Targeting NKG2DL with Bispecific NKG2D–CD16 and NKG2D–CD3 Fusion Proteins on Triple–Negative Breast Cancer

Triple–negative breast cancer (TNBC) is a particularly aggressive subtype of breast cancer with a poor response rate to conventional systemic treatment and high relapse rates. Members of the natural killer group 2D ligand (NKG2DL) family are expressed on cancer cells but are typically absent from healthy tissues; thus, they are promising tumor antigens for novel immunotherapeutic approaches. We developed bispecific fusion proteins (BFPs) consisting of the NKG2D receptor domain targeting multiple NKG2DLs, fused to either anti–CD3 (NKG2D–CD3) or anti–CD16 (NKG2D–CD16) Fab fragments. First, we characterized the expression of the NKG2DLs (MICA, MICB, ULBP1–4) on TNBC cell lines and observed the highest surface expression for MICA and ULBP2. Targeting TNBC cells with NKG2D–CD3/CD16 efficiently activated both NK and T cells, leading to their degranulation and cytokine release and lysis of TNBC cells. Furthermore, PBMCs from TNBC patients currently undergoing chemotherapy showed significantly higher NK and T cell activation and tumor cell lysis when stimulated with NKG2D–CD3/CD16. In conclusions, BFPs activate and direct the NK and T cells of healthy and TNBC patients against TNBC cells, leading to efficient eradication of tumor cells. Therefore, NKG2D–based NK and T cell engagers could be a valuable addition to the treatment options for TNBC patients.


Introduction
Invasive breast cancer (IBC) is the most common malignancy in women worldwide [1].The absence of estrogen receptor, progesterone receptor and human epidermal growth receptor 2 (HER2) is classified as triple-negative breast cancer (TNBC).TNBC accounts for up to 20% of all breast cancers, and remains one of the most difficult to cure.It also exhibits a particularly aggressive phenotype compared to other subtypes of IBC [2,3].TNBC patients typically receive conventional chemotherapy [4,5], but the recent introduction of PD1/PDL1 checkpoint inhibitors such as Pembrolizumab and Atezolizumab enhance T-cell-mediated anti-tumor effects and improve outcomes in TNBC.However, recurrence rates remain high, and the management of metastatic disease remains challenging [6].
The natural killer group 2D ligand (NKG2DL) family consists of the MHC class I-related chain (MIC) MICA and MICB and the UL16-binding proteins (ULBP) ULBP1-6.NKG2DLs are generally absent on healthy tissues, but are induced upon cellular stress such as malignant transformation or infection [7].NKG2DLs are potent inducers of antitumor immunity by activating natural killer (NK) cells and T cells after binding to the NKG2D receptor.This leads to the release of effector granules such as perforin and granzyme, thereby inducing apoptosis of the tumor cells [8].NKG2DL expression can be further enhanced by treatment with various DNA-damaging chemotherapeutic agents such as doxorubicin [9].It has been reported that all TNBCs express NKG2DL [10].Therefore, NKG2DLs are promising tumor antigens, and there are many therapeutic approaches targeting the NKG2DL/NKG2D axis are currently under clinical investigation.These include chimeric antigen receptor T cells (CAR-T) or chimeric antigen receptor NK cells (CAR-NK), or an anti-MICA/MICB monoclonal antibody, which have induced partial responses in leukemia and multiple myeloma patients (NCT04550663, NCT05117476, NCT05117476) [11,12].We have previously targeted NKG2DLs with NKG2D fusion proteins, in which the extracellular domain of the NKG2D receptor was fused to an antibody-dependent cellular cytotoxicity (ADCC)-optimized Fc-IgG1 moiety.With this construct, we were able to successfully induce NK cell reactivity and induce lysis of IBC cells [13].In general, monoclonal antibodies (mAbs) and IgG fusion proteins face two problems.First, they can only activate NK cells as effector cells, and second, their Fc parts can also bind to inhibitory Fcγ receptors, which reduces their cytotoxic effect [14].To overcome these hurdles, we improved the anti-tumor activity by creating bispecific fusion proteins (BFPs) with the NKG2D receptor domain linked to anti-CD16 or anti-CD3 Fab fragments to direct the NK cells or cytotoxic T lymphocytes against tumor cells.The main advantage of these two constructs is, on the one hand, that binding to inhibitory Fcγ receptors is prevented by the anti-CD16 part and, on the other hand, that T cells with higher cytotoxic potential, can be recruited by the anti-CD3 part.These BFPs have already been shown to have potent anti-tumor effects against soft tissue sarcomas and acute myeloid leukemia [15,16].
In this study, we evaluate the efficacy of NKG2D-CD16 and NKG2D-CD3 bispecific fusion proteins (BFPs) in the treatment of TNBC.We show that the BFPs effectively direct NK cells and T cells to the tumor cells and eliminate the TNBC cells.Furthermore, we report that NK cells and T cells from TNBC patients currently undergoing chemotherapy are able to efficiently eliminate TNBC tumor cells.

Characterization of NKG2DL Expression in Triple-Negative Breast Cancer Cells
We examined the mRNA expression of NKG2DL in eight TNBC cell lines and found that all of them expressed at least two different ligands at varying levels (Figure 1A).MICA showed the highest expression, while the expression of MICB and ULBP1-3 varied between different cell lines.ULBP4 was mostly absent in all cell lines tested.
Next, we determined the surface protein expression using specific mAbs against MICA, MICB, and ULBP1-4 (Figure 1B).MICA and ULBP2 showed the highest surface expression among cell lines tested, whereas ULBP4 could not be detected (Figure 1B).Six TNBC cell lines expressed four or five different NKG2DLs (37.5% each), except HS-578T, which expressed only two, and BT-549, which expressed three types of NKG2DLs (12.5% each) (Figures 1C and S1A).We examined the binding of all NKG2DLs on the surface with the extracellular NKG2D receptor domain fused to the Fc-IgG1 (Figure 1D).Based on our results, we selected BT-549, with the lowest, MDA-MB-468, with intermediate, and MDA-MB-231, with high NKG2DL expression for further experiments (Figures 1E and S1B).To evaluate the expression of NKG2DL in primary breast cancer material (n = 200), we used the TCGA database analysis and observed predominant expression of MICA, MICB and ULBP2 (Figure 1F).

Effector Cell Recruitment by NKG2D BFPs against TNBC Cells
We analyzed the ability of NKG2D-CD16 and NKG2D-CD3 to recruit NK cells and T cells, respectively, to the TNBC cells.MDA-MB-231 and MDA-MB-468 cells were incubated with peripheral blood mononuclear cells (PBMCs) from healthy donors in the presence or absence of NKG2D-CD16.Immune cell recruitment to TNBC was quantified visually by determining the number of colocalized effector and target cells per field of view (FoV), a representative example of which is shown for MDA-MB-231 (Figure 2A,B).We observed a significant increase in CD16 + cells interacting with TNBC cells when treated with NKG2D-CD16 (Figures 2C and S2A).A similar effect was observed for CD3 + cells when MDA-MB-231 or MDA-MB-468 cells were co-incubated with PBMCs from healthy donors in the presence of NKG2D-CD3 (Figures 2D-F and S2B), confirming the effective recruitment of NK cells and T cells induced by the respective BFPs.

Modulation of NK Cell and T Cell Reactivity against TNBC Cells with PBMCs from Healthy Donors
To analyze whether our constructs were able to induce effector cell reactivity against TNBC cells, we cocultured the PBMCs from healthy donors with BT-549, MDA-MB-231 and MDA-MB-468 cells in the presence or absence of BFPs.Application of NKG2D-CD16 or NKG2D-CD3 resulted in a significant induction of the activation marker CD69 on NK cells and T cells (Figure 2G-J).Similarly, the increased surface expression of CD107a confirmed that NKG2D-CD16/CD3 strongly induced the degranulation of NK cells and T cells, respectively (Figure 3A-D).Analysis of the supernatants by Legendplex assays showed a significant increase in IFNγ, TNF, granzyme A, perforin and granulysin secretion after treatment with NKG2D-CD16 (Figure 3E) or NKG2D-CD3 (Figure 3F).

Induction of Target Cell Lysis by Bispecific NKG2D Fusion Proteins
We then analyzed whether activation of effector cells resulted in cytotoxicity against tumor cells.Naturally, NK cells and T cells have different kinetics of activation and tumor cell killing [17].To investigate the intrinsic function of both types of effector cells, we performed cell lysis assays with different incubation times.Europium-based short-term cytotoxicity assays revealed the potency of both BFPs to lyse target cells after 2 h, but the more pronounced effect was observed with NKG2D-CD16 (Figures 3G,H and S3C).At 72 h, flow cytometry-based lysis assays showed much stronger lysis with NKG2D-CD3 (Figures 3I,J and S3D).The finding that the lysis ability of NKG2D-CD3 emerged after prolonged incubation time and then surpassed the effects of NKG2D-CD16 was also confirmed by live cell imaging over an incubation period of 120 h (Figures 3K,L and S3E).

Reactivity of NK Cells and T Cells from TNBC Patients against TNBC Cells
To understand whether there are disease-and treatment-related changes in the immune cell subsets of TNBC patients that could affect the treatment efficacy of our BFPs, we compared the PBMCs from healthy donors and TNBC patients.At the time of PBMC collection, all tested TNBC patients studied were receiving cytotoxic chemotherapy with or without added immunotherapy.The specific details of the clinical characterization and treatment regimens of all TNBC patients are shown in Table 1.In our TNBC cohort, we observed normal counts of lymphocytes, leukocytes, neutrophils, monocytes and platelets (Figures 4A and S4A).We evaluated the distribution of the PBMC subsets, including T cells (CD3 + CD4 + and CD3 + CD8 + ), monocytes (CD14 + ), dendritic cells (DCs, CD3 − CD56 − CD14 − CD19 − HLA-DR + ), B cells (CD19 + ), NKT (CD3 + CD56 + ) and NK cells (CD3 − CD56 + ), by performing flow cytometric analysis.TNBC samples showed fewer CD4 + T cells but the same amount of CD8 + T cells and an increase in DCs.No relevant difference was observed for other mononuclear cells compared to healthy donors (Figure 4B).To determine whether lymphocytes from TNBC patients could be stimulated by NKG2D-CD16 and NKG2D-CD3 in a similar manner compared to healthy donors, we used cocultures with TNBC cells.Treatment with both NKG2D-CD16 and NKG2D-CD3 significantly increased the activation of NK cells and CD4 + and CD8 + T cells, as determined by flow cytometry, for CD69 surface expression at 24 h (Figures 4C and S4B,C).In line, the detection of CD107a upregulation confirmed the effective degranulation of NK cells and T cells after treatment with BFPs (Figures 4D and S4D,E).Analysis of the coculture supernatants revealed a significant increase in the release of IFNγ, TNF, granzyme A, perforin and granulysin after treatment with both NKG2D-CD16 (Figure 4E) and NKG2D-CD3 (Figure 4F).Green fluorescent target cells were incubated with BFPs and PBMCs from healthy donors (n = 3) at an E:T ratio of 5:1 for 120 h.(K) Representative images at 0 h, 72 h and 120 h at 10x magnification and (L) data for separate TNBC cell lines and PBMCs from healthy donors (n = 3) are shown.All statistical tests were considered significant if the p-value was less than 0.05 (* p < 0.05, *** p < 0.001).p-values greater than 0.05 are indicated as not significant (ns).Finally, we evaluated the killing capacity of NK cells and T cells from TNBC patients against TNBC cell lines by live cell imaging (Figure 4G).While the strong lysis effect of NKG2D-CD3 was still present in TNBC patients, NKG2D-CD16 was only sufficient for the MDA-MB-468 target cell line (Figure 4G).

Discussion
The landscape of cancer treatment options has evolved rapidly in recent years.With the success of checkpoint inhibitors in TNBC, immunotherapy has become a standard of care, highlighting the strength of T-cell-based strategies for anti-tumor therapy.Despite these promising developments, survival rates are still lower than in other breast cancer subtypes, and the risk of recurrence remains a challenge [2].NKG2D-based NK cell and T cell engagers could be potential options for the treatment of TNBC.
In our present study, we observed effective lysis of TNBC cells by PBMCs from both healthy donors and TNBC patients after treatment with NKG2D-CD16 and NKG2D-CD3 BFPs.Increased surface expression of CD69 and CD107a on NK cells and CD4 + and CD8 + T cells indicated enhanced activation and degranulation accompanied by cytokine release upon NKG2D-CD3/CD16 BFP treatment.This effect was reflected in significant cytotoxicity against TNBC cells.
Recently, maturation of NK cells has been shown to be exclusively important in predicting their potential effect against TNBC tumors [18].In humans, terminally differentiated peripheral blood NK cells, exhibiting the CD56 dim CD16 + phenotype, are cytotoxic.In contrast, CD56 bright CD16 − NK cells, found in secondary lymphoid tissues, are considered to be immature and have reduced cytotoxic potential [19].Thacker et al. demonstrated an increased number of immature NK cells in various TNBC models and provided evidence for the pro-tumorigenic nature of NK cells within the tumor microenvironment (TME) of TNBC [18].Using the NKG2D-CD16 construct, we recruited cytotoxic CD16 + NK cells, potentially altering the balance between immature and cytotoxic NK cell populations.Another advantage of NKG2D-CD16 is that NK cell activation is not dependent on the NKG2D receptor, which is beneficial given that NKG2D expression on NK cells can be downregulated under certain circumstances [20].It is noteworthy that NKG2D-CD16 demonstrated superior cytotoxicity in the short-term assays, whereas NKG2D-CD3 required a longer period to achieve maximal anti-tumor activity.Although both NKG2D-CD16 and NKG2D-CD3 were effective, the killing ability of T cells stimulated by NKG2D-CD3 was more pronounced.This difference could be attributed to the fact that NKG2D-CD3 can induce T cell proliferation, whereas no proliferation was observed in NK cells activated by NKG2D-CD16 [21].
In terms of clinical implications, it is plausible to assume that the lower cytokine release upon treatment with NKG2D-CD16 compared to NKG2D-CD3 may lead to fewer side effects.This aspect qualifies NKG2D-CD16 as a promising therapy for elderly TNBC patients who may not tolerate immediate massive immune activation.In addition, our study showed that PBMCs from TNBC patients undergoing chemotherapy showed notable activation of NK cells and T cells upon BFP stimulation, suggesting the feasibility of combined immunotherapy.However, additional data are needed to fully understand the bioavailability of NKG2D BFPs in the TNBC TME in the clinical setting.
In conclusion, these results highlight the encouraging antitumor efficacy of NKG2D-CD3 and NKG2D-CD16 BFPs.PBMCs from TNBC patients showed a robust potential to targeting cancer cells, even during ongoing chemotherapy, suggesting that our novel approach could serve as stand-alone therapy or as an adjunct to chemotherapy.
NK cell and T cell activation and degranulation were assessed by flow cytometry.Briefly, 100,000 TNBC cells were cultured with allogeneic PBMCs from healthy donors or TNBC donors at an effector to target (E:T) ratio of 2.5:1 for 4 h and 24 h, followed by mAb staining.For flow cytometric evaluation of specific target cell lysis, TNBC cells were stained with 2.5 µM CellTrace™ Violet (Thermo Fisher Scientific, Waltham, MA, USA) and cocultured with PBMCs (E:T ratio of 5:1) in the presence of the BFPs (2.5 µg/mL) or control for 72 h.Latex beads (Sigma-Aldrich, Darmstadt, Germany) were used to ensure that equal volumes of cell suspension were analyzed.7-AAD was used to exclude dead cells from the analysis.PBMC subsets from healthy donors and TNBC patients were identified by counterstaining.
All measurements were performed using a FACS Canto II, FACS Fortessa or FACS Aria III (BD Biosciences, Heidelberg, Germany) and data were analyzed using FlowJo-V10 software (BD Biosciences, Heidelberg, Germany).

Analysis of Cytokine Secretion
To evaluate the cytokine release, supernatants from the coculture assays were analyzed at 4 h and 24 h using the Legendplex Human CD8/NK Panel (BioLegend, San Diego, CA, USA) according to the manufacturer's protocol.

Cytotoxicity Assay
Lysis of TNBC cells by PBMCs from healthy donors in the presence or absence of the BFP (2.5 µg/mL) was assessed by a 2 h Europium-based cytotoxicity assay, as previously described [15].Briefly, TNBC cells were labeled with DELFIA ® BATDA (Perkin Elmer,

Figure 1 .
Figure 1.Expression of NKG2DL in triple-negative breast cancer.(A) Expression of various NKG2DL mRNAs in TNBC cell lines was measured by qRT-PCR.(B) Surface expression of the indicated NKG2DL was stained with specific mAbs (10 µg/mL) and assessed by flow cytometry.Open histograms show isotype control staining, and filled histograms show the staining of the indicated NKG2DL.(C) The relative proportion of cell lines expressing the indicated amount of different NKG2DL on the surface is shown.(D) Schematic representation of NKG2DL staining with

Figure 2 .
Figure 2. Recruitment of effector cells to the TNBC cell lines and activation of NK cells and T cells by BFPs.(A,D) Schematic illustration of BFPs with the αCD16 (blue) (A) or αCD3 (green) (D) sFab linked with a CH2 domain of IgG1 to the extracellular receptor domain of NKG2D (created

Figure 2 .
Figure 2. Recruitment of effector cells to the TNBC cell lines and activation of NK cells and T cells by BFPs.(A,D) Schematic illustration of BFPs with the αCD16 (blue) (A) or αCD3 (green) (D) sFab linked with a CH2 domain of IgG1 to the extracellular receptor domain of NKG2D (created with BioRender.com).(B,C,E,F) TNBC tumor cells were incubated with PBMCs from a healthy donor

Figure 3 .
Figure 3. Reactivity of healthy donor T and NK cells against TNBC cell lines in response to NKG2D-CD16/CD3 treatment.PBMCs from healthy donors were cultured with TNBC cells at an

Figure 3 .
Figure 3. Reactivity of healthy donor T and NK cells against TNBC cell lines in response to NKG2D-CD16/CD3 treatment.PBMCs from healthy donors were cultured with TNBC cells at an E:T ratio of 2.5:1 (unless otherwise noted) with controls (gray) or NKG2D-CD16 (blue) NKG2D-CD3

Figure 4 .
Figure 4. Characterization of PBMCs from TNBC patients receiving chemotherapy or combined chemo-immunotherapy. (A) Leukocyte and lymphocyte counts from TNBC patients (n = 10) at the time of sample collection are shown.Dotted lines indicate the normal range for healthy individuals.(B) Peripheral mononuclear cells (MNCs) from TNBC patients (n = 19) and healthy donors (n = 7) were

Table 1 .
Clinical characteristics of TNBC patients and therapy line at the time of blood sampling.