P=0.01 for (OVA F1>F1) vs (None F1>F1). P=0.01 for (OVA F1>F1) vs (OVA BALB/c>F1). P=0.41 for (OVA BALB/c>F1) vs (None BALB/c>F1).

Full figure and legend (9K)

Taken together, the experiments using model one demonstrate that antigen exposure at the time of transplant leads to persistent expansion of antigen-specific CD4 cells in the post transplant repertoire. However, active GVHD impairs this antigen-driven early clonal expansion.

Model 2: Peritransplant antigen exposure increases NP366-specific CD8⁺ T-cell population in post transplant repertoire

We extended these studies to determine if early antigen exposure would have similar effects on CD8⁺ T cells. Here, we employed a model of influenza NP-peptide-specific CD8⁺ cells. Tetramer- and peptide-specific functional assays reagents were available to identify antigen-specific cells responding to the immunodominant CD8 epitope of the influenza NP, the main target of the anti-influenza CTL response. Lethally irradiated naïve C57BL/6 mice were transplanted with spleen and bone marrow cells from donor SW mice exposed to influenza virus. At 4 weeks after transplantation, BMT recipients with early antigen exposure (D-1) exhibited cytolytic activity to the NP366-labeled target cells (Figure 3a). BMT recipients that did not encounter antigen did not display any lytic activity despite receiving immune donor T cells.

Figure 3.

Early antigen exposure in BMT recipients maintains donor anti-NP CTL activity. SW mice were injected intraperitoneally (i.p.) with influenza virus-infected spleen cells twice at 2-week intervals and 2 weeks later were used as donors. Recipient C57BL/6 animals received BMT from influenza-immune or naïve control SW mice. In all experiments, recipients were killed 5 weeks after BMT. Spleen cells were restimulated with NP366-labeled naïve SW spleen cells in vitro and lysis of peptide-loaded targets was measured by ⁵¹Cr-release assay. Each point is the average of triplicate wells with the bars representing the s.e.m.'s. (a) Recipient animals were either exposed to antigen the day before transplantation, D-1, or were left unexposed. (b) In a separate experiment, recipient animals were either exposed to antigen on D-1 or +10 days after transplantation, D+10. Non-BMT naïve SW never exposed to NP and non-BMT SW immunized against NP were used as controls. (c–e) In a third experiment, recipient animals were either exposed to NP366 the day before transplantation (Ag, D-1) (shown in c) or 7 days after transplantation (Ag, D+7). Other irradiated recipients received DLI of 3 10⁶ nylon wool-purified immune or naïve donor T cells on D+8 (shown in d). Immune and naïve SW served as positive and negative controls, respectively (shown in e).

Full figure and legend (56K)

Based on the experiments in the OVA-Tg model in which antigen led to early clonal expansion, we hypothesized that the time at which antigen exposure occurred was critical. To test this hypothesis, subsequent experiments explored the timing of antigen exposure on later recovery of anti-NP366 reactivity (Figure 3b). BMT recipients of immune cells in which antigen exposure was delayed until 10 days after transplantation did not show any cytolytic antigen-specific activity, whereas animals exposed on day -1 had substantial anti-NP366 activity. Similar results were seen if antigen exposure was delayed until day +7 (Figure 3c).

One possible explanation for the ineffectiveness of delayed exposure is that antigen-specific donor T cells had died in the first week. An alternative hypothesis is that the donor T cells were still present in the recipient's repertoire following transplantation, but inefficient antigen presentation resulted in suboptimal expansion of NP366-specific T cells. To test these possibilities, early (D-1) and delayed (D+7) antigen exposure was studied with or without delayed infusion (D+8) of antigen-specific donor T cells. If the absence of anti-NP366 lytic activity were due to the disappearance of antigen-specific T cells rather than the suboptimal antigen presentation, infusion of these T cells immediately following the delayed D+7 antigen exposure should generate substantial antigen-specific immune activity. Recipients exposed to antigen on D+7 and then infused with a purified population of virus-specific donor cells on day 8 exhibited substantial activity against NP366-labeled targets (Figure 3d). Antigen encounter on day 7 (D+7) followed by transplantation of naïve donor T cells lacking anti-NP activity on D+8 was ineffective in inducing anti-NP366 reactivity (Figure 3d and e). These data suggest that the failure of delayed antigen exposure was not due to intrinsically poor antigen presentation.

These studies were extended by experiments designed to measure the number of antigen-specific T cells rather than functional activity of the population of T cells. CTL LDA was used to determine the precursor frequency of NP366-specific T cells in antigen exposed and unexposed cohorts. The CD8⁺ precursor frequencies in the antigen-exposed animals were approximately 10- to 20-fold higher than in unexposed or delayed exposure animals. After transplantation NP366-specific CTLs ranged from 1:30 000 to 50 000 CTLp in the groups exposed to antigen. In unexposed groups, CTL precursors were either undetectable or present at 1:300 000–500 000 cells, similar to CTLp levels found in naïve animals. LDA was complemented by NP366 class I tetramer studies because it is possible that the NP366-specific cells in the unexposed group were still present but were incapable of demonstrating cytolytic function measured in the LDA assay. BMT recipients transplanted with NP366-reactive T cells were either exposed to antigen (D-1) or left unexposed. At 1 week after transplantation, spleen cells were isolated and stained with NP-H2^b-tetramer specific for NP-responding cells and simultaneously tested for cytolytic activity against NP-peptide-labeled targets (Figure 4). The animals that had encountered antigen displayed a four-fold higher frequency of NP-tetramer binding with NP366-specific T cells accounting for 0.20% of the CD8⁺ T-cell repertoire. This was in sharp contrast to the animals unexposed to antigen in which the NP366-specific CD8⁺ T cells only accounted for 0.04% of the cells. In recipients exposed to antigen, the percentage of NP366-specific T cells were significantly higher (0.200.041, avgs.e.m.), while in unexposed animals, tetramer-positive cells made up only a small fraction (0.050.012) of total CD8⁺ T cells (Student's t-test, P<0.01). The number of NP-tetramer cells correlated with cytolytic activity against NP+target cells (insets, Figure 4). To determine whether the small subpopulation of cells present in unexposed animals (identified by tetramer assay) is functionally active, we made use of IFN- CFC. Splenocytes from antigen-exposed and -unexposed animals were isolated 2 weeks after stimulation and restimulated in vitro with 2 g of NP366-peptide. The rational for this strategy was to drive the proliferation of functional CD8⁺ T cells. If the cells were present but at frequencies below detectable levels in animals not exposed to antigen at the time of transplantation, in vitro restimulation should drive their numbers to detectable levels. Splenocytes were stimulated two times in vitro with NP-peptide at 7-day intervals and on day 15 examined for IFN- secretion. In the animals exposed to antigen early, the NP366-specific CD8⁺ cells could be detected at 30- to 50-fold higher numbers than their unexposed counterparts. In the latter group, prevalence of functional NP-specific CD8⁺ T cells were similar to those observed in naïve animals (Figure 5).

Figure 4.

Early antigen exposure at the time of transplantation increases the precursor frequency of the NP366-specific T cells. Donor SW mice were immunized three times by i.p. injection of influenza-infected spleen cells. Recipient C57BL/6 animals underwent BMT using influenza-immune donor spleen cells and bone marrow (10 10⁶ and 4 10⁶, respectively). Some recipients were exposed to influenza virus on day -1. At 10 days following BMT, the animals were killed and splenocytes were restimulated in vitro with NP-peptide. Cytolytic responses against NP-peptide-pulsed T2 target cells (+) and against T2 target cells without peptide (-) were also measured (insets). The percent lysis of targets is on the Y-axis. The effector:target ratio was 100:1. (+, dotted bar: NP-peptide-positive target cells) (-, solid bar: NP-peptide-negative target cells). Flow cytometry was performed using NP-366-tetramer for antigen-specific T cells. Flow results shown as gated on CD8⁺ on lymphocytes. The percent in upper right corner of each plot is the percent NP-tetramer-positive CD8 cells. The circle on the flow cytometry dot plot represents the region of CD8⁺, NP-tetramer-positive cells. Depicted are representative results from animals exposed to antigen the day before transplant (left) and not exposed to antigen (right).

Full figure and legend (67K)

Figure 5.

Antigen-specific CD8⁺ cells in antigen-exposed animals are functionally active. Donor SW mice were immunized three times by i.p. injection of virus-infected spleen cells. Recipient C57BL/6 animals underwent BMT using influenza-immune donors spleen cells and bone marrow (10 10⁶ and 4 10⁶, respectively). Some transplant recipients were exposed to influenza virus on day -1. At 10 days following BMT, the animals were killed and spleens were isolated and stimulated in vitro with 2 g of NP-peptide. After two rounds of stimulation, the splenocytes were assayed by intracellular cytokine assay for the secretion of IFN- by staining with PE-anti–IFN- and FITC-anti-CD8 antibodies. The results are representative of three experiments (total number n=4/experimental group).

Full figure and legend (66K)

Acute GVHD prevents clonal expansion of antigen-specific CD8⁺ T cells

Since acute GVHD impaired the effect of antigen exposure in the expansion of antigen-specific CD4⁺ T cells, we hypothesized that a similar effect would be seen with CD8⁺ cells. To create an acute GVHD environment, splenocytes from NP-immune C57BL/6 were transplanted into (BALB/c C57BL/6) F1 recipients. At 2 weeks after transplantation, splenocytes were isolated from experimental groups. Acute GVHD completely abrogated the expansion of NP-specific T cells as measured by cytolytic assays (Figure 6).

Figure 6.

Strong GVH reactions abrogate the effect of early antigen exposure on donor anti-NP CTL function. Donor C57BL/6 mice were immunized twice by i.p. injection of virus-infected spleen cells. Recipient BALB/cXC57BL/6 (F1) animals underwent BMT using influenza-immune donors spleen cells and bone marrow 2 weeks after the last donor immunization. Recipient animals were injected with antigen the day before transplantation, D-1. At 2 weeks after BMT, animals were killed and the spleen cells were restimulated with NP366-labeled naïve C57BL/6 spleen cells in vitro and lysis of peptide-loaded EL-4 targets was measured by ⁵¹Cr-release assay.

Full figure and legend (11K)

Discussion

After transplant, immune reconstitution occurs by both de novo generation of lymphocytes from hematopoietic stem cells and peripheral expansion of mature lymphocytes in the transplant product. The latter process is important for early immune function and recipients of T-cell-depleted transplants have more severe post transplant immune deficiencies. Our study confirms other observations that the engraftment of mature donor T cells is an inefficient process.^{2, 5} Our findings show that early antigen exposure can significantly enhance representation of desired CD4⁺ and CD8⁺ antigen-specific T cells in the post transplant immune repertoire. Antigen exposure at the time of transplant appears to induce a brief clonal burst in the first 2 days after transplant among the minority of transplanted T cells that successfully engraft in lymphoid organs. The enhanced representation is durable with increased numbers of functional, clonotypic T cells observed 1–2 months after transplant. Severe, active GVH reactions eliminate the beneficial effect of antigen exposure on donor T cells.

Antigen exposure is not absolutely required for persistence of T-cell memory and homeostatic proliferation of donor T cells does occur in the absence of specific antigen.^{7, 12} In the Tg-TCR CD4 OVA system in which 'unphysiologically' large clonal populations were transplanted, clonotypic T cells were clearly detectable in recipients that did not experience antigen exposure. Moreover, some clinical studies have demonstrated the transfer of donor T- or B-cell memory to transplant recipients.^{13, 14, 15, 16} However, our findings show that early antigen encounter can substantially increase the size of the specific T-cell population.¹⁷ This may be very important in sustaining antigen-specific T cells in which precursor frequency is 'physiologic' or small as was the case in the CD8 NP system. In this system, we were unable to detect NP-specific T-cell activity in transplant recipients not exposed to antigen. The physiologic effect of early antigen exposure may meaningfully enhance immune responses to infection or cancer cells. In other studies, we have observed that peritransplant exposure to the NP antigen decreased the number of pulmonary metastases of a tumor that employed NP as a model antigen in vivo.¹⁸ In addition, such exposure did not increase GVHD nor did it have adverse effects on engraftment of donor hematopoietic stem cells. Our results are compatible with a small clinical study in which pediatric allogeneic and autologous BMT patients mounted meaningful immune responses to varicella when immunized with inactivated varicella virus starting 1 month after transplant.¹⁹ Additional studies are needed to determine how this might be approached clinically. The systems and immunization methods used for the CD4 and CD8 models differed in several ways (eg precursor frequency, memory vs naïve cells, peptide vs complex virus immunization). Meaningful enhancement of an immune response in humans will likely require concerted manipulations of both CD4 and CD8 responses to the same target.

The mechanism of early loss of transplanted donor lymphocytes is not clear. If transplanted cells have a high rate of spontaneous apoptosis, early antigen encounter may interfere with this. However, we were unable to detect any difference between groups related to antigen exposure in the rate of apoptosis in flow-sorted clonotypic T-cell populations as measured by either annexin V or TUNEL assays (data not shown). Rather, our observations are more compatible with a model in which early antigen encounter induces a brief clonal burst of lymphocyte replication. These studies did not formally test this hypothesis using direct measures of cell replication such as BrdU incorporation or reductions in CFSE-related fluorescence intensity. However, the observation of a larger Tg T-cell population within 48 h, and lack of efficacy of antigen exposure at 7 days in the NP system both indicate that the increase in the relative size of the T-cell population occurs quite early after transplant. Experiments that directly measured the number of clonotypic T cells (using the TCR-specific monoclonal antibody for the CD4 cells, and the NP-tetramer for the CD8 cells) suggested that a single antigen exposure produced approximately a four-fold increase in the population. A larger difference (10- to 20-fold) was observed using NP-peptide-induced intracellular interferon production. This apparent difference might be explained by either activation of bystander nonclonotypic T cells in vitro or lower activation requirements for T cells after early antigen exposure. Our observation of a larger proportion of CD44-positive Tg cells in the antigen-treated recipients is consistent with the latter possibility.

One limitation of the current study is that it does not establish the mechanism by which GVH reactions eliminate the beneficial effect of early antigen exposure. It is not likely that alloantigen-specific donor T cells are killing other donor T cells. One hypothesis is that in lymphoid tissue microenvironments with active GVH cytokines are present that block clonal expansion or effective presentation of the pretransplant vaccine antigens. Current work is exploring this possibility. An alternate hypothesis for explaining the reduced frequency of antigen-specific T cells in the high GVHD environment is that they are diluted by the larger populations expanding in response to the alloantigens. While it is true that in the P>F1 models, the absolute number of splenocytes was increased compared to syngeneic transplants, this expansion does not account for the observed loss of effect of peritransplant antigen exposure in GVHD environments. In the CD4 model (Figure 2), the absolute number (not relative percentage) of OVA-specific T cells in the spleen was measured. In the CD8 model, 45% more splenocytes were recovered from the P>F1 mice. However, we observed a complete loss of antigen-specific cytolytic activity (Figure 6) that is not explained by a two-fold change in the effector to target ratio. Another limitation is that the experiments did not study the impact of prophylactic immunosuppressant agents on early transplant antigen exposure. Current work is exploring the hypothesis that this would blunt the desired clonal expansion. Preliminary experiments indicate that use of prophylactic tacrolimus reduces but does not eliminate the impact of early antigen exposure. If prophylactic immunosuppression does completely block the effect of early antigen exposure, it is conceivable that early antigen exposure could be employed to advantage in DLI following T-cell-depleted transplant in which prophylactic immunosuppression may be minimal. Current work is exploring this possibility.

Our findings have implications for post transplant adoptive immunotherapy for infectious diseases and cancer. In many studies, relatively rapid disappearance of T cells has been observed even in patients receiving simultaneous treatment with IL-2. Our models suggest that a strategy that combines active vaccination with antigen at the time of cellular therapy may help sustain the desired lymphocytes in the recipient's immune repertoire by expanding the number of tumor-specific T cells in vivo.^{20, 21, 22, 23} For transferring infectious pathogen-specific immunity, a strategy combining vaccination of the donor prior to cell donation plus peritransplant vaccination of the recipient could be studied.

References

References

Acknowledgements

This work was supported in part by a Research Scholar Grant (RSG-98-035-04-LIB) from the American Cancer Society (CAM) and grant support from the National Institutes of Health (5T32CA073954-05) (SM) (1 R01 CA10628-01) (CAM). We are grateful to Dr Cherylyn Savary for her generous assistance with flow cytometry.