Simultaneous targeting of IL-10 and PD-(L)1 pathways with a bispecific anti-PD-L1/IL-10-trap antibody enhances T cell responses Authors: Latika Singh, Natalia Zabavnik, Sambasivan Venkatasubramanian, Sagi Barzilai, Jing Ni, Andreas Evers, Olga Bogatyrova, Melissa G. Derner, Stefan Zielonka, Rinat Zaynagetdinov, Laura Helming Published: 2026 ABSTRACT Immune checkpoint inhibitors (ICIs) demonstrate clinical benefits with durable responses but only in a subset of patients with cancer. Understanding the mechanisms that limit T cell-mediated cytotoxicity of tumor cells and developing new therapies are both needed to overcome such limitations. Interleukin-10 (IL-10) is a cytokine with known pro- and anti-tumor roles. Here, we demonstrate that the nature of IL-10 activity depends on its concentration: very high exogenous IL-10 concentrations induce IFNγ production by T cells, while concentrations similar to those found in the serum of patients with cancer suppress T cell activity. To test if simultaneously trapping IL-10 while blocking PD-L1 could overcome these suppressive effects, which are hypothesized to limit the activity of anti-PD-L1 or drive adaptive resistance, we developed an anti-PD-L1/IL-10-trap antibody comprising a Fab paratope directed against programmed death ligand 1 (PD-L1) fused to a VHH domain targeting IL-10, which acts as an IL-10 “trap.” We found that treatment with anti-PD-L1/IL-10-trap protects T cells from suppressive IL-10 and drives superior activation compared to anti-PD-L1 or IL-10-trap alone. Furthermore, anti-PD-L1/IL-10-trap shifted macrophage polarization toward a more mature and activated phenotype and enhanced CD8+ T cell activation compared with anti-PD-(L)1, IL-10-trap, or their combination. Our data support the immunosuppressive role of IL-10 in tumors and demonstrate that simultaneous targeting of PD-L1 and IL-10 may provide superior T cell responses compared to anti-PD-(L)1 therapies alone or in combination with anti-IL-10 antibodies. Introduction Immune checkpoint inhibitors (ICIs), such as those directed against the programmed cell death 1 (PD-1) pathway, have demonstrated substantial clinical benefits and durable responses in patients with cancer within certain indications such as melanoma,1,2 lung cancer,3,4,5,6 Merkel cell carcinoma,7 urothelial cancer,8,9 and head and neck cancer,10,11 among other tumor types. However, the majority of patients either fail to respond to treatment or eventually develop resistance to these therapies.12 The PD-1 pathway, specifically the interaction between the PD-1 receptor and its ligand programmed death ligand 1 (PD-L1), plays a central role in maintaining immune tolerance by attenuating T cell responses, including T lymphocyte proliferation and cytotoxic effector functions. However, many tumors exploit this pathway to evade immune destruction by expressing PD-L1 and interacting with PD-1, which results in inhibition of cytotoxic T lymphocyte activity and primary resistance to treatments targeting this pathway.13,14 Predictably, increased expression of PD-L1 in tumors has been correlated with unfavorable clinical outcomes across multiple cancer types.13,15,16,17 Multiple mechanisms of resistance to ICIs have been identified, including defects in the antigen presentation machinery, poor T cell infiltration due to insufficient chemokine expression, tumor heterogeneity, and an immunosuppressive tumor microenvironment (TME) characterized by factors such as interleukin (IL)-10, IDO1, and TGF-β.18 IL-10 is a key immunoregulatory cytokine primarily produced by antigen-presenting cells (APCs), T cells including regulatory T cells (Tregs), and epithelial cells.19,20 Among APCs, dendritic cells (DCs) and macrophages are considered major producers of IL-10. In the TME, IL-10 primarily exerts immunosuppressive and pro-tumorigenic effects.21 However, some studies have reported an immunostimulatory role, with PEGylated IL-10 treatment in patients with cancer shown to restore the function of tumor-specific intratumoral CD8+ T cells and suppress tumor growth in murine models.22 While treatment with exogenous IL-10 may be immunostimulatory, it was shown that at concentrations similar to those measured in the serum of patients with cancer, IL-10 was associated with suppression of proinflammatory cytokine secretion, such as IFNγ by T cells, and with reduced cytokine production and antigen presentation by APCs, thereby impairing the induction of robust anti-tumor immune responses.23 In line with this, a high percentage of IL-10-positive tumor-associated macrophages (TAMs) has been associated with disease progression and poor prognosis in patients with non-small cell lung cancer (NSCLC).24 Similarly, in patients with various solid tumors and hematological malignancies, higher serum IL-10 levels have been associated with reduced overall survival.25 Early clinical studies suggest that PD-1 blockade may cause compensatory release of IL-10, which may represent a mechanism by which adaptive resistance develops against anti-PD-(L)1 immunotherapies in various cancer types.26,27 Thus, blocking IL-10 release, or otherwise removing IL-10 from the TME, could potentially prevent or reverse this resistance and enhance the effectiveness of PD-(L)1-targeted treatments. Dual inhibition of IL-10 and PD-L1 may thus be a promising therapeutic strategy, as these pathways independently and complementarily contribute to immunosuppression. Additionally, PD-L1 and IL-10 play distinct roles in tumor-intrinsic and tumor-extrinsic immune evasion, suggesting that their combined blockade may achieve greater anti-tumor activity than either monotherapy. Notably, simultaneous inhibition of IL-10 and PD-L1 has demonstrated superior restoration of antiviral T cell responses compared to targeting either pathway alone.28 IL-10 blockade has also been shown to enhance the effects of PD-1 inhibition on the expansion and function of tumor antigen-specific CD8+ T cells in patients with advanced melanoma.29 Given that tumor cells and tumor-associated suppressive myeloid cells produce and respond to IL-10, localized trapping of IL-10 in the TME, via interaction with anti-PD-L1/IL-10-trap with PD-L1 expressed on tumor cells, may more efficiently reverse IL-10 induced immunosuppression and enhance T cell responses relative to systemic IL-10 blockade. To combat some of the potential pitfalls of combination therapy, bifunctional fusion proteins are designed with synergy in mind, with the goal being a single bifunctional molecule that is more efficacious and has fewer off-target effects than the combination of the 2 individual molecule moieties.30 With this in mind, we developed anti-PD-L1/IL-10-trap (Figure 1), a bifunctional antibody comprising a paratope directed to block the PD-L1 pathway as described previously,7,31 fused to a humanized VHH domain (variable domain of the heavy chain of a heavy-chain-only antibody)32 specifically binding to IL-10 to inhibit its activity by sequestering soluble IL-10 (an IL-10 “trap”). In this work, we investigated whether simultaneous inhibition of IL-10 and PD-L1 pathways with anti-PD-L1/IL-10-trap induces stronger responses in human immune cells than treatment with either anti-PD-(L)1 or IL-10-trap alone or their combination.Figure 1Structural model of the anti-PD-L1/IL-10-trap bifunctional fusion proteinAnti-PD-L1/IL-10-trap is a bispecific antibody consisting of an IL-10-specific humanized VHH paratope and a fully human programmed death ligand 1 (PD-L1) IgG1 antibody. The VHH is fused via a linker to the N terminus of the lambda light chain of the PD-L1-directed IgG. The LALA mutations (L234A and L235A) were introduced to reduce Fc-mediated effector functions. Structural model of the anti-PD-L1/IL-10-trap bifunctional fusion protein Anti-PD-L1/IL-10-trap is a bispecific antibody consisting of an IL-10-specific humanized VHH paratope and a fully human programmed death ligand 1 (PD-L1) IgG1 antibody. The VHH is fused via a linker to the N terminus of the lambda light chain of the PD-L1-directed IgG. The LALA mutations (L234A and L235A) were introduced to reduce Fc-mediated effector functions. Results Generation of anti-PD-L1/IL-10-trap bifunctional antibody Anti-PD-L1/IL-10-trap is a bifunctional antibody, comprising a humanized anti-IL-10 VHH fused to the N terminus of the lambda light chain of an anti-PD-L1 antibody, separated by a flexible linker (Figure 1). Targeting PD-L1 expressed on suppressive myeloid cells and tumor cells, which also produce and respond to IL-10, allows for the localized sequestration of IL-10 in the TME. This approach also avoids the potential challenges of using a bispecific antibody to simultaneously target 2 different cellular receptors (PD-L1 and IL-10R), which could have different receptor densities and cellular locations. To construct the anti-PD-L1/IL-10-trap antibody, an IL-10-specific VHH was generated via camelid immunization combined with yeast surface display, as previously described by our group.33,34,35 We sought to engineer the bifunctional anti-PD-L1/IL-10-trap antibody based on IL-10-targeting single-domain antibodies, since these VHHs offer multiple reformatting options, including, most importantly, the ease of combination with existing antibodies such as an anti-PD-L1 antibody.36,37 The resulting VHH was subsequently humanized, as described elsewhere.32 The humanized IL-10 VHH was fused N-terminally via a flexible linker to the lambda light chain of the PD-L1-targeting antibody harboring LALA mutations (L234A and L235A) for the construction of the bifunctional trap protein, which enables bivalent targeting of both IL-10 and PD-L1 (Figure 1). The anti-PD-L1/IL-10-trap antibody was transiently produced in ExpiCHO cells and purified via protein A, followed by preparative size-exclusion chromatography (SEC). Ultimately, production yields (approximately 260 mg/L), analytical SEC after two-step purification (purity = 95.4%; Figure S1A), and thermal stability (Tonset = 60.4°C; Figure S1B) indicated favorable biophysical attributes of the anti-PD-L1/IL-10-trap antibody. Anti-PD-L1/IL-10-trap binds to PD-L1 and inhibits IL-10 signaling The PD-L1-binding arm of the anti-PD-L1/IL-10-trap antibody comes from a functionally active PD-L1 blocking antibody7 and retains full binding capacity within the bifunctional fusion protein, as anti-PD-L1/IL-10-trap bound to ectopically expressed PD-L1 on HEK293 cells with a profile comparable to that of an Fc-silenced version of the parent anti-PD-L1 antibody from which the PD-L1 binding arm was derived (half maximal effective concentration [EC50] = 4.3 nM, averaged from 2 experiments, and 2.4 nM, respectively). In addition, we assessed the binding of anti-PD-L1/IL-10-trap to PD-L1 expressed on the cell surface of human CD14+ myeloid cells from mixed lymphocyte reaction (MLR) co-cultures. Anti-PD-L1/IL-10-trap bound with high affinity to CD14+ myeloid cells (EC50 = 3.5 nM; Figures S2A and S2B; Table S2) compared to isotype control. Anti-PD-L1/IL-10-trap binding to human IL-10 was in the low nanomolar range (KD = 1.2 nM), as determined by biolayer interferometry (BLI) using the Octet system (Figure S2C), and anti-PD-L1/IL-10-trap demonstrated potent inhibition of IL-10 signaling using Hek-Blue IL-10 reporter cells (IC50 = 14.8 pM; Figure S2D). Dose-dependent effects of rhIL-10 on human CD8+T cells Previous studies have shown that IL-10 plays a dual role within the TME, both in immunostimulation and immunosuppression, with its effects dependent upon specific cell types and contexts.38 To investigate the impact of recombinant human IL-10 (rhIL-10) on T cells, human CD8+ T cells isolated from healthy donors were stimulated using plate-bound anti-CD3 and soluble anti-CD28 antibodies in the presence of increasing concentrations of rhIL-10 for 5 days. The effects of rhIL-10 were assessed by measuring IFNγ and granzyme B levels in the supernatants. Human CD8+ T cells from healthy donors displayed opposing responses to rhIL-10 based on its concentration. Treatment with concentrations of exogenous rhIL-10 similar to IL-10 levels reported in the serum of patients with cancer (≤0.02 ng/mL)25 significantly reduced CD8+ T cell activation as indicated by decreased granzyme B expression and IFNγ production compared to cells not treated with rhIL-10 (Figures 2A and 2B). In contrast, high concentrations of rhIL-10 (>20 ng/mL) promoted CD8+ T cell activation, with a trend toward increased granzyme B and IFNγ levels, although these changes did not reach statistical significance relative to untreated cells (Figures 2A and 2B).Figure 2Dose-dependent effects of rhIL-10 on human CD8+ T cell activityActivated CD8+ T cells were cultured with varying concentrations of rhIL-10. The concentrations of (A) granzyme B and (B) IFNγ were determined in supernatants collected from CD8+ T cell cultures. Data are presented as average ± SD from 2 or 3 technical replicates each (n = 4 separate donors), and one-way ANOVA with Tukey’s multiple-comparisons test was used to calculate p values (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and not statistically significant p > 0.05). Each color-coded dot represents an individual donor. The red box indicates a statistically significant change with treatment compared to the control. Dose-dependent effects of rhIL-10 on human CD8+ T cell activity Activated CD8+ T cells were cultured with varying concentrations of rhIL-10. The concentrations of (A) granzyme B and (B) IFNγ were determined in supernatants collected from CD8+ T cell cultures. Data are presented as average ± SD from 2 or 3 technical replicates each (n = 4 separate donors), and one-way ANOVA with Tukey’s multiple-comparisons test was used to calculate p values (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and not statistically significant p > 0.05). Each color-coded dot represents an individual donor. The red box indicates a statistically significant change with treatment compared to the control. Next, we investigated whether anti-PD-L1/IL-10-trap could reverse T cell suppression induced by a low dose of rhIL-10. Human CD8+ T cells from healthy donors were cultured and stimulated with plate-bound anti-CD3 and soluble anti-CD28 in the presence of low-dose rhIL-10. The cells were then treated with either an isotype control, anti-PD-L1 alone, IL-10-trap alone, anti-PD-L1/IL-10-trap, or a combination of anti-PD-L1 and IL-10-trap. Analysis of supernatants from cultures showed that treatment with anti-PD-L1/IL-10-trap significantly reversed CD8+ T cell suppression induced by a low dose of rhIL-10, as indicated by increased IFNγ levels compared to cultures treated with isotype control, anti-PD-L1, or IL-10-trap alone. The combination of anti-PD-L1 and IL-10-trap showed similar effectiveness to anti-PD-L1/IL-10-trap in this assay (Figures 3A and 3B).Figure 3Anti-PD-L1/IL-10-trap reverses rhIL-10-induced suppression of CD8+T cells(A) Representative histogram showing secreted IFNγ levels in activated CD8+T cells cultured in the presence of rhIL-10 and treated with isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap. The results are presented as average ± SD from 2–6 technical replicates. (B) Representative dot plot showing the fold change in IFNγ levels in CD8+ T cells relative to the isotype control (set to 1). Data are representative of 5 separate donors. A one-tailed paired t test was used to calculate p values (∗∗p < 0.01; ns, not statistically significant). Each symbol type represents an individual experiment. Anti-PD-L1/IL-10-trap reverses rhIL-10-induced suppression of CD8+T cells (A) Representative histogram showing secreted IFNγ levels in activated CD8+T cells cultured in the presence of rhIL-10 and treated with isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap. The results are presented as average ± SD from 2–6 technical replicates. (B) Representative dot plot showing the fold change in IFNγ levels in CD8+ T cells relative to the isotype control (set to 1). Data are representative of 5 separate donors. A one-tailed paired t test was used to calculate p values (∗∗p < 0.01; ns, not statistically significant). Each symbol type represents an individual experiment. Anti-PD-L1/IL-10-trap enhances activation of stimulated macrophages Several mechanisms have been proposed through which IL-10 can inhibit T cell-mediated anti-tumor immune responses within the TME, including suppression of IFNγ-dependent activation of macrophages and APCs, leading to reduced expression of MHC class II and CD86, as well as inhibition of the CD28 costimulatory pathway.21,39 To explore the impact of anti-PD-L1/IL-10-trap on macrophage activation during differentiation, human monocytes isolated from healthy donors were cultured in the presence of rhM-CSF for 7 days, followed by an additional 48 h with lipopolysaccharide (LPS) and rhIFNγ, along with isotype control, anti-PD-L1, IL-10-trap, or anti-PD-L1/IL-10-trap. The resulting myeloid cell population was subsequently assessed for expression of cell-surface markers and cytokine secretion. CD14+CD11c+M1 macrophages differentiated in the presence of anti-PD-L1/IL-10-trap displayed significant dose-dependent increases in CD83, CD86, and MHC class II surface expression on M1 macrophages compared to those treated with anti-PD-L1 alone. A similar trend was observed for all treatments that included blockade of IL-10, the IL-10-trap alone, or the combination of anti-PD-L1 and IL-10-trap at lower concentrations (Figures 4A–4F), suggesting that IL-10 blockade is likely responsible for the observed increase in cell-surface markers. In a separate experiment, we also examined the expression of CD163, CD86, and MHC class II on M1 macrophages. We observed that treatment with the anti-PD-L1/IL-10-trap antibody decreased CD163 expression while increasing CD86 and MHC class II expression on macrophages (Table S3), indicating higher expression of M1-related markers.Figure 4IL-10 blockade leads to macrophage activation(A, C, and E) Monocytes differentiated into macrophages in the presence of rhM-CSF were stimulated with LPS and rhIFNγ in the presence of varying concentrations of anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap. Representative graphs showing the mean fluorescence intensity (MFI) of cell-surface expression of CD86, CD83, and MHC class II on macrophages determined using flow cytometry. (B, D, and F) Representative data from 1 donor are shown. At least 2 independent experiments were performed with 2 donors. Monocytes were differentiated into macrophages, as described in (A), (C), and (E). The cells were treated with 115 pM of anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1+IL-10-trap, and dot plots show the MFI of cell-surface expression of CD86, CD83, and MHC class II on macrophages. (G, I, and K) The cells were differentiated into macrophages and treated as described in (A), (C), and (E). The concentrations of TNF-α, IL-12p70, and GM-CSF were measured in supernatants collected from macrophage cultures. (H, J, and L) Representative data from 1 donor are shown. At least 2 independent experiments were performed with 2 donors. The cells were treated with 694 pM of anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap, and proinflammatory cytokine levels in the supernatants were determined. One-way ANOVA followed by Tukey’s multiple-comparisons test was used to calculate p values, where ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, and ns, not statistically significant. IL-10 blockade leads to macrophage activation (A, C, and E) Monocytes differentiated into macrophages in the presence of rhM-CSF were stimulated with LPS and rhIFNγ in the presence of varying concentrations of anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap. Representative graphs showing the mean fluorescence intensity (MFI) of cell-surface expression of CD86, CD83, and MHC class II on macrophages determined using flow cytometry. (B, D, and F) Representative data from 1 donor are shown. At least 2 independent experiments were performed with 2 donors. Monocytes were differentiated into macrophages, as described in (A), (C), and (E). The cells were treated with 115 pM of anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1+IL-10-trap, and dot plots show the MFI of cell-surface expression of CD86, CD83, and MHC class II on macrophages. (G, I, and K) The cells were differentiated into macrophages and treated as described in (A), (C), and (E). The concentrations of TNF-α, IL-12p70, and GM-CSF were measured in supernatants collected from macrophage cultures. (H, J, and L) Representative data from 1 donor are shown. At least 2 independent experiments were performed with 2 donors. The cells were treated with 694 pM of anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap, and proinflammatory cytokine levels in the supernatants were determined. One-way ANOVA followed by Tukey’s multiple-comparisons test was used to calculate p values, where ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, and ns, not statistically significant. We also analyzed proinflammatory cytokine secretion in the cell culture medium of macrophages. Treatment with anti-PD-L1/IL-10-trap resulted in a significant, dose-dependent increase in the levels of TNF-α, IL-12p70, and GM-CSF, compared to IL-10-trap, anti-PD-L1, or the combination of anti-PD-L1 and IL-10-trap (Figures 4G–4L). We found that anti-PD-L1/IL-10-trap modulated the macrophage phenotype, promoting a more mature and activated phenotype, as determined by MHC class II, CD86, and CD83 cell-surface expression and increased levels of secreted TNF-α, IL-12p70, and GM-CSF. Anti-PD-L1/IL-10-trap enhances tumor cell killing by activated human T cells in vitro Next, we assessed whether treatment of human T cells with anti-PD-L1/IL-10-trap antibody augments T cell-mediated killing of PD-L1-positive tumor cells (MDA-MB-231). To this end, human peripheral blood mononuclear cell (PBMC)-derived T cells were co-cultured with MDA-MB-231 cells in the presence of the anti-PD-L1/IL-10-trap antibody, exogenous rhIL-10, and soluble anti-CD3/anti-CD28 antibody. Anti-PD-L1/IL-10-trap treatment enhanced activated T cell-mediated tumor cell killing compared to every single agent alone and showed a trend toward improved in vitro cytotoxicity of tumor cells relative to the combination of the 2 agents (Figure S3). Simultaneous blockade of IL-10 and PD-L1 with anti-PD-L1/IL-10-trap leads to enhanced T cell activation in MLR assay Previous studies have demonstrated that myeloid cells can suppress effector T cell responses through various mechanisms, including IL-10 secretion, PD-L1 expression, and cytokine-driven Treg conversion.39 We hypothesized that macrophage-derived IL-10 and PD-L1 may contribute to the suppression of T cell activation. Additionally, the induction of IL-10 by anti-PD-L1 may represent one of several potential mechanisms contributing to resistance to anti-PD-L1 immunotherapy.30,40 To assess the association between IL-10 expression and anti-PD-(L)1 immunotherapy, we examined PD-1 and PD-L1 expression on immune cells and measured IL-10 levels in the supernatants of a human two-way allogeneic MLR assay. Human PBMCs from 1 donor (responder) were co-cultured with PBMCs from another donor (stimulator) at a 1:1 ratio for 5 days. Both PD-1 and PD-L1 expression were elevated on CD8+ T cells on days 3 and 5 compared to the 4-h time point, and PD-L1 expression was also upregulated on CD14+ myeloid cells by days 3 and 5 compared to the 4-h time point (Figures 5A and 5B). Additionally, IL-10 levels in the MLR co-culture supernatants were increased on days 3 and 5 compared to day 2 in 3 donor pairs (Figure 5C). Anti-PD-L1 treatment resulted in a statistically significant increase in IL-10 production in MLR co-cultures compared to the isotype control in all 3 donor pairs. Treatment with anti-PD-L1/IL-10-trap or IL-10-trap alone resulted in a statistically significant decrease in free IL-10 levels relative to anti-PD-L1 alone. The combination of anti-PD-L1 and IL-10-trap also reduced free IL-10 levels relative to anti-PD-L1 treatment alone, and this effect was not significantly different compared to the anti-PD-L1/IL-10 trap treatment (Figure 5D). When MLR co-cultures were treated with anti-PD-L1/IL-10-trap, a dose-dependent increase in T cell activation (measured by IFNγ production) was observed, which was more effective than treatment with anti-PD-L1 or IL-10-trap alone (Figures 6A–6D) or with pembrolizumab (Figures 6E–6H). Our findings indicate that enhanced PD-L1 expression, along with increased IL-10 levels, may contribute to T cell suppression in MLR co-cultures and that dual blockade of PD-L1 and IL-10 enhanced the effector T cell response.Figure 5Anti-PD-L1 treatment increased IL-10 levels in MLR co-cultures(A and B) Representative flow cytometry dot plots showing cell-surface expression of PD-1 and PD-L1 on CD8+ T cells, as well as PD-L1 expression on CD14+ myeloid cells, measured in MLR co-cultures using flow cytometry. (C) Representative histogram illustrating free IL-10 levels in MLR co-cultures. The data are representative of 3 PBMC donor pairs. The results are expressed as average ± SD from 3 technical replicates in each experiment. (D) Representative dot plot showing the fold change in IL-10 levels in 3 different PBMC donor pairs relative to the isotype control (set to 1). p values were calculated using a one-way ANOVA followed by Tukey’s multiple-comparisons test, where ∗p < 0.05.Figure 6Anti-PD-L1/IL-10-trap treatment induces strong activation of human T cells in a two-way MLR assay(A–C) Representative line graphs showing IFNγ levels in MLR co-cultures treated with varying concentrations of isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap. Data are presented as average ± SD from 2 or 3 technical replicates each (n = 3 different PBMC donor pairs). (D) Results from 3 assays with 3 different donor pairs are plotted as fold changes over the isotype control (set to 1). The cells were treated with 2,200 pM of the isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap, and IFNγ concentration was determined in supernatants. In (E)–(G), PBMC co-cultures were set up as described in (A)–(C) in the presence of varying concentrations of isotype control, anti-PD-L1, pembrolizumab, IL-10-trap, or anti-PD-L1/IL-10-trap, and IFNγ concentration was determined in supernatants. Data are presented as average ± SD from 2 or 3 technical replicates each (n = 3 different PBMC donor pairs). In (H), results from 3 assays with 3 different donor pairs are plotted as fold change over isotype control (set to 1). The cells were treated with 57,000 pM of the isotype control, anti-PD-L1, pembrolizumab, IL-10-trap, or anti-PD-L1/IL-10-trap, and IFNγ concentration was determined in supernatants. p values were calculated using a one-way ANOVA followed by Tukey’s multiple-comparisons test, where ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. Anti-PD-L1 treatment increased IL-10 levels in MLR co-cultures (A and B) Representative flow cytometry dot plots showing cell-surface expression of PD-1 and PD-L1 on CD8+ T cells, as well as PD-L1 expression on CD14+ myeloid cells, measured in MLR co-cultures using flow cytometry. (C) Representative histogram illustrating free IL-10 levels in MLR co-cultures. The data are representative of 3 PBMC donor pairs. The results are expressed as average ± SD from 3 technical replicates in each experiment. (D) Representative dot plot showing the fold change in IL-10 levels in 3 different PBMC donor pairs relative to the isotype control (set to 1). p values were calculated using a one-way ANOVA followed by Tukey’s multiple-comparisons test, where ∗p < 0.05. Anti-PD-L1/IL-10-trap treatment induces strong activation of human T cells in a two-way MLR assay (A–C) Representative line graphs showing IFNγ levels in MLR co-cultures treated with varying concentrations of isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap. Data are presented as average ± SD from 2 or 3 technical replicates each (n = 3 different PBMC donor pairs). (D) Results from 3 assays with 3 different donor pairs are plotted as fold changes over the isotype control (set to 1). The cells were treated with 2,200 pM of the isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap, and IFNγ concentration was determined in supernatants. In (E)–(G), PBMC co-cultures were set up as described in (A)–(C) in the presence of varying concentrations of isotype control, anti-PD-L1, pembrolizumab, IL-10-trap, or anti-PD-L1/IL-10-trap, and IFNγ concentration was determined in supernatants. Data are presented as average ± SD from 2 or 3 technical replicates each (n = 3 different PBMC donor pairs). In (H), results from 3 assays with 3 different donor pairs are plotted as fold change over isotype control (set to 1). The cells were treated with 57,000 pM of the isotype control, anti-PD-L1, pembrolizumab, IL-10-trap, or anti-PD-L1/IL-10-trap, and IFNγ concentration was determined in supernatants. p values were calculated using a one-way ANOVA followed by Tukey’s multiple-comparisons test, where ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. In summary, we demonstrated that anti-PD-L1 treatment leads to increased release of IL-10, which may represent one of the mechanisms underlying resistance to anti-PD-L1 immunotherapy. Furthermore, our data support that anti-PD-L1/IL-10-trap may overcome the potential PD(L)-1 blockade/IL-10-associated resistance mechanism. Anti-PD-L1/IL-10-trap reduces LPS-induced IL-10 levels and immunosuppression in human whole blood To test whether anti-PD-L1/IL-10-trap can mitigate the suppressive effects of IL-10 in a human whole blood assay, we treated whole blood from healthy human donors with LPS, which is known to induce IL-10 production,41 followed by treatment with an isotype control, anti-PD-L1/IL-10-trap, anti-PD-L1, IL-10-trap, their combination, or pembrolizumab. As expected, LPS treatment increased IL-10 levels in the blood, which were significantly reduced by anti-PD-L1/IL-10-trap compared to the isotype control (Figure 7A). Additionally, anti-PD-L1/IL-10-trap treatment led to a significant increase in cytokine production, including IL-12p70, IFNγ, and TNF-α in LPS-treated blood, outperforming single-agent controls, their combinations, or pembrolizumab (Figures 7E–7G and 7B–7D). Importantly, anti-PD-L1/IL-10-trap did not induce cytokine secretion in whole blood in the absence of LPS, suggesting reduced risk for systemic intolerability (Figures 7A–7D).Figure 7Anti-PD-L1/IL-10-trap in combination with LPS, but not alone, induces cytokine production in human whole blood ex vivo(A–D) Representative histograms showing IL-10, IL-12p70, IFNγ, and TNF-α levels in whole blood from healthy individuals treated with isotype or anti-PD-L1/IL-10-trap (50 nM) in the presence or absence of LPS. (E–G) Representative histograms illustrating IL-12p70, IFNγ, and TNF-α levels in whole blood from healthy individuals treated with isotype, anti-PD-L1, pembrolizumab, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap at 1.2 nM in the presence of LPS. Data are presented as average ± SD from 3 different donors. p values were calculated using a one-way ANOVA followed by Tukey’s multiple-comparisons test, where ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, and ns, not statistically significant. Anti-PD-L1/IL-10-trap in combination with LPS, but not alone, induces cytokine production in human whole blood ex vivo (A–D) Representative histograms showing IL-10, IL-12p70, IFNγ, and TNF-α levels in whole blood from healthy individuals treated with isotype or anti-PD-L1/IL-10-trap (50 nM) in the presence or absence of LPS. (E–G) Representative histograms illustrating IL-12p70, IFNγ, and TNF-α levels in whole blood from healthy individuals treated with isotype, anti-PD-L1, pembrolizumab, IL-10-trap, anti-PD-L1/IL-10-trap, or anti-PD-L1 + IL-10-trap at 1.2 nM in the presence of LPS. Data are presented as average ± SD from 3 different donors. p values were calculated using a one-way ANOVA followed by Tukey’s multiple-comparisons test, where ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, and ns, not statistically significant. Discussion According to the current literature, IL-10 has been shown to play opposing roles in cancer immunity, with studies suggesting it exerts either immunostimulatory or immunosuppressive effects. For example, pegilodecakin, a PEGylated version of IL-10, showed promising immune activation in patients with cancer. However, it did not demonstrate a significant improvement in overall survival for cancers such as renal cell carcinoma and metastatic NSCLC.42 Moreover, several groups reported that IL-10 can facilitate immune evasion and tumor progression by hindering the maturation of APCs, inhibiting cytotoxic T lymphocytes, and suppressing anti-tumor cytokine secretion, thereby weakening their tumor-killing capacity.19,43,44 In line with this, a correlation between IL-10 levels and poor prognosis, higher tumor burden, and reduced immune infiltration was reported for some cancers, including breast cancer45 and Hodgkin’s lymphoma.46 Our data support this dual role of IL-10. We found that exogenous rhIL-10 concentrations comparable to serum concentrations observed in patients with cancer25 suppressed IFNγ production and granzyme B expression in T cells compared with control cells cultured without rhIL-10. In contrast, exogenous rhIL-10 at levels that are approximately 2,000-fold higher than those in serum of healthy individuals, similar to levels seen in the serum of patients with cancer who were treated with pegilodecakin,47 enhanced T cell activation, as measured by granzyme B expression, compared to T cells cultured in the absence of exogenous rhIL-10. These findings align with previous research demonstrating that high levels of exogenous IL-10 regulate cytotoxic enzymes and the Th1 cytokine IFNγ in CD8+ T cells.22,48 The pleiotropic effects of IL-10 may be attributable, at least in part, to the heterogeneous expression of its receptor across immune cell types. Limited response rates and the development of resistance are major challenges for anti-PD-(L)1 immunotherapies.49,50,51,52 The limited activity of anti-PD-(L)1 agents as monotherapies underscores the need for alternative therapeutic strategies, with simultaneous targeting of additional immunosuppressive pathways emerging as a promising approach to enhance immune checkpoint blockade. Early clinical studies suggest that IL-10 may contribute to the development of resistance to anti-PD-(L)1 therapies across multiple cancer types.27,29 We found that treatment of human PBMCs in an MLR co-culture with anti-PD-L1 increased IL-10 to levels similar to those reported in the serum of patients with cancer,25 concentrations that have been shown to promote immunosuppression. In line with that, simultaneous blocking of both IL-10 and PD-L1 using anti-PD-L1/IL-10-trap led to a significant increase in IFNγ secretion compared with treatment with single agents alone, which serves as a functional readout of T cell activation. IFNγ secretion following anti-PD-L1/IL-10-trap was not significantly different compared to a combination of the single agents. The immunosuppressive activity of IL-10 is not limited to a direct impact on T cells; it can also modulate properties of myeloid cells by downregulating MHC class II and/or costimulatory molecules (CD80/CD86) on the surface of DCs and macrophages, thereby impairing their capacity to activate T cells.19,44 In alignment with this, anti-PD-L1/IL-10-trap led to a more robust polarization of human macrophages toward the anti-tumor M1 phenotype compared to treatment with the anti-PD-L1 antibody alone or its combination with the IL-10-trap antibody, suggesting a dual modulatory role for anti-PD-L1/IL-10-trap on T cell- and myeloid cell-mediated activity. In the T cell activation assay, the combination of anti-PD-L1 and IL-10-trap demonstrated comparable effectiveness as measured by IFNγ production. However, in the macrophage polarization assay, the anti-PD-L1/IL-10-trap antibody demonstrated enhanced in vitro potency compared to the combination treatment, except for MHC class II expression, where the combination showed effectiveness comparable to that of anti-PD-L1/IL-10-trap. Macrophages were found to express PD-L1 and produce IL-10, which functions in an autocrine manner on APCs, and colocalizing the IL-10-trap to the cell membrane via PD-L1 binding likely enhances local concentrations and allows for the concurrent inhibition of pathways within the same cells, leading to a more robust and targeted immune activation. Additionally, colocalization of the IL-10-trap with APCs is thought to effectively compete with IL-10R for IL-10 binding, thereby more efficiently relieving the immunosuppressive effects of IL-10. Recent studies53 have reported differential expression of IL-10Rβ on monocytes and T cells, which may also account for the differing sensitivity of the anti-PD-L1/IL-10-trap observed in macrophages compared to T cells. Together, our findings suggest that levels of IL-10 similar to those seen in the serum of patients with cancer may limit activity of anti-PD-L1 targeting agents, perhaps by promoting adaptive resistance, and that simultaneous blocking of both IL-10 and PD-L1 may better activate anti-tumor T cell responses in patients with cancer. From this study, the proposed mechanism of action of anti-PD-L1/IL-10 trap involves PD-(L)1-induced suppression of T cell immune responses along with IL-10-mediated signaling in APCs, which inhibits antigen presentation and impairs their ability to activate T cells. In addition, IL-10 can directly affect T cells, inhibiting their anti-tumor immune functions. Blockade of PD-(L)1 and IL-10 signaling by a bifunctional anti-PD-L1/IL-10-trap antibody could potentially protect T cell responses by preventing suppression mediated by PD-L1 and IL-10. Safety aspects associated with anti-IL-10 therapies in the treatment of cancer are not well investigated. Anti-IL-10 antibody treatment was shown to be well tolerated in 6 patients with steroid-dependent systemic lupus erythematosus.54 A phase 1 clinical study (ClinicalTrials.gov: NCT02731742) testing anti-IL-10 monoclonal antibody MK-1966 in combination with in situ vaccination of a TLR9 agonist SD-101 in patients with advanced solid malignancies was halted due to the lack of efficacy55; however, no safety-related data have been reported. In our studies with human whole blood from healthy donors, treatment with anti-PD-L1/IL-10-trap alone did not induce production of cytokines associated with safety risks. In conclusion, we showed that simultaneous blocking of PD-L1 and IL-10 with anti-PD-L1/IL-10-trap leads to activation of myeloid cells, resulting in increased expression of costimulatory molecules and secretion of proinflammatory cytokines, which is enhanced relative to treatment with either arm alone or their combination. Furthermore, simultaneous blockade of both targets with anti-PD-L1/IL-10-trap induces T cell activation, as evidenced by increased IFNγ secretion and tumor cell killing compared with single agents alone, and exhibits a trend toward improved cytotoxicity compared with their combination. Thus, targeting both pathways with a bifunctional molecule may offer strategic advantages over traditional approaches that rely on combining separate single agents. Materials and methods PBMC isolation PBMCs were isolated from healthy human peripheral blood Leukopaks obtained from Research Blood Components (Watertown, MA, USA) using the Custom Leukopak PBMC isolation kit (Miltenyi Biotech, Auburn, CA, USA) via magnetic labeling and depletion of erythrocytes and granulocytes according to the manufacturer’s instructions. Generation of antibodies Anti-PD-L1/IL-10-trap is a bifunctional antibody composed of humanized IL-10-targeting VHH binders (IL-10-traps), which were fused to the N terminus of each of the lambda light chains of an anti-human PD-L1 IgG1 antibody via a linker. The anti-PD-L1 paratope of anti-PD-L1/IL-10-trap is based on avelumab (MSB0010718C) and is described in US patent no. US 9,624,298. LALA amino acid exchanges (L234A and L235A) were introduced into the Fc region to abolish Fc receptor binding. The isotype control, a human anti-HEL-IgG1-LALA, which is a modified version of a kappa IgG1 antibody that targets the hen egg lysozyme (HEL), and the IL-10-trap-only control, harboring the same IL-10 binders, were generated internally at Merck Healthcare KGaA (Darmstadt, Germany). Pembrolizumab is a humanized anti-PD-1 antibody purchased from Asaman (MFR: MERCK & CO., Rahway, NJ, USA) and was used as the anti-PD-1 control antibody. Anti-PD-L1/IL-10-trap was produced in ExpiCHO cells in a volume of 500 mL using pTT5 expression vectors56 according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). 10 days post-transfection, protein-containing supernatants were harvested by centrifugation and purified by protein A chromatography using HiTrap MabSelect SuRe resin (Cytiva, Wilmington, DE, USA) followed by preparative SEC (HiLoad Superdex 200 prep grade, Cytiva). Anti-PD-L1/IL-10-trap was concentrated using an Amicon Centrifugal Filter device (Millipore Sigma, Burlington, MA, USA). Sterile filtration was performed with Lyzer-CL GV 0.22 μm centrifugal devices (Millipore Sigma). Protein concentrations were determined using a Nanodrop ND-1000 (Peqlab, Fareham, UK). Protein purities were determined by analytical SEC on a TSKgel UP-SW3000 column (2 μm, 4.6 × 300 mm, Tosoh Bioscience) using an Agilent HPLC 1260 Infinity system. Assessment of the binding of anti-PD-L1/IL-10-trap to PD-L1 and IL-10 HEK293 cells ectopically overexpressing PD-L1 were incubated with serial dilutions of anti-PD-L1/IL-10-trap, anti-PD-L1, or IL-10-trap control antibodies, followed by a fluorophore-conjugated anti-human IgG secondary antibody. Binding signals were calculated using flow cytometry as mean fluorescence intensity (MFI). The affinity of anti-PD-L1/IL-10-trap for human IL-10 was determined via BLI using the Octet RED96 system (ForteBio, Pall Life Science, Menlo Park, CA, USA) at 25°C and 1,000 rpm agitation. Data were fitted and analyzed with ForteBio data analysis software 8.0 using a 1:1 binding model after Savitzky-Golay filtering. Anti-PD-L1/IL-10-trap was loaded on anti-hIgG Fc capture (AHC) biosensors for 180 s, followed by 60 s sensor rinsing in kinetics buffer (KB). Association to different concentrations of rhIL-10 in KB was measured for 300 s, followed by dissociation in KB for 600 s. Measurement of the soluble IL-10 sequestering effect of anti-PD-L1/IL-10-trap HEK-Blue IL-10 cells (cat# hkb-il10, InvivoGen, San Diego, CA, USA) were treated with increasing concentrations of anti-PD-L1/IL-10-trap (1.5–10,000 pM) in the presence of rhIL-10 (60 pM) (Peprotech, Cranbury, NJ, USA). rhIL-10-induced-STAT3 activation was measured using secreted embryonic alkaline phosphatase (SEAP) reporter activity in supernatants 24 h post-treatment (QUANTI-Blue, InvivoGen). Assessment of the effect of rhIL-10 on human CD8+ T cells CD8+ T cells were isolated from healthy human PBMCs using the EasySep Human CD8+ T cell Enrichment Kit (STEMCELL Technologies, Vancouver, BC, Canada) via negative selection, following the manufacturer’s instructions. Purified CD8+ T cells (1 × 105 cells per well) were cultured in cell culture medium in 96-well flat-bottom plates coated with plate-bound anti-CD3 (BioLegend, San Diego, CA, USA) and soluble anti-CD28 antibodies (BioLegend), in the presence of increasing concentrations (0.02–2,000 ng/mL) of rhIL-10. IFNγ and granzyme B levels in culture supernatants were determined using Mesoscale Discovery kits (Rockville, Maryland, USA) following the manufacturer’s protocol. In vitro activity of anti-PD-L1/IL-10-trap in reversing T cell suppression To evaluate the effectiveness of anti-PD-L1/IL-10-trap in reversing T cell suppression caused by rhIL-10, purified CD8+ T cells (1 × 105 cells per well) were cultured in 96-well flat-bottom plates coated with anti-CD3 and soluble CD28 antibodies in the presence of rhIL-10 (2.5 ng/mL) and isotype control (10 μg/mL), anti-PD-L1 (10 μg/mL), IL-10-trap (10 μg/mL), anti-PD-L1/IL-10-trap (10 μg/mL), or a combination of anti-PD-L1 (10 μg/mL) and IL-10-trap (10 μg/mL), then incubated at 37°C for 5 days. The levels of IFNγ in the culture supernatants were determined using Meso Scale Discovery kits according to the manufacturer’s protocol (Meso Scale Discovery). Modulation of proinflammatory cytokine secretion and costimulatory molecules in human macrophages Human monocytes isolated from healthy donors were cultured (1 × 106 cells per well in a 24-well plate) in the presence of 50 ng/mL rhM-CSF (Peprotech) for 7 days, followed by an additional 48 h with 100 ng/mL LPS (Sigma) and 50 ng/mL rhIFNγ (Peprotech), along with increasing concentrations (3.2–25,000 pM) of isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or a combination of anti-PD-L1 and IL-10-trap. The supernatants were collected, and proinflammatory cytokine levels were measured using ELISA according to the manufacturer’s instructions (R&D Systems). Human macrophages (2 × 105) from cultures were first stained with Fixable Viability Stain 510 (BD Biosciences, New Jersey, USA) in DPBS, followed by treatment with human Fc receptor blocking solution FcX (BD Biosciences). The cells were then stained with a cocktail of fluorescently conjugated antibodies (0.1 μg each of CD14, CD86, CD163, MHC class II, and CD83) in fluorescence-activated cell sorting (FACS) staining buffer on ice (Table S1). Samples were acquired with a Cytek flow cytometer using FACSDiva software (BD Biosciences), and flow cytometry data were analyzed using FCS Express Flow Cytometry Software (De Novo Software) and GraphPad Prism 10. Binding signals were quantified as MFI for each antibody. Anti-PD-L1/IL-10-trap promotes tumor cell killing by activated T cells MDA-MB-231 breast cancer cells (ATCC, Manassas, Virginia, USA) stably expressing luciferase were cultured in RPMI medium (Gibco, Life Technologies, Grand Island, New York, USA) supplemented with 10% heat-inactivated fetal bovine serum (Sigma, USA), 2 mM L-glutamine, 50 U/mL penicillin, 50 μg/mL streptomycin (both from Gibco, Life Technologies), and puromycin (Gibco, Life Technologies). Cells were maintained at 37°C in a humidified atmosphere containing 5% CO2. MDA-MB-231 luciferase tumor cells were co-cultured with CD3+ T cells in the presence of soluble anti-CD3 antibody (0.01 μg/mL), anti-CD28 antibody (1 μg/mL), and rhIL-10 (2.5 ng/mL). Co-cultures were treated with 57 nM of isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, or a combination of anti-PD-L1 plus IL-10-trap for 72 h. Tumor cell killing was quantified using the ONEGlo luciferase assay (Promega, Madison, WI, USA), and luminescence was measured with a luminometer (EnVision, PerkinElmer, Waltham, MA, USA). Human allogeneic two-way MLR assay Frozen PBMCs from 3 different donors were co-cultured at a 1:1 ratio for 4–5 days and treated with isotype control, anti-PD-L1, IL-10-trap, anti-PD-L1/IL-10-trap, pembrolizumab, or a combination of anti-PD-L1 and IL-10-trap, at increasing concentrations (18.2–57,000 pM). Culture supernatants were harvested on day 4 for ELISA analysis of IFNγ secretion according to the manufacturer’s instructions (Meso Scale Discovery). Measurement of surface expression of PD-1 and PD-L1 on T cells and PD-L1 on myeloid cells in MLR co-cultures The cells (2 × 105) from co-cultures were stained with Fixable Viability Stain ZOMbie NIR (BioLegend) in DPBS, followed by fluorescently conjugated antibodies (0.1 μg each of CD14, CD3, CD8, PD-1, and PD-L1) in FACS staining buffer on ice (Table S1) in the dark. All samples were acquired on a Fortessa flow cytometer using FACSDiva software (BD Biosciences). All flow cytometry data were analyzed using FlowJo software (FlowJo). Expression of PD-1 and PD-L1 was calculated as the frequency of cells showing a greater fluorescence signal when incubated with anti-PD-1 clone EH12.2H7 or PD-L1 clone 29E.2A3 than those incubated with the mouse IgG1 isotype control antibody. Measurement of proinflammatory cytokine release by anti-PD-L1/IL-10-trap in whole blood from healthy donors Peripheral blood from healthy human donors was treated with anti-PD-L1/IL-10-trap (50 nM or at EC50 = 1.2 nM in a cell-based assay) or control antibodies at an equimolar range in the presence of LPS for 24 h. IL-10, IFNγ, IL-12p70, and TNF-α concentrations were measured using the V-PLEX Proinflammatory Panel 1 Human Kit (Meso Scale Discovery) according to the manufacturer’s instructions. Statistical analysis The EC50, average, and standard deviation (SD) values were calculated using GraphPad Prism 10 (GraphPad). Nonlinear regression (4-parameter) fitting was used for graphing and calculation. Data were compared using a t test or one-way ANOVA followed by Tukey’s multiple-comparisons test. All statistical analyses were conducted using GraphPad Prism 10. Data were considered statistically significant at p < 0.05. Data and code availability Any requests for additional data by qualified scientific and medical researchers for legitimate research purposes will be subject to the healthcare business of Merck KGaA’s (CrossRef funder ID: 10.13039/100009945) data-sharing policy. All requests should be submitted in writing to the healthcare business of Merck KGaA’s data-sharing portal (https://www.emdgroup.com/en/research/ourapproach-to-research-and-development/healthcare/clinical-trials/commitment-responsible-data-sharing.html). When the healthcare business of Merck KGaA has a co-research, co-development, co-marketing, or co-promotion agreement, or when the product has been out licensed, the responsibility for disclosure might be dependent on the agreement between parties. Under these circumstances, the healthcare business of Merck KGaA will endeavor to obtain agreement to share data in response to requests.