IL-7/anti–IL-7 mAb complexes augment cytokine potency in mice through association with IgG-Fc and by competition with IL-7R

Interleukin-7 (IL-7) is a cytokine of central importance to the development and homeostasis of the adaptive immune system in mice and humans.^1-5  Among its pleiotropic effects, IL-7 determines the overall size of the resting T-cell pool as the prototypic survival factor for T cells.⁶  In lymphoid tissues, stromal cells constitutively produce the IL-7 relevant to T-cell homeostasis, and IL-7 levels are thought to be controlled through consumption by IL-7 receptor (IL-7R)–expressing cells.^6-8 

At supraphysiological levels, IL-7 is a potent mitogen. T cells that are adoptively transferred to lymphopenic hosts undergo slow IL-7–driven homeostatic proliferation.¹  Furthermore, treatment with exogenous IL-7 drives T-cell expansion in mice and primates.^9,10  The T-cell mitogenic properties of IL-7 have inspired clinical trials to explore the use of IL-7 as an adjuvant in suboptimal immune responses or as a means to reconstitute lymphodepleted individuals.¹⁰  One striking finding from these studies is that the calculated volume of distribution for IL-7 is massive.¹¹  This observation implies that an IL-7 sink exists in vivo which can rapidly absorb exogenous cytokine. Such a buffer system is consistent with the consumption model of IL-7 regulation, and IL-7 therapies might be further improved if the IL-7 sink is circumvented to deliver more cytokine on-target.

In mice, one method known to improve the potency of IL-7 treatment is to administer the cytokine as a prebound complex with a neutralizing anti–IL-7 monoclonal antibody (mAb), clone M25 (mouse IgG2b).^12,13  Indeed, IL-7/M25 complexes display in vivo biological potency that is 50- to 100-fold greater than that of IL-7 alone. A similar agonist effect has been reported for IL-2, IL-3, IL-4, and IL-6 in complex with their corresponding neutralizing anti-cytokine monoclonal immunoglobulin G (IgG).^12,14-17  It remains to be seen whether complexes of cytokine and mAb (cytokine/mAb) have agonistic effects in humans.

Despite their potential utility, the mechanism of action for agonist cytokine/mAb remains enigmatic. IgG, by virtue of its Fc domain, is endowed with unique pharmacokinetic properties, which it may impart to the associated cytokine. Cells such as macrophages and dendritic cells express cell-surface receptors for the IgG Fc domain, FcγR, which may capture and affect the distribution or presentation of cytokine/mAb.¹⁸  Additionally, cytokine/mAb complexes are likely to benefit from the actions of the neonatal Fc receptor, FcRn, which binds Fc in acidifying endosomes and recycles it back to the extracellular space, thereby avoiding degradation and prolonging the in vivo lifespan of rescued molecules.¹⁹  The importance of the Fc domain is evident in the diminished in vivo potency of cytokine complexes formed with F(ab′)₂ or Fab fragments of the anti-cytokine mAbs.^13,14  Nevertheless, cytokine/Fab fragment complexes still elicit stronger biological responses in vivo than cytokine alone, suggesting that the binding interaction between the mAb and cytokine may contribute to the phenomenon. An interesting feature observed among the cytokine/mAb pairs tested so far is that neutralizing mAbs are more effective than nonneutralizing mAbs in forming potent complexes.^12-14  Hence, the ability of the mAb to obscure its target cytokine from the receptor correlates with increased in vivo potency as a cytokine/mAb pair.

The paradox of a neutralizing antibody (Ab) augmenting the potency of the bound cytokine in vivo has been examined in detail only for IL-2/mAb.^20,21  We and others have demonstrated that IL-2/mAb potentiates IL-2 activity in vivo through a two-part mechanism that extends cytokine half-life and selectively focuses the cytokine to one of two forms of IL-2R.^20,21  Mechanistically, however, IL-2/mAb is a poor archetype cytokine/mAb because of the unique nature of IL-2R. IL-3, IL-4, and IL-6 are more similar to IL-7 in that there is only one type of receptor available to each cytokine. Therefore, the mechanism of IL-7/M25 necessarily cannot involve shunting cytokine to one of multiple receptor forms.

To better understand the potentiating effect of neutralizing Abs on their target cytokines, we examined the pharmacokinetic parameters of IL-7/M25 and dissected the individual contributions of the Fab and Fc domains to improving IL-7–driven CD8⁺ T-cell proliferation in vivo. Our results indicate that, despite a systemic delivery of the treatment, the majority of stimulation by IL-7/M25 is available to cells within the T-cell zones of secondary lymphoid tissue. We find that host expression of FcRn is crucial to maintaining the normal in vivo lifespan of M25 and likewise the full effect of IL-7/M25 treatment. A fusion protein of IL-7-Fc, in which IL-7 is covalently associated with IgG Fc, also exhibits potent mitogenic activity in vivo. However, IL-7-Fc is measurably less potent than IL-7/M25. Our findings suggest that, in addition to the longer half-life imparted by Fc and FcRn, the in vivo effect of IL-7 is enhanced because M25 restricts IL-7 binding to IL-7R.

C57BL/6 (referred to as B6.CD90.2⁺ or B6.CD45.2⁺), B6.PL-Thy1^a (B6.CD90.1⁺), B6.SJL-Ptprc^aPepc^b (B6.CD45.1⁺), and B6.129 × 1-Fcgrt^tm1Dcr (FcRn^−/−)²²  mice were obtained from The Jackson Laboratory. IL-7tg⁺ mice²³  (available from The Jackson Laboratory) were maintained on a B6.CD90.1⁺ background and were hemizygous for the transgene. Fcγ^−/−FcγRIIb^−/−24,25  and FcγRI^−/−26  mice were provided by Dr Jeffery Ravetch (Rockefeller University, New York, NY) and Dr Jeanne Baker (Elan Pharmaceuticals, San Francisco, CA), respectively, and crossed together to generate FcγR^−/− (ie, Fcγ^−/−FcγRIIb^−/−FcγRI^−/−) mice.²⁷  Mice were housed under specific pathogen-free conditions at The Scripps Research Institute or the La Jolla Institute for Allergy and Immunology and used at 2 to 6 months of age. Bone marrow (BM) chimeras were generated as described.²⁸  Where indicated, host mice were treated with FTY720 at a dose of 3 mg/kg dissolved in 1× phosphate-buffered saline + 0.1% bovine serum albumin by intraperitoneal injection on days −1 and 3 of 7-day experiments. Experiments involving the use of animals were approved by the Institutional Animal Care and Use Committees at The Scripps Research Institute and the La Jolla Institute for Allergy and Immunology.

Recombinant human IL-7 (rhIL-7) was obtained from the National Cancer Institute Biological Resources Branch. Where specified, recombinant mouse IL-7 (rmIL-7) was purchased from eBioscience (San Diego, CA). IL-7-Fc is a recombinant protein of hIL-7 fused to a mutant, nonlytic form of murine Fcγ2a (L235E, E318A, K320A, K322A) and was prepared as previously described.²⁹  Anti–IL-7 mAbs have been described previously.^13,30  Fab fragments of M25 (M25_Fab) were generated by using papain (Worthington Biochemical), and the IL-7 neutralizing capacity of M25_Fab was standardized to M25 by in vitro T-cell survival assay.

Freshly harvested B6 lymph node (LN) cells were cultured (5 × 10⁵/mL) in complete RPMI media with serial dilutions of IL-7 or IL-7-Fc. Forty-eight hours later, cells were washed and stained with 2 ng/mL propidium iodide, 2.4G2, and fluorescent Abs against CD4 and CD8. The frequency of surviving CD8⁺CD4⁻ T cells was determined by flow cytometry. To determine the relative IL-7 neutralizing activity of mAbs and Fab fragments, primary LN cells were cultured with a constant concentration of IL-7 (1 ng/mL) and titrating doses of mAb or Fab.

Lymphocytes were prepared from the pooled (inguinal, axillary, cervical, and mesenteric) LNs of donor mice. Where indicated, CD8⁺ T cells were enriched from donor LNs to >95% purity as previously described.^28,31  Treatment with pertussis toxin (PTX) was performed by incubating 4 × 10⁷ cells/mL in RPMI containing 0.1M N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid and 2% fetal calf serum with 100 ng/mL PTX (List Biological Laboratories, Campbell, CA) for 2 hours at 37°C, and washed >4 times. Donor cells were labeled with 5 μM of either carboxyfluorescein diacetate succinimidyl ester (CFSE) or CellTrace Violet (CTV) (both from Molecular Probes), as previously described,²⁸  and intravenously injected into host mice.

Cells (2 × 10⁶) from host spleen and pooled LNs were stained for flow cytometry, as previously described,^14,28,31  with mAbs from Biolegend and/or eBioscience. Data from the stained cells were collected with a BD LSR-II flow cytometer and analyzed by using FlowJo (Treestar). Statistical analyses by unpaired two-tailed t tests were conducted with Prism 5 (GraphPad Software).

T cells responding to endogenous IL-7 require access to the T-cell zones of secondary lymphoid tissue, the site of IL-7 production.^8,32  To determine whether a spatial requirement also exists for T-cell responsiveness to IL-7/M25, we analyzed the effect of interrupting T-cell trafficking, as induced by PTX and FTY720. PTX neutralizes Gα_i protein-coupled receptors such as CCR7 and prevents homing of T cells to the LNs and white pulp of the spleen.^32-34  Reciprocally, FTY720, a sphingosine analog, disrupts sphingosine-1-phosphate receptor function and prevents egress of T cells from lymphoid tissues.^35-37 

The effect of PTX was analyzed by measuring the response of T cells pretreated with PTX. Thus, B6.CD90.1⁺ LN cells were incubated in media with PTX, labeled with CFSE, and adoptively transferred to B6.CD90.2⁺-recipient mice. The hosts then received intraperitoneal injections of IL-7/M25 (1.5 μg/7.5 μg) on days 1, 3, and 5. The CFSE profiles of CD90.1⁺ CD8⁺ T cells from host spleens were analyzed on day 7. Control donor T cells pretreated in media alone proliferated vigorously in response to injections of IL-7/M25 (Figure 1A). In contrast, PTX-pretreated donor cells were found to undergo minimal cellular division. Such an impaired response to IL-7/M25 is not likely due to disruptions in IL-7 signaling pathways because PTX-treated T cells displayed normal responsiveness to the in vitro survival-promoting function of IL-7 (Figure 1B). Furthermore, it is known that PTX treatment does not impose a general restriction on T-cell proliferation,^32,33  although it inhibits T-cell homing into the LNs (Figure 1C). Similarly, CCR7^−/− donor CD8⁺ T cells, which have an impaired ability to home to host LNs, displayed a decreased responsiveness to IL-7/M25 (data not shown). Thus, the inability of CD8⁺ T cells to access the LN correlates with a greatly attenuated ability to respond to IL-7/M25 in vivo.

The above results support a model in which CD8⁺ T cells encounter the majority of IL-7/M25 in the T-cell zones. To further test this idea, we sought to determine the effect of obstructing T cells from exiting the T-cell zones by treating hosts with FTY720. In these hosts, IL-7/M25–induced proliferation of donor CD8⁺ T cells was only slightly diminished relative to hosts not receiving FTY720 treatment (Figure 1A, bottom). This finding applies to donor cells recovered from the spleen as well as from assorted LNs, indicating that the slight decrease in response to IL-7/M25 in the presence of FTY720 was uniform across the various host secondary lymphoid tissues (Figure 1D). Together with observations of PTX-treated donor cells, these data suggest that the stimulation induced by IL-7/M25 occurs more in the T-cell zones of secondary lymphoid organs than elsewhere.

The role of the Fc portion of M25 in mediating the strong potency of IL-7/M25 was assessed by using mice deficient in two types of Fc receptors, FcγR and FcRn. To test the putative roles of FcγR and FcRn, purified B6.CD90.1⁺ CD8⁺ T cells were labeled with CFSE and adoptively transferred to wild-type, FcγR^−/−, FcRn^−/−, or FcγR^−/−FcRn^−/− hosts treated with IL-7/M25. As observed previously, donor CD8⁺ T cells proliferate in wild-type hosts treated with IL-7/M25 (Figure 2A). The intensity of IL-7/M25–driven donor T-cell proliferation was minimally affected in FcγR^−/− hosts but was dramatically reduced in FcRn^−/− hosts (Figure 2A). These findings indicate that the various effector functions mediated by the FcγR are dispensable but that FcRn is essential to the potency of IL-7/M25, presumably by extending in vivo half-life.

FcRn prolongs the lifespan of circulating Abs.^38,39  Indeed, we observed greatly abbreviated serum persistence of M25 in FcRn^−/− vs wild-type mice (supplemental Figure 1B). However, to exclude the possibility that FcRn deficiency may exert a general dampening of the response to IL-7, we confirmed that T cells respond normally to other forms of IL-7 stimuli in FcRn^−/− mice. The response to increased levels of endogenous IL-7 was assessed by adoptively transferring CFSE-labeled T cells into irradiated hosts and was found to be equally efficient in irradiated FcRn^−/− and wild-type mice (supplemental Figure 1C). Donor T-cell response to a high dose of exogenous IL-7 treatment was also not significantly different between FcRn^−/− and wild-type hosts (Figure 2B). Moreover, injections of IL-7 complexes generated with M25_Fab, which drive a low amount of CD8 T-cell proliferation,¹³  induced an equal amount of donor T-cell expansion in FcRn^−/− and wild-type hosts (supplemental Figure 1D). Thus, the relatively weak potency of IL-7/M25 in FcRn^−/− mice is specific to IL-7/M25 and not due to a generalized defect in recognition of IL-7 in FcRn^−/− mice.

To assess the discrete contributions of FcRn expressed by hematopoietic or nonhematopoietic cell types, BM chimera mice were generated with FcRn expression restricted to either BM-derived (wtBM→FcRn^−/−) or to radioresistant cells (FcRn^−/−BM→wt). Control and chimeric BM-reconstituted mice served as hosts to compare the potency of IL-7/M25, as reflected by the proliferation of donor CD8⁺ T cells. As expected, IL-7/M25 induced greater CD8⁺ T-cell proliferation in control FcRn-sufficient (ie, wtBM→wt) than in FcRn-deficient (ie, FcRn^−/−BM→FcRn^−/−) hosts (supplemental Figure 1E). Chimeric hosts with FcRn expression restricted to either the BM-derived or radioresistant compartments both supported an intermediate level of donor cell expansion in response to IL-7/M25 (supplemental Figure 1E). Thus, maximum potency of IL-7/M25 requires contributions from FcRn expressed by cells of both the hematopoietic and radioresistant compartments.

Since the in vivo lifespan of IL-7/M25 is much greater than that of IL-7,^13,40  it is possible that simply extending the duration of IL-7 availability in vivo is sufficient to match the potency of IL-7/M25. To probe this hypothesis, we compared the donor CD8⁺ T-cell proliferative response in hosts treated with one dose of IL-7/M25 to that induced by multiple injections of IL-7 administered over 24 hours (ie, 12 doses at 2-hour intervals). Analysis was conducted 5 to 6 days after the initiation of treatment. A single injection of IL-7/M25 stimulated a modest but detectable response of donor CD8⁺ T cells at a dose of 1.5 μg IL-7 plus 7.5 μg M25 and a substantial response at a higher dose of 10 μg IL-7 plus 50 μg M25 (Figure 3A, middle). In contrast, proliferation of donor cells in mice receiving 12 injections of IL-7 alone was modest whether the hosts were treated with 1.5 μg or 10 μg IL-7 per injection (Figure 3A, bottom). These results indicate that merely prolonging the availability of free IL-7 is insufficient to recapitulate the proliferative effect of IL-7/M25 in vivo.

M25_Fab fragments are less effective than whole M25 in enhancing the in vivo potency of the bound IL-7, presumably because of their inability to benefit from the FcRn-mediated extension of Ab half-life (Figure 2 and Boyman et al¹⁴ ). To determine whether prolonged in vivo availability of IL-7/M25_Fab results in improved potency, we analyzed the proliferation of donor CD8⁺ T cells in hosts treated with multiple injections of IL-7/M25_Fab administered every 2 hours for 24 hours. As controls, some hosts were treated with only 1 dose of IL-7/M25 or IL-7/M25_Fab. Predictably, a single injection of IL-7/M25 (3 μg/15 μg) stimulated greater donor CD8⁺ T-cell proliferation than a single injection of IL-7/M25_Fab (3 μg/equivalent of 15 μg of M25) (Figure 3B). Although the effect of a single injection of IL-7/M25_Fab was barely detectable, treatment with a series of 12 injections of IL-7/M25_Fab elicited a donor cell response of a magnitude similar to that of a single injection of IL-7/M25 (Figure 3B). Therefore, in contrast to the response to repeated injections of free IL-7, extending the availability of IL-7/M25_Fab through multiple injections effectively recapitulated the effect of a single injection of IL-7/M25. This finding likely reflects a slightly longer in vivo lifespan of IL-7/M25_Fab relative to that of free IL-7, perhaps because the smaller IL-7 was more prone to glomerular filtration or because M25_Fab interferred with another clearance mechanism unrelated to the kidney.

To better understand the kinetics of T-cell stimulation by IL-7/M25, we analyzed the surface expression of the IL-7Rα chain, CD127, which is downregulated in the presence of IL-7 and recovers to normal levels 12 hours after the removal of IL-7 stimulus.⁴¹  Control injection with a low dose (1.5 μg) of IL-7 induced a small but significant downregulation of CD127 on LN CD8⁺ T cells by 12 hours; in contrast, injection of 1.5 μg /7.5 μg IL-7/M25 elicited a stronger CD127 downregulation by 6 hours and reached a nadir at 12 hours followed by a prolonged period of gradual return to normal levels for the next 24+ hours (Figure 3C). Interestingly, IL-7 injected at a high dose (10 μg) resulted in rapid and profound CD127 downregulation for 12 hours with a complete return to normal levels between 12 and 24 hours posttreatment. These results are consistent with an early and transient tempo of availability when IL-7 is administered as a free cytokine and a delayed kinetics of availability for IL-7 bound to M25.

To probe the extent to which the Fc portion accounts for the mitogenic activity of IL-7/M25, the potency of the complexes was compared with an IL-7-Fc fusion protein.²⁹  In order to achieve a fair comparison between IL-7/M25 and IL-7-Fc, the relative activity of IL-7-Fc and IL-7 was determined by an in vitro T-cell survival assay (data not shown). The in vivo potency of the fusion protein was then compared with a functionally equivalent dose of IL-7, either as free cytokine or as bound to M25, by measuring the induced proliferation of CD8⁺ donor T cells in wild-type hosts. Although injections of IL-7 alone failed to induce a response, efficient proliferation of donor CD8⁺ T cells was observed in hosts treated with either IL-7/M25 or IL-7-Fc, with the complexes stimulating faster proliferation than the fusion protein (Figure 4A). Moreover, a significantly greater number of donor CD8⁺ T cells were recovered from hosts treated with IL-7/M25 than from those treated with IL-7-Fc (Figure 4B, left). As expected, the potency of both IL-7/M25 and IL-7-Fc was dependent on host expression of FcRn (Figure 4B). Thus, associating an IgG Fc domain with IL-7 serves to greatly improve the in vivo potency of the cytokine, albeit slightly less than that achieved by IL-7/M25.

The greater mitogenicity of IL-7/M25 relative to IL-7-Fc is consistent with M25 contributing to improved IL-7 potency in vivo by engaging additional mechanism(s) besides the involvement of FcRn. One apparent difference between the IL-7/M25 and the fusion protein is that the epitope/paratope interaction between IL-7 and M25 is absent in IL-7-Fc. Thus, we asked whether binding M25 to the IL-7 moiety of IL-7-Fc (ie, in the form of IL-7-Fc/M25) might enhance the in vivo potency of the fusion protein. Surprisingly, in stark contrast to IL-7-Fc, injections of IL-7-Fc/M25 nearly completely failed to induce proliferation of donor CD8⁺ T cells in wild-type as well as FcγR^−/− hosts (Figure 5A and data not shown).

The observation that M25 binding abrogates the in vivo activity of IL-7-Fc may reflect the capacity of M25 to neutralize IL-7 signaling.³⁰  To confirm that M25 interferes with IL-7 signaling in IL-7/M25, we compared CD8⁺ T-cell proliferation in hosts injected with IL-7/M25 formulated at higher molar ratios of M25. Although IL-7/M25 complexes mixed at the typical 2:1 molar ratio (ie, 1:5 mass ratio) of IL-7 to M25 are potent in vivo, increasing the molarity of M25 resulted in a dose-dependent decrease in donor CD8⁺ T-cell proliferation (supplemental Figure 2A). Similarly, the survival-promoting function of IL-7 in vitro was attenuated by the addition of M25 to the culture media (supplemental Figure 2B). These data confirm that M25 neutralizes IL-7 activity in vivo and in vitro and are an indication that binding to M25 precludes IL-7 from interacting with IL-7R.

M25 also appears to be unique among anti–IL-7 mAbs in its ability to both abrogate IL-7R interaction at excess doses and to enhance IL-7 potency at low doses.¹²  For instance, the nonneutralizing anti–mIL-7 mAb MAB407 was ineffective at enhancing the in vivo mitogenicity of IL-7 in the form of rmIL-7/MAB407 (supplemental Figure 2C). We also tested the neutralizing anti–hIL-7 mAb MAB207, although we found that it was not as effective as M25 at neutralizing IL-7 in vitro (supplemental Figure 2B). Moreover, unlike M25, MAB207 did not enhance in vivo IL-7 potency in the form of IL-7/MAB207 (supplemental Figure 2C). MAB207 and M25 also bind to IL-7 noncompetitively, since the pair can be used to detect rhIL-7 in a sandwich enzyme-linked immunosorbent assay (supplemental Figure 2D). Thus, the potential of an mAb to enhance IL-7 potency in vivo appears to be inversely correlated with its ability to block IL-7R access to IL-7 and hence to its fine specificity.

The observation that high concentrations of M25 neutralize IL-7 activity in vivo and in vitro suggests that M25 must eventually dissociate from IL-7 in order to realize the potent activity of IL-7/M25. IL-7-Fc, which is a dimeric form of IL-7, is likely to bind with a greater avidity to M25 than monomeric free IL-7. Thus, we speculated that the impotency of IL-7-Fc/M25 in vivo might be due to a highly stable interaction between the fused IL-7 and M25. To test this idea, we analyzed whether complexes of IL-7-Fc and M25_Fab would display increased potency. Indeed, we found that IL-7-Fc/M25_Fab complexes were able to drive strong proliferation of donor CD8⁺ T cells (Figure 5B). Moreover, IL-7-Fc/M25_Fab complexes were slightly more potent than IL-7-Fc both in terms of the rate of proliferation, with IL-7-Fc/M25_Fab driving a response similar to that of IL-7/M25, and also in terms of donor cell recoveries (Figure 5B-D). These data are consistent with epitope/paratope interactions between IL-7 and M25_Fab providing a contribution to the mechanism of IL-7/M25 that is complementary to and in addition to the effect of Fc/FcRn interactions.

Agonist cytokines/mAbs represent an approach to greatly boosting the efficacy of administered cytokines; however, their mechanism of action remains to be fully explained. The findings presented in this report indicate that IL-7/M25 enhance the in vivo potency of IL-7 by synergistic effects of two separate interactions—one between the Fc of M25 and host FcRn and a second between the M25_Fab and IL-7.

Endogenous IL-7 is produced in multiple tissues and by many cell types.^42-46  Fibroblastic reticular cells residing within the T-cell zones of secondary lymphoid tissue are considered to be the primary source of IL-7 essential for the survival and gradual turnover of mature T cells.⁸  Accordingly, depriving T cells of access to secondary lymphoid tissue impairs their ability to survive or proliferate in response to endogenous homeostatic factors.^32,33  Interestingly, the results in this article suggest that access to the T-cell zones is also a requisite to optimally respond to exogenous IL-7. Although intraperitoneal injection results in a systemic distribution of the inoculum, the effect of the IL-7/M25 was largely targeted to T cells that have direct access to the T-cell zones. Furthermore, T cells retained within the secondary lymphoid organs by FTY720 appeared to receive the majority of the stimulation by IL-7/M25. Nonetheless, these observations do not exclude the possibility that T cells encounter IL-7/M25 outside of lymphoid tissues. These results could also reflect the requirement of contact with self-peptide/major histocompatibility complex ligands for IL-7–driven T-cell proliferation, which would occur much more readily in the T-cell zones than in circulation.¹³ 

Provided that T cells have access to intact secondary lymphoid tissue, the majority of IL-7/M25 activity can be attributed to the interaction between FcRn and Fc. In the absence of either host FcRn or the Fc portion of M25, the potent activity of IL-7/M25 was equally decreased. Furthermore, by directly associating Fc with IL-7, the IL-7-Fc fusion protein displayed substantial host-FcRn–dependent activity. Besides FcRn, no other known Fc-binding protein is likely to be involved in supporting the activity of IL-7/M25. In addition to the data presented from FcγR^−/− hosts, we also observed normal responses to IL-7/M25 treatment in C1q^−/− hosts or in hosts treated with cobra venom factor to deplete complement (data not shown). It should also be noted that the IL-7-Fc fusion protein we analyzed has mutations that preclude binding to FcγR and C1q, supporting the notion that Fc enhances IL-7 potency in vivo independently of these Fc-binding proteins.²⁹  Taken together, these data indicate that the Fc, via the actions of FcRn, accounts for approximately 80% of the in vivo T-cell mitogenic effect of IL-7/M25 treatment.

FcRn captures Fc in acidifying endosomes, thereby preventing its degradation and prolonging Ab lifespan.¹⁹  When injected into rodents, IL-7 has a serum half-life of only ∼2 hours, whereas IL-7/M25 complexes induce T-cell proliferation for roughly 24 hours following injection.^13,40  Moreover, IL-7/M25 induced a much stronger and more prolonged signaling through IL-7R than through free IL-7, as indicated by the pattern of IL-7–mediated downregulation of CD127. IL-7 injected at multiple times during 24 hours displayed much lower potency than one injection of IL-7/M25, implying that it is a much less effective approach than treatment with IL-7/M25. Unexpectedly, we found that multiple injections of IL-7/M25_Fab were sufficient to match the effect of one dose of IL-7/M25, indicating that the fate of exogenous IL-7 is altered when it is bound to M25_Fab. Nevertheless, a 12-fold increase in the cumulative IL-7 dose was required for multiple injections of IL-7/M25_Fab to match the proliferative effect of one injection of IL-7/M25, underscoring the essential role of the Fc–FcRn interaction in maximizing the potency of the complexes.

FcRn is known to perform functions besides prolonging the in vivo lifespan of Fc-associated moieties.¹⁹  Indeed, FcRn is functionally expressed by dendritic cells and macrophages. If IL-7/M25 delivery to T-cell zones by such migratory hematopoietic cells were a major mechanism whereby FcRn augments IL-7/M25 potency, then FcRn expression should be more important in these cell types and correspondingly less important in radioresistant cell types. However, maximal potency of IL-7/M25 treatment required that both cellular compartments were FcRn sufficient, and an equal defect was observed when either radioresistant or BM-derived cell types lacked FcRn expression. Equal contributions from stromal and hematopoietic FcRn is also a hallmark of Ab half-life.³⁹  Therefore, our observations are most consistent with the simple explanation that FcRn–Fc interactions benefit IL-7/M25 by extending their in vivo lifespan.

It is well known that the pharmacokinetics of most proteins can be improved by covalent linkage with an IgG Fc. Although this also applies to IL-7, IL-7-Fc was found to be less effective in driving T-cell proliferation than IL-7/M25, suggesting an additional mechanism that is independent of the FcRn–Fc interaction. This observation parallels that for IL-2/mAb complexes, which are more potent in vivo than IL-2-Fc fusion proteins.²¹  For IL-2, this difference was explained by the finding that anti–IL-2 mAb prevents IL-2 from being sequestered by the high-affinity IL-2R, thereby circumventing an IL-2 sink.^21,47-49  As mentioned above, this mechanism was thought to be unique to IL-2/mAb complexes because of the existence of two forms of IL-2R. However, despite IL-7 being a single-receptor cytokine, our results indicate that a similar mechanism is involved in the activity of IL-7/M25. First, M25 is unique among three tested anti–IL-7 mAbs in its ability to neutralize IL-7 binding to IL-7R at high molar excess and to boost the mitogenicity of IL-7 at lower molar ratios. Second, binding M25_Fab to IL-7-Fc raised its potency to the level of IL-7/M25. Collectively, these related findings provide strong support for the idea that the unique IL-7 neutralizing specificity of M25 accounts for 15% to 20% of the mitogenic activity stimulated by IL-7/M25 and IL-7-Fc/M25_Fab.

How does the neutralizing specificity of M25 augment IL-7 potency? We believe that the neutralizing aspect of M25 sequesters IL-7 to create a depot for the cytokine. IL-7 levels are tightly regulated in vivo by IL-7R⁺ cells.⁷  However, the presence of the neutralizing M25 may shift the IL-7 binding equilibrium away from IL-7R. Thus, although IL-7 or IL-7-Fc would be subject to unfettered consumption by an IL-7 sink, IL-7/M25 and IL-7-Fc/M25_Fab are resistant to binding by IL-7R and may serve to protect IL-7 from receptor-mediated consumption. The exact details of the mechanism are enigmatic at this point. However, a close examination of our data indicates that competition between M25 and IL-7R results in prolonged interaction between IL-7 and IL-7R over time and/or space. Such an effect is suggested by the difference in the proliferation histograms of CD8⁺ T cells from mice treated with IL-7-Fc or IL-7-Fc/M25_Fab. Although both treatments stimulate a similar maximum number of divisions, the response of the donor CD8⁺ T cells as a population is more uniform in hosts treated with IL-7-Fc/M25_Fab than with IL-7-Fc alone. Thus, there is a more even distribution of cytokine signals by IL-7-Fc/M25_Fab than the unprotected IL-7-Fc, which is consumed disproportionately, resulting in a large fraction of cells that did not divide at all. Additionally, the kinetics of CD127 downregulation demonstrate that IL-7/M25 and high-dose IL-7 both reduced receptor expression to half-normal levels. However, the effect of high-dose IL-7 was much more immediate and short-lived than that of IL-7/M25, perhaps indicating that all of the IL-7 bound to M25 is not immediately available to the T cells. Although these data provide only indirect clues to the in vivo fate of exogenous IL-7 when it is not protected by M25, they clearly indicate that a greater amount of the cytokine is registered by CD8⁺ T cells when IL-7 is administered as bound to M25.

In summary, the data presented herein collectively indicate that IL-7/M25 complexes exhibit enhanced mitogenicity largely because of the effector functions of Fc and FcRn, which extend the lifespan of the complexes, and also in part because of IL-7/M25 interactions, which increase cytokine availability by competing with IL-7R. Thus, despite the differences in IL-2R and IL-7R biology, this two-part mechanism is similar to that described for IL-2/mAb.²¹  It is probable that the other reported agonist cytokine/neutralizing Ab complexes also function as cytokine depots.^12,20  Although counterintuitive, the recurring theme of the depot effect supports speculation that the in vivo potential of many ligands in addition to cytokines may be improved through methods that temporarily or partially obstruct ligand–receptor interactions.

The authors thank Gregg Silverman for providing C1q^−/− mice, Hugh Rosen for generously supplying FTY720, and Onur Boyman for critical reading of this manuscript.

This work was supported by a grant from the National Institutes of Health (RO1 AI075164-04) for manuscript number 21900 from The Scripps Research Institute.

Contribution: C.E.M., E.M.M.v.L., and C.D.S. designed the research. C.E.M., E.M.M.v.L., and S.J.I. performed research. Y.-C.S. and D.C.R. contributed vital new tools. C.E.M. and C.D.S. wrote the manuscript.

Conflict-of-interest disclosure: C.D.S. is a shareholder of Nascent Biologics, Inc. The remaining authors declare no competing financial interests.

Correspondence: Charles D. Surh, Academy of Immunology and Microbiology, Institute of Basic Science, Pohang University of Science and Technology, San 31, Hyoja-Dong, Nam-Gu, Pohang, 790-784, Republic of Korea; e-mail: csurh@ibs.re.kr.