Human embryonic kidney 293 (HEK293) and 293A, murine colon carcinoma cell line MC38, murine melanoma cell line B16-F10, murine mammary carcinoma cell line 4T1 and murine lymphoma cell line were obtained from American Type Culture Collection (ATCC). HEK293, 293A and A20 cell lines were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine (FBS) and maintained at 37°C in 5% CO₂, while B16-F10, MC38 and 4T1 cell lines were cultured in Roswell Park Memorial Institute 1640 (RPMI 1640) supplemented with 10% FBS in the same condition.

In a meticulously orchestrated process, a fragment of the engineered SIRPα variant fused with the Fc gene (hereafter referred to as SA) was synthesized by the adept hands of GENEWIZ Biotech, located in Suzhou, China. The plasmid pDC316-hTERT-E1AE1B (pDC316-oAd), a critical component of our experimental framework, underwent precise enzymatic digestion with SalI. This enzymatic action was followed by a strategic ligation of the digested product with the SA fragment, culminating in the creation of the recombinant plasmid pDC316-oAd-SA. The pDC316-oAd-SA plasmid, now a beacon of our genetic engineering prowess, was then allied with an adenoviral backbone plasmid, pBHBlox(delta)E1-3cre, and the union was transfected into the robust HEK293A cell line. This transfection set the stage for the viral drama to unfold, with the cells diligently working to produce viral progeny. After a period of approximately 7 to 14 days post transfection, the diligent observation of the cellular culture yielded the sighting of multiple viral plaques within the HEK293A cells—a testament to the successful replication of our engineered virus. These plaques were harvested with care and underwent rigorous verification through the gold-standard polymerase chain reaction (PCR) analysis to confirm their identity. The recombinant adenovirus, now confirmed and proudly named oAd-SA, was allowed to propagate in the hospitable environment of HEK293 cells. Subsequently, it was subjected to a purification ritual involving ultracentrifugation through discontinuous cesium chloride (CsCl) gradients—a process that separated the viral particles from the cellular debris and other impurities with high fidelity. Consistent with our commitment to methodological rigor, other viruses utilized in our study were packaged and treated with an equivalent level of precision and care. After determining the viral titer, which is a measure of virus concentration, the viruses were meticulously portioned and conserved in the icy embrace of -80°C, ensuring their potency and readiness for future applications.

4T1 and MC38 cells were plated in 24-well plates and infected with oAd-SA compared to oAd-ON at an MOI of 0, 1, 5,10,20 and 50. The cells were stained with crystal violet staining solution (Sigma) for 5 min after the different time of the infection.

4T1 and MC38 cells were plated in 96-well plates and infected with oAd-SA compared to oAd-ON at an MOI of 0, 1, 5,10,20 and 50. Cytotoxicity was evaluated by using CCK8 after the different time of the infection.

The purified virus was used to infect 4T1, MC38, CT26, and B16-F10 cells at an MOI of 20. After 2 days, cell pellets were collected and washed twice with PBS. The pellets were then divided into two portions. One portion was lysed with TRIzol for 30 minutes to extract RNA from the cell sediment. The other portion was lysed with RIPA buffer containing protease inhibitors for 30 minutes to extract proteins from the cell sediment. This process will facilitate the subsequent detection of the target gene SIRPα-Fc expression.

Cells (A20, 4T1, and MC38) were individually collected into flow cytometry tubes, washed with 1× PBS, resuspended in 100 μl of 1× PBS, and incubated with CD47 flow cytometry antibody. After antibody removal by washing, CD47 expression on tumor cells was measured using flow cytometry. Concurrently, cells were seeded at 1×10⁵ cells/well in a 24-well plate, treated with oAd-ON or oAd-SA viruses at MOI 50, and incubated for 2 days prior to cell collection. Meanwhile, CD47 expression on tumor cells was measured using WB.

Euthanize a C57 or BLAB/c mouse using cervical dislocation. Remove the hind legs with sterile scissors and tweezers, ensuring to remove attached muscles. Wash the bones twice with 5 ml of ice-cold sterile 1x PBS. Use a 1 ml syringe to flush cells from the femur and tibia to obtain a cell suspension. Filter the cell suspension through a 70 μM filter and centrifuge at 450 x g for 10 minutes to collect the cell pellet. Resuspend the cell pellet in 5 ml of red blood cell lysis buffer and incubate for 2 minutes until the suspension loses its red color. Add 10 ml of complete DMEM medium to stop the lysis, then centrifuge at 450 x g for 10 minutes. Discard the supernatant. Resuspend the cell pellet in 20 ml of pre-warmed 10% FBS 1640 medium, and transfer to a 10 cm culture dish. Incubate in a 5% CO₂ incubator. Change the medium after 3 days of culture and continue to culture for an additional 2 days. Perform flow cytometry antibody detection to verify cell purity and phenotype before use.

Bone marrow derived macrophage (BMDM) was isolated from femur and tibia of Balb/c mice and confirmed by flow cytometry. 24h before treatment, 1x10⁵ A20, 4T1 and MC38 cells were seeded in 12-well plate, three wells for each cell. The cells were infected by Ad-ON or Ad-SA with multiplicity of infection of 50 (MOI=50). Wells with medium only was taken as control. 2 days later, above tumor cells were collected, stained with PKH26, reseeded in 24-well plate. Then, 5x10⁴ BMDM stained with Crystal Field Stabilization Energies (CFSE) was added into each well. This was followed by incubation for 3.5 h in the incubator. Finally, the cells mixture was collected, and phagocytosis was evaluated via flow cytometry and confocal microscopy. After co-culturing tumor cells with macrophages, F4/80 was used to label the macrophages to assess the effect of oAd-SA on macrophage polarization, with CD86 serving as a marker for M1-type macrophages.

The 6-8 weeks old female Balb/c and C57BL/6 mice were purchased from Huafukang Bioscience (Beijing, China). All the animal experiments were approved by Institutional Animal Care and Use Committee Sichuan University. Briefly, 1×10⁶ 4T1 or A20 cells was implanted subcutaneously on the right flank of Balb/c mice and 1×10⁶ MC38 cells was inoculated at the same position of C57BL/6 mice to establish xenografts, respectively. The mice were randomly divided into different groups based on experimental arrangement before receiving any treatment. In brief, mice from Ad-SA group were intratumorally injected 5×10⁸ PFU Ad-SA in 50ul PBS every three days when tumor reached 50-100mm³ on average, 3 doses were administrated in total. Mice from Ctrl and Ad-ON group received same volume of PBS or same dosage of Ad-ON, respectively. During the experiment, tumor size was measured by a caliper and determined according to formula:

Where L and W represent the length and width of the tumor, respectively.

Animals were euthanized 4 days after the administration of adenovirus. Tumor and spleen were harvested. For tumor cells analysis, about 100mg of tumor was minced into small pieces and submerged in RPMI-1640 (Gibco) with 0.1% (w/v) IV collagenase and 1% FBS at 37°C for one and a half hours with agitation, filtered with 70-um sieve, washed and resuspended in PBS. Then stained with Fixable Viability Stain 620 to exclude dead cells. The cells were blocked by Fc-block (BD biosciences) and stained with antibodies. For nuclear factors detection like FoxP3, fixation and permeabilization kit (eBioscience) was used. For spleen cells analysis, the spleen was grinded through 70-um sieve, then cells were extracted with lymphocyte separation medium (BD bioscience) according to manufacture instructions and stained with antibodies. For intracellular factors analysis such as IFNγ, fixation and permeabilization solution (BD bioscience) was utilized. Cells were analyzed on a NovoCyte flow cytometer. Antibodies to CD3 (145-2C11), CD4 (RM4-5), CD8 (53-6.7), CD45 (30-F11), CD11b (M1/70), F4/80 (BM8), CD206, Gr1 (RB6-8C5), CD25 (3C7), FoxP3 (MF-14), CD86 (GL-1), IFN-γ and TNF-α were acquired from Biolegend.

Frozen sections of tumor tissue were obtained, with nuclei stained using DAPI, CK19⁺ cells labeled in green fluorescence, and F4/80⁺ cells in red fluorescence. Similar to the in vitro phagocytosis assay, yellow fluorescence (resulting from the overlap of red and green) indicates active phagocytosis. Immunofluorescence was utilized to observe in vivo phagocytosis.

Data were analyzed using GraphPad Prism. Statistical significance was determined using unpaired t-tests. Animal survival was illustrated using Kaplan-Meier survival curves and analyzed using the log-rank test. Immunohistochemistry and immunofluorescence images were quantified using Image Pro Plus 6.0. Data are presented as mean ± SEM. A p-value of <0.05 was considered statistically significant. In figures, symbols denoting significance are: *p < 0.05, **p < 0.01, ***p < 0.001, and ns (no statistical significance).

RT-PCR analysis revealed a prominent band at 612 bp for the SA gene, as shown in Figure 1A . Interestingly, the reverse transcription amplification from CT26 cells exhibited a weaker band compared to the others. Figure 1B displays the results of a Western blot analysis on the cell pellets of 4T1, MC38, B16-F10, and CT26, indicating the presence of SIRPα-mIgGFc protein at around 50 kDa. The intensity of the protein bands was strongest in 4T1 cells, followed by B16-F10 and MC38, while CT26 showed no significant protein band. This weaker expression in CT26 may be attributed to its lower sensitivity to adenovirus, resulting in fewer viral particles entering the cells and consequently a diminished expression of the target gene. Additionally, nucleic acid detection methods generally exhibit higher sensitivity compared to protein detection.

In an in vitro virus infection assay, we found that 4T1 and MC38 cells demonstrated strong sensitivity to adenovirus, leading us to construct tumor models for subsequent in vivo experiments. The oAd-SA demonstrated a significant inhibitory effect on tumor cell growth, with infection of tumor cells at varying multiplicities of infection (MOI = 0, 1, 5, 10, 20, 25) resulting in a dose-dependent decrease in both cell density and viability, as illustrated in Figures 1C, D .

A subsequent study was conducted in order to confirm the results of the initial investigation, which had revealed the presence of high levels of CD47 expression in normal mouse tissues ( Supplementary Figure S1C ), as well as in mouse tumor tissues. This was achieved by obtaining relevant data from the MGI database (50), specifically 4T1, A20, and MC38, as vividly depicted in Figure 2A . This initial observation laid the groundwork for our subsequent inquiry into the functionality and efficacy of the SIRPα mutant secreted by tumor cells infected with the engineered oAd-SA. Employing the sophisticated technique of flow cytometry, we observed a remarkable reduction in CD47 expression on the aforementioned tumor cell lines post-infection with oAd-SA, in stark contrast to their untreated counterparts or those exposed to the control virus oAd-ON, with the 4T1 cell line exhibiting particularly pronounced effects as illustrated in Figures 2B, C . Further exploration through immunofluorescence and flow cytometry unveiled that infection with oAd-SA significantly bolstered the phagocytic activity of macrophages against the A20 tumor cells when juxtaposed with the control virus oAd-ON ( Figures 2D, E ). This enhancement in phagocytosis is a testament to the biological impact of the SIRPα mutant. The collective findings from these assays converge to suggest that the SIRPα mutant released by oAd-SA-infected tumor cells not only retains its functionality but also exerts a potent effect on modulating the tumor-immune cell interaction. The A20 and macrophage co-culture experiments demonstrated that oAd-SA infection resulted in an augmentation of the proportion of M1-type macrophages ( Figure 2F ).

To meticulously assess the therapeutic potency of oAd-SA, we meticulously established subcutaneous tumor models using the A20, MC38, and 4T1 cell lines, which are representative of the malignancies under investigation. The experimental mice were subjected to a regimented treatment protocol as delineated in Figure 3A , ensuring a standardized approach to evaluating the intervention. In a stark contrast to the cohorts that received either PBS or the control virus oAd-ON, the intratumoral administration of oAd-SA was observed to markedly decelerate the progression of A20 subcutaneous tumors. This intervention also yielded a significant extension in the survival span of the treated animals, as evinced in Figures 3B, C . While all mice in the PBS and oAd-ON groups succumbed to the disease within a 57-day period post tumor cell inoculation, a notable subset of the oAd-SA treated group, specifically 3 out of 8 animals, remained alive up to 66 days post inoculation, as depicted in Figure 2C . Furthermore, oAd-SA demonstrated an enhanced capacity to curb the growth of MC38 and 4T1 subcutaneous tumors in comparison to the oAd-ON treatment group, as illustrated in Figures 2D, F . However, this therapeutic impact did not translate into a statistically significant difference in the survival outcomes for these tumor models, as indicated in Figures 3E, G . The aggregate data from these experiments substantiates the superior efficacy of oAd-SA in tumor inhibition when juxtaposed with the control virus oAd-ON, thereby highlighting its potential as a promising therapeutic agent in oncology. It is imperative to note that, thus far, this treatment has not exhibited any signs of tissue toxicity. Furthermore, histopathological analysis of tissue sections from mice revealed the absence of significant lesions ( Supplementary Figure S2 ), thereby substantiating the remarkable safety profile of the method.

for the purpose of investigating anticancer mechanism of oAd-SA, T cells profiles in MC38 tumor and spleen tissue was determined by flow cytometry. The results demonstrated that oAd-SA increased CD3⁺, CD4⁺ and CD8⁺ T cells percentage in all viable cell digested from MC38 tumor tissue compared with mice giving PBS. Nevertheless, oAd-SA displayed slightly higher T cells percentage than oAd-ON, no statistical significance was achieved. Moreover, it is found that oAd-SA was able to shift CD4⁺ T cells to CD8⁺ phenotype in CD3⁺ T cells. OAd-ON showed the same trend ( Figure 4A ).

T cells play an important role in overall immune response against tumor. Hence, we stimulated spleen cells with PMA and Golgi blockers for 2 h and checked IFN-γ secretion of CD4⁺ and CD8⁺ T cells from spleen. The result showed oAd-SA induced expansion IFN-γ-producing CD4⁺ and CD8⁺ T cells compared with control virus ( Figures 4B–E ), which suggests oAd-SA can activate overall immune response against tumor cells. Since antitumor effects of cytotoxic T cell could be compromised by expression of immune checkpoint molecule like PD-1, Tim-3 and CTLA-4. We further checked PD-1 expression on CD3⁺ and CD8⁺ T cells from tumor tissue. It was found that both oAd-ON and oAd-SA are capable of downregulating PD-1 expression on CD3⁺ and CD8⁺ T cells ( Figures 4F–I ). And oAd-SA can further suppress PD-1 expression on CD8⁺ T cells compare to oAd-ON ( Figures 3H, I ). These data tell oAd-SA can stimulate overall immune response against tumor and inhibit the expression of immune checkpoint on infiltrated T cells.

Given the result that oAd-SA is in position to enhance phagocytosis of macrophages against malignant cells in vitro, we explored the percentage and phenotypes of macrophage in tumor microenvironment with flow cytometry. The results demonstrated that oAd-SA significantly decreased the ratio of TAMs (F4/80⁺) in MC38 tumor tissue compared to mice receiving PBS and oAd-ON ( Figures 5A, C ). Most importantly, oAd-SA downregulated the percentage of M2 phenotypes (CD206⁺) of TAMs ( Figures 5B, C ). However, data from A20 tumor tissue displayed opposite result that oAd-SA increased TAMs percentage in tumor microenvironment compare with mice treated with PBS or control virus. Nevertheless, the ratio of M2 phenotype was similar among three groups. oAd-SA significantly increased the percentage of M1 phenotype which resulting in elevated M1/M2 ratio ( Figures 5D, E ). These data suggested that oAd-SA reprogrammed tumor microenvironment via altering the phenotype of macrophage. Then, we investigated if oAd-SA infection can augment phagocytosis of TAM in vivo. OCT frozen tissue sections of A20 tumor were stained by murine CK19 and F4/80 antibodies. Immunofluorescence result indicated that oAd-SA substantial improved phagocytotic effects of TAMs compare to oAd-ON ( Figure 5F ).

There are studies suggest that tumor cells would upregulate PD-L1 expression following oncolytic virus infection with the aim of escaping immune clearance. Therefore, we investigate PD-L1 level post administration of oAd-SA. Compared to animals from PBS and oAd-ON group, oAd-SA did improve PD-L1 expression in tumor microenvironment ( Figures 6A–D ). Hence, for achieving best therapeutic effect, we treated mice with oAd-SA in combination with PD-L1 antibody. Yet, the results were rather disappointing. In MC38 subcutaneous xenografts, animals were sensitive to PD-L1 antibody monotherapy. Mice administrated with combination treatment of oAd-SA and PD-L1 antibody displayed no gain on both tumor suppression and survival compared to ones receiving PD-L1 antibody alone ( Figures 6E, F ). In 4T1 model, despite combination of two agents showed superior tumor inhibition capability compared with oAd-SA or PD-L1 monotherapy, no improvement was achieved on animal survival. ( Figures 6G, H ).