Nanoparticle delivery of VEGF-B mRNA promotes T cell infiltration within tumor and triggers robust antitumor immunity

Abstract

The advancement of mRNA-based cancer immunotherapies has gained significant momentum, particularly after the success of mRNA vaccines during the COVID-19 pandemic and the recognition of mRNA vaccine development with the 2023 Nobel Prize. mRNA encoding cytokines, antibodies, and chimeric antigen receptor T cells has demonstrated substantial therapeutic potential in both preclinical models and clinical trials. Previous study identified vascular endothelial growth factor B (VEGF-B) as a metabolic regulator that controls lipid synthesis and maintains mitochondrial membrane integrity, essential for the survival of activated T cells. In this study, we demonstrate that mRNA encoding VEGF-B, delivered to tumors via lipid nanoparticles, effectively controls tumor growth in both subcutaneous and lung metastasis tumor models. Combination with programmed death-1 blockade significantly amplified therapeutic efficacy, leading to complete tumor regression in the lung metastasis model. Immune profiling revealed that nanoparticle delivery of VEGF-B mRNA reprograms the tumor microenvironment by increasing CD8⁺ T cell infiltration and enhancing the expression of effector molecules, including interferon-γ, tumor necrosis factor alpha, and granzyme B, while downregulating the exhaustion molecule programmed death-1. These findings highlight the considerable promise of mRNA-based therapies in reshaping the tumor microenvironment and enhancing cancer immunotherapy outcomes.

Graphical abstract

Lipid nanoparticle delivery of VEGF-B mRNA reprograms tumor microenvironment, enhancing CD8⁺ T cell infiltration and effector functions while reducing T cell exhaustion, and thereby effectively controls tumor growth in both subcutaneous and lung metastasis models.

Introduction

Since the first identification of vascular endothelial growth factor B (VEGF-B) in 1996,¹ its role in angiogenesis has been the focus of ongoing debate. Due to its high sequence homology with vascular endothelial growth factor A and their similar receptor-binding patterns,² VEGF-B was initially assumed to function as an angiogenic factor. However, research has produced conflicting results. While some studies suggest that VEGF-B promotes angiogenesis,^3,4,5,6 others have shown it to inhibit angiogenesis^7,8 and suppress tumor growth.^8,9,10 Beyond its debated role in angiogenesis, VEGF-B has been implicated in lipid metabolism^11,12 and metabolic disorders such as diabetes.^13,14 Notably, several studies highlight a protective role for VEGF-B, particularly in preventing apoptosis and promoting cell survival under stress conditions.^{5,6,15,16,17,18,19,20}

The rise of cancer immunotherapy has transformed cancer treatment, offering unprecedented efficacy and personalized therapeutic options, especially for malignancies that were previously refractory to conventional therapies. One of the most successful strategies is immune checkpoint blockade (ICB), where monoclonal antibodies are employed to block inhibitory receptors, such as programmed death-1 (PD-1), cytotoxic T-lymphocyte-associated protein 4, and lymphocyte activation gene 3. By inhibiting these inhibitory signals, ICB therapies reinvigorate T cells, enabling sustained antitumor activity. Another major advancement is chimeric antigen receptor (CAR) T cells therapy, where autologous T cells are engineered to express CARs that specifically recognize tumor-associated antigens.²¹ These engineered T cells are reintroduced into the patient, where they selectively target and eliminate cancer cells. CAR T therapy has shown remarkable efficacy, particularly in hematologic cancers, yielding durable responses in some patients. In addition to these cell-based therapies, cytokines serve a crucial role in cancer immunotherapy. These signaling molecules regulate immune cell activity, proliferation, and survival, helping coordinate the immune response. Cytokines such as interleukin (IL)-2 and interferon-α have been approved for clinical use,²² where they enhance immune activity against tumors. When used alone or in combination with other treatments, cytokines can modulate the tumor microenvironment (TME), boosting the overall effectiveness of immunotherapy.

The success of mRNA vaccines in addressing COVID-19 has revitalized interest in mRNA as a platform for the delivery of therapeutic proteins. Compared to DNA-based therapeutics, mRNA offers several advantages, including the elimination of concerns about genomic integration, thus avoiding unwanted mutagenesis. mRNA can be produced in cell-free systems, allowing rapid, scalable, and cost-effective manufacturing. Its transient expression profile strikes a favorable balance between therapeutic efficacy and safety, as demonstrated in multiple studies where it shows potent therapeutic effects with an excellent safety profile. To date, mRNA has been explored as cancer vaccines,²³ cytokines for immunotherapy,²⁴ tumor suppressor genes to inhibit tumor growth,²⁵ CARs for engineered T cell therapies,^26,27 and genome editing for gene therapy.^28,29 Prior research has demonstrated that VEGF-B signaling is essential for sustaining lipid synthesis and mitochondrial functionality, facilitating T cell-mediated immune responses.¹² Here, we report that nanoparticle-delivered mRNA encoding VEGF-B effectively controls tumor growth across various tumor models by reprogramming the TME. This reprogramming is characterized by increased infiltration and enhanced polyfunctionality of CD8⁺ T cells and downregulation of inhibitory receptor PD-1. Combining VEGF-B mRNA with PD-1 blockade significantly improved therapeutic efficacy in the lung metastasis model, resulting in complete tumor regression. Our results reveal that VEGF-B mRNA not only fosters an immunostimulatory TME but also augments the effectiveness of ICB therapies, presenting a novel and promising approach for cancer treatment.

Results

Ψ-modified VEGF-B mRNA with co-transcriptionally tailing showed the most abundant expression in vitro

For mRNA-based protein replacement therapeutics, reducing immunogenicity and increasing protein expression efficiency are essential considerations.³⁰ Early research showed that unmodified in vitro transcription (IVT) mRNA is highly immunogenic, but modified nucleosides such as pseudouridine (Ψ), N6-methyladenosine (m6A), 5-methylcytidine (m5C), and 5-methyluridine (m5U) can improve mRNA stability and translation efficiency while reducing immunogenicity.^31,32

To enhance translation efficiency, we synthesized VEGF-B mRNA with an hemagglutinin (HA)-tag by IVT and included different modified nucleotides and tailing method: N1-methyl-pseudouridine triphosphate (m1Ψ) or Ψ, E. coli Poly(A) polymerase (EPAP) tailing, or co-transcriptionally tailing. These constructs were transfected into HEK293T and B16F10 cells, and VEGF-B expression and secretion were evaluated through immunoprecipitation (IP)-western blot analysis of whole-cell lysates and culture supernatants (Figures 1A and 1B). Our results demonstrated that VEGF-B mRNA with Ψ modification and co-transcriptionally tailing exhibited the highest expression levels, consistent with previous findings where Ψ-modified mRNA delivered intravenously showed improved translation efficiency compared to uridine-containing mRNA.³³

Figure 1.

Characterization of VEGF-B mRNA modifications and mRNA-LNPs

(A and B) IP-western blot analysis of VEGF-B protein expression in HEK293T and B16F10 cells after transfection of VEGF-B mRNA with different modifications and tailing method. Non, non-modified; m1Ψ, m1Ψ-modified and EPAP tailing; Ψ1, Ψ-modified and EPAP tailing; Ψ2, Ψ-modified and co-transcriptionally tailing. (C) Transmission electron microscopy observation of VEGF mRNA-LNP; scale bars, 200 μm. (D) Size distribution of VEGF mRNA-LNP and control mRNA-LNP detected by Malvern dynamic light scattering Zetasizer. (E) VEGF-B protein secretion levels in B16F10 cells transfected with VEGF-B mRNA-LNP, detected by ELISA. Cells in a six-well plate were transfected with 2 μg of the modified mRNA per well. Data are presented as the median ± standard error of the mean (SEM). Significance was determined with ordinary one-way ANOVA in (E). ∗∗∗∗p < 0.0001.

Next, we encapsulated the VEGF-B mRNA with such modification in lipid nanoparticles (LNPs), with EGFP mRNA of similar length as control. Electron microscopy and particle size analysis revealed that both VEGF-B mRNA-LNP and EGFP mRNA-LNP exhibited a particle size distribution of approximately 80 nm (Figures 1C and 1D). Following transfection into B16F10 cells, elevated levels of VEGF-B protein were detected in the culture supernatant (Figure 1E). No inhibitory effect was observed following VEGF-B overexpression in B16F10 and MC38 cell lines (Figure S1).

Nanoparticle delivery of mRNA leads to localized translation and secretion

To assess the efficiency of IVT mRNA expression in vivo, we constructed Ψ-modified luciferase mRNA, encapsulated it in LNPs, and injected intratumorally into MC38 subcutaneous tumors. Bioluminescence was monitored at the tumor site using in vivo imaging system (IVIS) at 6, 24, 48, 72, and 96 h post-injection. Peak bioluminescence occurred at 6 h, followed by a logarithmic decline every 24 h, returning to baseline by 96 h (Figures 2A and 2B).

Figure 2.

In vivo validation of mRNA delivery and VEGF-B expression

(A) In vivo bioluminescence of Ψ-modified luciferase mRNA shows luciferase activity after nanoparticle delivery to subcutaneous MC38 tumors. (B) Attenuation of localized bioluminescence intensity in tumors following a single intratumoral injection of luciferase mRNA-LNP. (C and D) Flow cytometry was performed to assess EGFP expression in B16F10 tumors following injection with EGFP mRNA-LNP. Left: representative flow cytometry plots showing EGFP expression. Right: quantification of EGFP⁺ cells. (E) Immunohistochemistry analysis of VEGF-B in subcutaneous MC38 tumors 6 h after intratumoral injection of VEGF mRNA-LNP. Scale bars, 50 μm. (F and G) VEGF-B protein levels were measured in the tumor interstitial fluid of subcutaneous B16F10 tumors and serum 48 h after intratumoral injection of VEGF mRNA-LNP. Mice was intratumorally injected with 5 μg of luciferase mRNA, EGFP mRNA, or VEGF-B mRNA encapsulated in LNPs. Significance was determined with ordinary one-way ANOVA in (D), (F), and (G). Data are presented as the median ± SEM. ∗∗p < 0.01; ∗∗∗∗p < 0.0001.

To identify the types of cells transfected by the mRNA, we harvested tumors 6 h after injecting EGFP mRNA-LNP and processed them into single-cell suspensions. Tumor tissues contained primarily tumor cells, immune cells, and epithelial cells.³⁴ Epithelial cells (CD326⁺) and immune cells (CD45⁺) were labeled to determine which cell populations had taken up and expressed EGFP mRNA. Our analysis, using VEGF-B mRNA-LNP as control, indicated that both tumor cells (CD326⁻ CD45⁻) and immune cells (CD326⁻ CD45⁺) were able to uptake and express the mRNA injected intratumorally (Figures 2C and 2D), which is consistent with previous studies.^24,35,36 Immunohistochemistry (IHC) further confirmed high levels of VEGF-B expression in tumors following intratumoral injection of VEGF-B mRNA-LNP (Figure 2E).

As VEGF-B is a secreted protein, its expression and secretion post-injection are critical for its therapeutic efficacy. To confirm whether nanoparticle delivery of VEGF-B mRNA leads to expression and secretion, we isolated native, undiluted extracellular fluid from the tumors, referred to as tumor interstitial fluid (TIF), using tissue centrifugation.^37,38 We measured VEGF-B protein levels in TIF using ELISA. Results showed high level of VEGF-B protein in both the TIF and the serum of treated mice (Figures 2F and 2G), confirming successful expression and secretion of the VEGF-B mRNA-LNP in vivo.

Expression of mRNA encoding VEGF-B inhibits tumor growth in vivo

With the feasibility of nanoparticle delivery of VEGF-B mRNA established, we next assessed its therapeutic efficacy in melanoma cell line B16F10- and colon adenocarcinoma cell line MC38-derived subcutaneous tumor model. Multiple intratumoral injections of VEGF-B mRNA-LNP significantly inhibited tumor growth (Figures 3B and 3E). At the conclusion of the study, the tumor weight in the group treated with VEGF-B mRNA-LNP was markedly reduced compared to that in the control group (Figures 3C and 3F); increased VEGF-B mRNA-LNP dosage resulted in better tumor control in the B16F10 tumor model (Figure S2). Survival analysis showed that mice treated with VEGF-B mRNA-LNP had a longer survival time compared to controls (Figure 3I). The treatment was well-tolerated, with no significant changes in body weight observed (Figures 3D and 3G); histological examination of major organs revealed no notable pathological changes (Figure 3H). Spleen B cells and CD4⁺ and CD8⁺ T cells were analyzed to evaluate potential systemic immune effects; no immune shifts was observed (Figure S3). Together, these data suggest that nanoparticle delivering VEGF-B mRNA inhibits tumor growth with a favorable safety profile in vivo.

Figure 3.

VEGF-B mRNA treatment inhibits subcutaneous tumor growth and improves survival

(A) Schematic representation of the experimental design for the subcutaneous tumor model. (B and E) Tumor growth curve of MC38 and B16F10 tumor treated with saline, control mRNA-LNP, or VEGF mRNA-LNP. Subcutaneous tumors were established 7 days prior to treatment with 2 μg of VEGF mRNA-LNP or control mRNA-LNP every 2 days via intratumoral injection; black arrows indicate injection time points. (C and F) Tumor weights of MC38 or B16F10 tumor after treatment with saline, control mRNA-LNP, or VEGF mRNA-LNP. (D and G) Average tumor-bearing mice body weight over time. (H) Representative hematoxylin and eosin (H&E) staining images of the heart, spleen, liver, lung, and kidney indicated treatment groups are shown. Scale bars, 50 μm. (I) Overall survival of subcutaneous B16F10 tumor-bearing mice treated with saline, VEGF mRNA-LNP, or control mRNA-LNP. Two-way ANOVA with Sidak’s correction was used for comparison between VEGF or control in (B) and (E). Ordinary one-way ANOVA in (C) and (F). Log rank (Mantel-Cox) test for survival comparison between VEGF or control in (I). Data are presented as the median ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

VEGF-B mRNA enhances therapeutic efficacy of PD-1 blockade

Building on the findings from subcutaneous tumor experiments, we sought to explore a more clinically relevant scenario using an orthotopic tumor model. The Lewis lung carcinoma (LLC) model was selected for its well-documented reproducibility and suitability for modeling lung metastasis,³⁹ enabling us to further investigate the therapeutic effects of VEGF mRNA-LNP. Additionally, we considered the widespread clinical use of PD-1 blockade in non-small cell lung cancer and the finding that the relative abundance of CTLA-4^hiPD-1^hi CTLs predicts response to anti-PD-1 therapy.⁴⁰ These factors prompted us to combine a lung metastasis model with PD-1 blockade in our exploration. Several delivery vehicles have been reported for mRNA targeting the lungs,^41,42,43,44 enabling its use in treating various lung-associated diseases. To evaluate the antitumor activity of VEGF-B mRNA-LNP in lung cancer, we constructed a cell line LLC stably expressing luciferase (LLC-Luc) using lentiviral infection, which allowed us to track tumor burden in the lung via bioluminescence. In this model, 1 × 10⁶ luciferase-labeled LLC-Luc cells were injected intravenously, followed by random grouping and lung-targeted intravenous injections of VEGF-B mRNA (selective organ targeting-lipid nanoparticles, SORT-LNP)⁴³ and/or intraperitoneal injections of anti-PD-1 antibody (Figure 4A).

Figure 4.

VEGF-B mRNA treatment reduces lung metastases and improves survival

(A) Schematic representation of the experimental design for the lung metastases model. (B) In vivo bioluminescence imaging was conducted on a lung metastasis tumor model in mice subjected to different treatments as indicated. (C) Quantification of bioluminescence signals in mouse lungs at day 10, 15, 19, and 23. (D) Overall survival in the LLC-Luc lung metastasis tumor model treated with saline, control mRNA-LNP, VEGF mRNA-LNP, anti-PD-1, and anti-PD-1 with VEGF mRNA-LNP. Each mouse received 2 μg of mRNA or equal volume of saline by intravenous injection. Anti-PD-1 antibody was intraperitoneally injected three times for a total of 300 μg. Two-way ANOVA with Sidak’s correction was used for comparison between anti-PD-1 and anti-PD-1 + VEGF in (C). Log rank (Mantel-Cox) test was performed for survival comparison between anti-PD-1 and anti-PD-1 + VEGF in (D). Data are presented as the median ± SEM. ∗∗p < 0.01; ∗∗∗p < 0.001.

By day 23 post-inoculation, bioluminescence intensity in the lungs of mice treated with the combination of VEGF-B mRNA-LNP and PD-1 blockade was significantly lower compared to those treated with either PD-1 blockade or VEGF-B mRNA-LNP alone (Figures 4B and 4C), indicating better tumor control. Survival was extended in the combination group compared to the other treatment groups (Figure 4D); 2 out of 7 mice in the combination group achieved complete tumor regression. Together, these data indicated that combination with VEGF-B mRNA-LNP enhances the antitumor efficacy of PD-1 blockade.

Treatment with VEGF-B mRNA promotes T cell infiltration and enhances T cell polyfunctionality

To gain a deeper understanding of how VEGF-B mRNA influences the TME, we conducted immunofluorescence microscopy on B16F10 tumors after multiple intratumoral VEGF-B mRNA treatments. The VEGF-B mRNA treatment group exhibited enhanced infiltration of both CD4⁺ and CD8⁺ T cells (Figure 5A); flow cytometry confirmed increased CD4⁺ and CD8⁺ T cells frequency (Figure 5B). Potential shifts in immune composition within the TME were analyzed. No significant impact was observed in regulatory T cell (Treg), B cell, natural killer cell, and macrophage population (Figures S4A–S4E); notably, we saw trends toward increased neutrophils and myeloid-derived suppressor cells (MDSCs) and reduction of conventional dendritic cells (cDCs) in the mRNA-LNP-treated group, potentially due to the innate immune response and toxicity caused by LNPs (Figures S4F–S4I). Intracellular analysis of CD8⁺ T cell demonstrated elevated expression and population of tumor necrosis factor alpha (TNF-α), interferon-γ (IFN-γ), and granzyme B (GZMB), indicating enhanced polyfunctionality of T cells (Figures 5C, 5D, and 5E). Surprisingly, we observed a reduction in both the population and expression levels of the inhibitory receptor PD-1 on CD8⁺ T cells, both within the TME and in the draining lymph nodes (dLNs) (Figures 5F and 5I). Prolonged treatment reduced PD-1⁺ cell population and expression levels (Figure S5). Additionally, we observed a reduction in Annexin V⁺ apoptotic CD4⁺ and CD8⁺ T cells in the treated mice, along with an increased number of CD4⁺ and CD8⁺ T cells in the dLNs (Figures 5G and 5H). Together, these data demonstrate that mRNA encoding VEGF-B reprogrammed the TME by increased infiltration and enhanced polyfunctionality of CD8⁺ T cells and downregulation of inhibitory receptor PD-1.

Figure 5.

VEGF-B mRNA enhances T cell infiltration and polyfunctionality in the TME

(A) Tumor-infiltrating CD4⁺ and CD8⁺ T cell assessed by immunofluorescence microscopy. Scale bars, 100 μm. (B) Representative flow cytometry plots of tumor-infiltrating CD4⁺ and CD8⁺ T cells with quantification. Gated: single cell/live/CD45⁺. (C–F) Detailed tumor-infiltrating CD8⁺ T cells profiling expressing IFN-γ, TNF-α, GZMB, and PD-1. Gated: single cell/live/CD45⁺/CD3⁺/CD8⁺. (G–I) Flow cytometry analysis of CD4⁺ and CD8⁺ T cell number and Annexin V⁺ and PD-1⁺ population within tumor-draining lymph node. Gated: single cell/live/CD45⁺/CD3⁺. Significance was determined with ordinary one-way ANOVA. Data are presented as the median ± SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

Discussion

In addition to the well-established mRNA vaccines designed for infectious diseases such as COVID-19, an increasing body of research is investigating the potential of mRNA-based therapies for cancer and various other medical conditions. These therapies include mRNA encoding tumor antigens for cancer vaccines,²³ cytokines for immunotherapy,²⁴ tumor suppressor genes to inhibit tumor growth,²⁵ CARs for engineered T cell therapies,^26,27 and genome editing for gene therapy.^28,29

This study focuses on mRNA-based protein replacement therapy. Unlike mRNA vaccines, the advancement of mRNA-based protein replacement therapies encounters additional challenges related to achieving sufficient protein expression levels, ensuring targeted delivery, and minimizing immunogenicity. To address these challenges, we compared the in vitro expression and secretion levels of VEGF-B mRNA with non-modification, m1Ψ-modified with EPAP tailing, Ψ-modified with EPAP tailing, and Ψ-modified by co-transcriptionally tailing. Among these, VEGF-B mRNA with Ψ-modification with co-transcriptionally tailing demonstrated the highest expression and secretion levels. Choosing a modification that promotes higher expression levels may help reduce the necessary dosage and associated side effects, enhancing the potential clinical applicability of mRNA-based protein replacement therapies.

Using the MC38 and B16F10 subcutaneous tumor models, we showed that local delivery of VEGF-B mRNA effectively suppressed tumor growth and prolonged the survival of tumor-bearing mice, with minimal toxicity. Treatment with VEGF-B mRNA was associated with enhanced infiltration of both CD4⁺ and CD8⁺ T cells within the TME. Flow cytometry results of tumor-infiltrating CD8⁺ T cells revealed an increase in polyfunctional T cells that express IFN-γ, TNF-α, and GZMB. Notably, IFN-γ signatures are related with clinical responses to PD-1 blockade.⁴⁵ A reduction in both the population and expression levels of the inhibitory receptor PD-1 in CD8⁺ T cells was observed, both within the TME and in the dLNs. The overexpression of VEGF-B in the TME appears to improve the overall fitness of CD8⁺ T cells.

Overexpression of VEGF-B also leads to a significant reduction in Annexin V⁺ CD4⁺ and CD8⁺ T cells, along with an increased number of these cells in the dLNs of treated mice, which correlated with a higher frequency of CD4⁺ and CD8⁺ T cells within tumors, as well as an increase in IFN-γ⁺ and TNF-α⁺ CD8 T cells. These findings suggest that lymph node residency may be essential in maintaining the frequency and polyfunctionality of tumor-infiltrating T cells. Additionally, PD-1⁺ population and expression levels in tumor-infiltrating CD8⁺ T cells decrease but remain high after 3-dose treatment, which provide a therapeutic window for PD-1 blockade and suggest a potential synergistic effect. Results from lung metastasis tumor model demonstrated considerable amplification of therapeutic efficacy. PD-1 reduction trend was also observed in the dLNs, which is important considering evidence that dLNs are essential for the maintenance of precursor exhausted T cells and their response to ICB.⁴⁶ Another study found that dLNs play a crucial role in boosting CAR T cell responses using mRNA vaccine strategies.⁴⁷

The interplay between mRNA-LNP administration and immune cell dynamics reflects complex immunological mechanisms. Notably, mRNA-LNP treatment is associated with increase in neutrophil and MDSC populations, alongside decrease in cDCs. The inflammatory response elicited by mRNA-LNP is characterized by a robust and transient influx of neutrophils, activation of multiple inflammatory pathways, and the production of pro-inflammatory cytokines such as IL-1β and IL-6.^48,49,50 Research indicates that neutrophil infiltration is a common response to LNP delivery, suggesting that the observed increase in neutrophils is a direct consequence of the immune system’s reaction to mRNA-LNP.^51,52 Furthermore, the secretion of IL-1β, IL-6, and granulocyte-macrophage colony-stimulating factor (GM-CSF) has been documented to impact MDSC accumulation and mobilization, providing a potential explanation for the rise in MDSCs following mRNA-LNP treatment.^53,54,55 Reduction on cDCs was also observed in mRNA-LNP-treated groups, consistent with findings reported in multiple studies.^51,56,57 This trend may potentially stem from activation or stress responses to mRNA-LNP toxicity that led to apoptosis reducing the number of cells.⁵⁶ It is also worth noting that cDCs are among the most efficient immune cells in taking up mRNA-LNP,⁵⁷ which may contribute to their declining population post-treatment.

Lung-targeted mRNA-based protein replacement therapies have been investigated for treating a variety of lung diseases, including cystic fibrosis,⁵⁸ primary ciliary dyskinesia,⁵⁹ α-1 antitrypsin deficiency,⁶⁰ asthma,⁶¹ and lung cancer.⁶² Delivery methods for these therapies include the use of nebulization of hyperbranched poly(β-amino ester)-based polymers^42,63 or LNPs^64,65,66 and systemic delivery of organ-targeted LNPs.^41,43,44 These approaches aim to provide targeted and effective delivery of therapeutic proteins to the lungs, offering a potential avenue for treating lung-specific diseases. Our study explored the feasibility of utilizing the SORT system⁴³ to deliver VEGF-B mRNA to the lungs for lung cancer treatment. VEGF-B mRNA treatment has shown a decrease in PD-1⁺ population in CD8⁺ T cell (Figure 5F), but it remains high after 3-dose treatment (Figure S5) in a subcutaneous tumor model, which provides a therapeutic window for PD-1 blockade and suggests a potential synergistic effect. The lung metastasis model demonstrates inspiring results; combination of VEGF-B mRNA and PD-1 blockade significantly amplified therapeutic efficacy of PD-1 blockade, leading to complete regression.

However, the role of CD8+ T cells in the overall antitumor efficacy remains unclear. While we validated the therapeutic benefits of VEGF-B mRNA and its impact on CD8⁺ T cells in subcutaneous and lung metastasis tumor models, the antitumor activity is likely multifactorial and may not solely be driven by T cell infiltration and activity. Though no inhibitory effect is observed in VEGF-B-overexpressing cancer cell lines in vitro (Figure S1), VEGF-B has been reported as an endogenous inhibitor of angiogenesis by suppressing the FGF2/FGFR1 pathway.⁷ The impact of VEGF-B mRNA on other immune and non-immune cells within the TME requires further exploration to fully understand its broader modulatory effects.

In the lung metastasis tumor model, single-agent treatment was ineffective in inhibiting tumor growth. Despite the clinical success of cancer immunotherapy in extending survival across various cancers, only a limited subset of patients derives significant benefit from these treatments. This limited response is closely associated with the characteristics of the TME, particularly whether the tumor is classified as “hot” or “cold” in terms of responsiveness to ICB. The LLC cell line used in this study is characterized by low T cell infiltration and a high presence of myeloid-derived suppressor cells, classifying it as a “cold” tumor.⁶⁷ The combination therapy of VEGF-B mRNA and PD-1 blockade may enhance T cell infiltration and polyfunctionality while alleviating immune suppression. This suggests that the combination therapy may have reprogrammed the TME from an immune-resistant “cold” state to a more immunogenic “hot” state, thereby enhancing the antitumor response.

Additionally, mRNA-encoded cytokines have been studied for their ability to modulate the TME through autocrine and paracrine signaling. In several preclinical models, mRNA-based cytokine therapies have demonstrated potent antitumor effects, both as standalone treatments and in combination with other therapies.^24,68,69 However, conventional mRNA therapies still face challenges such as achieving therapeutic concentrations at the target site, lipid accumulation, off-target effects, and adverse immune-related events due to repeated dosing.^70,71 To address these issues, newer mRNA types, such as self-amplifying mRNA, trans-amplifying mRNA, and circular mRNA, have been developed. These variants offer longer durations of protein expression, increased stability, and greater resistance to RNA degradation, making them promising candidates for long-term therapeutic applications.^72,73

In summary, our study demonstrated that nanoparticle delivery of mRNA encoding VEGF-B effectively suppresses tumor growth by eliciting T cell-mediated responses across multiple tumor models while exhibiting minimal toxicity. VEGF-B overexpression promoted an immunostimulatory TME, characterized by increased CD8⁺ T cell infiltration, enhanced effector molecule expression, and reduced exhaustion. Using the SORT system to deliver mRNA specifically to the lungs, therapeutic effect of VEGF-B mRNA was amplified when combined with PD-1 blockade, leading to complete tumor regression in a lung metastasis model.

Given the rapid and cost-effective production capabilities of mRNA platform for a diverse range of therapeutic molecules, our findings hold promise for the future development of mRNA-based immunotherapies. However, further research is needed to fully elucidate the applications of mRNA-based therapeutics, enhance their efficacy and safety, and guarantee their affordability and accessibility for all patients in need. Despite these challenges, the remarkable potential of mRNA-based cancer therapies continues to drive rapid advancements in the field, paving the way for future breakthroughs.

Materials and methods

Animals

Six-week-old female C57BL/6 mice were obtained from Shanghai Lingchang Biotechnology and kept at the specific pathogen-free animal facility of the Department of Laboratory Animal at Shanghai Jiao Tong University School of Medicine (SJTUSM). The animal experiment protocol was approved by the Experimental Animal Ethical Committee at SJTUSM.

Cells

The human embryonic kidney cell line HEK293T, murine melanoma cell line B16F10, murine colon adenocarcinoma cell line MC38, and murine LLC cell line were obtained from the American Type Culture Collection. All cell lines were cultured with high-glucose Dulbecco’s modified Eagle’s medium (DMEM; HyClone) with 10% fetal bovine serum (FBS; Gibco) in 5% CO₂ 37°C incubator. 100 U/mL penicillin and 100 μg/mL streptomycin were added.

Preparation and optimization of mRNA

An HA tag was incorporated at the 3′ terminus of the VEGF-B open reading frame, which was subsequently inserted downstream of the T7 promoter within the pcDNA3.1 vector. Universal primers were designed based on the vector sequence to generate linear DNA templates suitable for IVT. The DNA template was obtained through PCR, and IVT was carried out using T7 RNA polymerase. After template digestion with DNA enzymes, poly(A) tailing was carried out using EPAP. For co-transcriptional tailing, it is accomplished by using a DNA template encoding a poly(A) stretch. IVT mRNA was purified with RNA Clean & Concentrator-25 kit (Zymo Research), followed by verification by agarose gel electrophoresis. Chemical modification was introduced to improve protein expression, which includes N1-methyl-pseudouridine triphosphate and pseudouridine. The introduction of chemically modified nucleotides was 100% substitution.

Construction and characterization of LNP

The ionizable lipid (DLin-MC3-DMA), cholesterol, 1,2-distearoyl-sn-glycero-3-phosphocholine, and polyethylene glycol-modified lipid (PEG2000-DMG) were dissolved in ethanol at a molar ratio of 50:10:38.5:1.5. mRNA was diluted in citrate disodium hydrogen phosphate buffer (pH = 4). The lipids dissolved in ethanol were assembled with in vitro-transcribed mRNA at a ratio of mRNA to cationic lipids of 1:3 (molar ratio) to encapsulate them into LNPs using the NanoAssemblr platform (Precision NanoSystems). The encapsulation efficiency and concentration of the LNPs were subsequently assessed using the RiboGreen assay, while the particle size distribution of the LNPs was analyzed with a Malvern dynamic light scattering Zetasizer.

Subcutaneous mouse tumor model for antitumor efficacy study

A total of 1 × 10⁶ B16F10 or MC38 cells in 100 μL Dulbecco's Phosphate Buffered Saline (DPBS) were subcutaneously implanted into the right flank of mice. Measurements of the tumor were taken every 2 days by the experimental endpoint. The formula used to calculate tumor volume was: v = a × b²/2, where a exceeds b.

Expression kinetics of mRNA encapsulated in LNPs

The subcutaneous tumor model was established by inoculating MC38 cells (1 × 10⁶ cells/mouse) into the right flank of C57BL/6 mice. After 10 days, luciferase mRNA-LNP was injected intratumorally. D-Luciferin was injected intraperitoneally at various time intervals (6, 24, 48, 72, and 96 h) following administration, and the bioluminescence intensity was measured and documented utilizing the IVIS Spectrum (PerkinElmer, USA).

Isolation of serum and TIF

Mice were anesthetized through inhalation of 2% isoflurane, after which blood was drawn by cardiac puncture employing a 1 mL syringe, resulting in roughly 400 μL of whole blood. This blood was allowed to clot at room temperature for 1 hour and then centrifuged at 1,500 × g and 4°C for 20 min to isolate the serum. The mouse was euthanized by CO₂ inhalation, and the tumor was excised, washed in PBS, and placed on a 40 μm filter. Following centrifugation at 4°C and 110 × g for 10 min, the TIF was harvested, immediately frozen in liquid nitrogen, and subsequently stored in a −80°C freezer.

Construction of LLC-Luc stable expression cell line

LLC cells were infected with lentivirus packaging the pLenti-CMV-Luc-Neo plasmid (provided by Prof. Yingjie Xu) to establish a stable LLC cell line overexpressing luciferase. The infection process involved incubating cells with virus concentrate and infection enhancer polybrene at 37°C 1,000 × g for 1 h to facilitate infection, followed by incubation for 12 h and subsequent culture in DMEM with 400 μg/mL G418 for 14 days to select stable clones. Clones with luminescence intensity above 100 lum/cell were identified using the PerkinElmer IVIS Spectrum imaging system.

Pulmonary metastasis mouse tumor model for antitumor efficacy study

The LLC-Luc cell line was cultured to a stable state, and 1 × 10⁶ cells were injected intravenously into mice. After 1 week, pulmonary bioluminescence signals were monitored using the IVIS system, and mice were randomly divided into groups. For the VEGF-B mRNA-LNP group, 10 μg of VEGF-B mRNA-LNP was administered every 2 days via intravenous injection (SORT LNP), totaling five doses. Control groups received EGFP mRNA-LNP or saline in an equivalent volume. In anti-PD-1 or combination treatment groups, 100 μL of anti-PD-1 antibody (1 μg/μL) was injected intraperitoneally every 4 days, totaling three doses (300 μg).

Immunoprecipitation

The culture supernatants from HEK293T or B16F10 cells were mixed with anti-HA magnetic beads and incubated overnight at 4°C. Following incubation, the beads were rinsed three times with ice-cold Radio-Immunoprecipitation Assay (RIPA) buffer. Subsequently, 20–40 μL of loading buffer was added to the beads, and the mixture was heated at 98°C for 10 min. The resulting supernatant was then collected and subjected to SDS-PAGE analysis.

Bioluminescence imaging

Bioluminescence imaging was conducted to quantify the photon flux emitted by LLC-Luc metastatic tumors, followed by comprehensive data analysis. Prior to imaging, mice were administered an intraperitoneal injection of d-luciferin (Yeasen) at a concentration of 20 mg/mL, with a dosage of 150 mg per kilogram of body weight, 10 min before the procedure. Once anesthetized with isoflurane, the mice were positioned within an IVIS imaging chamber (PerkinElmer). The bioluminescence intensity was captured and subsequently quantified using Living Image software.

Histology

For immunofluorescence staining, tumor tissues were fixed in 4% paraformaldehyde (PFA), dehydrated using a 30% sucrose solution, and subsequently embedded and frozen in optimal cutting temperature compound. The frozen samples were sectioned at a thickness of 10 μm and mounted onto adhesive glass slides (ServiceBio) for further analysis. Slides were washed with PBS and blocked with 3% BSA for 2 h before staining. Samples were incubated with antibodies targeting CD4 (GB15064, ServiceBio) and CD8 (GB15068, ServiceBio) for 2 h at room temperature protecting from light. After staining, samples were washed 3 times with PBS and mounted using Prolong glass antifade mountant (Thermo Fisher Scientific) and imaged using C1 laser scanning confocal microscope (Nikon). For IHC, tumors were fixed in 4% PFA, embedded in paraffin, and sectioned at 10 μm. To facilitate antigen retrieval, tissue sections were treated with citrate buffer adjusted to a pH of 6.0. Subsequently, the sections were subjected to a blocking step using a 3% hydrogen peroxide solution. Tissue sections were treated with an anti-HA-tag antibody (M180-3, MBL) during incubation. This was succeeded by the application of a biotin-conjugated secondary antibody and streptavidin-horseradish peroxidase. The resultant signal was then made visible through the use of diaminobenzidine as a chromogenic substrate. For H&E staining, harvested organs were preserved in 4% PFA. The preserved specimens were then encased in paraffin, sliced into 10-micrometer sections, and affixed to adhesive glass slides (ServiceBio) for staining with H&E. The prepared slides were subsequently examined and photographed using a Nikon E100 microscope.

ELISA

The concentration of VEGF-B protein was quantified through ELISA with the mouse VEGF-B ELISA kit (KOA0852, Rockland Immunochemicals Inc.), adhering strictly to the protocol provided by the manufacturer.

Surface and intracellular antibody staining

Cell surface staining was conducted for 30 min at 4°C using PBS enriched with 2% FBS, referred to as FACS buffer, in the presence of the following antibodies: CD3 (100222, BioLegend), CD4 (100422, BioLegend), CD8a (162305, BioLegend), CD45 (109828, BioLegend), CD11b (101216, BioLegend), NK1.1 (156504, BioLegend), CD25 (102051, BioLegend), PD-1 (135224, BioLegend), F4/80 (123135, BioLegend), B220 (103212, BioLegend), Ly-6C (128005, BioLegend), Ly-6G (127624, BioLegend), I-A/I-E (107607, BioLegend), and CD11c (117317, BioLegend). For Annexin V, samples were washed 3 times with Annexin V binding buffer and then stained with fluorescein isothiocyanate Annexin V (640922, BioLegend) after surface staining. Before surface staining, samples were preceded by a 15-min incubation with TruStain FcX PLUS (anti-mouse CD16/32) antibody (156604, BioLegend) and Zombie fixable viability dye (BioLegend). For intracellular cytokine staining, samples were ex vivo restimulated with cell activation cocktail (423303, BioLegend) for 5 h; fixed and permeabilized using the Cyto-Fast Fix/Perm buffer set (426803, BioLegend); and stained with antibodies against IFN-γ (505830, BioLegend), TNF-α (506306, BioLegend), and Granzyme B (372208, BioLegend).

Flow cytometry

Flow cytometry was conducted with BD Fortessa or Beckman CytoFLEX S/LX, and the data were processed using FlowJo 10. Graphs and statistical analyses were performed with GraphPad Prism 10.

Statistics

Unpaired two-tailed Student’s t tests were used for comparisons between two groups, while one-way or two-way ANOVA followed by Holm-Šídák, Bonferroni post hoc tests were employed for analyses involving more than two groups. Survival analysis was determined using the log rank (Mantel-Cox) test. At least three independent experimental results are presented. SEM and p values were computed using GraphPad Prism 10, with p < 0.05 being statistically significant.

Data availability

All data can be accessed by contacting the corresponding author, Jinke Cheng, with a reasonable request.

Acknowledgments

The authors acknowledge the support of lab members, Qiuju Fan, Guoyuan Peng, Prof. Hongsheng Tan, Prof. Rong Cai, Prof. Tianshi Wang, Prof. Jiao Ma, Prof. Yong Zuo, Dr. Yalan Chen, Kexin Liu, and Yirong Zhang, from the Department of Biochemistry and Molecular Cell Biology of SJTUSM. All experiments conducted on mice complied with the Guide for the Care and Use of Laboratory Animals, receiving approval from the Experimental Animal Ethical Committee at SJTUSM.

This research received financial support from the National Key Research and Development Program of China (grant no. 2020YFA0803600), National Natural Science Foundation of China (grant no. 32170773), and the Science and Technology Commission of Shanghai Municipality (grant no. 22ZR1435700).

The graphical abstract and Figures 3A and 4A were created with BioRender.com with licensing rights for publication.

Author contributions

G.Z. designed and carried out the experiments, analyzed the results, prepared the figures, and drafted the manuscript. J.T., H.T., J.H., and Q.F. assisted in enhancing the experimental process. G.P. provided assistance with the subcutaneous tumor model and flow cytometry. Yirong Zhang and M.Z. contributed to the lung metastasis model. Yu Zhang and Y.X. prepared mRNA-LNPs and helped revise the manuscript. J.C. helped with experiment design and manuscript revision. All authors read, verified, and approved the manuscript.

Declaration of interests

All authors declare no competing interests.