Renca (ATCC® CRL-2947™) was purchased and sub-cultured at 37°C, 5% CO₂. The cell line was maintained in a RPMI-1640 complete medium containing 10% FBS, sodium pyruvate, non-essential amino acid, l-glutamine, and penicillin streptomycin. The complete medium used in vitro culture with lymphocytes was supplemented with 50 µM 2-mercaptoethanol. Trypsin-EDTA solution was utilized to detach Renca cells following ATCC®‘s cell passage protocol. Renca cells from passage 3 to 14 were used in the present study.

Eight-to twelve-week-old female BALB/c mice were purchased from Hilltop laboratory Animals (Scottdale, PA) and housed in the Duquesne University Animal Care Facility. Animals were handled in accordance with protocols approved by the Duquesne University Institutional Animal Care and Use Committee. Hairs on the upper right abdomen were shaved prior to cell inoculation and injections. A P0 BALB/c mouse was inoculated with 2 × 10⁶ Renca cells that were in the proliferation state. The primary tumor was collected 15–17 days after inoculation and processed for further in vivo inoculation (see Tumor Dissociation method in the Supplementary Data Sheet S1). Processed tumor cells (1–2 × 10⁶) were suspended in HBSS and inoculated subcutaneously (s.c.) to an average set of 10 BALB/c mice.

The first dose of treatment or control was injected peritumorally (s.c.) on day 3 after inoculation, following by two more doses on day 6 and day 10. Full treatment components included 0.2 mg aPD1 (10 mg/kg equivalent), 0.2 mg ADA (5 mg/ml), and 0.117 mg Z15_EAK (5 mg/ml). Tumors (width and length) were measured every 2–4 days with an electronic caliper. Tumors and inguinal lymph node were harvested 2 days after the last dose. Specimens were weighed immediately after collection (Metter Toledo ME54E balance). End-point tumors were placed immediately in RNAlater solution and stored at -20°C until PCR analysis. Animals with tumors showing signs of infection were euthanized and excluded from the analysis.

The inguinal lymph node is responsible for draining the site of injection. dLN were stored in RPMI complete media, on ice, and weighed immediately upon collection. dLN were crushed against a 40 µm strainer using a sterile syringe plunger. The cell lysate was then centrifuged and resuspended for cell counting. The number of cells were equilibrated across the samples for flow cytometry staining and ex vivo cell culture plating.

dLN cell staining was performed in low-retention microcentrifuge tubes and with reagents provided in the mouse Regulatory T Cell Staining Kit (eBioscience™). Anti-mouse CD16/CD32 was first used to block non-specific Fc binding. Following procedure described in the kit protocol, the cells were stained with 0.125 µg anti-mouse CD4 FITC (RM-5) and 0.5 µg anti-mouse/rat FoxP3 PE (FJK-16s). Samples in the study were analyzed with the Attune NxT flow cytometer (Thermo Fisher). On average, 10⁵ events were run, with gates to exclude debris and doublet cells. AbC™ Total Antibody Compensation Bead Kit (Invitrogen) was employed to adjust color compensation.

Renca and dLN cells were seeded at the ratio of 1:10 in a 96-well plate with a final volume of 250 µL. Each well was supplemented with 20 ng/ml recombinant mouse IL-2 (R&D systems) to stimulate lymphocyte proliferation. After an overnight incubation at 37°C, 5% CO₂, the cell suspension was centrifuged, and the supernatant was analyzed with an IFNɣ mouse uncoated ELISA kit (Invitrogen). Each cell culture sample was tested as triplicates in the ELISA assay and detected with the TECAN infinite M1000 microplate reader (Männedorf, Switzerland). The concentration interpolation was obtained from an 8-point or 5-point standard curve (15–2,000 pg/ml).

RNA was isolated using a TRIzol Extraction kit. Whole tumors were homogenized in TRIzol solution using a Tumor Homogenization Kit (GentleMACS, Miltenyi Biotech). RNA extraction was performed according to manufacturer protocol. The RNA concentration and quality were analyzed using an RNA nano 6000 kit (Agilent Technologies). Each RNA sample was run in triplicate, and the results from the Agilent 2100 Bioanalyzer were averaged to obtain a final concentration and RNA Integrity Number (RIN). Samples with a RIN >8 were considered to have good quality RNA. 1 µg of RNA was reverse transcribed using a cDNA Synthesis Kit (SuperScript VILO IV Master Mix, ThermoFisher Scientific). The resultant cDNA was diluted to a final concentration of 5 ng/uL. Samples were analyzed on the Mx3000P Real Time Cycler, with a Taqman Real-Time PCR Master Mix (Applied Biosystems #4304437) and inventoried TaqMan probes from ThermoFisher. 5 µL (25 ng) of each cDNA sample were run singleplex for 100 cycles. ACTB was used as a normalizing gene.

Extracted RNA were quantitated using the Qubit™ RNA BR Assay Kit (Thermo Fisher Scientific) followed by an RNA quality check using the Fragment Analyzer (AATI). Sequencing was performed using NovaSeq 6000 platform (Illumina) to an average of 40M 101PE reads, on the NovaSeq SP-200 flowcell. Gene counts were run through edgeR’s calcNormFactors function, specifying the Trimmed Mean of M-values (TMM) algorithm as a means of normalization. Differential gene expression (DGE) analysis was performed with responder status, treatment, cohort, and sequencing batch listed as covariates in a generalized linear model evaluated by a quasi-likelihood F-test in edgeR. p-Values were adjusted using the Benjamini–Hochberg procedure to control false discovery rate (FDR). Genes were considered differentially expressed if they met the following criteria:|log₂| > 1 and adjusted p-value < 0.05. More details on the processing of raw data and analysis methodology can be found in the Supplementary Data Sheet S1.

Statistical analysis was performed in GraphPad Prism 8.0. Data plots shown represent mean and standard mean error. Nonparametric, unpaired t-test was conducted for comparison between two groups. Experiments with three tested groups were analyzed using an ordinary one-way ANOVA followed by Dunnett’s multiple comparison tests. Significant values were computed based on two-tailed assumption and marked with asterisks and represented the following: nonsignificant (ns) p ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. A total of 60 BALB/c mice were involved in the analysis. Mice failed to develop palpable tumors after inoculation within the time frame of the experiment were excluded.

We hypothesized that locoregional delivery of aPD1 would lead to the induction of a Th1 T cell response in Renca tumors and in tumor draining lymph nodes (dLN). Hydrogels have been used previously to concentrate cytokines and antibodies in tumors to enhance antitumor efficacy (Lv et al., 2017; Chao et al., 2020). In our prior work, we shown that IgG tethered with the SAP EAK could accumulate in epithelial tumors for at least 7 days (Wen et al., 2013). Repeated injections of Z15_EAK and IgG did not elicit acute inflammation or toxicities in mice (Pham et al., 2019). Further, perfusion of IgG complexed with Z15_EAK over Renca cells in vitro led to significant accumulation of the antibody on the monolayer (Supplementary Figure S1).

The effects of aPD1 antibody (IgG1) and ADA admixed with Z15_EAK (aPD1/ADA gel) on Renca tumors were investigated in Balb/c mice (Figure 1). The tumors used in the experiments were passaged in vivo such that each inoculum contained a single cell suspension recovered from a Renca tumor. The in vivo passage served to increase tumorigenic properties (Thirunarayanan et al., 2007; Lacoste et al., 2017). Each mouse was injected subcutaneously with 1.5–2 × 10⁶ in vivo passaged Renca cells and aPD1/ADA gel was injected subcutaneously adjacent to the inoculation site. The rationale for injecting the gel near the inoculating site (“peritumoral”) is that aPD1 and ADA would slowly diffuse into the tumors and target T cells trafficking to and from dLN (Meng et al., 2020). Altogether, six cohorts of Balb/c mice were inoculated and treated to allow comparison between treatments: aPD1/ADA gel against saline, aPD1/ADA gel against aPD1 gel, and aPD1/ADA gel against aPD1/ADA formulated in saline (aPD1/ADA) (Supplementary Table S1). Each cohort of mice included two to three groups for practical reasons; all the mice in a given cohort received the same cells processed from the same in vivo passaged tumor to minimize heterogeneity. For this reason, we believe it is more valid to compare the treatment groups within cohorts. The outcomes of the tumor and lymph nodes analyses were used to stratified into “immune-stimulatory” (IST) or “immune-suppressed” (ISU) as described in the Supplementary Material.

The formulations were evaluated first for their ability to alter tumor growth kinetics. Tumor size was monitored in mice inoculated with Renca cells isolated from in vivo passage. Beginning on day three after tumor inoculation, gels containing 0.2 mg of aPD1 and 0.2 mg of ADA (aPD1/ADA gel) were injected subcutaneously around the tumor inoculation site for three doses, with each given 3 days apart (Figure 1). The repeated injections did not induce overt, acute toxicities, and the mice maintained their body weights throughout the experimental periods (Supplementary Figure S2). Given that mice were inoculated with cells harvested from different in vivo passaged tumor cells, experiments were conducted in cohorts of mice inoculated with cells isolated form the same passaged tumors in which consistent engraftment could be assumed; each cohort contained mice divided into two to three treatment groups (n = 5 mice each) and independently analyzed. In the experiment designated B7, aPD1/ADA gel-treated mice exhibited close to a 2-day delay in emergent of palpable tumors compared to aPD1 gel-treated animals (Figure 2A). In experiment B3, tumors treated with aPD1/ADA gel were smaller from day 6 and onward compared to those treated with saline (Figure 2B). In experiment B8 where mice received either aPD1/ADA gel or aPD1/ADA formulated in saline, the tumors in the former group were significantly smaller toward the end of the monitoring period (Figure 2C). Collectively, these results indicate that ADA and Z15_EAK augmented aPD1 in delaying the growth of Renca tumors.

The nature of T cells infiltrated into the tumors were characterized using qRT-qPCR. Higher expressions of CD8α and IFNɣ were observed in tumors treated with aPD1/ADA gel compared to those treated with saline among the 16 tumors collected from experiments B3 and B9 (Figures 3A,B). The relative expression of IFNɣ to FoxP3 increased significantly (Figure 3C), but no significant difference was observed in the expression of IL-17A or IL-12a (Supplementary Figure S3A; RNA quality in the third cohort B6, which included aPD1/ADA gel vs. saline treated mice, was exceptionally low therefore the samples were not included in the analysis). Contributions of ADA in the formulation were examined in three cohorts of mice treated with either aPD1/ADA gel or aPD1 gel. In two of the three cohorts, aPD1/ADA gel-treated mice exhibited a trend of increased in IFNɣ expression compared to aPD1 gel (Figure 3D) and a greater shift from FoxP3 to IFNɣ was observed (Figure 3E). In the third cohort, no effect was seen with the addition of ADA (Supplementary Figure S3B). In cohort B7, ADA significantly enhanced the expression of IL-17A (Figure 3F). No difference in CD8α expression was observed between treatment and controls in cohorts treated with aPD1/ADA gel, aPD1 gel, or aPD1/ADA (Supplementary Figures S3C,D). The results suggest potential immune-activating functions of ADA, but additional studies are necessary to establish decisive impact.

We next analyzed the T cell subsets developed in the tumor draining lymph nodes in response to the formulations. Inguinal dLN recovered from mice received aPD1/ADA gel weighed five times more than the mice received saline, suggesting an expansion of immune cells (Figure 4A). In delineating the T cell subsets in these dLN, we found lower frequencies of Tregs in the nodes treated with aPD1/ADA gel compared to those received saline (Figure 4B), suggesting a reversal of immune suppression. In cohort B6, higher levels of IFNɣ were detected in ex vivo cultures of dLN isolated from aPD1/ADA gel-treated mice than those in mice received saline (Figure 4C). The results suggest that peritumoral injection of aPD1/ADA gel skewed the T cell response in the dLN towards a tumor-rejecting phenotype.

We also compared the effects of aPD1/ADA gel and aPD1 gel on dLN T cells. dLN from mice treated with aPD1/ADA gel were larger than those received aPD1 gel (Figure 4A), indicating that ADA altered the local immune milieu. Both treatments generated approximately the same levels of Tregs and INFɣ in ex vivo cultured dLN (Figures 4D,E). The impact of formulating aPD1 and ADA with Z15_EAK on T cell response in the dLN was inconclusive; it could be that peak concentrations in dLN rather than overall exposure of aPD1 at the dose given (0.2 mg) is the decisive factor, and that the depot effect enhances long-term antitumor immunity, which was not measured in the current study. Another possibility is that a stronger gel than Z15_EAK was necessary to enhance retention. This can be accomplished by intermixing Z15_EAK with EAK using a co-assembly strategy (Wen et al., 2014). Nevertheless, while the intended effects of ADA and Z15_EAK on the T cell response were not detected, the enzyme and the gelation biomaterials did delay tumor growth. Taken together, the T cell subsets developed in tumors and dLN and the observed tumor growth point to at least two polarized immune responses, referred in the following narrative as “immune-stimulatory” (IST) and “immune-suppressed” (ISU).

RNA-seq was used to delineate the transcriptomic features differentiating tumors responded to aPD1/ADA gel and those did not, and tumors which received the formulation versus saline. Tumors were classified as IST or ISU based on the abundance of CD4+FoxP3+ Tregs, CD8⁺ T cells, and IFNɣ production, as classification metrics, regardless of treatment (Supplementary Figure S4). The RNA-Seq analyses, reported here in dendrograms and plots, revealed unique gene expression patterns from which responding tumors could be defined based on gene expression patterns. The purpose of the analysis was therefore to validate the internal consistencies of the classifications and explore unique signatures of localized aPD1 therapies in the Renca model.