Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 May 13.
Published in final edited form as: Nat Biomed Eng. 2021 Apr 26;5(9):1038–1047. doi: 10.1038/s41551-021-00712-1

Inhibition of post-surgery tumour recurrence via a hydrogel releasing CAR-T cells and anti-PDL1-conjugated platelets

Quanyin Hu 1,2,3,4,9, Hongjun Li 1,2,5,6,9, Edikan Archibong 3,7, Qian Chen 1,2,3, Huitong Ruan 1,2, Sarah Ahn 7, Elena Dukhovlinova 7, Yang Kang 1,2, Di Wen 1,2, Gianpietro Dotti 7,✉, Zhen Gu 1,2,3,5,6,8,✉
PMCID: PMC9102991  NIHMSID: NIHMS1799463  PMID: 33903744

Abstract

The immunosuppressive microenvironment of solid tumours reduces the antitumour activity of chimeric antigen receptor T cells (CAR-T cells). Here, we show that the release—through the implantation of a hyaluronic acid hydrogel—of CAR-T cells targeting the human chondroitin sulfate proteoglycan 4, polymer nanoparticles encapsulating the cytokine interleukin-15 and platelets conjugated with the checkpoint inhibitor programmed death-ligand 1 into the tumour cavity of mice with a resected subcutaneous melanoma tumour inhibits the local recurrence of the tumour as well as the growth of distant tumours, through the abscopal effect. The hydrogel, which functions as a reservoir, facilitates the enhanced distribution of the CAR-T cells within the surgical bed, and the inflammatory microenvironment triggers platelet activation and the subsequent release of platelet-derived microparticles. The post-surgery local delivery of combination immunotherapy through a biocompatible hydrogel reservoir could represent a translational route for preventing the recurrence of cancers with resectable tumours.

Despite multiple treatment modalities involving surgery, radiotherapy and chemotherapy, tumour relapse frequently occurs in patients with resectable solid tumours1,2. One reason accounting for tumour recurrence is the post-surgery inflammation that can trigger tumour growth and metastasis3–5. Immunotherapy with immune-checkpoint blockade reactivates T cells at the tumour site by blocking the immune-checkpoint pathways, and may prevent tumour recurrence after surgery6–8. However, systemic administration of checkpoint inhibitors promotes sustained clinical responses in less than 20% of patients with immunogenic tumours9,10. Furthermore, checkpoint blockade is ineffective in tumours that are characterized by a low burden of somatic mutations that generates neoantigens and endogenous T-cell responses11,12. Additionally, side effects including autoimmune diseases secondary to checkpoint blockade remain a concern13,14.

An alternative way to provide tumour-specific T cells relies on the adoptive transfer of engineered T cells15,16. Adoptive transfer of T cells expressing a CAR17,18 has been shown to be particularly effective in patients with B-cell derived malignancies19. By contrast, the application of CAR-T cells in solid tumours remains challenging20,21, at least in part because the tumour microenvironment in solid tumours is highly immunosuppressive and induces the exhaustion of CAR-T cells22,23. Thus, the combination of CAR-T cells with immune checkpoint blockade could be critical to enhance CAR-T-cell activity in solid tumours.

In this Article, we developed a biodegradable hydrogel reservoir that encapsulates CAR-T cells targeting the human chondroitin sulfate proteoglycan 4 (CSPG4; CSPG4.CAR), for implantation into the tumour-resection cavity in a melanoma tumour-bearing mouse model (Fig. 1a). We selected CSPG4 as a target antigen because it is highly expressed in human melanoma, yet its expression is restricted in normal cells24,25. The implantable gel acts as a reservoir to concentrate and gradually release CAR-T cells after surgical resection of the tumour. To maintain the activity and proliferation ability of CAR-T cells, the cytokine IL-15 was also encapsulated in nanoparticles loaded in the hydrogel26. Furthermore, human platelets conjugated with the anti-PDL1 blocking antibody (aPDL1) were also harboured in the hydrogel to block the PD1/PDL1 pathway. The inflammation that occurs post-surgery triggers the activation of platelets27,28, leading to the formation of platelet-derived microparticles (PMPs) that release aPDL1 antibodies that bind to the tumour cells and block PDL1. Taken together, the developed hydrogel creates a favourable milieu in which CAR-T cells eradicate residual tumour cells post-surgery and prevent tumour recurrence.

Fig. 1 |. Characterization of the engineered hydrogel-based cell delivery.

Fig. 1 |

Open in a new tab

a, Schematic of the tumour resection model and implantation of the engineered HA hydrogel. Platelets activated during the wound healing process after surgery release aPDL1 in the form of PMP-aPDL1. MHC, major histocompatibility complex; TCR, T-cell receptor. b, Confocal imaging of CAR-T cells and P–aPDL1 encapsulated in the hydrogel (CAR-T-P–aPDL1@gel). CAR-T cells and platelets were labelled with CellTracker Green and rhodamine B, respectively. Hoechst 33342 was used to stain the nuclei. Cell number, 2 × 106 CAR-T cells; platelet number, 1 × 107. This experiment was performed three times with similar results. Scale bar, 100 μm. c, Cryo-scanning electron microscopy imaging of CAR-T-P–aPDL1@gel. Cell number, 2 × 106 CAR-T cells; platelet number, 1 × 107. Scale bars, 10 μm. This experiment was performed three times with similar results. d, Confocal imaging of the live/dead assay of CAR-T cells released from the hydrogel. Live cells and dead cells were labelled with green fluorescence and red fluorescence, respectively. Scale bar, 100 μm. This experiment was performed three times with similar results. e,f, Human IFNγ (e) and IL-2 (f) released by CAR-T cells encapsulated in the hydrogel. Data are mean ± s.d. n = 3 biologically independent samples. Statistical analysis was performed using unpaired two-tailed Student t-tests. g, In vivo degradation of the HA hydrogel labelled with Cy5.5 at weeks 0 and 8. n = 3 mice per group.

Results

Engineering the hydrogel for cell delivery.

The aPDL1 antibody was covalently conjugated to the cell surface of human platelets (designated P–aPDL1). The binding of aPDL1 to platelets was visualized by colocalization of the fluorescein-labelled aPDL1 antibody and rhodamine B-labelled platelets (Supplementary Fig. 1) and confirmed by flow cytometry (Supplementary Fig. 2). Conjugation of aPDL1 did not affect platelet functionality, as evidenced by the conserved collagen-binding activity of aPDL1-conjuated platelets (Supplementary Fig. 3). The release of aPDL1 was accelerated after platelet activation by thrombin and the formation of PMPs29 (Supplementary Fig. 4). The interaction between activated platelets and melanoma cells was investigated using a Transwell assay (1 μm micropores) that allows free transportation of PMPs, but not intact platelets. After incubation with thrombin30, wheat germ agglutinin (WGA)-Alexa Fluor 594-labelled platelets in the top chamber were activated and generated PMPs that migrated to the bottom chamber in which WGA-Alexa Fluor 488-labelled WM115 melanoma cells were seeded. Colocalization of red fluorescence and green fluorescence demonstrated the binding of PMP-aPDL1 and melanoma cells and relocation of aPDL1 antibodies on the surface of melanoma cells (Supplementary Fig. 5). By contrast, negligible binding of aPDL1 antibodies to the tumour cells was observed in the control platelets.

The hydrogel was generated using the acrylate-group-modified hyaluronic acid (HA) that can be cross-linked by ultraviolet irradiation with cross-linker and photo initiator31. The HA hydrogel was lyophilized for further storage and use (Supplementary Fig. 6). CAR-T cells targeting the CSPG4 antigen and the P–aPDL1 were loaded into the hydrogel (Supplementary Fig. 7). To support the viability and proliferation of CAR-T cells in the hydrogel, the cytokine IL-15 was also encapsulated using PLGA nanoparticles (IL-15 NPs) that were prepared in water-in-oil-in-water emulsion32. IL-15 NPs displayed a size of around 180 nm and a spherical morphology as assessed using transmission electronic microscopy (TEM) imaging (Supplementary Fig. 8a,b), and released IL-15 in a sustained manner with more than 60% release at 120 h (Supplementary Fig. 8c). No substantial leakage of either CAR-T cells or IL-15 NPs loaded into hydrogel was found, as evidenced by 97.3% and 94.6% loading efficacy of CAR-T cells and IL-15 NP, respectively (Supplementary Fig. 9). The distribution of P–aPDL1 and CAR-T cells in the hydrogel was demonstrated by confocal microscopy and cryo-scanning electron microscopy (Fig. 1b,c). The hydrogel formulation did not cause toxic effects on CAR-T cells (Fig. 1d). Furthermore, the HA hydrogel did not induce non-specific activation of CAR-T cells, as evidenced by negligible production of IFNγ and IL-2 in the absence of antigen stimulation (Fig. 1e,f). Taken together, these results demonstrate that the HA hydrogel acts as a reservoir for both CAR-T cells and platelets carrying the aPDL1, and can integrate molecules delivered by NPs. We further studied the degradation of the HA hydrogel in vivo using the In Vivo Imaging System (IVIS). As shown in Fig. 1g and Supplementary Fig. 10, the HA hydrogel was gradually degraded by hyaluronidase (Supplementary Fig. 11) after implantation and minimal signal was detected after 8 weeks, demonstrating the optimal biocompatibility and degradability of the HA hydrogel.

In vitro CAR-T-cell antitumour effects.

We investigated the release profile of both CAR-T cells and platelets encapsulated within the hydrogel. As shown in Fig. 2a, both CAR-T cells and platelets were released in a sustained manner. However, platelets showed faster release; at 96 h, nearly all platelets were released, as compared to 50% release of CAR-T cells. The faster release of platelets is probably due to their pronounced motility within the hydrogel as demonstrated by confocal microscopy (Fig. 2b). The average quantitative migration distance of platelets was fourfold longer than that of CAR-T cells (Fig. 2c). The hydrogel also supported the gradual release of CAR-T cells in vivo because 30% of CAR-T cells were released within 1 week (Supplementary Fig. 12).

Fig. 2 |. CAR-T cells encapsulated in the hydrogel target WM115 melanoma cells in vitro.

Fig. 2 |

Open in a new tab

a, The release profile of CAR-T cells and platelets from the hydrogel (CAR-T-P–aPDL1@gel) in vitro. Cell number, 1 × 107 CAR-T cells; platelet number, 1 × 107. Data are mean ± s.d. n = 3 biologically independent samples. b, Movement track of platelets and CAR-T cells in the hydrogel as imaged using confocal microscopy. Each colour represents the movement track of one cell. The movement of CAR-T cells and platelets was monitored every 30s for 30min. Cell number, 2 × 106 CAR-T cells; platelet number, 1 × 107. Scale bar, 100 μm. This experiment was performed three times with similar results. c, The migration distance of platelets and CAR-T cells in the hydrogel after 30 min. Data are mean ± s.d. n = 400 independent cells. Statistical analysis was performed using an unpaired two-tailed Student t-test; ***P < 0.001. d–f, Flow cytometry plots (d) and quantification of tumour cells (e) and CAR-T cells (f) in a coculture of tumour cells and T cells for 72 h. T@gel, control T cells in the hydrogel; T-platelet@gel, control T cells and platelets in the hydrogel; CAR-T@gel, CAR-T cells in the hydrogel; CAR-T-P–apDL1@gel, CAR-T cells and aPDL1-loaded platelets in the hydrogel. Data are mean ± s.d. n = 3. Cell number, 3.3 × 105 CAR-T cells and 1 × 106 WM115 cells; platelet number, 1 × 107 platelets. The amount of aPDL1 was 1 μg. g,h, Human IFNγ (g) and IL-2 (h) released by CAR-T cells during the cocultures shown in d. Data are mean ± s.d. n = 3 biologically independent samples. Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Tukey’s honest significant difference (HSD) post hoc test. *P = 0.0426 (g); *P = 0.0229 (h).

The antitumour activity of CAR-T cells was investigated in a coculture assay33. The hydrogel loaded with P–aPDL1 and CAR-T cells was placed in the top cell strainer (pore size, 40 μm) and GFP-labelled WM115 tumour cells were seeded in the bottom chamber. After 72 h, the percentage of WM115 and T cells was not significantly changed in the hydrogels loaded with either control T cells or control T cells and platelets (Fig. 2d). By contrast, the percentage of WM115 cells was decreased (12.7 ± 5.8%) and the percentage of T cells increased (76.8 ± 7.7%) when CAR-T cells were plated in the hydrogel (Fig. 2d–f). The percentage of WM115 cells further decreased when CAR-T cells were combined with P–aPDL1 (Fig. 2d–f). CAR-T cells loaded in the hydrogel released IFNγ and IL-2, and the cytokine levels were further increased in the presence of P–aPDL1 (Fig. 2g,h). CAR-T cells also proliferated in response to tumour cells (Supplementary Fig. 13). As shown in Supplementary Fig. 14, WM115 cells that were exposed to CAR-T cells or CAR-T cells and P–aPDL1 showed a substantial increase in PDL1 expression compared with the other control groups. Collectively, these data demonstrate that the combination of CAR-T cells and P–aPDL1 exhibits higher T-cell activation and cytokine release in vitro.

In vivo antitumour activity of the CAR-T-P–aPDL1@gel.

To validate the in vivo antitumour effects of the engineered hydrogel, NOD-scid Il2rgnull (NSG) mice were inoculated subcutaneously with the luciferase-labelled WM115 tumour cells. After engraftment, WM115 tumour cells retained the expression of the target antigen CSPG4 and displayed PDL1 expression as detected using confocal microscopy (Supplementary Fig. 15). PDL1 expression by tumour cells in vivo was also confirmed by flow cytometry (Supplementary Fig. 16). When the size of the tumours reached ~150 mm3, tumour masses were partially removed and the sizes of remaining tumours did not show a significant difference on the basis of the bioluminescence intensity (Supplementary Fig. 17). The mice were next treated with saline, non-transduced T cells and platelets coencapsulated in the hydrogel (T-platelet@gel), P–aPDL1-loaded gel (P–aPDL1@gel), CAR-T cells and isotype-antibody-conjugated platelets encapsulated in the hydrogel (CAR-T-P-isotype-antibody@gel), free CAR-T cells, free CAR-T cells and P–aPDL1, CAR-T@gel, CAR-T-loaded gel and P–aPDL1 (CAR-T@gel + P–aPDL1), CAR-T-loaded gel and systemic injection of aPDL1 (CAR-T@gel + systemic aPDL1), and CAR-T-P–aPDL1@gel. Tumour recurrence was monitored by measuring tumour bioluminescence signals. As shown in Fig. 3a,b and Supplementary Fig. 18, the saline and T-platelet@gel groups showed negligible therapeutic effects. Whereas mice that were treated with free CAR-T cells with or without P–aPDL1 achieved only moderate anti-tumour effects, mice treated with CAR-T@gel + P–aPDL1 showed the most prominent protection from tumour recurrence. CAR-T-P–aPDL1@gel displayed higher anti-tumour activity even compared with CAR-T@gel + systemic aPDL1. Quantitative bioluminescence intensity of the tumours treated with CAR-T-P–apDL1@gel at week 3 was 6.4-fold lower compared with the CAR-T@gel + P–aPDL1 group and greater than 60-fold lower than other treatment groups (Fig. 3c). Direct measurement of the tumour volume showed that mice that were treated with CAR-T-P–aPDL1@ gel at week 3 displayed significantly smaller tumours compared with the other treatment groups (Fig. 3d,e).

Fig. 3 |. CAR-T cells encapsulated in the hydrogel control WM115 melanoma growth in vivo.

Fig. 3 |

Open in a new tab

a, Representative tumour bioluminescence. P–aPDL1@gel, P–aPDL1 encapsulated in the hydrogel; CAR-T, CAR-T cells directly inoculated into the resection cavity; CAR-T + P–aPDL1, CAR-T cells and a PDL1 directly inoculated into the resection cavity; CAR-T@gel, CAR-T cells encapsulated in the hydrogel; CAR-T@gel + P–aPDL1, CAR-T cells encapsulated in the hydrogel and P–aPDL1; CAR-T-P–aPDL1@gel, CAR-T cells and P–aPDL1 co-encapsulated in the hydrogel. In all of the experimental groups, IL-15 NPs (IL-15, 1 μg) were included. Cell number, 2 × 106 CAR-T cells; platelet number, 1 × 107. The amounts of aPDL1 and IL-15 were both 1 μg. b, Region-of-interest analysis of tumour bioluminescence intensities. c, Comparison of tumour bioluminescence intensities at week 3 after treatment. Data are mean ± s.d. n = 6 mice per group. Statistical analysis was performed using one-way ANOVA followed by Tukey’s HSD post hoc test; **P = 0.0071. d, Summary of the tumour volume at week 3 after treatment. Data are mean ± s.d. n = 6 mice per group. Statistical analysis was performed using one-way ANOVA followed by Tukey’s HSD post hoc test; *P = 0.0486, ***P < 0.001. e, Representative tumours after 3 weeks. Groups:1, saline; 2, P–aPDL1@gel; 3, CAR-T; 4, CAR-T + P–aPDL1; 5, CAR-T@gel; 6, CAR-T@gel + P–aPDL1; and 7, CAR-T-P–aPDL1. Scale bar, 1 cm.

In vivo CAR-T proliferation and abscopal treatment efficacy.

To investigate the underlying mechanism of the superior anti-tumour efficacy of CAR-T-P–aPDL1, CAR-T cells were labelled with luciferase to track their persistence in vivo (Supplementary Fig. 7). As shown in the Fig. 4a, the bioluminescence of free CAR-T cells rapidly disappeared, whereas the bioluminescence of CAR-T cells loaded in the hydrogel persisted as demonstrated by an increase of greater than threefold compared with the other treatments (Fig. 4b). Furthermore, we monitored the long-term persistence of CAR-T cells using in vivo imaging. CAR-T-P–aPDL1 showed a detectable signal up to week 4, whereas all of the other treatment groups displayed negligible signals (Supplementary Fig. 19). As shown in Fig. 4c, T-cell bioluminescence correlated with a higher detection of tumour-infiltrating CAR-T cells. Furthermore, PMPs were detected within the tumour, indicating the undergoing activation of transferred platelets due to the local inflammatory environment (Fig. 4c). Cytokine levels within the tumour were also examined to further demonstrate that CAR-T cells released IFNγ and IL-2 locally after encountering antigens (Fig. 4d,e). Moreover, IL-15 was detectable within the tumour throughout the time-course treatment and the levels of which were significantly higher compared with in untreated tumours (Supplementary Fig. 20). To investigate whether CAR-T cells that are locally delivered at the tumour site through CAR-T-P–aPDL1@gel cause systemic tumour protection, we established a double-tumour model in which tumour cells were implanted into both flanks of NSG mice. When the tumour size reached a size of around 100 mm3, surgery was performed on the tumour located on right side and hydrogel was implanted. As shown in Fig. 5a and Supplementary Fig. 21, CAR-T-P–aPDL1@ gel displayed an abscopal effect, causing inhibition of the tumour growth at the contralateral site (Fig. 5b,c). To explain the underlying mechanism of the abscopal effects, we first analysed the chemokine gradients within the tumour before and after surgical resection. The majority of chemokines included in the chemokine array used showed significantly higher levels after surgery (Supplementary Fig. 22). Moreover, CAR-T cells delivered through CAR-T-P–aPDL1@gel in the resection cavity reached the blood stream because CAR-T cells were also detectable in the peripheral blood (Supplementary Fig. 23a,b) and at the distant tumour site (Supplementary Fig. 23c–e). Collectively, the combination of a favourable chemokine milieu within the tumour created by the surgical resection and the recirculation of CAR-T cells support the control of tumour growth at distant sites by locally delivered CAR-T cells.

Fig. 4 |. CAR-T cells encapsulated in the hydrogel persist and expand in vivo.

Fig. 4 |

Open in a new tab

a, Representative T-cell bioluminescence. CAR-T, CAR-T cells directly inoculated into the resection cavity; CAR-T + P–aPDL1,CAR-Tcells and P–aPDL1 directly inoculated into the resection cavity; CAR-T @gel, CAR-T cells encapsulated in the hydrogel; CAR-T@gel + P–aPDL1, CAR-T cells encapsulated in the hydrogel and P–aPDL1; CAR-T-P–aPDL1@gel, CAR-T cells and P–apDL1 co-encapsulated in the hydrogel. In all of the experimental groups, IL-15 NPs (IL-15, 1 μg) were included. Cell number, 2 × 106 CAR-T cells; platelet number, 1 × 107. The amounts of aPDL1 and IL-15 were both 1 μg. b, Region-of-interest analysis of T-cell bioluminescence. Data are mean ± s.d. n = 6 mice per group. Statistical analysis was performed using one-way ANOVA followed by Tukey’s HSD post hoc test; **P = 0.0064 (comparison between CAR-T@gel + P–aPDL1 and CAR-T-P–aPDL1@gel),**P = 0.0013 (comparison between CAR-T@gel and CAR-T-P–aPDL1@gel). c, Confocal imaging of CAR-T-P–aPDL1@gel 72h after treatment. Left: distribution of CAR-T cells and P–aPDL1. Right: enlarged image to show the activation of platelets. Arrows represent the PMPs. CAR-T cells and platelets were labelled with CellTracker Orange and WGA-Alexa Fluor 488, respectively. Cell number, 2 × 106 CAR-T cells; platelet number, 1 × 107. The amounts of aPDL1 and IL-15 were both 1 μg. Scale bars, 100 μm (left) and 20 μm (right). This experiment was performed three times with similar results. d, Human IL-2 detected in vivo within the tumour at week 1 after treatment. Data are mean ± s.d. n = 6 biological independent samples. Statistical analysis was performed using one-way ANOVA followed by Tukey’s HSD post hoc test; **P = 0.0037 (comparison between CAR-T@gel + P–aPDL1 and CAR-T-P–aPDL1@gel), ***P = 0.0004 (comparison between CAR-T@gel and CAR-T-P–aPDL1@gel). e, Human IFNγ detected in vivo within the tumour at week 1 after treatment. Data are mean ± s.d. n = 6 biological independent samples. Statistical analysis was performed using one-way ANOVA followed by Tukey’s HSD post hoc test; *P = 0.0162 (comparison between CAR-T@gel+P–aPDL1 and CAR-T-P–aPDL1@gel), **P = 0.0023 (comparison between CAR-T@gel and CAR-T-P–aPDL1@gel).

Fig. 5 |. Engineered hydrogel cell delivery promotes abscopal antitumour effects.

Fig. 5 |

Open in a new tab

a, Representative tumour bioluminescence images. CAR-T+P–aPDL1, CAR-T cells and P–aPDL1 directly inoculated into the resection cavity; CAR-T-P–aPDL1@gel, CAR-T cells, platelets and P–aPDL1 co-encapsulated in the hydrogel. In all of the experimental groups, IL-15 NPs (IL-15, 1 μg) were included. Cell number, 2 × 106 CAR-Tcells; platelet number, 1 × 107. The amounts of aPDL1 and IL-15 were both 1 μg. b, Tumour growth measured on the left side after treatment of the primary tumour. Data are mean ± s.d. n = 6 mice per group. Statistical analysis was performed using two-way ANOVA; **P = 0.0023. c, Representative tumours at day 18. Scale bar, 1 cm.

Discussion

We have shown that a biocompatible hydrogel can intratumourally deliver CAR-T cells and enhance T-cell persistence compared with the free injection of CAR-T cells. We have also shown that platelets can act as bioresponsive cells and intratumourally release aPDL1 antibodies upon activation. This in turn sustains CAR-T-cell antitumour activity, protecting the cells from exhaustion.

Resection of the tumour mass is frequently only partially effective, and adjuvant chemotherapy and radiotherapy are used to attempt to eradicate the tumour. In situ delivery may be a more specific and effective approach to prevent tumour recurrence. Here we propose that in situ drug delivery can be adapted to cell therapies and, in particular, to adoptive T-cell therapies. Preclinical and clinical studies have shown the feasibility of the intratumoural delivery of CAR-T cells with the intent to concentrate T cells at the tumour site, overcoming the challenge of T-cell biodistribution within the tumour and the potential side effects encountered when CAR-T cells are targeting antigens shared by normal tissues34,35. Biomaterials can play a critical role, enabling the encapsulation of cells that maintain their viability and functional characteristics36,37. We found that the HA-based hydrogel preserves fully functional CAR-T cells and also enables the encapsulation of nanoparticles carrying growth factors for CAR-T cells, representing another advantageous layer to enhance their antitumour activity.

In addition to highly functional effector T cells concentrated at the tumour site, a curative immune response should also reverse the immunosuppressive tumour microenvironment. We have found that the HA-based hydrogel is a highly flexible delivery system that enables the encapsulation of more than one functional cell type. Specifically, we have developed a strategy that involves the local delivery of platelets in combination with CAR-T cells. Platelets can be decorated with antibodies, enabling, for example, the local biodistribution of a checkpoint inhibitor. Furthermore, platelets are responsive to inflammation secondary to the wound healing process and form microplatelets that promote effective biodistribution of the loaded checkpoint inhibitors to tumour cells. Finally, activated platelets are a source of ligands, such as CD40L, and of chemokines that boost CAR-T cells immunity and recruit other immune cells, further amplifying the antitumour effects38–40. Collectively, these findings substantiated the rationale of combining platelets and CAR-T cells as antitumour agents.

We demonstrated that CAR-T cells and platelets could be combined and delivered making use of an optimized HA hydrogel formulation that can be implanted directly into the tumour bed after surgical resection. The hydrogel supports T-cell viability by introducing nanoparticles releasing IL-15, while platelets delivering aPDL1 antibodies synergize in protecting the T cells from exhaustion. As a result, the formulated hydrogel remarkably controlled local tumour recurrence and triggered an abscopal effect inhibiting distant tumour growth. It remains to be demonstrated whether the proposed cell-delivery strategy can be applied to other human cancers that can be resected, such as pancreatic cancer and breast cancer. It also remains to be investigated whether host platelets can affect the functional activity of platelets encapsulated within the hydrogel, but in our experimental models we did not observe any negative impact as encapsulated platelets effectively deliver the aPDL1 antibodies.

To facilitate the translational capability of the delivery approach, the duration and controllability of T-cell release from the hydrogel could be further improved by increasing the cross-linking density of the hydrogel network and by incorporating an internal or external stimulus. Additional studies will be needed to demonstrate any beneficial effects of the proposed strategy in metastatic and inoperable tumour models. Furthermore, the hydrogel reservoir can be manipulated to incorporate other therapeutic bioparticulates and to form a local ‘immune cell factory’ for the elimination of residual tumour cells.

Methods

Cell lines and cells.

The human melanoma WM115 cells were tagged with the fusion protein eGFP–firefly-luciferase. Human T cells were engineered with CSPG4.CAR as previously described41. Here we used a CSPG4-specific CAR that encodes the CD28 endodomains with enhanced functionality. WM115 cells were maintained in Dulbecco’s modified Eagle’s medium (Gibco; Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen), 100 U ml−1 penicillin (Invitrogen) and 100 U ml−1 streptomycin (Invitrogen). CAR-T cells were cultured in complete medium containing 45% RPMI 1640 and 45% Click’s medium (Irvine Scientific) with 10% FBS (HyClone), 2 mmoll−1 GlutaMAX, human recombinant IL-7 (5 ng ml−1, Pepro Tech) and human recombinant IL-15 (10 ng ml−1, Pepro Tech). Cells were cultured in an incubator (Thermo Fisher Scientific) at 37 °C under an atmosphere of 5% CO2 and 90% relative humidity. The cells were subcultivated approximately every 2–3 d at 80% confluence at a split ratio of 1:3.

Antibodies.

The antibodies used in this study were as follows: APC-anti-human CD3 (BioLegend, HIT3a, 300312), BV241-anti-human PDL1 (BioLegend, 29E.2A3, 329714), PE-anti-human PDL1 (BioLegend, 29E.2A3, 329706), PE-anti-human CD4 (BioLegend, A161A1, 357404), FITC-anti-human CD8 (BioLegend, SK1, 344704), FITC-anti-human CD45 (BioLegend, 2D1, 368508), PE-anti-CSPG4 (MCSP), human (Miltenyi Biotech, EP-1, 130–091-225), FITC-anti-melanoma (MSCP), human (Miltenyi Biotech, EP-1, 130–098-794), PE-anti-human IgG (BioLegend, M1310G05, 410708) and PE-anti-mouse IgG (BioLegend, Poly4053, 405307). All antibody dilutions were performed according to the manufacture’s guidance. All flow cytometry data were analysed using Cytoexpert and FlowJo software.

Fabrication and characterization of IL-15 NPs.

PLGA (10 mg) was dissolved in 2 ml CH2Cl2 to prepare the oil phase and 0.4 ml 1% w/v polyvinyl alcohol (PVA) solution containing 50 μg ml−1 IL-15 was prepared for the water phase. The CH2Cl2 solution was slowly added to the PVA solution followed by sonication. Thereafter, 10 ml 2% (w/v) PVA solution was added to the previous emulsion mixture and sonicated to obtain double-emulsion solution. IL-15-NP solution was applied to the rotary evaporator for dichloromethane evaporation. Next, the emulsion was centrifuged to collect IL-15 NPs. The hydrodynamic size of IL-15 NPs was studied by Zetasizer (Nano ZS, Malvern). The morphology of IL-15 NPs was characterized using TEM (JEM-2000FX, Hitachi) after being stained with 1% (w/v) phosphotungstic acid. The amount of IL-15 loaded in the hydrogel was 1 μg.

To study in vitro the release of IL-15 from IL-15 NPs, 1 ml IL-15 NPs was incubated with phosphate-buffered saline (PBS) (pH 7.4) in a shaker (New Brunswick Scientific). At predetermined time points, the level of IL-15 released was determined using a specific enzyme-linked immunosorbent assay (ELISA) kit (Abcam).

Preparation of platelets loaded with the aPDL1 antibody.

Human platelet concentrate was purchased from Zen-Bio. The human platelet concentrate was centrifuged for 20 min at 100g and then at 800g for 20 min to collect the platelets. The platelet pellet was collected and resuspended in PBS containing 1 μM PGE1 for further use. Platelets were counted using a haemocytometer under a microscope. The platelet solution was centrifuged at 800g for 20 min and washed with PBS to remove the PGE1 for the platelet-activation study.

The conjugation of the aPDL1 to the surface of platelets was performed as previously described28. In brief, platelets were incubated with Traut’s reagent to generate thiol groups. The aPDL1 was reacted with sulfo-SMCC at a molar ratio of 1:1.2 for 2h at 4 °C. The excess SMCC linker was discarded by centrifugation in an ultrafiltration tube (molecular weight cut-off, 3 kDa). The SMCC–aPDL1 was added to the platelets and stirred for 2h at room temperature. aPDL1 was purified by repeatedly washing with PBS and centrifuging at 800g. To determine the amount of conjugated aPDL1 on the surface of platelets, aPDL1 was lysed with ultrasonication in 0.1% Triton-X100 buffer and the released aPDL1 was measured using a specific ELISA kit (Human IgG Total ELISA Kit, Abcam). The release profile of aPDL1 was investigated with addition of thrombin (0.5 U ml−1). In brief, 1 × 108 P–aPDL1 platelets were added with thrombin and maintained at 37 °C. At predetermined time points, the released aPDL1 in the supernatant was measured using an ELISA kit after centrifugation at 800g for 20 min.

To demonstrate the existence of the aPDL1 on the surface of platelets, aPDL1 was labelled with fluorescein isothiocyanate-NHS (FITC) before reacted with rhodamine-B-stained platelets. Next, the fluorophores-labelled P–aPDL1 was visualized using confocal microscopy. Furthermore, the binding of aPDL1 to platelets was also analysed by flow cytometry after staining with FITC-labelled anti-human IgG antibodies. The functionality of P–aPDL1 was examined using collagen-binding experiments. In brief, human collagen type IV (2 mg ml−1) was added to a confocal dish and maintained at 4 °C for overnight. Next, the confocal dish was washed with PBS and further blocked by 2% bovine serum albumin for 2h. The confocal dish without collagen pretreatment was set as a control. 1 × 106 P–aPDL1 platelets or naive platelets stained with wheat germ agglutinin Alexa Fluor 594 were added to the dishes. Then, 1 min later, the dishes were washed with PBS and analysed using confocal microscopy imaging.

Preparation of cell-loaded HA hydrogel.

The HA was modified with the methacrylic anhydride (MA) to form double bonds as previously described31. To form HA hydrogel, 400 μl HA solution (4% w/v) together with N,N-methylenebisacrylamide (MBA) (MBA:HA, 1:5, w/w) and a photo-initiator Irgacure 2959 (0.1%, w/v) were added into a 48-well plate. After radical polymerization using ultraviolet radiation for 25 min using a BlueWave 75 UV Curing Spot Lamp (DYMAX), the obtained hydrogel was frozen and processed using a lyophilizer for 48 h. To load the cells into the hydrogel, the lyophilized HA hydrogel was placed on ice and 1 × 106 CAR-T cells and 1 × 107 platelets conjugated with aPDL1 were seeded. Thereafter, the hydrogel was maintained on ice for 15–20 min to ensure sufficient inclusion of cells.

Characterization of the cells loaded into the hydrogel.

To visualize the distribution of cells in the hydrogel, CAR-T cells were labelled with Cell Tracker green and Hoechst 33342. Platelets were labelled with rhodamine B. CAR-T cells (2 – 106 cells) and platelets (1 – 107 cells) were seeded into the hydrogel, embedded with optimal cutting temperature (OCT) compound and immerged in liquid nitrogen. Slides obtained from frozen material were analysed by confocal microscope. Morphology was characterized by cryo-SEM (JEOL 7600F, Gatan Alto). Functionality and viability of CAR-T cells in the hydrogel were examined after degradation of the HA hydrogel. In brief, the hydrogel containing 1 × 107 CAR-T cells was placed into a cell strainer with 40 μm pore size and embedded in a six-well plate. After 48 h, IL-2 and IFNγ released in the medium were measured using an ELISA kit (Abcam). For CAR-T-cell viability, the hydrogel was treated with HAase (1 mg ml−1) for 1 h to digest the HA matrix. CAR-T cells released were collected by centrifugation at 500g for 5 min and stained for live/dead cell assay kit (Thermos Fisher Scientific) and imaged by fluorescence microscope.

To measure the hydrogel loading efficiency, CAR-T cells and IL-15 with Cy5.5 fluorophore were loaded in the hydrogels in a 48-well plate. After 15 min, hydrogels with the CAR-T cells (2 × 106) and IL-15 NPs were transferred to another well. The fluorescence intensity in the original wells was measured by imaging to evaluate cells/NPs leaked out from the hydrogel.

To study the degradation in vivo of the HA hydrogel, the hyaluronidase concentration at the tumour site was first investigated. Tumour and skin were isolated from the mice and then mechanically homogenized. The hyaluronidase concentration was detected with ELISA kit according to the manufacturer’s instructions. Next, the HA hydrogel labelled with Cy5.5 was implanted into the NSG mouse subcutaneously. Fluorescence imaging was recorded using the IVIS (Perkin Elmer).

Release and mobility of the cells within the hydrogel.

Hydrogel containing 1 × 107 CAR-T cells and 1 × 107 platelets was placed into a cell strainer with a 40 μm pore size. The strainer was embedded in a six-well plate with 5 ml of medium without any cytokine. At predetermined time points, CAR-T cells and platelets were counted with hemocytometer. To evaluate the release of CAR-T cells in vivo, CAR-T cells were stained with CellTracker deep red for 30 min. The hydrogel loaded with 2 × 106 fluorophore-labelled CAR-T cells was implanted into the tumour resection cavity. The hydrogel was then removed every day for a week, and CAR-T cells released in the resection cavity were quantified using IVIS. The analysis of signals was performed using Living Image Software v.4.3.1.

Cell mobility within the hydrogel was characterized by labelling CAR-T cells with CellTracker Orange CMRA and platelets with CellTracker Green CMFDA (Thermo Fisher Scientific). Cells were imaged using a confocal microscope equipped with a humidified environmental chamber (37 °C, 5% CO2), and images were recorded every 30s for 30 min. The cell track was analysed using Imaris imaging analysis software.

CAR-T and tumour cells coculture study.

To examine the killing effects of CAR-T cells, the T cells in various formulations—including T cells loaded in the hydrogel (T@gel; T-cell number, 3.3 × 105), T cells and platelets loaded in the hydrogel (T-platelet@gel; T-cell number, 3.3 × 105; platelet number, 1 × 107), CAR-T cells loaded in the hydrogel (CAR-T@gel; CAR-T-cell number, 3.3 × 105), and CAR-T cells and platelets loaded in the hydrogel (CAR-T-P–aPDL1@gel; CAR-T-cell number, 3.3 × 105; platelet number, 1 × 107; aPDL1, 1 μg)—were placed into a strainer with 40 μm pore size and embedded in a six-well plate seeded with 1 × 106 GFP-labelled WM-115 cells. After 72 h, cells were trypsinized and collected for flow cytometry analysis. IL-2 and IFNγ released in the medium were measured using ELISA kits.

Proliferation of CAR-T cells was measured using the carboxyfluorescein succinimidyl ester (CFSE) staining assay. In brief, T cells were stained with CFSE (5 μM) according to the manufacturer’s instructions, seeded in the hydrogel and cocultured with WM115 melanoma cells. After 96 h, cells were collected after trypsinization, stained with the anti-CD3 antibodies and analysed using flow cytometry. PDL1 expression of WM115 cells after coculture with CAR-T cells was analysed using flow cytometry. All in vitro CAR-T-cell studies were performed without the addition of the IL-15 NPs.

In vivo anti-tumour efficacy.

The NSG mice (female, aged 6–8 weeks) were purchased from Jackson laboratory. All of the animal studies strictly followed the animal protocol approved by the Institutional Animal Care and Use Committee at the University of North Carolina at Chapel Hill, North Carolina State University and UCLA. To establish the in vivo human WM115 melanoma model, NSG mice were injected with 5 × 106 WM115 cells subcutaneously. After 2 weeks, the tumours were partially removed and embedded in OCT to obtain frozen sections. Slides of tumour sections were stained with the anti-PDL1 and anti-CSPG4 antibodies and Hoechst 33342, and then analysed using a confocal microscope. Control WM115 tumours and CAR-T@gel-treated (CAR-T-cell number, 2 × 106) tumours were digested by collagenase, mechanically disrupted into single-cell suspension, stained with BV421-anti human PDL1 (29E.2A3) antibodies and analysed using flow cytometry. An irrelevant antibody was used as the isotype control.

For in vivo anti-tumour studies, NSG mice were injected with 5 × 106 luciferase-labelled WM115 cells subcutaneously. When the tumour size reached ~150 mm3, tumours were partially removed, leaving ~5% of the tumour mass. The bioluminescence signal of remaining tumour tissue was detected using the IVIS. Resection cavities were filled with various formulations, including saline, T-platelet@gel, P–aPDL1@gel, CAR-T, CAR-T-P-isotype antibody@gel, CAR-T + P–aPDL1, CAR-T@gel, CAR-T@gel + P–aPDL1, CAR-T@gel + systemic aPDL1, CAR-T-P–aPDL1@gel (2 × 106 CAR-T cells, 1 × 107 platelets, 1 μg aPDL1). The dose and frequency for systemic injection of aPDL1 were 10 μg every day for 3 d through tail vein injection. In all of the experimental groups, IL-15 NPs (IL-15, 1 μg) were included. To visualize the luciferase signal, mice were injected intraperitoneally with d-luciferin at a dose of 150 mg kg−1 in 100 μl PBS. Mice were imaged after 5 min with 1 s of acquisition time. Mice were imaged at 1, 2 and 3 weeks after hydrogel implantation. Bioluminescence signals were analysed using Living Image Software v.4.3.1 (PerkinElmer). After 3 weeks, tumours were removed and imaged. Tumour volumes were recorded using a digital calliper and calculated according to the formula: length × width2 × 0.5.

To investigate the abscopal effects of CAR-T-P–aPDL1@gel, the double-tumour-bearing NSG mouse model was built by injecting 5 × 106 luciferase-labelled WM115 cells into the abdomen site subcutaneously. When the tumour size reached 100 mm3, the tumour on the right side was surgically removed (leaving ~5% residual tumour) and resection cavities were filled with different formulations, including saline, CAR-T, CAR-T@gel, CAR-T + P–aPDL1 and CAR-T-P–aPDL1@gel (CAR-T, 2 × 106; platelets, 1 × 107; aPDL1, 1 μg). In all of the experimental groups IL-15 NPs (IL-15, 1 μg) were included. Mice were imaged at weeks 1 and 2 to monitor tumour recurrence on the surgical side, while the growth of the untouched tumours on left side was monitored using a digital calliper. Mice were also bled to evaluate the redistribution of CAR-T cells in the peripheral blood. A blood sample (100 μl) was collected and the red blood cells were lysed. CAR-T cells in the peripheral blood were detected using FITC-labelled anti-human CD45 and APC-labelled anti-human CD3 antibodies and analysed using flow cytometry. To analyse the presence of CAR-T cells in the distant tumour, tumours were removed, digested with collagenase for 30 min, mechanically disrupted and filtered through a cell strainer with a pore size of 40 μm. Cell suspensions were stained with APC-labelled anti-human-CD3 antibodies and analysed using flow cytometry. Tumours were also embedded with OCT and immerged in liquid nitrogen. Slides obtained from frozen tissues were analysed using confocal microscopy after staining with Hoechst 33342, PE labelled anti-human-CD4 and FITC-labelled anti-human-CD8 antibodies.

In vivo CAR-T-cell expansion.

The resection tumour model was established as previously described. After surgery, the resection cavity was treated with CAR-T, CAR-T + P–aPDL1, CAR-T@gel, CAR-T@gel + P–aPDL1, CAR-T-P–aPDL1@ gel (CAR-T-cell number, 2 × 106; platelet number, 1 × 107; aPDL1, 1 μg). In all of the experimental groups IL-15 NPs (IL-15, 1 μg) were included. CAR-T cells were cotransduced with a vector encoding the fusion protein GFP–firefly luciferase for in vivo imaging. Bioluminescence signals were recorded on day 1 after implantation and then at days 4, 7, 14, 21 and 28. The analysis of signals was performed using Living Image Software v.4.3.1.

To study the concentration of IL-15, tumour tissues were collected after 1, 2, 3 and 4 weeks after hydrogel implantation. After tissue homogenization, IL-15 concentration was detected using an ELISA kit according to the manufacturer’s instructions. IL-15 concentration in untreated tumours was also detected.

To investigate the tumour distribution of both CAR-T cells and platelets, CAR-T cells were stained with CellTracker Orange and platelets were labelled with WGA-Alexa Fluor 488. After tumour removal, the CAR-T-P–aPDL1@gel was implanted and the tumour tissue was taken out after 72 h to generate a frozen section. After staining with Hoechst 33342, the slide was analysed using confocal microscopy for observation. Cytokine levels in the tumours were measured using the LEGENDplex human and mouse cytokine detection kits. Briefly, tumours were collected 7 d after treatment, weighed and mechanically disrupted. After homogenizing in cold PBS, the cell mixture was filtered through the cell strainer and, after centrifugation, was analysed by flow cytometry using the software provided by the manufacturer.

To investigate the chemokine levels in the WM115 tumour after surgery, when the tumour size reached ~150 mm3, tumours were partially removed leaving ~5% of the tumour mass. After 2d, the surgical tumours and complete tumour were collected and mechanically disrupted. After homogenizing in cold PBS, the cell mixture was filtered through the cell strainer and, after centrifugation, was analysed by flow cytometry using the software provided by the manufacturer. The chemokine was measured using the LEGENDplex human chemokine detection kit.

Statistics.

All results are presented as mean ± s.d. Statistical analysis was evaluated using GraphPad Prism (v.7.0). One-way ANOVA followed by Tukey’s post hoc test was performed for multiple-group analyses and unpaired Student t-tests were performed for two-group analyses. The differences between experimental groups and control groups were considered to be statistically significant at P < 0.05; *P < 0.05, **P < 0.01, ***P < 0.001.

Reporting Summary.

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Supplementary Material

SI

Acknowledgements

This work was supported by grants from the Jonsson Comprehensive Cancer Center at UCLA, the Alfred P. Sloan Foundation (Sloan Research Fellowship), NIH 1R01CA234343-01A1, a pilot grant from the UNC Cancer Center and the start-up package from Zhejiang University to Z.G.

Footnotes

Competing interests

Patents describing the drug-delivery system documented in this Article have been filed with the US Patent Office. Q.H. and Z.G. are listed as inventors on the provisional patent application (provisional patent application no. 63/055,738). Z.G. is the co-founder of Zencapsule Inc., and the other authors declare no competing interests.

Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41551-021-00712-1.

Peer review information Nature Biomedical Engineering thanks the anonymous reviewers for their contribution to the peer review of this work.

Reprints and permissions information is available at www.nature.com/reprints.

Data availability

The main data supporting the results in this study are available within the paper and its Supplementary Information. All data generated in this study, including source data and the data used to make the figures, are available from Figshare (https://figshare.com/s/fa6578df11fba2539c13).

References

  • 1.Demicheli R, Retsky M, Hrushesky W, Baum M & Gukas I The effects of surgery on tumour growth: a century of investigations. Ann. Oncol. 19, 1821–1828 (2008). [DOI] [PubMed] [Google Scholar]
  • 2.Lukianova-Hleb EY et al. Intraoperative diagnostics and elimination of residual microtumours with plasmonic nanobubbles. Nat. Nanotechnol. 11, 525–532 (2016). [DOI] [PubMed] [Google Scholar]
  • 3.Demicheli R, Retsky MW, Hrushesky WJ & Baum M Tumour dormancy and surgery-driven interruption of dormancy in breast cancer: learning from failures. Nat. Rev. Clin. Oncol. 4, 699–710 (2007). [DOI] [PubMed] [Google Scholar]
  • 4.Baker D, Masterson T, Pace R, Constable W & Wanebo H The influence of the surgical wound on local tumour recurrence. Surgery 106, 525–532 (1989). [PubMed] [Google Scholar]
  • 5.Ceelen W, Pattyn P & Mareel M Surgery, wound healing, and metastasis: recent insights and clinical implications. Crit. Rev. Oncol. Hematol. 89, 16–26 (2014). [DOI] [PubMed] [Google Scholar]
  • 6.Pardoll DM The blockade of immune checkpoints in cancer immunotherapy. Nat. Rev. Cancer 12, 252–264 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Topalian SL, Drake CG & Pardoll DM Immune checkpoint blockade: a common denominator approach to cancer therapy. Cancer Cell 27, 450–461 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Topalian SL, Taube JM, Anders RA & Pardoll DM Mechanism-driven biomarkers to guide immune checkpoint blockade in cancer therapy. Nat. Rev. Cancer 16, 275–287 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Sharma P & Allison JP The future of immune checkpoint therapy. Science 348, 56–61 (2015). [DOI] [PubMed] [Google Scholar]
  • 10.Marquez-Rodas I et al. Immune checkpoint inhibitors: therapeutic advances in melanoma. Ann. Transl. Med. 3, 267 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Jenkins RW, Thummalapalli R, Carter J, Cañadas I & Barbie DA Molecular and genomic determinants of response to immune checkpoint inhibition in cancer. Annu. Rev. Med. 69, 333–347 (2018). [DOI] [PubMed] [Google Scholar]
  • 12.Nowicki TS, Hu-Lieskovan S & Ribas A Mechanisms of resistance to PD-1 and PD-L1 blockade. Cancer J. 24, 47–53 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Michot J et al. Immune-related adverse events with immune checkpoint blockade: a comprehensive review. Eur. J. Cancer 54, 139–148 (2016). [DOI] [PubMed] [Google Scholar]
  • 14.Sharma P & Allison JP Immune checkpoint targeting in cancer therapy: toward combination strategies with curative potential. Cell 161, 205–214 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Rosenberg SA, Restifo NP, Yang JC, Morgan RA & Dudley ME Adoptive cell transfer: a clinical path to effective cancer immunotherapy. Nat. Rev. Cancer 8, 299–308 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Restifo NP, Dudley ME & Rosenberg SA Adoptive immunotherapy for cancer: harnessing the T cell response. Nat. Rev. Immunol. 12, 269–281 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Jackson HJ, Rafiq S & Brentjens RJ Driving CAR T-cells forward. Nat. Rev. Clin. Oncol. 13, 370–383 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fesnak AD, June CH & Levine BL Engineered T cells: the promise and challenges of cancer immunotherapy. Nat. Rev. Cancer 16, 566–581 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Lee DW et al. T cells expressing CD19 chimeric antigen receptors for acute lymphoblastic leukaemia in children and young adults: a phase 1 dose-escalation trial. Lancet 385, 517–528 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Newick K, O’Brien S, Moon E & Albelda SM CAR T cell therapy for solid tumours. Annu. Rev. Med. 68, 139–152 (2017). [DOI] [PubMed] [Google Scholar]
  • 21.Abken H Adoptive therapy with CAR redirected T cells: the challenges in targeting solid tumours. Immunotherapy 7, 535–544 (2015). [DOI] [PubMed] [Google Scholar]
  • 22.John LB, Kershaw MH & Darcy PK Blockade of PD-1 immunosuppression boosts CAR T-cell therapy. Oncoimmunology 2, e26286 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.John LB et al. Anti-PD-1 antibody therapy potently enhances the eradication of established tumours by gene-modified T cells. Clin. Cancer Res. 19, 5636–5646 (2013). [DOI] [PubMed] [Google Scholar]
  • 24.Pluschke G et al. Molecular cloning of a human melanoma-associated chondroitin sulfate proteoglycan. Proc. Natl Acad. Sci. USA 93, 9710–9715 (1996). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Rivera Z et al. CSPG4 as a target of antibody-based immunotherapy for malignant mesothelioma. Clin. Cancer Res. 18, 5352–5363 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kochenderfer JN et al. Lymphoma remissions caused by anti-CD19 chimeric antigen receptor T cells are associated with high serum interleukin-15 levels. J. Clin. Oncol. 35, 1803–1813 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Wang C et al. In situ activation of platelets with checkpoint inhibitors for post-surgical cancer immunotherapy. Nat. Biomed. Eng. 1, 0011 (2017). [Google Scholar]
  • 28.Hu Q et al. Conjugation of haematopoietic stem cells and platelets decorated with anti-PD-1 antibodies augments anti-leukaemia efficacy. Nat. Biomed. Eng. 2, 831–840 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Gemmell CH, Ramirez SM, Yeo EL & Sefton MV Platelet activation in whole blood by artificial surfaces: identification of platelet-derived microparticles and activated platelet binding to leukocytes as material-induced activation events. J. Lab. Clin. Med. 125, 276–287 (1995). [PubMed] [Google Scholar]
  • 30.Kahn ML et al. A dual thrombin receptor system for platelet activation. Nature 394, 690–694 (1998). [DOI] [PubMed] [Google Scholar]
  • 31.Jiang T, Mo R, Bellotti A, Zhou J & Gu Z Gel–liposome-mediated co-delivery of anticancer membrane-associated proteins and small-molecule drugs for enhanced therapeutic efficacy. Adv. Funct. Mater. 24, 2295–2304 (2014). [Google Scholar]
  • 32.Feczkó T, Tóth J, Dósa G & Gyenis J Optimization of protein encapsulation in PLGA nanoparticles. Chem. Eng. Process. 50, 757–765 (2011). [Google Scholar]
  • 33.Caruana I et al. Heparanase promotes tumour infiltration and antitumour activity of CAR-redirected T lymphocytes. Nat. Med. 21, 524–529 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Stephan SB et al. Biopolymer implants enhance the efficacy of adoptive T-cell therapy. Nat. Biotechnol. 33, 97–101 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Smith TT et al. Biopolymers codelivering engineered T cells and STING agonists can eliminate heterogeneous tumours. J. Clin. Invest. 127, 2176–2191 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Dawson E, Mapili G, Erickson K, Taqvi S & Roy K Biomaterials for stem cell differentiation. Adv. Drug Deliv. Rev. 60, 215–228 (2008). [DOI] [PubMed] [Google Scholar]
  • 37.Orive G et al. Cell encapsulation: promise and progress. Nat. Med. 9, 104–107 (2003). [DOI] [PubMed] [Google Scholar]
  • 38.Mause SF, von Hundelshausen P, Zernecke A, Koenen RR & Weber C Platelet microparticles: a transcellular delivery system for RANTES promoting monocyte recruitment on endothelium. Arterioscler. Thromb. Vasc. Biol. 25, 1512–1518 (2005). [DOI] [PubMed] [Google Scholar]
  • 39.Curran KJ et al. Enhancing antitumour efficacy of chimeric antigen receptor T cells through constitutive CD40L expression. Mol. Ther. 23, 769–778 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Henn V et al. CD40 ligand on activated platelets triggers an inflammatory reaction of endothelial cells. Nature 391, 591–594 (1998). [DOI] [PubMed] [Google Scholar]
  • 41.Pellegatta S et al. Constitutive and TNFα-inducible expression of chondroitin sulfate proteoglycan 4 in glioblastoma and neurospheres: implications for CAR-T cell therapy. Sci. Transl. Med. 10, eaao2731 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SI
NIHMS1799463-supplement-SI.pdf (2.7MB, pdf)

Data Availability Statement

The main data supporting the results in this study are available within the paper and its Supplementary Information. All data generated in this study, including source data and the data used to make the figures, are available from Figshare (https://figshare.com/s/fa6578df11fba2539c13).

RESOURCES