Interleukin-7 Modulates Anti-Tumor CD8+ T Cell Responses via Its Action on Host Cells

The adoptive transfer of antigen-specific CD8+ T cells is a promising approach for the treatment of chronic viral and malignant diseases. In order to improve adoptive T cell therapy (ATT) of cancer, recent strategies aim at the antibody-based blockade of immunosuppressive signaling pathways in CD8+ T cells. Alternatively, adjuvant effects of immunostimulatory cytokines might be exploited to improve therapeutic CD8+ T cell responses. For example, Interleukin-7 (IL-7) is a potent growth, activation and survival factor for CD8+ T cells that can be used to improve virus- and tumor-specific CD8+ T cell responses. Although direct IL-7 effects on CD8+ T cells were studied extensively in numerous models, the contribution of IL-7 receptor-competent (IL-7R+) host cells remained unclear. In the current study we provide evidence that CD8+ T cell-mediated tumor rejection in response to recombinant IL-7 (rIL-7) therapy is strictly dependent on IL-7R+ host cells. On the contrary, CD8+ T cell expansion is independent of host IL-7R expression. If, however, rIL-7 therapy and peptide vaccination are combined, host IL-7R signaling is crucial for CD8+ T cell expansion. Unexpectedly, maximum CD8+ T cell expansion relies mainly on IL-7R signaling in non-hematopoietic host cells, similar to the massive accumulation of dendritic cells and granulocytes. In summary, we provide evidence that IL-7R+ host cells are major targets of rIL-7 that modulate therapeutic CD8+ T cell responses and the outcome of rIL-7-assisted ATT. This knowledge may have important implications for the design and optimization of clinical ATT protocols.


Introduction
The size of the peripheral T cell pool is remarkably stable throughout life. Although infections can cause a strong increase in T cell numbers, they usually return to steady-state levels after pathogen clearance. This indicates that self-regulatory mechanisms maintain T cell numbers [1]. A central factor controlling peripheral T cell homeostasis is IL-7. It acts as a growth and survival signal for T cells, which express the IL-7R and constitutively consume IL-7 [2]. Consequently, the size of the peripheral T cell pool becomes self-limiting as soon as IL-7 production and consumption reach an equilibrium [1]. Due to the lack of IL-7 consumption by T cells, IL-7 availability is increased in lymphopenic humans [3] and mice [4]. Lymphopenia-associated IL-7 overabundance contributes to the activation of naïve T cells, which undergo homeostatic or lymphopenia-induced proliferation (LIP) and convert into memory-like cells, which express high levels of CD44 and IFNγ [5].
The adoptive transfer of antigen-specific T cells is an important therapeutic option for the treatment of viral infections and cancer and has been performed successfully in animal models as well as in the clinic [6,7]. It is well established that the efficacy of adoptive T cell therapy (ATT) can be improved if recipient T cells are depleted by chemotherapy or irradiation prior to T cell transfer [6,8]. This positive effect of lymphodepletion results from the increased availability of T cell growth and survival factors such as IL-7 [9,10].
From our own experiments we know that thymus, lymph nodes, skin and intestine are the major sources of IL-7 in the mouse [11,12]. Nevertheless, steady-state IL-7 production is not sufficient for effective anti-tumor T cell responses under non-lymphopenic conditions. The injection of recombinant IL-7 (rIL-7) circumvents this problem and boosts anti-tumor T cell responses [13,14]. Since IL-7 promotes T cell survival [15,16], activation [17,18], proliferation [19] and memory T cell (T M ) formation [20] its direct action on T cells is supposed to be the major cause for its potent anti-tumor effects [21]. For the effective treatment of viral infections and cancer by ATT high numbers of adoptively transferred CD8 + cells are required in vivo [7]. Their longevity and subsequent accumulation can be improved by rIL-7 therapy suggesting that this approach can be used to improve ATT [21]. Importantly, the adjuvant effects of rIL-7 correlate with tumor growth delay rather than complete rejection [13,22,23]. Given that i) regulatory immune cells such as dendritic cells (DCs) and granulocytes expand in response to elevated IL-7 levels [24,25] and ii) non-hematopoietic cells such as fibroblasts and intestinal epithelial cells express functional IL-7 receptors (IL-7R) [12,26], we hypothesized that IL-7R + host cells might modulate anti-tumor CD8 + T cell responses.
In the current study we asked whether and how host IL-7R signaling affects ATT efficacy. For this purpose we established an ATT model, which enabled us to discriminate between direct and indirect effects of rIL-7 therapy on tumor-specific CD8 + T cells. Our data demonstrate, that LIP of CD8 + T cells and subsequent T M differentiation are promoted by rIL-7 in a host IL-7R-independent manner. However, tumor rejection strictly requires host IL-7R expression. Furthermore, we show that IL-7R + non-hematopoietic host cells are crucial for maximum CD8 + T cell expansion and T M differentiation if rIL-7 therapy is combined with peptide vaccination. Importantly, despite efficient CD8 + T cells expansion, peptide vaccination deteriorates rIL-7-dependent ATT efficacy. In summary, we provide evidence that host IL-7R signaling modulates multiple aspects of CD8 + T cells activation and T M differentiation and can promote tumor rejection in a context-dependent fashion.

Generation of bone marrow (BM) chimeras and bioluminescence detection
BM recipients were anesthetized (Ketamin/Rompun i.p.), irradiated lethally and injected with BM cells i.v. 6-18 hours later. Donor BM was isolated from femur and tibia. BM from one donor was used to reconstitute 3 recipients. BM chimeras received antibiotics via the drinking water for 3-4 weeks and were used for experiments earliest 6 weeks after BM injection. To visualize luciferase activity in live animals, bioluminescence intensities (BLI) were measured using the IVIS Imaging system (Xenogen) as described before [11,27].
Tumor cell challenge EG7 lymphoma cells produce chicken ovalbumin (OVA) and are targets of CD8 + OT-I T cells. EG7 cells were cultured in RPMI+10% FCS medium with 0.4 mg/ml G418. 1 x 10 6 cells were injected s.c. in the right flank of the indicated mice. Mice with tumors >250 mm 3 were scored as tumor positive. Tumor growth was monitored every 2-3 days. Mice were euthanized by cervical dislocation when tumors reached a diameter of 10-15 mm or when showing the following signs: hunched posture, inactivity, worsening body condition, rough coat, orbital tightening or abnormal breathing. Body weight was not assessed in this study.
In order to investigate whether IL-7 therapy promotes tumor rejection, groups of OT-Ireconstituted Rag -/mice received rIL-7 every 3-4 days for 18 days starting one day prior to adoptive T cell transfer. To improve its function, rIL-7 was complexed with IL-7-specific antibodies (αIL-7) prior to injection as described before [29]. 7/12 Rag -/mice receiving OT-I T cells plus IL-7 therapy completely rejected EG7 lymphoma cells demonstrating that IL-7 therapy strongly enhances OT-I-dependent tumor rejection in our experimental system ( Fig 1A). Importantly, rIL-7 treatment did not affect primary EG7 growth in either host (data not shown). Several studies provided evidence that rIL-7 promotes activation, survival, function of CD8 + T cells [15][16][17][18][19]30] and memory T cell (T M ) formation [20]. So far, however, these effects were considered to result from direct effects of rIL-7 on CD8 + T cells. Besides CD8 + T cells, however, numerous hematopoietic and non-hematopoietic cells express the IL-7 receptor (IL-7R) [31]. Hence, it remained unclear whether the success of IL-7-assisted adoptive T cell transfers relies on IL-7R signaling in CD8 + T cells and/or host cells. In order to address this question, Rag -/mice lacking the IL-7Rα chain (Rag -/-IL-7R -/-) were reconstituted with OT-I T cells, received IL-7 therapy or PBS and were challenged with EG7 lymphoma cells. This approach allowed us to separate direct from indirect effects of rIL-7 on CD8 + T cell-mediated lymphoma rejection. As compared to untreated controls, OT-I cells strongly delayed tumor growth in Rag -/-IL-7R -/recipients ( Fig 1B) similar to what we had observed in Rag -/mice ( Fig 1A). This demonstrates that host IL-7R-deficiency does not limit LIP-associated CD8 + T M differentiation (S1 Fig) and subsequent anti-tumor immunity ( Fig 1B). In contrast to Rag -/mice, rIL-7 treatment of OT-Ireconstituted Rag -/-IL-7R -/mice did not improve anti-tumor CD8 + T cell responses ( Fig 1B) indicating that direct effects of rIL-7 on CD8 + T cells are not sufficient for successful tumor rejection. In fact, IL-7R signaling by host cells is crucial for tumor rejection after CD8 + T cell transfer and rIL-7 therapy. Next we studied whether impaired tumor rejection in Rag -/-IL-7R -/mice resulted from reduced LIP. Rag -/and Rag -/-IL-7R -/mice were reconstituted with CD8 + OT-I T cells and spleen cells were analyzed 21-27 days later. CD44 levels (S1A Fig) and recovery rates (S1B Fig) did not differ between OT-I cells from Rag -/and Rag -/-IL-7R -/mice. In conclusion, LIP-driven expansion and T M formation are independent of host IL-7R signaling.
To analyze the impact of host IL-7R signaling on rIL-7 therapy-related T M differentiation and expansion, Rag -/and Rag -/-IL-7R -/mice were reconstituted with OT-I T cells and treated with rIL-7 (+IL-7) or PBS (-IL-7) as described above. In accordance with S1B Fig, similar numbers of OT-I cells were recovered from spleens of PBS-treated Rag -/and Rag -/-IL-7R -/mice 21-25 days after adoptive transfer ( Fig 1C). IL-7 signaling suppresses IL-7Rα chain (CD127) expression by naive T cells [32]. Accordingly, a strong and host-independent down-modulation of CD127 was observed 5 days after adoptive transfer and rIL-7 treatment ( Fig 1D). Importantly, OT-I T cells expressed less CD127 in PBS-treated Rag -/-IL-7R -/than in Rag -/mice arguing for elevated steady state IL-7 levels in the latter. Nevertheless, this did not affect the long-term abundance ( Fig 1C) or early rIL-7 responsiveness of OT-I T cells ( Fig 1D).
In conclusion, host IL-7R signaling is dispensable for rIL-7-driven CD8 + T cell expansion ( Fig 1C) but not for subsequent tumor rejection. This highlights that high numbers of therapeutic CD8 + T cells do not necessarily correlate with therapeutic success.
Host IL-7R signaling is not required for rIL-7-induced T M differentiation but for granulocyte and DC expansion CD8 + T M express high levels of CD127 and the anti-apoptotic molecule B cell lymphoma protein-2 (Bcl-2) [33], which can be up regulated by IL-7 [34]. In PBS-treated Rag -/mice CD127 hi and Bcl-2 hi OT-I T M were less frequent than in Rag -/-IL-7R -/mice (Fig 2A and 2B). In response to rIL-7 treatment, however, resulting OT-I T M expressed elevated levels of CD127 and Bcl-2 in both hosts 21-25 days after transfer (Fig 2A and 2B). Hence, despite early CD127 down-regulation (Fig 1D), IL-7 therapy promoted the generation of CD127 hi Bcl-2 hi OT-I T M irrespective of the host. IL-7 therapy also promoted the generation of CD62L hi OT-I cells ( Fig  2C) expressing low levels of KLRG-1 ( Fig 2D). Furthermore, a high percentage of OT-I cells recovered from both hosts rapidly produced IFN-γ after short-term in vitro re-stimulation ( Fig 2E). Hence, rIL-7 therapy favors the generation of functional CD8 + T M cells with a CD127 hi Bcl-2 hi CD62L hi KLRG-1 lo phenotype in a host-independent fashion. Interestingly, despite their high numbers ( Fig 1C) and favorable phenotype (Fig 2), rIL-7-induced OT-I T M cells failed to reject tumors in Rag -/-IL-7R -/mice ( Fig 1B). This suggested a contribution of rIL-7-responsive host cells to EG7 elimination.
Host IL-7R signaling promotes CD8 + T cell expansion and modulates T M differentiation in response to peptide vaccination and IL-7 therapy IL-7 administration improves vaccination-induced T cell responses [13,22]. To test whether this also requires host IL-7R signaling, Rag -/and Rag -/-IL-7R -/mice were reconstituted with OT-I cells, immunized with SIINFEKL and treated with rIL-7. Peptide-vaccinated mice receiving PBS served as controls. As shown in Fig 4A rIL-7 treatment strongly increased spleen cell numbers in Rag -/but not Rag -/-IL-7R -/mice. Furthermore, rIL-7 treatment promoted OT-I cell expansion in Rag -/mice (Fig 4B). Although to a much lesser extent, this was also observed in rIL-7-treated Rag -/-IL-7R -/mice ( Fig 4B). Thus, IL-7R signaling in CD8 + T cells promotes some degree of rIL-7-induced CD8 + T cell expansion although the full-blown response requires host IL-7R expression.
To test whether IL-7R + host cells also modulate T M differentiation the phenotype of peripheral blood OT-I cells was determined by flow cytometry. In Rag -/and Rag -/-IL-7R -/mice, frequencies of CD62L hi OT-I cells were similar after vaccination (Fig 4C). This was further increased by rIL-7 and was independent of host IL-7R expression (Fig 4C). OT-I cells expressed similar levels of CD127 after vaccination of Rag -/and Rag -/-IL-7R -/mice. To our surprise, rIL-7 therapy further promoted CD127 expression by OT-I cells only in Rag -/mice ( Fig 4D). KLRG-1 hi OT-I cells were more frequent in vaccinated Rag -/-IL-7R -/than in Rag -/mice. In both hosts their abundance decreased in response to rIL-7, though to a lesser extent in Rag -/-IL-7R -/mice ( Fig 4E). After vaccination, OT-I proliferation was lower in Rag -/mice as shown by low numbers of Ki67 hi cells (Fig 4F). Bcl-2 expression was similar in vaccinated Rag -/and Rag -/-IL-7R -/mice and was up regulated by rIL-7 only in the latter (Fig 4G). In summary, the rIL-7-induced up-regulation of CD62L by OT-I cells was largely independent of IL-7R + host cells. In contrast, host cells promoted the expansion of OT-I cells (Fig 4B) and modulated their expression of CD127, KLRG-1 and Bcl-2 in response to rIL-7 therapy (Fig 4D, 4E and 4G).
After vaccination, OT-I cells isolated from Rag -/mice produced high levels of IFN-γ ( Fig  4H). This response was less pronounced in Rag -/-IL-7R -/mice but could be improved by rIL-7 treatment (Fig 4H). Importantly, rIL-7 treatment could not further promote the generation of IFN-γ hi OT-I cells in Rag -/mice (Fig 4H). OT-I cells producing high levels of TNF-α were similarly frequent in Rag -/and Rag -/-IL-7R -/mice and further expanded in both hosts in response to rIL-7 treatment (Fig 4I). The expression of PD-1, a marker for dysfunctional T cells, was comparable for OT-I cells recovered from mice of both strains. In agreement with a previous study [35], rIL-7 administration reduced PD-1 expression on OT-I cells in Rag -/mice ( Fig 4J). Surprisingly, this was not the case in Rag -/-IL-7R -/mice ( Fig 4J) indicating that IL-7R + host  cells determine the levels of PD-1 expression by CD8 + T cells. Hence, immunomodulation by rIL-7 relies on complex interactions between IL-7R + host and therapeutic CD8 + T cells.

IL-7R + non-hematopoietic cells are crucial for CD8 + T cell expansion in response to vaccination and rIL-7 therapy
Prolonged exposure to elevated levels of IL-7 down modulates il-7 gene activity in non-BMderived cells [4,12], which are the major source of IL-7 in vivo [20]. To determine whether non-hematopoietic cells respond to rIL-7 therapy, IL-7 reporter mice expressing luciferase under the control of the il-7 promoter [11] were treated with rIL-7. As a readout for il-7 gene activity, bioluminescence intensities (BLI) were determined before and after rIL-7 treatment. As shown in S2 Fig, rIL-7 administration reduced il-7 promoter activity significantly demonstrating that IL-7-producing non-BM-derived cells indeed respond to rIL-7 therapy.
Similar to DCs, OT-I accumulation in the spleen was independent of IL-7R expression by BM derived cells if non-BM-derived cells expressed the IL-7R (Fig 5E; R ! R vs RR ! R). In accordance with Fig 4B, OT-I responses were least efficient in chimeras lacking the IL-7R on BM-derived and non-BM-derived cells (Fig 5E; RR ! RR). Importantly, IL-7R expression by BM-derived cells alone was insufficient to fully recover OT-I accumulation (Fig 5E; R ! RR). IL-7 treatment is known to alter homing patterns of CD8 + T cells [36]. To exclude migration-related differences in splenic OT-I cell numbers, BM chimeras were reconstituted with renilla luciferase-transgenic (ChRluc tg ) CD8 + OT-I T cells [27], vaccinated and treated with

IL-7R + non-BM cells modulate CD8 + T M differentiation in response to vaccination and rIL-7 therapy
Having shown that non-BM cells promote the expansion of CD8 + T cells, we analyzed their phenotype. Three weeks after adoptive transfer, OT-I cells isolated from the spleens of BM chimeras were analyzed by flow cytometry. In agreement with Fig 4E, KLRG-1 hi OT-I cells were more frequent in RR ! RR than in R ! R chimeras (Fig 6A) further emphasizing that T M differentiation into KLRG-1 hi cells is modulated by IL-7R + host cells. These cells appear to be of BM and non-BM origin as shown by the fact that R ! RR chimeras contained most KLRG-1 hi OT-I cells. However, their numbers were reduced in RR ! RR chimeras and lowest in R ! R and RR ! R chimeras (Fig 6A).
The frequencies of CD62L hi OT-I cells were identical in R ! R and RR ! RR chimeras suggesting their host IL-7R-independence (Fig 6B). However, CD62L hi OT-I cells were less abundant in RR ! R and R ! RR (Fig 6B). This indicates that IL-7R + BM-and non-BM-derived cells are part of a complex network exerting opposing functions on CD62L hi T M differentiation.
Highest levels of CD127 were found on OT-I cells from R ! R and RR ! R chimeras and lowest on those from RR ! RR chimeras (Fig 6C) demonstrating that IL-7R expression by non-BM cells is a prerequisite for maximum CD127 expression by CD8 + T cells. If only BMderived cells expressed the IL-7R (R ! RR) CD127 levels were significantly higher than in RR ! RR chimeras but still below those found on OT-I cells primed in RR ! R chimeras ( Fig  6C). Thus, IL-7R + non-BM cells are major regulators of CD127 expression by OT-I cells.
The genes encoding Bcl-2 and CD8 are direct targets of IL-7 and other cytokines utilizing the IL-2Rγ c (CD132) for signal transduction [37]. CD127 lo OT-I cells from RR ! RR chimeras expressed highest levels of CD132, Bcl-2 and CD8 (Fig 6D-6F) suggesting that other CD132-dependent cytokines than IL-7 caused the differentiation of CD127 lo CD132 hi Bcl-2 hi CD8 hi T M cells in RR ! RR chimeras. In contrast, R ! R and RR ! R chimeras rather contained CD127 hi CD132 lo Bcl-2 lo CD8 lo OT-I cells (Fig 6C-6F). However, if only BM-derived cells expressed the IL-7R (R ! RR) we observed an intermediate OT-I phenotype (Fig 6C-6G  and 6F).
In response to rIL-7 treatment PD-1 was downregulated only on OT-I cells primed in Rag -/mice but not on those primed in Rag -/-IL-7R -/mice (Fig 4J). In agreement with this, PD-1 expression was most pronounced in RR ! RR chimeras (Fig 6G). PD-1 levels were significantly lower in both groups of chimeras expressing IL-7R on non-BM-derived cells and (R ! R and RR ! R). In contrast, OT-I cells from R ! RR chimeras showed an intermediate phenotype (Fig 6G) indicating that IL-7R + non-BM-derived cells are major regulators of PD-1 expression by CD8 + T cells.
In summary, IL-7R + host cells promoted the expansion of CD8 + T M cells in response to rIL-7 therapy (Fig 4B-4J). Surprisingly, regulatory host cells were mainly of non-BM origin (Figs 5E, 5F and 6).
The combination of rIL-7 therapy and peptide vaccination impairs T celldependent tumor rejection in Rag -/mice CD8 + CD62L hi KLRG-1 lo IFN-γ hi PD-1 lo T M cells are well suited to provide long-term protection against chronic infections and tumors [7,38,39]. Peptide vaccination and rIL-7 therapy induced the generation of such T M cells in Rag -/mice (Fig 4B-4J), a process that was mainly controlled by IL-7R + non-hematopoietic cells (Figs 5E, 5F and 6). However, their therapeutic potential remained unclear. To test this, Rag -/mice were reconstituted with CD8 + OT-I T cells and vaccinated with SIINFEKL one day later. Additionally, mice received rIL-7 or PBS according to the scheme described above. To ensure appropriate OT-I expansion and differentiation, mice were challenged with EG7 cells 21 days after adoptive T cell transfer. In untreated Rag -/control mice, EG7 tumors grew rapidly (Fig 7A). In contrast, 6/12 peptide-vaccinated Rag -/mice remained tumor free ( Fig 7A). Surprisingly, however, only 2/12 Rag -/mice rejected EG7 lymphomas after rIL-7 therapy (Fig 7A). Importantly, this was not due to impaired DC expansion in rIL-7-treated Rag -/mice ( Fig 7C). Hence, the protective effect of peptide vaccination was blunted by rIL-7.

Discussion
The major goal of our study was to clarify whether and how host IL-7R signaling contributes to rIL-7-driven anti-tumor CD8 + T cell responses. For this purpose, SIINFEKL-specific CD8 + OT-I cells were transferred into Rag -/and Rag -/-IL-7R -/mice, which were then treated with rIL-7. As shown in Fig 1A, OT-I-reconstituted Rag -/mice receiving IL-7 therapy rejected SIINFEKL-expressing EG7 lymphoma cells. This was not the case for Rag -/-IL-7R -/mice demonstrating that the success of rIL-7 therapy is dependent on host IL-7R expression in our experimental system. However, CD8 + T cell expansion and differentiation were largely independent of host IL-7R signaling. Irrespective of host IL-7R expression nearly identical numbers of OT-I cells were recovered from spleens after rIL-7 treatment. Similarly, the differentiation of CD8 + T M with a CD127 hi Bcl-2 hi CD62L hi KLRG-1 lo IFN-γ hi phenotype was induced in rIL-7-treated Rag -/and Rag -/-IL-7R -/mice. Hence, our data show that rIL-7-induced CD8 + T cell expansion and subsequent T M differentiation are not affected by host IL-7R expression and related differences in IL-7 availability (Fig 1D). In agreement with previous studies [21], the beneficial effects of rIL-7 on CD8 + T cell function, differentiation and survival appear to result mainly from IL-7R signaling in CD8 + T cells. Nevertheless, it is important to stress that effective CD8 + T cell expansion and differentiation do not necessarily correlate with tumor rejection. This conclusion is supported by the fact that rIL-7-induced OT-I expansion and subsequent T M differentiation occurred efficiently in Rag -/-IL-7R -/mice while tumor rejection failed. This finding emphasizes the importance of IL-7-responsive host cells for rIL-7-assisted ATT in our model system.
Dendritic cells promote CD8 + T cell responses under lymphopenic conditions [40], crosspresent tumor-derived antigens [41] and expand in response to rIL-7 treatment [24], which promotes T-DC interactions [42]. Additionally, rIL-7 stimulates myelopoiesis and the subsequent accumulation of CD11b + cells [25], which can cross-present tumor antigens and promote CD8 + T cell-mediated tumor rejection [43]. After rIL-7 treatment, DCs and granulocytes accumulated only in IL-7R-competent mice. However, this accumulation was not required for expansion and functional maturation of OT-I cells but correlated positively with tumor rejection. This suggests that rIL-7-expanded DCs and granulocytes support CD8 + T cell responses in the late effector phase rather than in the early phase after adoptive transfer.
IL-7R signaling in combination with TCR stimulation boosts CD8 + T cell responses in multiple experimental systems. TCR signal strength and the timing of IL-7R signaling appear to be important to achieve optimal IL-7 effects [44]. To generate maximum CD8 + T cell responses, we reconstituted Rag -/and Rag -/-IL-7R -/mice with OT-I cells and vaccinated them with SIIN-FEKL. Unlike after rIL-7 treatment alone, host IL-7R signaling was crucial for maximum OT-I expansion in response to vaccination and rIL-7. Additionally, OT-I cells up-regulated CD127 and down-modulated PD-1 in a host-IL-7R-dependent fashion. This was not the case for all other markers tested. The host-dependent modulation of CD8 + T cell differentiation was confirmed in BM chimeras. Surprisingly, we identified IL-7R + radio-resistant host cells as major regulators of CD8 + T cell expansion and differentiation. For maximum levels of CD127 expression and restriction of KLRG-1, CD132, Bcl-2, CD8, and PD-1, IL-7R expression by radioresistant host cells was sufficient. CD127 expression is modulated by multiple intracellular signaling events in T cells. Whether non-BM-derived host cells affect these or other signaling pathways remains open.
Besides OT-I expansion and differentiation, IL-7R + non-BM-derived cells also controlled granulocyte and DC expansion. Even if donor BM was devoid of the IL-7R, its expression by radio-resistant cells supported the expansion of CD11b + cells and DCs. However, IL-7R expression by BM-derived cells was largely sufficient to rescue the expansion of CD11b + cells and CD8 -DCs. Surprisingly, this was not the case for CD8 + DCs. Their accumulation depended more on IL-7R expression by non-BM-than by BM-derived cells. Together, we demonstrate that IL-7R signaling in BM-and non-BM-derived cells contributes to rIL-7-driven DC expansion. This might help to reconcile apparently conflicting results regarding the relative importance of cell autonomous IL-7R signaling for DC generation obtained in different experimental systems [24,45].
Fibroblastic reticular cells (FRCs) and lymphoid endothelial cells (LECs) are major sources of IL-7 in secondary lymphoid organs [46]. Furthermore, other cells of non-BM origin such as intestinal epithelial cells (IECs) [47], keratinocytes [48], hepatocytes [49] and fibroblasts [50] were shown to produce IL-7 in vivo. The maintenance of CD8 + T cell homeostasis requires IL-7R signaling in CD8 + T cells which is triggered by IL-7 from non-BM derived cells [20]. However, only little is known about the consequences of IL-7R signaling in non-BM-derived cells and subsequent immune modulation [51]. As we have shown in S2 Fig, the application of rIL-7 leads to the systemic down-modulation of il-7 gene activity supporting previous reports demonstrating that il-7 gene activity is regulated in an IL-7-mediated negative feedback loop [4,12]. Given that IL-7R signaling modulates gene expression profiles in multiple non-BM-derived cell types [4,12,26], long-term rIL-7 therapy would not only affect immune cell but also nonimmune cell homeostasis. For instance, prolonged IL-7 overabundance promotes IEC expansion, survival and subsequent alterations in intestinal physiology [12]. Since many cell types of non-BM-derived origin can express the IL-7R [31], the local down regulation of endogenous IL-7 production and alterations in tissue homeostasis might be as yet underestimated side effects of rIL-7 therapy. Whether i) non-BM-derived cells located in the tumor stroma and/or other tissues are the main targets of rIL-7, and ii) whether rIL-7 signaling in such cells promotes or suppresses rIL-7-assisted ATT remains to be shown.
IL-7R + host cells appear to promote antigen-dependent CD8 + T cell function also in an IL-7-independent fashion. For example, successful peptide vaccination and subsequent tumor rejection, in the absence of rIL-7 treatment, strictly required host IL-7R expression (Fig 7A and  7B). Importantly, however, peptide vaccination blocked the therapeutic effect of rIL-7 (Figs 1A and 7A) in Rag -/mice although IL-7R-dependent DC expansion was normal (Fig 7C). It has been reported that TCR signaling can interfere with the beneficial effects of IL-7 on T cells [52]. Hence, impaired tumor rejection in rIL-7-treated, peptide vaccinated Rag -/mice might have been a result of altered T cell rather than host cell function.
In summary, our data provide evidence for the complex interplay between IL-7R + host and CD8 + T cells in the course of anti-tumor CD8 + T cell responses. While productive host-CD8 + T cell interactions can be promoted by rIL-7 therapy, the inappropriate combination with other immune stimuli can cause adverse effects.