The First-In-Class Anti-AXL×CD3ε Pronectin™-Based Bispecific T-Cell Engager Is Active in Preclinical Models of Human Soft Tissue and Bone Sarcomas

Simple Summary Sarcomas are a group of heterogeneous diseases with a poor prognosis and scarce therapeutic options. Innovative approaches based on novel therapeutic targets are eagerly awaited. AXL, a TAM family tyrosine kinase receptor, recently emerged as an interesting target for several type of sarcomas. Here, we propose an innovative immunotherapeutic strategy based on the targeting of AXL, using a first-in-class Pronectin™-based Bispecific T-Cell Engager (pAXL×CD3ε) for the treatment of sarcomas. Our results demonstrate that pAXL×CD3ε redirects T cells toward AXL-expressing sarcoma cell lines, leading a dose-dependent and T cell-mediated cytotoxicity in vitro. Moreover, pAXL×CD3ε inhibits the in vivo growth of human sarcoma xenografts and improves survival in immunocompromised mice, thus representing a new-generation strategy for the treatment of a still-incurable disease. Abstract Sarcomas are heterogeneous malignancies with limited therapeutic options and a poor prognosis. We developed an innovative immunotherapeutic agent, a first-in-class Pronectin™-based Bispecific T-Cell Engager (pAXL×CD3ε), for the targeting of AXL, a TAM family tyrosine kinase receptor highly expressed in sarcomas. AXL expression was first analyzed by flow cytometry, qRT-PCR, and Western blot on a panel of sarcoma cell lines. The T-cell-mediated pAXL×CD3ε cytotoxicity against sarcoma cells was investigated by flow cytometry, luminescence assay, and fluorescent microscopy imaging. The activation and degranulation of T cells induced by pAXL×CD3ε were evaluated by flow cytometry. The antitumor activity induced by pAXL×CD3ε in combination with trabectedin was also investigated. In vivo activity studies of pAXL×CD3ε were performed in immunocompromised mice (NSG), engrafted with human sarcoma cells and reconstituted with human peripheral blood mononuclear cells from healthy donors. Most sarcoma cells showed high expression of AXL. pAXL×CD3ε triggered T-lymphocyte activation and induced dose-dependent T-cell-mediated cytotoxicity. The combination of pAXL×CD3ε with trabectedin increased cytotoxicity. pAXL×CD3ε inhibited the in vivo growth of human sarcoma xenografts, increasing the survival of treated mice. Our data demonstrate the antitumor efficacy of pAXL×CD3ε against sarcoma cells, providing a translational framework for the clinical development of pAXL×CD3ε in the treatment of human sarcomas, aggressive and still-incurable malignancies.

sarcoma. AXL inhibitors were shown to affect the viability of Ewing sarcoma cells [31]. In addition, a high expression of AXL gene was found in leiomyosarcoma, and its activity was suppressed by two different multi-tyrosine kinase inhibitors (Crizotinib and Foretinib) [20]. Finally, other studies have found that osteosarcoma cells highly express AXL [32], of which the inhibition significantly reduces lung metastases [21]. Despite all these promising findings, an effective anti-AXL treatment is still not available for these aggressive malignancies.
A novel class of non-immunoglobulin, single-domain therapeutic proteins (Pronectins™), based on "antibody mimics" technology, has been just developed with the aim of providing a novel platform for the treatment of various diseases, including cancer. Pronectins™ were isolated from synthetic human libraries, built upon the 14th domain of Fibronectin III (14FN3) scaffold, which is selected by a bioinformatic approach, and on advanced complementarity-determining region (CDR) diversity of more than 25 billion loop sequences [33]. Since Pronectins™ mimic the natural human repertoire, they are poorly immunogenic [34]. Several pharmacological properties are associated to the fibronectin III scaffold, such as high stability, tissue penetration, and low cost of production. Furthermore, they are smaller than a conventional mAb, representing a favorable feature for the local delivery to the solid tumor mass [35]. Starting from Pronectins™, it is possible to generate multimers, fusion proteins, bispecifics or constructs with site-specific modifications for tailored therapy [36]. Bispecific T-Cell Engagers (BTCEs) emerged as a novel promising strategy for hematologic malignancies, but their application in solid tumors is highly challenging, due to the paucity of selective tumor-associated antigens (TAAs) and the struggle in penetrating the solid tumor mass [37]. In this scenario, a Pronectin™-based BTCE (pBTCE) can help overcome these limitations.
Based on this rationale, we investigated the in vitro and in vivo activity of a first-inclass pBTCE targeting AXL (pAXL×CD3ε) as a potential immunotherapeutic agent for the treatment of sarcomas.

Generation and Development of pAXL×CD3ε
A highly specific anti-AXL Pronectin™, a non-immunoglobulin and single-domain protein, has been isolated from synthetic libraries based on the human scaffold of the 14th domain of fibronectin III (14FN3), as previously described [38]. By bioinformatic analysis aimed to select the best candidate within the amino acid loop diversity and minimize or prevent immunogenicity, 6 Pronectins™ with a KD < 10 nM were identified and AXL54 was chosen for targeting purposes (KD = 8 nM). This Pronectin™ was used to develop a first-in-class BTCE (AXL54 (Pronectin™)-linker-scFV CD3, pAXL×CD3ε), for investigation as an anti-tumor novel agent. The linker is made of a single unit of Gly4-Ser (GGGGS) [38].

Transduction of Sarcoma Cell Lines
Sarcoma cells were plated at 1 × 10 5 cells/mL in 6-well plate and incubated O/N. To obtain sarcoma cells stably expressing green fluorescent protein (GFP) transgene, a lentiviral GFP-encoding vector was added according to the manufacturer's instruction (SBI System Biosciences, Mountain View, CA, USA). Polybrene (Sigma-Aldrich, Saint Louis, MO, USA) was also used to a final concentration of 8 µg/mL. Two days after transduction, cells were selected using DMEM, supplemented with 20% FBS, containing 1 mg/mL puromycin (Sigma Aldrich). After antibiotic selection, puromycin-resistant transduced cells were assessed for the expression of GFP by flow cytometry, using Attune NxT Flow cytometer (Thermo Fisher Scientific) and microscopy (Thunder Imaging Systems, Leica Microsystems, Wetzlar, Germany).

Detection of AXL Expression and Target Quantification
AXL expression was analyzed on each sarcoma cell line by flow cytometry. Cells were incubated with FITC-conjugated AXL antibody (#MAB154-100, R&D Systems, Minneapolis, MN, USA) for 15 min at RT in the dark. The tubes were washed in PBS 1X and centrifuged 400× g for 5 min, resuspended in 500 µL of PBS 1X and analyzed by a flow cytometer.
To quantify AXL expression on sarcoma cell lines, calibrated microspheres (Quantum Simply Cellular, Bangs Laboratories Inc., Fishers, Castenaso, BO, Italy) were used according to the manufacturer's protocol. Briefly, saturating amounts of FITC-conjugated AXL antibody were added to one drop of each microbead suspension, and the final mixes were incubated for 30 min at RT in the dark. Samples were washed twice using PBS 1X (2500× g), resuspended in 500 µL of PBS 1X and analyzed by a flow cytometer. Simultaneously, each cell line was stained with FITC-conjugated AXL antibody, as previously described. The analysis was performed maintaining the same instrument setting used for QSC beads. A QuickCal ® spreadsheet, provided by Bangs Laboratories, was used to convert the main fluorescence intensity (MFI) from microspheres to antibody-binding capacity (ABC) values.

Redirected T-Cell Cytotoxicity Assay
PBMCs were isolated from at least 3 donors and labeled with CellTrace™ Violet viable marker (Invitrogen, Waltham, MA, USA), according to the manufacturer's instructions, and co-cultured with sarcoma cell lines (CAL-72, ESS-I, HT-1080, SAOS-2, Rh-30, SW982 or RD-ES) at different effector-to-target-cell (E:T) ratio, in the presence of increasing concentrations of pAXL×CD3ε (0.1 µg/mL, 1 µg/mL and 2.5 µg/mL) or anti-B-cell maturation antigen (BCMA) Pronectin™-based BTCE, pBCMA×CD3ε (2.5 µg/mL), as a negative control. BCMA is in fact highly restricted to hematopoietic B cells and is not expressed by solid tumors, therefore representing a suitable negative control in our case. Cells were incubated for 72 h at 37 • C and 5% CO 2 , and finally stained with 7-AAD (BD Biosciences, La Jolla, CA, USA). The cytotoxic effect on sarcoma cell lines was detected by flow cytometry and reported as the percentage of 7-AAD + /CellTrace™ Violetcells. The 10:1 E:T ratio was selected because it allowed for the highest toxicity.
Cells stably expressing GFP gene were co-cultured with PBMCs from at least 3 donors at 10:1 E:T ratio, in the presence of pAXL×CD3ε (2.5 µg/mL). Cells were incubated for 72 h at 37 • C and 5% CO 2 . Cytotoxicity was assessed by flow cytometry monitoring MFI in GFP-positive cells.
For microscope imaging, co-cultured cells were plated on a round cover glass (Fisher Scientific) above 24 wells, and fixed using 4% paraformaldehyde (PFA) for 15 min. Sections were washed three times with PBS 1X, mounted in Vectashield with DAPI (Vector Lab, Newark, CA, USA) and analyzed using Thunder Imaging Systems (Leica, Wetzlar, Germany).

Cell Viability Assay
Cells were plated in 96 wells treated with different concentrations of pAXL×CD3ε, and cell viability was evaluated by Cell Titer-Glo Luminescent Assay (CTG; Promega, Madison, WI, USA), as previously reported [41].

Western Blot
Whole-cell protein extracts were obtained using NP40 lysis buffer containing Halt Protease and Phosphatase Inhibitor cocktail (Invitrogen, Thermo Fisher Scientific), separated using 4-12% Novex Bis-Tris SDS-acrylamide gels (Invitrogen), and transferred on nitrocellulose membranes (Bio-Rad, Hercules, CA, USA), as previously reported [42]. Nitrocellulose membranes were incubated O/N at 4 • C with primary antibody. In detail, anti-AXL (#4566) by Cell Signaling Technology (Danvers, MA, USA) and anti-GAPDH (sc-25778) by Santa Cruz (Dallas, TX, USA) were used for Western blotting (WB) procedures. The membrane was washed thrice with PBS-Tween and incubated with the secondary antibody (anti-rabbit IgG HRP-linked antibody #7074S, Cell Signaling Technology) for 1 h at RT. Chemiluminescence was recorded using SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Scientific). Densitometric analysis of blots was performed using LI-COR Image Studio Digits Ver 5.0 (Bad Homburg, Germany).

RNA Isolation and Quantitative Real-Time PCR
The WizPrep™ Total RNA Mini Kit (Wizbiosolutions, Seongnam, South Korea) was used, according to the manufacturer's guidelines, to extract purified RNA from sarcoma cell lines. The RNA quantity and quality were assessed by NanoDrop ® (ND-1000 Spectrophotometer). cDNA was obtained from the reverse transcription of total RNA, using the "high-capacity cDNA reverse transcription kit" (Applied Biosystems, Foster City, CA, USA). Taq-Man ® assay (Life Technologies, Carlsbad, CA, USA) was used to detect and quantify AXL (Hs01064439_m1), and GAPDH (Hs03929097_g1) was considered to normalize the recorded threshold cycle values. qRT-PCR was performed in triplicate and relative expression was obtained through the comparative cross threshold method on a ViiA7 System (Thermo Fisher Scientific, Waltham, MA, USA).

Analysis of the Activity of pAXL×CD3ε in Combination with Chemotherapeutic Drugs
SAOS-2 were plated in 24 wells and co-cultured at selected 10:1 E:T ratio with PBMCs from at least 3 donors, labelled with CellTrace™ Violet. Cells were treated with pAXL×CD3ε (1 µg/mL), trabectedin (0.2 nM) or their combination (pAXL×CD3ε + tra-bectedin). After 72 h of incubation, cells were stained using 7-AAD and analysis of positive cells were performed through flow cytometry.

In Vivo Studies
In vivo experiments were performed according to standard guidelines and approved protocols by the National and Institutional Animal Committee (483/2020-PR, 18 May 2020). Four-to-six-week-old male NSG (NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ) mice were purchased from Charles River Laboratories (Wilmington, MA, USA). Animals were regularly monitored and euthanized when signs of disease-related symptoms or graft-versus-host disease (GvHD) developed.
To obtain a subcutaneous (sc) xenografted in vivo model, 10 mice were inoculated in the dorsal right flank with HT-1080 cells (3 × 10 6 ) resuspended in 100 µL of PBS 1X. On day 4, 10 × 10 6 PBMCs from healthy donors were intraperitoneally (ip) injected into each mouse. The same day, mice were randomized in 2 groups (5 mice for each group), and 0.1 mg/kg pAXL×CD3ε or vehicle were ip injected for 15 consecutive days. Tumor sizes were measured with a digital caliper. The tumor volume (tv) was calculated using the formula: where W is the tumor width and L is the tumor length, as previously described [43]. Mice were sacrificed when the tv reached >2000 mm 3 . At the time of sacrifice, blood samples were collected. Red blood cell lysis was performed, and cells were stained with anti-human CD45 BV510 and CD3 PerCP-Cy5.5 to evaluate PBMCs engraftment. Explanted tumors were analyzed by WB, as previously described, using anti-caspase-3 (#9668, Cell Signaling Technology) and anti-PARP (#9532, Cell Signaling Technology), and by immunohistochemistry (IHC) using anti-CD3 antibody (#GA503, Agilent Dako, Glostrup, Denmark).

Statistical Analysis
Statistical evaluations were carried out using a parametric Student's t-test by the Graph-Pad software (www.graphpad.com, accessed on 10 February 2023). Graphpad Prism version 6.0 was used to make graphs. Only results with a p value < 0.05 were accepted as statistically significant. Each value is reported as the mean of at least 2 experiments ± SD/SEM.

Evaluation of AXL Expression on Sarcoma Cell Lines
To investigate the expression of AXL on sarcoma cells ( Figure 1A Figure 1B,C). This trend was confirmed performing quantitative analysis of antigen expression density for each cell line, using calibrated microspheres to assess the antibody-binding capacity (ABC). As reported in Figure 1D, AXL expression was in a range between 21,000 and 4200 antigen molecules on CAL-72 and SW982 cells, respectively. Through qRT-PCR analysis, we assessed the AXL mRNA expression in sarcoma cells ( Figure 1E). Our findings revealed a different pattern of target expression, which was in accordance with data retrievable by cBioPortal for the Cancer Genomics dataset (cbioportal.org, accessed on 10 February 2023) and Cancer Cell Line Encyclopedia (CCLE) dataset (https://depmap.org/portal/interactive, accessed on 10 February 2023). Further Western blot analyses were performed to investigate the expression of AXL protein in sarcoma cells, reporting a clear difference in the band intensity between various cell lines ( Figure 1F). According to our results, AXL expression is not correlated to specific sarcoma sub-types. retrievable by cBioPortal for the Cancer Genomics dataset (cbioportal.org, accessed on 10 February 2023) and Cancer Cell Line Encyclopedia (CCLE) dataset (https://depmap.org/portal/interactive, accessed on 10 February 2023). Further Western blot analyses were performed to investigate the expression of AXL protein in sarcoma cells, reporting a clear difference in the band intensity between various cell lines ( Figure  1F). According to our results, AXL expression is not correlated to specific sarcoma subtypes. Our data demonstrate that AXL is highly expressed in sarcoma cells, therefore representing a potential mean for selective targeting. The interaction of pAXL×CD3ε with sarcoma cells was assessed through indirect staining, using an anti-human IgG secondary antibody (Jackson ImmunoResearch, West Grove, PA, USA) ( Figure S1).

T-Cell Mediated Cytotoxicity Is Induced by pAXL×CD3ε In Vitro
To assess the activity of pAXL×CD3ε, sarcoma cells with different expression levels of AXL were co-cultured with purified human T cells (E:T ratio selected at 10:1) from healthy donors, in the presence of three different concentrations of pAXL×CD3ε (0.1 µg/mL, 1 µg/mL, and 2.5 µg/mL) for 72 h. Increasing concentrations of pAXL×CD3ε produced T-cell-mediated cytotoxicity on sarcoma cells except for the RD-ES because of its low expression of the target antigen (Figure 2A,B). In particular, the cytotoxic activity of pAXL×CD3ε followed a binding-response effect leading to around 50% of cell death on CAL-72, Rh-30, and HT-1080, 40% on SAOS-2, 30% and 20% in SW982 and ESS-I, respectively, at 2.5 µg/mL. As a negative control, we performed a cytotoxicity assay using a Pronectin™-based BTCE binding a different target expressed by B cells only, the BCMA, not expressed by sarcoma cells (pBCMA×CD3ε).
produced T-cell-mediated cytotoxicity on sarcoma cells except for the RD-ES because of its low expression of the target antigen (Figure 2A,B). In particular, the cytotoxic activity of pAXL×CD3ε followed a binding-response effect leading to around 50% of cell death on CAL-72, Rh-30, and HT-1080, 40% on SAOS-2, 30% and 20% in SW982 and ESS-I, respectively, at 2.5 µg/mL. As a negative control, we performed a cytotoxicity assay using a Pronectin™-based BTCE binding a different target expressed by B cells only, the BCMA, not expressed by sarcoma cells (pBCMA×CD3ε). As shown in Figure 2B, 2.5 µg/mL of pBCMA×CD3ε did not induce cytotoxicity in two different sarcoma cell lines, such as HT-1080 and Rh-30, when it was used at a higher dose. To exclude direct cytotoxicity of pAXL×CD3ε on sarcoma cells, we performed a cell viability assay that allows the evaluation of metabolically active cells, in the absence of effector cells. We found that different concentrations of pAXL×CD3ε did not alter cell growth capability of cancer cells ( Figure 2C), indicating that T lymphocytes are indeed required to induce the redirected cytotoxicity of sarcoma cell lines. Furthermore, we performed co-culture experiments on sarcoma cells stably expressing GFP in the presence of 2.5 µg/mL of pAXL×CD3ε. As expected, we observed a strong reduction of GFP signal via imaging analysis, confirming the cytotoxic activity of pAXL×CD3ε observed in our experimental models (Figrues 2D and S2A). Moreover, the percentage of viable cells and fluorescence quantification evaluated by flow cytometry led to the same result (Figrues 2E and S2B).
Taken together, our findings demonstrate that pAXL×CD3ε has an antitumor effect through the recruitment of cytotoxic T lymphocytes.

pAXL×CD3ε Triggers T-Lymphocyte Activation against Sarcoma Cells
Functional effects on PBMCs co-cultured with sarcoma cells at 10:1 E:T ratio, in the presence of increasing concentrations of pAXL×CD3ε or vehicle, were also evaluated after 72 h of treatment. As shown in Figure 3  Consistent with their cytotoxic function, T lymphocytes were also positive for CD107a degranulation marker ( Figure 4A,B).
These data indicate that pAXL×CD3ε produces a dose-dependent activation of T lymphocytes against AXL-positive sarcoma cells.

pAXL×CD3ε Increases Cytotoxicity Induced by Trabectedin
To verify if pAXL×CD3ε could make tumor cells more sensitive to conventional chemotherapeutic drugs, SAOS-2 was selected as the cell model to investigate the effect of redirected T-cell toxicity. Cells were co-treated with pAXL×CD3ε (1 µg/mL) and trabectedin (0.2 nM). After 72 h of treatment, an enhanced cytotoxic effect was observed for pAXL×CD3ε plus trabectedin, as compared to the effect induced by the single agents. In detail, pAXL×CD3ε increased cell death >20% in SAOS-2 cells compared to the effect induced by trabectedin alone ( Figure 5A). The dot plots in Figure 5B provide a graphical overview of the reduction in cell viability (%). These data suggest a potential advantage induced by the combination of pAXL×CD3ε with chemotherapeutics commonly used for sarcoma therapy. These data indicate that pAXL×CD3ε produces a dose-dependent activation of T lymphocytes against AXL-positive sarcoma cells.

pAXL×CD3ε Increases Cytotoxicity Induced by Trabectedin
To verify if pAXL×CD3ε could make tumor cells more sensitive to conventional chemotherapeutic drugs, SAOS-2 was selected as the cell model to investigate the effect of redirected T-cell toxicity. Cells were co-treated with pAXL×CD3ε (1 µg/mL) and trabectedin (0.2 nM). After 72 h of treatment, an enhanced cytotoxic effect was observed for pAXL×CD3ε plus trabectedin, as compared to the effect induced by the single agents. In detail, pAXL×CD3ε increased cell death >20% in SAOS-2 cells compared to the effect induced by trabectedin alone ( Figure 5A). The dot plots in Figure 5B provide a graphical overview of the reduction in cell viability (%). These data suggest a potential advantage induced by the combination of pAXL×CD3ε with chemotherapeutics commonly used for sarcoma therapy.

pAXL×CD3ε In Vivo Activity
The in vivo antitumor efficacy of pAXL×CD3ε was validated against human HT-1080 cell xenografts in NSG-immunocompromised mice ( Figure 6A). A total of 10 xenografted mice were randomized to receive pAXL×CD3ε (0.1 mg/kg, five mice) or the vehicle alone (VEH, five mice) as the control group. A significant reduction of tumor growth was observed in NSG mice treated with pAXL×CD3ε as compared to VEH ( Figure 6B). After 20 days from the cell engraftment, mice treated with pAXL×CD3ε showed a tumor volume of about 630 mm 3 versus 1200 mm 3 in the VEH-only group. This effect translated into a prolonged survival of treated animals ( Figure 6C). To demonstrate the engraftment of human T-lymphocytes in these immunocompromised mice, flow cytometry analyses were performed on peripheral blood samples collected from mice on the day of sacrifice. An anti-CD3-fluorochrome-conjugated antibody was used for the staining, and T-cell engraftment was confirmed both in pAXL×CD3ε and VEH groups ( Figure 6D). Retrieved xenografts from mice were homogenized and WB analysis was performed on the wholecell protein extracts. The analysis revealed the induction of apoptotic processes, which was demonstrated by the increase of cleaved PARP and cleaved caspase-3 in treated mice as compared to VEH ( Figure 6E). IHC analyses also highlighted the infiltration of CD3+

pAXL×CD3ε In Vivo Activity
The in vivo antitumor efficacy of pAXL×CD3ε was validated against human HT-1080 cell xenografts in NSG-immunocompromised mice ( Figure 6A). A total of 10 xenografted mice were randomized to receive pAXL×CD3ε (0.1 mg/kg, five mice) or the vehicle alone (VEH, five mice) as the control group. A significant reduction of tumor growth was observed in NSG mice treated with pAXL×CD3ε as compared to VEH ( Figure 6B). After 20 days from the cell engraftment, mice treated with pAXL×CD3ε showed a tumor volume of about 630 mm 3 versus 1200 mm 3 in the VEH-only group. This effect translated into a prolonged survival of treated animals ( Figure 6C). To demonstrate the engraftment of human T-lymphocytes in these immunocompromised mice, flow cytometry analyses were performed on peripheral blood samples collected from mice on the day of sacrifice. An anti-CD3-fluorochrome-conjugated antibody was used for the staining, and T-cell engraftment was confirmed both in pAXL×CD3ε and VEH groups ( Figure 6D). Retrieved xenografts from mice were homogenized and WB analysis was performed on the whole-cell protein extracts. The analysis revealed the induction of apoptotic processes, which was demonstrated by the increase of cleaved PARP and cleaved caspase-3 in treated mice as compared to VEH ( Figure 6E). IHC analyses also highlighted the infiltration of CD3+ cells in tumor xenografts from mice treated with pAXL×CD3ε, thus demonstrating the effective engagement of T-lymphocytes at the tumor site ( Figure 6F). cells in tumor xenografts from mice treated with pAXL×CD3ε, thus demonstrating the effective engagement of T-lymphocytes at the tumor site ( Figure 6F). Based on these findings, pAXL×CD3ε demonstrates promising antitumor activity against sarcoma xenografts in vivo.

Discussion
Cancer immunotherapy based on T-cell engagement is a valuable therapeutical option and is in an advanced phase of clinical evaluation for different hematological malignancies [44][45][46]. While conventional mAbs bind the same antigen with both fragment antigen-binding (Fab) arms [47], BTCEs simultaneously bind a TAA on cancer cells and the epsilon (ε) subunit of CD3 on the T lymphocytes and, therefore, can efficiently trigger redirected T-cell cytotoxicity in an MHC-independent fashion [48,49]. This simultaneous engagement of the antigen on tumor cells and effector cells leads to an Based on these findings, pAXL×CD3ε demonstrates promising antitumor activity against sarcoma xenografts in vivo.

Discussion
Cancer immunotherapy based on T-cell engagement is a valuable therapeutical option and is in an advanced phase of clinical evaluation for different hematological malignancies [44][45][46]. While conventional mAbs bind the same antigen with both fragment antigen-binding (Fab) arms [47], BTCEs simultaneously bind a TAA on cancer cells and the epsilon (ε) subunit of CD3 on the T lymphocytes and, therefore, can efficiently trigger redirected T-cell cytotoxicity in an MHC-independent fashion [48,49]. This simultaneous engagement of the antigen on tumor cells and effector cells leads to an immunological synapse, resulting in T-cell activation and subsequent release of inflammatory cytokines and cytolytic molecules that lead to the killing of cancer cells [39,40,50].
Despite their demonstrated efficacy in patients with hematological malignancies, no BTCEs have been approved so far for the treatment of solid tumors [51]. There are, in fact, some main hurdles that can hamper the use of BTCEs in solid tumors: (i) on-target offtumor toxicities due to the absence of specific TAAs; (ii) impaired anti-cancer activity due to the hostile and immunosuppressive tumor microenvironment (TME) that antagonizes T-cell infiltration into the tumor mass; (iii) reduced bioavailability and scarce penetration within a solid tumor mass [52,53]. In this scenario, the TME may play a relevant role in cancer progression and can influence the clinical management of these diseases [54][55][56], since it includes immune cells and stromal cells interacting with malignant cells through contact mechanisms or cytokines and subcellular structures, inducing both pro-tumor and antitumor activity. Among them, CD4+ and CD8+ T cells, together with NK cells, dendritic cells, and M1 tumor-associated macrophages (TAMs), promotes cell killing [57][58][59]. Novel approaches would aim to restore the immune function overcoming cancer suppressive effects [60]. In this light, different strategies are emerging by innovative protein-based scaffolds and novel targets [39,40,61].
Here, we assessed that AXL is highly expressed among a variety of sarcomas, confirming previous data performed on primary sarcoma samples [20,22,31,32,62,63] and representing a promising target for the development of innovative immunotherapeutic approaches, especially in chemo-refractory disease. Previous studies, using mAbs against AXL, have reported activity and manageable toxicity in sarcoma patients, suggesting its targeting potential also in the clinical setting [22,64]. Recently, different strategies based on mAbs and CAR-T cells showed encouraging results against AXL-expressing sarcomas and some of them are currently under clinical investigation. The safety and tolerability of CCT301-38 CAR-modified autologous T cells are being investigated in subjects with r/r sarcomas (NCT05128786). Patients with AXL gene alterations were also recruited for another phase I study to determine the safety, tolerability, pharmacokinetics, and antitumor effects induced by Mipasetamab Uzoptirine (ADCT-601) alone, or in combination with other anti-cancer drugs (NCT05389462). The immunogenicity and antitumor efficacy of BA3011, a conditionally active biologic (CAB) AXL-targeted antibody drug conjugate (CAB-AXL-ADC), is being investigated in a phase I/II study in different sarcoma subtypes, in monotherapy or combined with a PD-1 inhibitor (NCT03425279). Finally, a trial has been completed on different tumor types, including sarcoma, to investigate the safety and efficacy of Enapotamab Vedotin (HuMax-AXL-ADC), an AXL-specific antibody drug conjugate (NCT02988817).
On these premises, we focused on AXL as an immunotherapeutic target to be exploited in sarcoma treatment by BTCE-based strategy. We used an emerging protein therapeutic class, called Pronectins™ [36], taking advantage of their small size and low molecular weight to reach a higher concentration within the tumor tissue. Consistently, here we demonstrated that pAXL×CD3ε indeed redirects T cells toward AXL-expressing sarcoma cells, leading to a dose-dependent T-cell activation, with a consequent release of inflammatory cytokines and cytolytic molecules. Our results are in accordance with data recently reported, showing that pAXL×CD3ε exhibits cytotoxic effects on AXL-positive MDA-MB-231 cells and minimal cytotoxicity on AXL-negative CHO cells [38]. Moreover, we found enhanced cytotoxic effects as a function of increased concentrations of pAXL×CD3ε, which was highly promising taking into account the presence of immune cells in the TME of sarcomas. Furthermore, we demonstrated that the combination of pAXL×CD3ε with trabectedin, a conventional active chemotherapeutic drug, improved the cytotoxicity of sarcoma cells. Even if the use of conventional chemotherapeutics is under interindividual variability term of efficacy and toxicity [65][66][67], our data suggest the feasibility of combinatorial treatments and are consistent with preliminary reports, which showed anti-sarcoma activity of immunotherapy/chemotherapy combination [68]. Importantly, in our in vivo model, pAXL×CD3ε guaranteed the recruitment of T lymphocytes in the tumor site and significantly inhibited the growth of sarcoma xenografts, suggesting that this strategy has the potential to also control the fast-growing tumor cells in patients. These findings are of translational relevance, since conventional approaches are still largely unsuccessful, and the only favorable strategy at the present, for the treatment of this incurable disease, is represented by surgery in combination with pre-or post-surgery therapies.
Overall, we demonstrate that the first-in-class pAXL×CD3ε-based immunotherapy exerts significant anti-sarcoma activity in vitro and in vivo, and therefore represents a promising tool to be developed in the clinical setting, offering a novel opportunity to overcome the unmet need of long-term control of drug refractory disease.

Conclusions
Despite the identification of molecular mechanisms driving sarcoma genesis, as well as the discovery of key transcription factors, sarcoma treatment still represents a great challenge. The variability in response to current therapies can be ascribed to their heterogeneity and aggressive behavior. Taken together, our results indicate that AXL-targeting by the Pronectin™-based BTCE platform may represent a new-generation strategy for the treatment of this still-incurable disease.