RNF8 enhances the sensitivity of PD-L1 inhibitor against melanoma through ubiquitination of galectin-3 in stroma

doi:10.21203/rs.3.rs-2353854/v1

Download PDF

Article

RNF8 enhances the sensitivity of PD-L1 inhibitor against melanoma through ubiquitination of galectin-3 in stroma

https://doi.org/10.21203/rs.3.rs-2353854/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The failure of melanoma immunotherapy can be mediated by immunosuppression in the tumor microenvironment (TME) and insufficient activation effector T cells to against tumor. Here, we show that inhibition of galectin-3 (gal-3) enhances the infiltration of T cells in TME and improves the sensitivity of anti-PD-L1 therapy. We identify that RNF8 downregulated the expression of gal-3 by K48-polyubiquitination and promoted gal-3 degradation via ubiquitin proteasome system. RNF8 deficiency in host but sufficiency in implanted melanoma results immune exclusion and tumor progression due to up-regulation of gal-3. Up-regulation of gal-3 decreased the immune cell infiltration by restricting IL-12 and IFN-γ. Inhibition of gal-3 reverses immunosuppression and induces immune cells infiltration in tumor microenvironment. Moreover, gal-3 inhibitor treatment can increase the sensitivity of PD-L1 inhibitors via increasing immune cells infiltration and enhancing immune response in tumor. This study reveals a previously unrecognized immunoregulation function of RNF8 and provides a promising strategy for therapy of “cold” tumors. Tremendous effects of melanoma treatment can be achieved by facilitating immune cell infiltration combined with anti-PD-L1 treatment.

Biological sciences/Immunology/Immune evasion

Health sciences/Diseases/Skin diseases/Skin cancer/Melanoma

Tumor microenvironment

immunotherapy

melanoma

RNF8

galectin-3

PD-L1

Targeted PD-1/PD-L1 therapy is an effective treatment for melanoma¹. PD-1/PD-L1 blocking antibodies serve as key factors in TME, by re-invigorating exhausted T cells and thereby reviving pre-existing anti-tumor immunity². However, more than half of the patients failed to achieve the therapeutic effect, due to the extremely complicated internal environment and variable objective response rate. Researches of the failure on elimination tumor suggest that it is mainly because of the insufficient immune response against tumor³. Thus, strategies to enhance immune response and remove immunosuppression have been considered as the central goal to improve therapeutic effect.

TME is characterized as the multiple complex components and signal crosstalk, which is composed of multiple cells and extracellular matrix (ECM) providing structural support for tumor cells, energy source and signal transduction. In solid tumors, insufficient infiltration of cytokines and immune cells restricted by ECM components result in poor immune response and tumor immune escape, which is closely related to the immunotherapy effect⁴.

Gal-3 presents high binding affinity for β-galactosidase and glycosylated molecules. Studies have shown that gal-3 is increased in various tumors, and it’s up-regulation in TME is closely related to immune response and tumor progression⁵. Emerging evidences suggest that TME may induce gal-3 expression to keep cellular homeostasis and promote cell survival⁶. Gal-3 could suppress activation of cytotoxic T cells, and shield ligands on the surface of tumor cells, thus leading to immunosuppression⁷. Gal-3 has been considered to be a potential target for clinical therapy since blocking gal-3 could improve efficiency of tumor treatment⁸. Nevertheless, the underlying mechanisms that gal-3 regulating immunosuppression in TME remain substantially unelucidated.

RNF8 is a ubiquitin E3 ligase expressed ubiquitously in multiple tissues. Some previous works have improved that RNF8 plays essential roles in histone ubiquitination in DNA damage response⁹, spermatogenesis⁵, neurodegeneration¹⁰ and epithelial-mesenchymal transition (EMT)¹¹. In this study we found that RNF8 plays an essential role in maintaining stability of TME. We demonstrated that depletion of RNF8 in vivo induced gal-3 expression and suppressed immune cells infiltration as well as the cytokines recruitment in TME, leading to increase the tumor progression compared with RNF8 wildtype (RNF8^+/+) mice. Gal-3 inhibited immune cells infiltration in TME through binding with cytokines IL-12 and IFN-γ, and the internal environment of tumor regulated gal-3 expression and distribution. Moreover, RNF8 regulated gal-3 K48 linked polyubiquitination to promote gal-3 degradation via ubiquitin-proteasome system. Gal-3 intervention collaborated with PD-L1 inhibition improved antitumor effect significantly, which provided a promising therapeutic strategy against tumor through converting immune excluded TME into immune infiltrated TME.

Cells lines and cells culture

A375, A549, B16F10 and 293T cells were obtained directly from Procell, Wuhan, China. A375 and 293T cells were cultured at 37°C with 5% CO₂ in complete DMEM medium (HyClone, SH30022.01) supplemented with 10% FBS (Gibco, 10100147C) and 1% penicillin/ streptomycin (Solarbio, P1400). A549 and B16F10 cells were cultured at 37°C with 5% CO₂ in complete 1640 medium (Cytiva, SH30809.01) supplemented with 10% FBS and 1% penicillin/streptomycin.

Animals

The 3-month-old RNF8^−/− and RNF8^+/+ male mice were used in our studies. The RNF8^−/− mice were provided by Xiaochun Yu as described previously¹². The gene deficient embryonic stem (ES) cell clone RRR260 was used to produce the RNF8^−/− mice. We used RNF8^+/− mice to produce RNF8^−/− mice. And the expression of RNF8 was detected in supplementary Fig. 1C. All animals were maintained in the SPF laboratory at the Lanzhou University Animal Center. The animals were fed under standard housing conditions and given free access to food and water. The experimental protocols were approved by the Committee on the Ethics of Animal Experiments of Lanzhou University.

RNA and plasmids interference

Lentiviral vectors encoding human RNF8 and scrambled control shRNA were obtained from Genechem company. The specific sequences of LV-RNF8-RNAi was shown in Table S1 of supplementary materials. We transfected the A375 and 293T cells by using hitransG P (GeneChem Co., Shanghai, China). The cDNA sequences of LGALS3 and UBB gene were delivered into GV362-CMV × 3(Flag, HA or HIS)-Neo and GV658-CMV × polyA-puro vector for conditional expression (GeneChem Co., Shanghai, China). The mutant construct of LGALS3 (K176R) was generated by site-directed mutagenesis and cloned into GV362-CMV × 3Flag-Neo vector (GeneChem Co., Shanghai, China). Lipofectamine 3000 regents (Invitrogen, L3000001) was used for transient transfection according to the manufacturer’s instructions. The transfection efficiency was detected by fluorescent microscopy (supplementary Fig. 1A-B) after 48 h.

Mouse xenograft tumor models

The 3-month-old RNF8^−/− and RNF8^+/+ male mice were used to build tumor model. Then, 5×10⁶ B16F10 cells were subcutaneously injected into the flank of RNF8^+/+ and RNF8^−/− mice respectively. Tumor size were recorded every day from palpating the tumor. LacNAc (FuShen, FS0234), the inhibitor of gal-3 was injected into the tumor center at a concentration of 0.1 µmol per tumor from Day 7, and injected every two days. PD-L1 inhibitor (AbMole, PD-1/PD-L1 inhibitor 1, M8959) was injected into the abdomen at a concentration of 1 µmol per mouse after one weeks. And one weeks later, the mice accepted again PD-L1 inhibitor therapy. These treatments were showed in supplementary Fig. 1D. Animals were sacrificed and subcutaneous tumors were harvested at the end of experiments.

Mass cytometry

For CyTOF experiments, the mouse intracellular cytokine I panel kit was purchased from Fluidigm, the CyTOF antibodies in mouse lymphocytes phenotyping panel was purchased pre-conjugated directly from the supplier (PLTTECH, China). And the CyTOF antibodies and intracellular cytokine panel were listed in Table S2 and Table S3. TILs in melanoma were isolated using mouse tumor-infiltrating tissue lymphocyte separation medium (Solarbio, P9000). After passing through a 100 µm filter membrane, the cells were washed twice with PBS and prepared for staining. 3×10⁶ cells were prepared for staining and test. EQ four element calibration beads (Fluidigm) were used to normalize the signal intensity according to manufacturer’s instructions. The files were normalized and analyzed using Cytobank platform on website to gate the cell populations and handle high dimensional data.

t-distributed Stochastic Neighbor Embedding (t-SNE) Analysis

A random subset of 10⁴ cells was selected for each sample and incorporated into a single expression matrix prior to t-SNE analysis. In addition to protein markers in t-SNE analysis, beads, event length, intercalator, viability, center, offset, residual, and time channel were removed from the expression matrix. A total of 10⁵ cells and 8 markers (Table S2) were used to create a t-SNE map of the TILs.

Spanning Tree Progression of Density Normalized Events (SPADE) Analysis

Our data types were stored as flow Cytometry standard (FCS) files, compatible with many analysis packages, and were represented as a data matrix with rows corresponding to individual cells and columns corresponding to markers of interest (CD3⁺ CD4⁺, CD3⁺ CD8⁺, NK). The generation of optimal SPADE input data required the selection of live and singlet events or cells, using DNA gates and event length basing constraints for large-scale cell assays or some form of cell viability probe.

Co-immunoprecipitation

A375 and 293T cells were lysed in IP binding buffer (BeaverBeads protein A Immunoprecipitation Kit, 22202-20). After centrifugation, the supernatants were collected and incubated with protein A/G magnetic beads and 5 mg antibodies for 4 h at 4^oC followed by washing three times with IP buffer. The components of immunoprecipitate were investigated by SDS-PAGE gels, then followed by transferring onto the PVDF membranes and analyzed by corresponding antibodies. In the experiment of IP ubiquitin, proteasome inhibitors (MG132- HY-13259, MCE; Epoxomicin- HY-13821, MCE) and deubiquitinase inhibitors (PR-619- HY-13814, MCE) were additional needed.

Galectin-3-coated beads and in vitro bioassays

Dynabeads M-280 Streptavidin (11-205-D, Invitrogen) were incubated with saturating amounts of recombinant human gal-3 (10289-H08H1, sinobiological) overnight at 4°C. Then, beads were washed three times with PBS (containing 2 mM EDTA and 0.05% Tween-20). Under adding lactose (Solarbio, L8911), LacNAc or PBS, different amounts of gal-3-coated beads were incubated with the cytokines (IFN-γ, 11725-HNAS, sinobiological; IL-12, 10052-H02H, sinobiological) for 1 h at room temperature on a shaker. After magnetic sorting of the beads, the cytokines present in the supernatants were measured by ELISA (IFN-γ-JL12152, JiangLai; IL-12- JL29081, JiangLai), and the technical route was shown in Fig. 3F.

Cone beam CT (CBCT) scanning and 3D imaging

Mice with subcutaneous melanoma were scanned by X-RAD SmART device (Precision X-Ray, INC) at Experimental Center of Lanzhou University before they sacrificed. Firstly, CBCT Scout was choosed to acquire a rapid, low dose and resolution image, then a high resolution CBCT was runed. The DCMFiles were opened by radiantviewer. And the software was used for 3D modeling of subcutaneous tumor in mice.

Statistical analysis

Statistical analysis was performed by GraphPad Prism 8 software. Data are presented as the mean ± standard deviation for each sample. Student’s t test was used to determine the statistical differences between two groups. The data of integral optical density was obtained by ImageJ software. P < 0.05 was considered as statistically significant.

RNF8 deficiency in stroma accelerated progression of melanoma and suppressed immune cells infiltration in TME.

To investigate the effect of RNF8 deficiency in tumor stroma on TME, RNF8^+/+ and RNF8^−/− mice were inoculated with melanoma B16F10 cells respectively. Knockout of RNF8 in host mice significantly accelerated progression of melanoma (Fig. 1A-B). HE staining showed that fewer lymphocyte cells were observed in TME of RNF8^−/− mice (Fig. 1C-D). S100 and HMB45 positive cells in melanoma were enhanced in RNF8^−/− mice (Supplementary Fig. 2A-B), which implied that RNF8 deficiency in TME promoted melanoma progression.

Subsequently, we detected the immune cells infiltration of melanoma in RNF8^+/+ and RNF8^−/− mice by mass cytometry (CyTOF). SPADE analysis distinctly identified the variant immune cell populations in TME between RNF8^+/+ and RNF8^−/− group (Fig. 1E). In-depth analysis of immune cells by t-SNE identified CD3⁺ T cells, CD3⁺ CD4⁺ T cells, CD3⁺ CD8⁺ T cells and NK cells clusters, moreover, these clusters were decreased in melanoma-beared RNF8^−/− mice (Fig. 1F-G). For investigation differential pattern of immunity in melanoma, we defined 75 percent in central area as the tumor center, and the rest as tumor periphery (Supplementary Fig. 2C). Results revealed that CD3 expression was significantly higher in tumor periphery than center, moreover, CD3 expression in both tumor center and periphery from RNF8^−/− mice was more obvious lower than from RNF8^+/+ mice (Fig. 1H-I, Supplementary Fig. 2D-E).

RNF8 mediated infiltration of T cells in TME through IL-12/IFN-γ axis.

By post-translational modifications omics of ubiquitination, we found that immune response of related signal molecules was abnormal remarkably, in which the cytokines, IL-12 and IFN-γ, caused attention due to their immune regulation effect (Fig. 2A-C). Studies have indicated that IL-12 and IL-2 could activate T cells and NK cells to promote IFN-γ generation by helper T-cells¹³. The IL-12 and IFN-γ expression in melanoma from RNF8^−/− mice were reduced (Fig. 2D). Furthermore, we investigated expression difference of IL-12 and IFN-γ in tumor center and periphery. Surprisingly, expression of IL-12, IFN-γ, CXCL-9 and CXCL-10 in tumor center was dramatically decreased than tumor periphery (Fig. 2E-I). Consistently, IHC staining of IL-12 and IFN-γ in tumor got the similar results (Fig. 2J-M). The remarkable distribution difference was more typical in RNF8^−/− mice, cytokines in tumor center performed remarkable decreased. The specific distribution pattern of cytokines exactly conformed with the infiltration of T cells in TME. Therefore, RNF8 could mediate T cells infiltration by regulating IL-12/IFN-γ axis expression in TME.

The distribution and content of IL-12 and IFN-γ was regulated by gal-3 in TME.

Cytokines are identified as the key mediators of intercellular communication and immune activation in TME, which play important roles in immune response and anti-tumor. The expression dysregulation of cytokines has been indicated in all the tumor types that have been detected. Based on the remarkable distribution and expression difference of IL-12 and IFN-γ in melanoma particular in RNF8^−/− mice, we investigated ubiquitination modifications omics data of control and RNF8-knockdown cells (Fig. 3A-C). Analysis revealed that RNF8 deficiency related to immunity, ubi-conjugation pathway and intracellular signal transduction. Among these genes, gal-3, a lectin that involved in bind with various proteins, was chosen for further investigation. Western blot detection showed that gal-3 was increased in tumor-beared RNF8^−/− mice (Fig. 3D). Gal-3 was highly expressed in tumor center than periphery. Moreover, tumors from RNF8^−/− mice showed more gal-3 expression than from RNF8^+/+− mice (Fig. 3E). Gal-3 has been reported to increase in multiple tumors and bind with glycoproteins by forming lattices, such as cytokines which exist in glycosylated form mostly in tumor¹⁴. Then we tested the interaction of gal-3 with cytokines IL-12 and IFN-γ. The dynabeads were incubated with recombinant human gal-3, then IL-12 and IFN-γ were incubated with gal-3-coated beads respectively with lactose or LacNAc (LAC), the content of IL-12 and IFN-γ in solution was detected by ELISA for evaluating the capture status (Fig. 3F). Results indicated that IL-12 and IFN-γ were bound by gal-3 in a dose-dependent manner, and were released under the condition of lactose or LAC presence due to competitive binding (Fig. 3G-H).

Next, we investigated the gal-3 generation in tumor. Hypoxia could enhance transcription of gal-3¹⁵. Due to the rapid growth of tumors, blood vessels formation near tumor center is commonly deficient, leading to insufficient energy supply of ischemia and hypoxia, and increasing hypoxia inducible factor-1α (HIF-1α) expression. Na₂SO₃ was used for establishing hypoxic cell model. After intervention for 0, 12 and 24 h in A375 and A549 cells, HIF-1α in cytoplasm and nucleus was detected (Fig. 3I-K). Results indicated that HIF-1α nuclear translocation was elevated in a time-dependent manner, consist with the conclusion that HIF-1α regulates gene transcription via nuclear translocation from cytoplasm¹⁶. NF-κB is another important regulator of gal-3 contributed to tumor growth and metastasis¹⁷. Our results demonstrated that phosphor-nuclear factor kappa B (p-NF-κB) was increased in nucleus with time, meanwhile, Phospho-Inhibitory Subunit of NF Kappa B Alph (p-IκBα) in cytoplasm was enhanced (Fig. 3I-K). The expression of HIF-1α in tumor center was increased compared with tumor periphery, however, the expression of HIF-1α was no significant change in tumor-beared RNF8^+/+ and RNF8^−/− mice (Fig. 3L). Collectively, the different distribution of gal-3 in tumor was caused by multiple factors (Fig. 3M). These results suggest that the expression of gal-3 is mainly regulated by RNF8, thus led to the unbalance of cytokines and immune cells in TME.

RNF8 interacted with gal-3.

As shown in Fig. 4A, a remarkable enhancement of gal-3 was observed after RNF8 knockdown. In contrast, RNF8 over-expression led to a depression of gal-3 (Fig. 4B). The result was confirmed by immunofluorescent counterstaining of RNF8 and gal-3 (Fig. 4C-D). In addition, immunofluorescent staining showed that RNF8 and gal-3 was co-localized in cytoplasm (Fig. 4E), indicating the potential links between RNF8 and gal-3. An interaction between endogenous RNF8 and gal-3 was observed (Fig. 4F-G), this was confirmed by interaction between Flag-RNF8 and HA-gal-3 over-expressed in both 293T and A375 cells (Fig. 4H-I). These data clearly suggested that gal-3 interacted with RNF8 and was negatively regulated by RNF8.

RNF8 mediated degradation of gal-3 via promoting K48-linked ubiquitination.

The cycloheximide (CHX) chase assays indicated that RNF8 supplementary (Flag-RNF8) significantly enhanced gal-3 (HA-gal-3) degradation in a time-dependent manner (Fig. 5A). The proteasome inhibitor MG132 treatment partially enhanced endogenous ubiquitinated gal-3 in 293T and A375 cells respectively (Fig. 5B, Supplementary Fig. 3A). Consistently, ubiquitinated HA-gal-3 was increased obviously once supplemented exogenous ubiquitin (His-ub) in two cell lines (Fig. 5C, Supplementary Fig. 3B). Next, exogenous His-ub and HA-gal-3 was used to detect the level of ubiquitinated gal-3. The ubiquitinated exogenous gal-3 was reduced in RNF8-KD (Flag-shRNF8) cells but elevated after RNF8 over-expressed (Fig. 5D, Supplementary Fig. 3C). The whole ubiquitination level of gal-3 by endogenous ubiquitin revealed similar results (Fig. 5E, Supplementary Fig. 3D).

Then we compared the K48 and K63 ubiquitination level of gal-3 in RNF8-KD and RNF8 over-expressed cells. The K48-ubiquitinated gal-3 was significantly reduced in RNF8-KD cells compared with control, and RNF8 over-expression enhanced the K48 polyubiquitination of gal-3 (Fig. 5F, Supplementary Fig. 3E). However, the K63-ubiquitination of gal-3 showed no response to the intervention (Fig. 5G, Supplementary Fig. 3F). To explore the necessary residue required for gal-3 ubiquitination, we constructed the gal-3 mutant according to LC-MS/MS analysis of purified gal-3 protein (Supplementary Fig. 3G), in which lysine residue was replaced with arginine residue. The cycloheximide chase assays performed that K176R significantly reversed HA degradation in 293T cells (Fig. 5H). As expected, K176 mutation abolished the K48-ubiquitination of gal-3 in both 293T and A375 cells (Fig. 5I-K, Supplementary Fig. 3H-I). Collectively, RNF8 regulated gal-3 K48 linked polyubiquitination to promote gal-3 degradation via ubiquitin-proteasome system.

Intervention of gal-3 remolded the tumor immune microenvironment.

It has been reported that LacNAc (LAC) plays a significant affinity with gal-3¹⁸. In this study, B16F10 cells were injected subcutaneously into RNF8^+/+ and RNF8^−/− mice, and then LAC was injected into tumor center. Results showed that long-term treatment with LAC delayed melanoma growth and progression in both RNF8^+/+ and RNF8^−/− groups (Fig. 6A-B). HE exhibited that LAC treatment improved immune cells infiltration in TME and suppressed tumor growth (Fig. 6C-D, Supplementary Fig. 4A). IHC staining of CD3⁺ T cells infiltration in melanoma was increased after LAC intervention, particularly in RNF8^−/− mice, which performed fewer immune cells infiltration in TME (Supplementary Fig. 4B-C).

Subsequently, immune cells subsets in tumor of RNF8^+/+ and RNF8^−/− mice were investigated by CyTOF. The t-SNE analysis was presented for visualized observation of the immune cell populations changes of LAC treated melanoma in RNF8^+/+ and RNF8^−/− group (Supplementary Fig. 4D). SPADE analysis presented the subsets ratio change in RNF8^+/+ and RNF8^−/− mice treated with LAC, compared with that in treated with DMSO (Fig. 6E-F). The CD3⁺CD4⁺, CD3⁺CD8⁺ and NK cells in tumor from RNF8^−/− mice were obviously lower than RNF8^+/+ mice, LAC treatment significantly increased infiltration of CD3⁺CD4⁺, CD3⁺CD8⁺ and NK cells in tumors both in RNF8^+/+ and RNF8^−/− mice (Supplementary Fig. 5). Consistently, CyTOF analysis performed that the pro-inflammatory cytokines IL-2 and IFN-γ expression was decreased, and anti-inflammatory cytokines IL-4 and IL-10 level was increased in RNF8^−/− mice compared with RNF8^+/+ mice, interestingly, LAC treatment improved the infiltration of cytokines in RNF8^−/− mice (Supplementary Fig. 6).

IHC staining analysis demonstrated that IL-12 and IFN-γ positive cells in tumor center were increased in RNF8^+/+ and RNF8^−/− mice after LAC treatment (Fig. 6G-H, Supplementary Fig. 7A-B). Western blot presented that LAC intervention for a long-term partially reduced gal-3 expression in tumor center and periphery. Surprisingly, the increasing of CD3, IL-12, IFN-γ, CXCL-9 and CXCL-10 expression in LAC treated tumors was much higher than tumors treated with DMSO (Fig. 6I, Supplementary Fig. 8). These data suggest that inhibition of gal-3 could relieve immunosuppression through improving cytokines and immune cells infiltration even in the harsh TME.

Inhibition of gal-3 enhanced sensitivity of PD-L1 inhibitor for melanoma.

The immune checkpoint blockade has been verified the effectiveness on clinical treatment, however, due to the poor immune infiltration, more than half of the patients fail to benefit from the therapy such as anti-PD-1, anti-LAG-3 and anti-CTLA-4. Based on the previous results, we treated melanoma with combined intervention including gal-3 inhibitor and PD-L1 inhibitor to study the effect on immune remodeling and tumor progression. The results showed that combined therapy of PD-L1 inhibitor and LAC decreased tumor growth dramatically than single using of PD-L1 inhibitor or LAC (Fig. 7A-D). IHC staining performed that the Ki-67 positive cells in tumor were less in combined therapy group compared with single treating group (Fig. 7E-F). Consistently, HE analysis revealed that the infiltration of immune cells was increased in combined therapy group (Fig. 7G-H). Western blot data also indicated that cytokines content in tumor was remarkably increased after combined therapy compared with other groups, including IL-12, IFN-γ, CXCL-9 and CXCL-10 (Fig. 7I). Collectively, gal-3 inhibition combined with PD-L1 inhibition treatment performed a prominent efficacy of melanoma through enhancing immune infiltration and removing immunosuppression in TME.

In respect of recognition, responding and elimination tumors, immune cells play a decisive role. These immune cells may process tumor-antagonizing or tumor-promoting functions. Although the tumor-antagonizing immune cells target and kill tumor in the early stage of tumorigenesis, multiple cellular molecular have changed in TME to help tumor cells adapt to the microenvironment and maintain growth, which lead to tumor escape and treatment resistant¹⁹. ²⁰. Current studies mainly focus on remodeling TME, blockading immunosuppression and enhancing immune cells response to improve effect of tumor treatment. Recently, the immune checkpoint modulators (represented by anti-CTLA4 and anti-PD-1/PD-L1 antibodies) have presented unexpected antitumor effect²¹. This study focused on TME component and indicated a new mechanism about regulating TME to prevent melanoma progression. We identified that RNF8-mediated ubiquitination was a crucial post-translational modification of gal-3. We investigated regulation mechanism of immune exclusion by RNF8 mediated gal-3 ubiquitination via ubiquitin-proteasome system in TME, and the effect of gal-3 inhibition combined with immune checkpoint inhibitor treatment on melanoma.

Actually, most of the components in TME generate mainly from the host besides the tumor cells. And the components of TME are of importance in regulating immune escape and tumor progression. The current studies explore the effect of RNF8 on progression and therapy resistant of tumor cells, based on the mechanisms related to DNA damage repair and genomic instability²². In this study, we demonstrated that RNF8 is an important regulator of gal-3, RNF8 deficiency of host led to increase the gal-3 level in TME and promote implanted tumor progression (Fig. 1, Fig. 3). Posttranslational modifications such as phosphorylation, ubiquitination, methylation and proteolytic cleavage play an indispensable role in regulating the protein functions. Most studies related to gal-3 focused on its inhibition and proteolytic cleavage, few studies had been conducted on ubiquitination of gal-3 and its function changes^{23, 24}. Previous study showed that Calpain4 was critical to gal-3 tyrosine phosphorylation and ultimate secretion from the cell²⁵. A recent study proposed that gal-3 could be efficiently cleaved by purified extracellular 20S proteasome and thus disable function²⁶. Our results indicated that RNF8 interacted with gal-3, whereas gal-3 was negatively regulated by RNF8 (Fig. 4). The ubiquitination assay in vitro confirmed that gal-3 was degraded through ubiquitin proteasome system by RNF8 (Fig. 5). Furthermore, we found that knockdown of RNF8 reduced gal-3 ubiquitination level especially the K48-linked ubiquitination, but showed no effect on K63-linked ubiquitination (Fig. 5). Dramatically, mutation the specific site of gal-3 K176 abolished the K48-linked ubiquitination, and also showed no effect on K63-linked ubiquitination. These results illustrated a clear regulatory relationship, RNF8 mediated gal-3 K48 linked polyubiquitination and promoted gal-3 degradation via ubiquitin proteasome system.

The matrix remodeling enzymes, cytokines and metabolites keep changing constantly and interact with each other, affect tumor progression and therapy cooperatively²⁷. Cytokines can be generated commonly by cells such as endothelial cells and fibroblasts for recruiting immune cells²⁸. Nevertheless, components in TME could sequester cytokines, chemokines and other molecules to contact with immune cells and establish an immune excludes environment. An important regulatory element is glycoprotein gal-3, which has been suggested to influence immune cells and regulate their function negatively²⁹. Gal-3 can be secreted and dispersed to the cell surface and ECM, then binds with the glycosylated proteins to form lattices and restrains the diffusion of the glycosylated soluble factors in TME³⁰. The restraint of cytokines by gal-3, such as IL-2 and IFN-γ, reduced downstream signal molecules production and function, thus resulting immune exclusion in TME³¹. Moreover, the dense collagen deposition in TME has been proposed as an obstacle for by immune cells infiltration into the tumor, gal-3 accumulation contribute to collagen deposition and thus intercept immune cells entry into TME³². In this study, we demonstrated that RNF8 deficiency induced gal-3 high expression in TME, and accelerated melanoma growth with a deteriorated TME. In vitro experiment indicated that gal-3 could bind with IL-12 and IFN-γ in a dose dependent manner, and lactose (LAC) intervention could competitively bind to gal-3 therefore release the cytokines (Fig. 3). Consistently, RNF8 knockout of host resulted in significantly decreased IL-12 and IFN-γ in tumor, which was directly related to gal-3 elevation (Fig. 2). Nevertheless, LAC injection in tumor effectively liberated IL-12 and IFN-γ and thus elevating downstream chemokines CXCL-9 and CXCL-10, particularly in RNF8 deficient mice (Fig. 6).

Several studies have suggested that IL-12 related cytokine therapy combined with immunotherapy present effective synergistic antitumor effect³³. In clinical melanoma treatment, delivery of IL-12 gene into the tumor elicited partially complete regression and tumor suppression in metastatic melanoma³⁴. Recent studies also proposed that effective antitumor responses required sufficient IL-12 produced by dendritic cells upon sensing IFN-γ released by T cells, a crosstalk between DC and T cells, to stimulate T cell immunity³⁵. Our data confirmed that decreased pro-inflammatory cytokines (IL-2, IFN-γ, IL-12) and elevated anti-inflammatory cytokines (IL-4, IL-10) in tumor-beared RNF8^−/− mice, yet inhibition gal-3 treatment effectively increased pro-inflammatory cytokines and decreased anti-inflammatory cytokines infiltration in TME (Supplementary Fig. 6). Furthermore, inhibition gal-3 could release of more IL-12 and IFN-γ in TME, which enhances the sensitivity of PD-L1 inhibitor treatment for melanoma and presents a more significant antitumor effect.

Gal-3 has been suggested to inhibit T cell activation and function through promoting TCR down-regulation, and knockout of gal-3 increased TCR and IFN-γ expression in CD4⁺ T cells³⁶. Studies indicated that gal-3 could bind with LAG-3 on CD8⁺ T cells and inhibit the cell function³⁶. Moreover, co-culture gal-3 deficient melanoma cells with CD8⁺ T cells significantly induced IFN-γ level in CD8⁺ T cells, and promoted tumor reactive T cells expansion³⁷. In this study, we found that the infiltration of fewer immune cells in melanoma-beared RNF8 deficient host compared with control host, including CD4⁺ T, CD8⁺ T and NK cells (Fig. 2). Likewise, intervention gal-3 obviously increased immune cells infiltration in tumors in both RNF8^+/+ and RNF8^−/− mice (Supplementary Fig. 5). These results suggested that RNF8 played a pivotal role for maintaining immune response in TME, through restricting gal-3 expression to ensure the cytokines and chemokines and immune cells infiltration.

Studies have indicated that the center and periphery of tumor perform high epigenetic heterogeneity and transcriptional heterogeneity. It is widely known that hypoxia is a critical element to affect tumor diversity and progression, which is associated with energy metabolism, oxygen supply and tumor angiogenesis³⁸. Actually, hypoxia occurs in most solid tumor interior during the development because of the rapid growth and energy competition of tumor cells. Meanwhile, the development of hypoxic regions within rapidly growing tumors could also improve T-cell suppression via the increased production of adenosine in such areas³⁹. It has been suggested that a lower methylation in tumor core is relating to hypoxia, speculating that hypoxia in TME contributes to epigenetic changes and is related to survival of patients⁴⁰. Clearly, the heterogeneity in tumors will determine tumor adaptability and progression, and even affect tumor therapy effect⁴¹. We investigated the expression difference of tumor center and periphery. Exactly, gal-3 expression in tumor center was higher than that in periphery especially in RNF8^−/− host (Fig. 3). As well, distribution pattern of cytokines and chemokines in tumor performed site dependent, low central expression and high peripheral expression (Fig. 2), which was opposite to gal-3 distribution. Then we detected correlation between hypoxia and gal-3 expression by cell and mice models. The nuclear translocation of gal-3 regulators, HIF-1α and p-NF-κB, was increased in a time dependent manner under condition of hypoxia. What ' s more, HIF-1α level in tumor center was increased compared with tumor periphery, yet it was no significant difference between RNF8^+/+ and RNF8^−/− mice (Fig. 3). It suggests that the high expression of gal-3 in tumor center and periphery from RNF8^−/− mice is related with the degradation of gal-3 via ubiquitin-proteasome system.

By exploring the RNF8/gal-3/cytokine axis action in TME, we acquired a better understanding in tumor immune exclusion, and designed effective combination therapeutic strategy for melanoma. We treated melanoma with gal-3 inhibitor and PD-L1 inhibitor, which exhibited superior effect than separately treatment in suppression of tumor development (Fig. 7), consistent with the previously studies on tumor therapy by inhibition gal-3 and immune checkpoint synchronously⁴². In this study, combination therapy recovered IL-12/IFN-γ positive feedback mechanisms and induced immune cells infiltration in TME, through removing restriction of cytokines by suppressing gal-3, meanwhile, shutting down the PD-L1 function of tumor cells by inhibition of immune checkpoint. This strategy promoted remodeling of immune microenvironment and enhanced immune response in melanoma, and effectively inhibited tumor progression. Collectively, these results strongly indicated the effectiveness of this “cocktail strategy”, which provided a potential platform for therapy of “cold” tumors.

In summary, the present study indicated an essential regulatory role of RNF8 mediated ubiquitination degradation of gal-3 in maintaining immune response in TME. Due to the crucial effect of gal-3 on restriction of cytokines and subsequent immune infiltration, we propose that targeting gal-3 and combining with immune checkpoint inhibition may offer a potential strategy for melanoma immunotherapy.

Authors Contributions

Yanan Guo, Rong Shen conceived and performed experiments, analyzed data, and wrote the manuscript. Keren Yang, Yutong Wang and Haoyun Song performed experiments. Xiangwen Liu, Xin Cheng and Rile Wu analyzed data. Yanfeng Song and Degui Wang conceived and oversaw the study.

Competing interests

The authors declare no conflicts of interests.

Data Availability Statement

All relevant data can be obtained from the corresponding author upon request.

Funding

This work was financially supported by National Natural Science Foundation of China (No. 82071695, 82060535, 82203630), Natural Science Foundation of Gansu Province (No. 21JR7RA450).

Arozarena I, Wellbrock C. Phenotype plasticity as enabler of melanoma progression and therapy resistance. Nature reviews. Cancer 2019;19:377-91.
Dammeijer F, van Gulijk M, Mulder EE, Lukkes M, Klaase L, van den Bosch T, van Nimwegen M, Lau SP, Latupeirissa K, Schetters S, van Kooyk Y, Boon L, et al. The PD-1/PD-L1-Checkpoint Restrains T cell Immunity in Tumor-Draining Lymph Nodes. Cancer Cell 2020;38:685-700.e8.
Robert C, Ribas A, Schachter J, Arance A, Grob JJ, Mortier L, Daud A, Carlino MS, McNeil CM, Lotem M, Larkin JMG, Lorigan P, et al. Pembrolizumab versus ipilimumab in advanced melanoma (KEYNOTE-006): post-hoc 5-year results from an open-label, multicentre, randomised, controlled, phase 3 study. The Lancet. Oncology 2019;20:1239-51.
Flavahan WA, Gaskell E, Bernstein BE. Epigenetic plasticity and the hallmarks of cancer. Science 2017;357.
Guo Y, Song Y, Guo Z, Hu M, Liu B, Duan H, Wang L, Yuan T, Wang D. Function of RAD6B and RNF8 in spermatogenesis. Cell cycle 2018;17:162-73.
Kuang J, Li L, Guo L, Su Y, Wang Y, Xu Y, Wang X, Meng S, Lei L, Xu L, Shao G. RNF8 promotes epithelial-mesenchymal transition of breast cancer cells. Journal of experimental & clinical cancer research : CR 2016;35:88.
Li L, Guturi KKN, Gautreau B, Patel PS, Saad A, Morii M, Mateo F, Palomero L, Barbour H, Gomez A, Ng D, Kotlyar M, et al. Ubiquitin ligase RNF8 suppresses Notch signaling to regulate mammary development and tumorigenesis. The Journal of clinical investigation 2018;128:4525-42.
Lee HJ, Li CF, Ruan D, Powers S, Thompson PA, Frohman MA, Chan CH. The DNA Damage Transducer RNF8 Facilitates Cancer Chemoresistance and Progression through Twist Activation. Molecular cell 2016;63:1021-33.
Nowsheen S, Aziz K, Aziz A, Deng M, Qin B, Luo K, Jeganathan KB, Zhang H, Liu T, Yu J, Deng Y, Yuan J, et al. L3MBTL2 orchestrates ubiquitin signalling by dictating the sequential recruitment of RNF8 and RNF168 after DNA damage. Nat Cell Biol 2018;20:455-64.
Ouyang S, Song Y, Tian Y, Chen Y, Yu X, Wang D. RNF8 deficiency results in neurodegeneration in mice. Neurobiology of aging 2015;36:2850-60.
Perez-Guijarro E, Yang HH, Araya RE, El Meskini R, Michael HT, Vodnala SK, Marie KL, Smith C, Chin S, Lam KC, Thorkelsson A, Iacovelli AJ, et al. Multimodel preclinical platform predicts clinical response of melanoma to immunotherapy. Nature medicine 2020;26:781-91.
Guo Z, Tian Y, Guo Y, Li B, Liu X, Xie K, Song Y, Wang D. RAD6B Plays a Critical Role in Neuronal DNA Damage Response to Resist Neurodegeneration. Frontiers in cellular neuroscience 2019;13:392.
Greten FR, Grivennikov SI. Inflammation and Cancer: Triggers, Mechanisms, and Consequences. Immunity 2019;51:27-41.
Ozga AJ, Chow MT, Luster AD. Chemokines and the immune response to cancer. Immunity 2021;54:859-74.
Tait Wojno ED, Hunter CA, Stumhofer JS. The Immunobiology of the Interleukin-12 Family: Room for Discovery. Immunity 2019;50:851-70.
Kaczanowska S, Beury DW, Gopalan V, Tycko AK, Qin H, Clements ME, Drake J, Nwanze C, Murgai M, Rae Z, Ju W, Alexander KA, et al. Genetically engineered myeloid cells rebalance the core immune suppression program in metastasis. Cell 2021;184:2033-52 e21.
Garris CS, Arlauckas SP, Kohler RH, Trefny MP, Garren S, Piot C, Engblom C, Pfirschke C, Siwicki M, Gungabeesoon J, Freeman GJ, Warren SE, et al. Successful Anti-PD-1 Cancer Immunotherapy Requires T Cell-Dendritic Cell Crosstalk Involving the Cytokines IFN-γ and IL-12. Immunity 2018;49:1148-61.e7.
Gollob JA, Mier JW, Veenstra K, McDermott DF, Clancy D, Clancy M, Atkins MB. Phase I trial of twice-weekly intravenous interleukin 12 in patients with metastatic renal cell cancer or malignant melanoma: ability to maintain IFN-gamma induction is associated with clinical response. Clinical cancer research : an official journal of the American Association for Cancer Research 2000;6:1678-92.
Lei X, Lei Y, Li J-K, Du W-X, Li R-G, Yang J, Li J, Li F, Tan H-B. Immune cells within the tumor microenvironment: Biological functions and roles in cancer immunotherapy. Cancer letters 2020;470:126-33.
Petitprez F, Meylan M, de Reyniès A, Sautès-Fridman C, Fridman WH. The Tumor Microenvironment in the Response to Immune Checkpoint Blockade Therapies. Front Immunol 2020;11:784.
Dai E, Zhu Z, Wahed S, Qu Z, Storkus WJ, Guo ZS. Epigenetic modulation of antitumor immunity for improved cancer immunotherapy. Molecular cancer 2021;20:171.
Kim SJ, Kang HG, Kim K, Kim H, Zetterberg F, Park YS, Cho HS, Hewitt SM, Chung JY, Nilsson UJ, Leffler H, Chun KH. Crosstalk between WNT and STAT3 is mediated by galectin-3 in tumor progression. Gastric cancer : official journal of the International Gastric Cancer Association and the Japanese Gastric Cancer Association 2021;24:1050-62.
Song L, Tang JW, Owusu L, Sun MZ, Wu J, Zhang J. Galectin-3 in cancer. Clinica chimica acta; international journal of clinical chemistry 2014;431:185-91.
Vuong L, Kouverianou E, Rooney CM, McHugh BJ, Howie SEM, Gregory CD, Forbes SJ, Henderson NC, Zetterberg FR, Nilsson UJ, Leffler H, Ford P, et al. An Orally Active Galectin-3 Antagonist Inhibits Lung Adenocarcinoma Growth and Augments Response to PD-L1 Blockade. Cancer research 2019;79:1480-92.
Gilson RC, Gunasinghe SD, Johannes L, Gaus K. Galectin-3 modulation of T-cell activation: mechanisms of membrane remodelling. Progress in lipid research 2019;76:101010.
Kouo T, Huang L, Pucsek AB, Cao M, Solt S, Armstrong T, Jaffee E. Galectin-3 Shapes Antitumor Immune Responses by Suppressing CD8+ T Cells via LAG-3 and Inhibiting Expansion of Plasmacytoid Dendritic Cells. Cancer immunology research 2015;3:412-23.
Vasiukov G, Novitskaya T, Zijlstra A, Owens P, Ye F, Zhao Z, Moses HL, Blackwell T, Feoktistov I, Novitskiy SV. Myeloid Cell-Derived TGFβ Signaling Regulates ECM Deposition in Mammary Carcinoma via Adenosine-Dependent Mechanisms. Cancer research 2020;80:2628-38.
Demotte N, Bigirimana R, Wieers G, Stroobant V, Squifflet JL, Carrasco J, Thielemans K, Baurain JF, Van Der Smissen P, Courtoy PJ, van der Bruggen P. A short treatment with galactomannan GM-CT-01 corrects the functions of freshly isolated human tumor-infiltrating lymphocytes. Clinical cancer research : an official journal of the American Association for Cancer Research 2014;20:1823-33.
Farhad M, Rolig AS, Redmond WL. The role of Galectin-3 in modulating tumor growth and immunosuppression within the tumor microenvironment. Oncoimmunology 2018;7:e1434467.
Hwang MP, Fecek RJ, Qin T, Storkus WJ, Wang Y. Single injection of IL-12 coacervate as an effective therapy against B16-F10 melanoma in mice. Journal of controlled release : official journal of the Controlled Release Society 2020;318:270-78.
Hicks KC, Chariou PL, Ozawa Y, Minnar CM, Knudson KM, Meyer TJ, Bian J, Cam M, Schlom J, Gameiro SR. Tumour-targeted interleukin-12 and entinostat combination therapy improves cancer survival by reprogramming the tumour immune cell landscape. Nature communications 2021;12:5151.
Smith LK, Boukhaled GM, Condotta SA, Mazouz S, Guthmiller JJ, Vijay R, Butler NS, Bruneau J, Shoukry NH, Krawczyk CM, Richer MJ. Interleukin-10 Directly Inhibits CD8(+) T Cell Function by Enhancing N-Glycan Branching to Decrease Antigen Sensitivity. Immunity 2018;48:299-312 e5.
Gordon-Alonso M, Hirsch T, Wildmann C, van der Bruggen P. Galectin-3 captures interferon-gamma in the tumor matrix reducing chemokine gradient production and T-cell tumor infiltration. Nature communications 2017;8:793.
Bao Y, Wang Z, Liu B, Lu X, Xiong Y, Shi J, Li P, Chen J, Zhang Z, Chen M, Wang L, Wu Z. A feed-forward loop between nuclear translocation of CXCR4 and HIF-1α promotes renal cell carcinoma metastasis. Oncogene 2019;38:881-95.
Wang L, Li YS, Yu LG, Zhang XK, Zhao L, Gong FL, Yang XX, Guo XL. Galectin-3 expression and secretion by tumor-associated macrophages in hypoxia promotes breast cancer progression. Biochemical pharmacology 2020;178:114113.
Fischöder T, Laaf D, Dey C, Elling L. Enzymatic Synthesis of N-Acetyllactosamine (LacNAc) Type 1 Oligomers and Characterization as Multivalent Galectin Ligands. Molecules 2017;22.
Bumba L, Laaf D, Spiwok V, Elling L, Křen V, Bojarová P. Poly-N-Acetyllactosamine Neo-Glycoproteins as Nanomolar Ligands of Human Galectin-3: Binding Kinetics and Modeling. Int J Mol Sci 2018;19.
Kubli SP, Berger T, Araujo DV, Siu LL, Mak TW. Beyond immune checkpoint blockade: emerging immunological strategies. Nature reviews. Drug discovery 2021;20:899-919.
Young A, Mittal D, Stagg J, Smyth MJ. Targeting cancer-derived adenosine: new therapeutic approaches. Cancer Discov 2014;4:879-88.
Pitt JM, Marabelle A, Eggermont A, Soria JC, Kroemer G, Zitvogel L. Targeting the tumor microenvironment: removing obstruction to anticancer immune responses and immunotherapy. Annals of oncology : official journal of the European Society for Medical Oncology 2016;27:1482-92.
Riera-Domingo C, Audige A, Granja S, Cheng WC, Ho PC, Baltazar F, Stockmann C, Mazzone M. Immunity, Hypoxia, and Metabolism-the Menage a Trois of Cancer: Implications for Immunotherapy. Physiological reviews 2020;100:1-102.
Gupta R, Somyajit K, Narita T, Maskey E, Stanlie A, Kremer M, Typas D, Lammers M, Mailand N, Nussenzweig A, Lukas J, Choudhary C. DNA Repair Network Analysis Reveals Shieldin as a Key Regulator of NHEJ and PARP Inhibitor Sensitivity. Cell 2018;173:972-88 e23.

(Not answered)

Supplementaryinformation.docx

Download PDF

Reviewers invited by journal
27 Jan, 2023
Submission checks completed at journal
19 Dec, 2022
Editor assigned by journal
18 Dec, 2022
First submitted to journal
18 Dec, 2022

You are reading this latest preprint version

RNF8 enhances the sensitivity of PD-L1 inhibitor against melanoma through ubiquitination of galectin-3 in stroma

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Cells lines and cells culture

Animals

RNA and plasmids interference

Mouse xenograft tumor models

Mass cytometry

t-distributed Stochastic Neighbor Embedding (t-SNE) Analysis

Spanning Tree Progression of Density Normalized Events (SPADE) Analysis

Co-immunoprecipitation

Galectin-3-coated beads and in vitro bioassays

Cone beam CT (CBCT) scanning and 3D imaging

Statistical analysis

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1