LAMTOR1 degrades MHC-II via the endocytic in hepatocellular carcinoma

Abstract

Tumor cell surface antigen recognition is a major hallmark of cancer therapy, and loss of major histocompatibility complex class I (MHC-I) is the most common mechanism that impairs tumor cell surface antigen processing and expression. In addition to this, MHC-II regulates antigen presentation in CD4⁺ T cell immune responses involved in tumor killing by CD8⁺ T cells, whereas the regulation of endocytosis regulating MHC-II antigen presentation has not been reported. Therefore, the regulation of the endocytosis pathway on the expression of MHC-II surface level and antitumor T cell response remains to be explored. In this experiment, we found that LAMTOR1 regulates the endocytic pathway through the GTPase domain of DNM2 and triggers the formation of autophagosomes. We performed flow cytometry and western blotting analyses, which revealed that the expression of MHC-II molecules on the surface of cells is influenced by LAMTOR1 through the endocytic pathway. We showed that the expression of MHC-II molecules, which recognize CD4⁺ T cells on the surface of cells, was regulated by LAMTOR1 through an endocytic pathway. By coculture experiments, we showed that CD8⁺/CD4⁺ T cells exhibit substantially higher levels of tumor cell apoptosis than those observed when hepatocellular carcinoma (HCC) cells were cocultured with CD8⁺ T cells alone. This study revealed that LAMTOR1 decreases the expression levels of MHC-II on cell surfaces in order to reduce antigen expression, leading to a decrease in antitumor T cell responses.

Graphical Abstract

LAMTOR1 promotes immune escape by targeting MHC-II molecules via endocytic degradation in HCC (graphical_abstract). Therefore, LAMTOR1 may be used clinically as a specific biomarker for screening HCC or as a potential target for individualized patient therapies.

Introduction

Hepatocellular carcinoma (HCC) is one of the most common cancer types in the world (1). Major histocompatibility complex class I (MHC-I) is an important target for tumor-specific CD8⁺ cytotoxic T lymphocytes. Loss of major histocompatibility complex class I (MHC-I) is the most common mechanism of immune escape of tumor cells. In addition to this, MHC-II regulates antigen presentation in CD4⁺ T cell immune responses and is involved in tumor killing by CD8⁺ T cells. However, the effect of MHC-II on CD4⁺ T cell immune responses affecting tumor presentation of antigens has not been recognized until recently (2–5).

Cells exhibit endocytic responses to immune evasion by delivering MHC-II to lysosomes for degradation (6,7). Endocytosis and autophagy can occur synergistically to promote their optimal behavior (8, 9). Both processes may be involved in lysosomal degradation through the formation of amphisomes (10). Endocytosis begins with the internalization of plasma membrane cargo into endosomes, labeled by specific ligand proteins and small Rab GTPases (11,12). Various autophagic molecules can be localized to recycling endosomes (13–16). In addition, studies have established that the function of endosome-associated genes is associated with the regulation of autophagic vesicle maturation, closure and merging with lysosomes (17–19). It is also reported that the formation of endosomes and autophagosomes, which induces the production of amphisome bodies (20). Several studies have implicated the role of LAMTOR1 in cellular metabolism, which induced autophagy formation and the anchoring of the MEK-ERK pathway to late endosomes (21,22). Therefore, we aimed to study that LAMTOR1 promotes immune escape by targeting MHC-II molecules via endocytic degradation in HCC.

Materials and methods

Cell culture

The cell lines L02, Huh7, HepG2 and mouse Hepa1-6 were procured from the American Type Culture Collection (ATCC, Manassas, VA). The cell lines L02, Huh7, HepG2 and mouse Hepa1-6 were cultured by DMEM (Gibco, Grand Island, USA) supplemented with 10% fetal bovine serum and penicillin (10 000 U/mL), streptomycin (0.1 mg/mL) medium and grown in a constant temperature humidified incubator at 37°C and 5% CO2. All cell experiments were validated using Huh7, HepG2 and Hepa1-6 cell lines, while Hepa1-6 cell lines were used in the animal experiments. The cells were cytogenetically tested and authenticated using the short tandem repeat method. All cell-based experiments were carried out on cells that have been tested for <12 weeks. The cell lines were lasted tested in March 2022.

Antibodies and reagents

The antibodies used were: anti-LAMTOR1 (catalog #8943S;CST), anti-dynamin 2 (DNM2; catalog #ab3457;Abcam), anti-RAB5 (catalog #E6N8S;CST), anti-TSG101 (catalog #GTX70255;GeneTex), anti-RAB35 (catalog #GTX120249;GeneTex), anti-RAB7 (catalog #E9O7A;CST), anti-VPS26A (catalog #ab181352;Abcam), anti-P62 (catalog #D5L7G;CST), anti-LC3B (catalog #E5Q2K;CST), anti-HLA-A (catalog #ab52922;Abcam), anti-HLA-B (catalog #ab193415;Abcam), anti-HLA-C (catalog #ab126722;Abcam) and anti-MHC-II (catalog #ab55152;Abcam).

The reagents used were: 1 μM Dynasore (catalog #HY-15304;MedChemExpress), 5 mM 3-methyl adenine (3-MA;autophagy inhibitor) (catalog #HY-19312; MedChemExpress) and 10 nM mammalian target of rapamycin (RAP;HY-10219; MedChemExpress).

Mouse experiments

Mice were anesthetized and 1 × 10⁶ Hepa1-6 cells were suspended in Hank’s balanced salt solution and injected into the left lobe of the mouse liver with an insulin syringe. Then, the peritoneum was sutured and the skin was closed. Mice were euthanized after 2–3 weeks and samples were collected after transcardial perfusion with phosphate-buffered saline (PBS).

Cell transfection

Cell lines with LAMTOR1 overexpression (LAMTOR1-OE) and short hairpin RNA (shRNA; sh-LAMTOR1) knockdown were generated by lentiviral systemic culture; lentiviral transfection was also performed, targeting the shRNA of RAB7A and ATG7. To generate stably expressed cell lines, puromycin screening (2 µg/mL) was performed after cell transfection:

sh-LAMTOR1-1: 5ʹ-CCCATCCCGTTCTCTGATTTG-3ʹ

sh-LAMTOR1-2: 5ʹ-CATGGAGCAGCATGAGTACAT-3ʹ

RAB7A: 5ʹ-AGGCTAGTCACAATGCA-GATA-3ʹ

ATG7: 5ʹ-CTGCTGAGGAGCTCTCCAT-3ʹ

Immunoblotting

Total proteins were separated by SDS-PAGE electrophoresis, transferred to polyvinylidene difluoride (PVDF) membranes, blocked bovine serum albumin (BSA), incubated with antibody.

Immunofluorescence assay

Cells were cultured on coverslips for 1–2 days. After apposition, cells were washed, fixed, blocked and incubated with antibody and nuclei were stained using DAPI, placed on slides and Nikon A1R confocal microscope with 60× and 20× objectives were observed.

Histology and immunohistochemistry

Human hepatocellular carcinoma specimens (HCC) tumor samples (n = 45) were collected from the patients who received surgery at China Medical University and Xuchang Central Hospital, Henan University of Science and Technology from November 2019 to January 2020. All patients had histologically confirmed HCC. Patients with neoadjuvant treatment, inflammatory diseases, or active infection were excluded. Informed consent was obtained before sample collection. The study was approved by the Committee for the Ethical Review of Research, China Medical University.

All protocols utilized for hepatocellular carcinoma specimens were approved by Chinese Medical University. Hepatocellular carcinoma tissues were stained by using LAMTOR1 (1:500;CST,3868) and CD4 (1:400;ab133616;Abcam) and determine protein expression.

RNA sequencing analysis

Total RNA was extracted from HCC cells and sequenced on an Illumina HiSeq 2500 System (San Diego, CA, USA). Enrichment analysis of sh-LAMTOR1 with sh-Ctrl in Huh7 cells was performed using the Metascape platform (fold-change > 1.0, adjusted P-value < 0.05).

TCGA dataset analysis

We retrieved 371 HCC samples with detailed gene expression data and clinical features from The Cancer Genome Atlas (TCGA) database and analysed the relationship between LAMTOR1 and related functions from these samples.

Co-immunoprecipitation

Co-immunoprecipitation (co-IP) kits (catalog #ab206996, Abcam) were used to study the relationship between LAMTOR1 and DNM2. LAMTOR1 protein (2 µg) was incubated with DNM2 (5 µg) or its deletion constructs 1/2/3 in an in vitro co-IP assay. LAMTOR1 was then detected with antibody LAMTOR1 (D1H6) and the deletion structure of the protein was detected with a 6xHis antibody.

Tumor and T cell coculture experiments and detection of apoptotic tumor cells

CD3⁺ T cells were harvested from splenocytes of mice by flow cytometry and sorted for CD4⁺ T as well as CD8⁺ T cells. Hepa1-6 cells and CD8⁺ T and CD4⁺ T cells were cultured with 50% Dulbecco's modified Eagle's medium (DMEM) and 50% RPMI-1640,10% FBS. Hepa1-6 cells were detected in the apoptosis rate by flow cytometry.

Adeno-associated virus serotype 8 (AAV8) production

Three plasmids were transfected with HEK293T cells to prepare the virus, including rAAV-shRNA-EGFP, an AAV helper plasmid (containing Rep and Cap genes) and an adenovirus helper plasmid (pAD). All recombinant AAV vectors were derived from AAV8. AAV8 vectors use the H1 promoter to express LAMTOR1 siRNA or interfering control siRNA, while the cytomegalovirus virus promoter drives EGFP reporter expression.

Statistical analyses

All experimental results were repeated at least 3 times and statistical significance was calculated using ANOVA. For ANOVA, nonparametric tests were performed by comparing the median of all groups with P < 0.05.

Results

Association of LAMTOR1 with the clinical pathology in hepatocellular carcinoma

To screen key vesicle trafficking factors regulating HCC, we analysed the function of five gene sets using KEGG datasets, in which LAMTOR1 was identified as a signature factor (Figure 1A). Several studies have implicated the role of LAMTOR1 in cellular metabolism, including the anchoring of the MEK-ERK pathway to late endosomes via LAMTOR1 (21,22). Although its mechanism has not been elucidated, LAMTOR1 is highly expressed in HCC according to the TCGA dataset (P<0.01) (Figure 1B). To determine the clinical significance of LAMTOR1, we found that the positive relationship between LAMTOR1 and stage (P = 0.001) as well as prognostic (P = 0.002) (Figure 1C–E). Fluorescence microscopy images showed endogenous LAMTOR1 protein distribution in human HCC tissues (Figure 1F). The expression of LAMTOR1 was increased in Huh7, HepG2 and Hepa1-6 cell lines compared with L02 cell lines (P < 0.05) (Figure 1G). Moreover, the expression of LAMTOR1 was higher in HCC tissues compared with adjacent tissues (n = 45) (P < 0.05) (Figure 1H). We performed a univariate analysis using information from pan-cancer entries in TCGA datasets and found that high expression of LAMTOR1 was a risk factor for overall survival (OS) (P < 0.05) and disease-free survival (DSS) in HCC (P < 0.05) (Figure 1I and J). The levels of LAMTOR1 were evaluated in HCC tissues (n = 45) and compared with patients’ OS (P = 0.011) (Supplementary Figure S1K and L available at Carcinogenesis Online). We also found that LAMTOR1 expression was significantly related to the stage (P = 0.0006) (Table 1). These findings revealed that LAMTOR1 is an oncogenic gene in HCC.

Figure 1.

Upregulated expression of LAMTOR1 in relation with oncogenes in hepatocellular carcinoma (HCC). (A) Venn diagram showing the function of five gene sets in vesicle trafficking based on Kyoto Encyclopedia of Genes and Genomes (KEGG) dataset. (B) LAMTOR1 protein levels analysed by The Cancer Genome Atlas (TCGA) database in HCC and adjacent tissues. (C–E) Correlation of LAMTOR1 with clinical characteristics, area under the curve evaluation (AUC) and prognostic value of overall survival (OS) in patients with HCC based on the TCGA dataset. (F) Fluorescence microscopy images showing the LAMTOR1 protein distribution in human HCC and adjacent tissues. Scale bar: 20 μm. (G) LAMTOR1 protein expression in HCC cell lines compared with normal liver cell lines. (H) LAMTOR1 protein expression in HCC and adjacent tissues. (I, J) Overall survival (OS) and disease-free survival (DFS) of patients with pan-cancer according to the TCGA datasets. (K, L) Representative graphs of immunohistochemical (IHC) staining for LAMTOR1 expression in 45 HCC patients (scale bar: 1.25 mm; scale bar: 200 μm).

Table 1.

Association between the expression of LAMTOR1 and clinicopathological features of HCC patients

Characteristics	No.	High LAMTOR1	Low LAMTOR1	P-value
		(Q Score ≥ 150; n = 20)	(Q Score < 150; n = 25)
Age(y)				0.4227
<60	21	8	13
≥60	24	12	12
Sex				0.5403
Female	18	7	11
Male	27	13	14
Tumor size (cm)				0.0662
<5.0	18	5	13
≥5.0	27	15	12
Tumor differentiation				0.2426
Well	2	1	1
Moderate	18	8	10
Poor	25	11	14
Stage				0.0006*
I–II	23	4	19
III–IV	22	16	6
Vessel infiltration				0.6067
Negative	32	15	17
Positive	13	5	8

All statistical tests are two-sided.

Regulation of the endocytic pathway by LAMTOR1 is dependent on DNM2

LAMTOR1 controls cell growth by regulating the balance of energy metabolism (23). The potential pathways imply that high LAMTOR1 expression related with endocytic pathways (Figure 2A). We also analysed LAMTOR1 dependent transcriptome in parental and LAMTOR1 knockout (KO) Huh7 cells through RNA sequencing (RNA-Seq) (Figure 2B). Given that endosomes are one of the hallmarks of endocytic pathways, the expression of several markers was tested as shown in the results. HCC tissues showed a punctate cytoplasmic distribution of LAMTOR1 colocalized with early endosomes (RAB5) and late endosomes (RAB7) (Supplementary Figure S2A and C available at Carcinogenesis Online). Moreover, LAMTOR1 knockdown resulted in a decreased number of late endosomes (Figure 2D) and decreased perinuclear internalization of EGFR (P < 0.05) (Supplementary Figure S2B and C available at Carcinogenesis Online).

Figure 2.

Regulation of endocytosis by LAMTOR1 is DNM2-dependent. (A) KEGG/Gene Ontology (GO) analysis shows signaling pathways that were substantially affected by LAMTOR1 in the TCGA database. (B) KEGG/GO analysis shows the signaling pathways affected by LAMTOR1 in RNA sequences. (C) Confocal microscopy of endogenous LAMTOR1 and RAB7 (red) in HCC and adjacent tissues. Scale bar: 20 μm. (D) Immunofluorescence detection of RAB7 of control (short hairpin [sh]-control [Ctrl]) and LAMTOR1 knockdown (sh-LAMTOR1) groups as observed by confocal microscopy. Scale bar: 20 μm. (E) Correlation of LAMTOR1-related genes was analysed based on TCGA datasets. (F) Expression of LAMTOR1 based on RNA sequences using heat map showing the expression of six genes in HCC. (G) Huh7, HepG2 and Hepa1-6 cells were transfected with sh-Ctrl and sh-LAMTOR1 and western blot was performed, which showed endocytosis hub gene expression. (H) sh-Ctrl and sh-LAMTOR1 groups were incubated with pH rodo™ Red dye for time-lapse real-time imaging. (I) Detection by co-immunoprecipitation between LAMTOR1 and DNM2 interaction. (J) Interaction between LAMTOR1 and DNM2 was constructed by detailed mapping of DNM2. (K) D-terminal GTPase region of DNM2 interacting with LAMTOR1 was assessed using in vitro immunoprecipitation assay.

The endocytic pathway refers to the regulation of substance internalization through the formation of 60–120 nm vesicles by invagination of the cytoplasmic membrane, which wraps external substances and imports them into the cell (24–27). We investigated the mechanisms underlying LAMTOR1-induced endocytosis signaling in HCC and identified 17 genes that were substantially differentially expressed. We clustered these hub genes and found that DNM2, RAB5, RAB7, RAB35, VPS26A and TSG101 were influenced by LAMTOR1 in the TCGA database (Figure 2E). We verified these results in parental and LAMTOR1 KO Huh7 cells by RNA sequencing (RNA-Seq) (Figure 2F). Immunoblotting was also performed to demonstrate that LAMTOR1 knockdown decreased RAB7, RAB35 and TSG101 (P < 0.05) (Figure 2G). Time-lapse microscopy indicated that sh-LAMTOR1 decreased endocytic speed (Figure 2H), confirming that LAMTOR1 induces the endocytic process.

DNM2 GTPase is a key step in the regulation of the separation of endosomal vesicles from the plasma membrane (28). We found that prolonged Dynasore inhibition leads to increased expression of early endosomes (RAB5) and late endosomes (RAB7) (P < 0.05) (Supplementary Figure S2D available at Carcinogenesis Online). To further explore the association of LAMTOR1–DNM2 interaction, co-immunoprecipitation was performed on a panel of HCC cell lines (Figure 2I). For a more detailed mapping of the D-terminal region of DNM2, three domains were expressed: construct 1: full-length domain, construct 2: partial deletion in GTPase domain and construct 3: full GTPase domain deleted (Figure 2J). The binding of these constructs to LAMTOR1 was assessed using in vitro immunoprecipitation, and deletion of part of the GTPase domain (23 amino acids) or full GTPase domain abrogated binding to LAMTOR1 (Figure 2K). These results suggest that LAMTOR1 regulates endocytosis by interacting with the GTPase domain of DNM2.

LAMTOR1 induces the endocytic pathway to form amphisomes

Enrichment of intracellular autophagy and lysosomes can be observed following increased activity of LAMTOR1 (29). We found that LC3-I expression levels increased (P < 0.05) and P62 expression levels remarkably decreased (P < 0.05) in the sh-LAMTOR1 group (Supplementary Figure S3A available at Carcinogenesis Online). Furthermore, transmission electron microscopy revealed a decrease in the number of autophagosomes after treatment with sh-LAMTOR1 (Supplementary Figure S3B available at Carcinogenesis Online). In response to stress, cells initiate endocytosis and autophagy to maintain cellular homeostasis (30,31). This leads to the formation of amphisomes, which are hybrid organelles that play a key role in the cross-talk between endosomes and autophagy (32). We investigated whether LAMTOR1 induced the fusion of endosomes and autophagosomes. It was revealed that CD63, and TSG101 were strongly colocalized with LC3B in sh-Ctrl but not sh-LAMTOR1 (P < 0.05) (Supplementary Figure S3C and E available at Carcinogenesis Online). Confocal images showed higher LC3B colocalization with CD63 and TSG101 in HCC tissues (Supplementary Figure S3D and F available at Carcinogenesis Online). We also found that LC3 and CD63 were strongly colocalized with LAMP1 in sh-Ctrl but not in sh-LAMTOR1 cells (P < 0.05) (Supplementary Figure S3G and H available at Carcinogenesis Online). LAMP1 colocalization with LC3 and LAMP1 in HCC tissues than adjacent tissues (Supplementary Figure S4A available at Carcinogenesis Online). Confocal images showed higher CD63 colocalization with LAMP1 in HCC tissues (Supplementary Figure S4B available at Carcinogenesis Online). These results suggest that activation of LAMTOR1 triggers the formation of endosomes and autophagosomes, which induces the production of amphisome bodies.

LAMTOR1 regulates cell adhesion molecules by endocytosis

Endocytosis contributes to the characterization of the migratory behavior of tumor cells (33–36). We used the TCGA database to analyse the correlation between endocytosis and cell adhesion molecules, immunity, cell stemness and extracellular vesicles. We extracted the molecules related to endocytosis in the KEGG database using R, and analysed the correlation with cell adhesion molecules (Supplementary Figure S5A–D available at Carcinogenesis Online). To analyse the correlation of endocytosis with other signaling pathways, we showed that endocytosis was remarkably associated with cell adhesion molecules, suggesting a potential role of endocytosis in the regulation of tumor cell proliferation and migration (Supplementary Figure S6A and B available at Carcinogenesis Online). The Gene Set Enrichment Analysis showed that LAMTOR1 expression was correlated with cell adhesion molecules (Supplementary Figure S6C available at Carcinogenesis Online). We further examined the potential role of LAMTOR1 in cell adhesion molecule gene sets using Gene Ontology (GO) analysis. The results showed that LAMTOR1 associated with MHC class II protein complex, T cell costimulation and antigen processing and presentation (Supplementary Figure S6D available at Carcinogenesis Online). This suggests that LAMTOR1 has positive effects on cell adhesion molecules of antigen processing. LAMTOR1 was found to be associated with cell adhesion molecules (Supplementary Figure S5E available at Carcinogenesis Online) and revealed CD4 based on GSE140228 datasets (Supplementary Figure S5F available at Carcinogenesis Online). We also evaluated the role of LAMTOR1 on the migration and proliferation of HCC cell lines using Transwell and 5-ethynyl-2ʹ-deoxyuridine (EdU) assays. Notably, treatment with sh-LAMTOR1 decreased tumor growth (P < 0.05) (Supplementary Figure S6E and F available at Carcinogenesis Online). Confocal imaging showed higher MHC-II (red) colocalization with CD63 (green) in HCC tissues compared with adjacent tissues (Supplementary Figure S6G available at Carcinogenesis Online). Confocal imaging further showed higher MHC-II (green) colocalization with LC3B (green) in HCC tissues compared with adjacent tissues (Supplementary Figure S6H available at Carcinogenesis Online). These results showed that activation of LAMTOR1 regulated MHC-II molecules through endosome and autophagosome formation.

LAMTOR1 degrades plasma surface MHC-II by endocytosis

We found that LAMTOR1 inversely regulates the expression of MHC-II (P<0.05) (Supplementary Figure S7A and B available at Carcinogenesis Online). MHC-I is regulated by autophagy, while MHC-II is regulated by autophagy as well as endocytosis, demonstrating that MHC-II degradation pathway different with MHC-I (P < 0.05) (Supplementary Figure S7C and D available at Carcinogenesis Online). To gain insights into LAMTOR1 is involved in MHC-II degradation, We constructed LAMTOR1, RAB7 and ATG7 stable transduction cell lines to clarify the whole cell expression of MHC-II regulation by LAMTOR1 through RAB7 (endosomal) and ATG7 (autophagy). We found that the degradation of MHC-II molecules by LAMTOR1 through endocytosis and autophagy (Figure 3A). To explore LAMTOR1 exchange plasma membrane MHC-II levels in HCC cells, we separated MHC-II molecules on the cell membrane surface by sucrose density gradient centrifugation and detected them by western blotting (P < 0.05) (Figure 3B). Next, to test the expression of MHC-II molecules on the plasma membrane, we performed flow cytometric analysis to demonstrate that LAMTOR1 regulates changes in the levels of MHC-II molecules on the cell membrane surface through endocytic pathways (Figure 3C and D). Immunofluorescence staining showed drastically attenuated colocalization of LAMTOR1 with MHC-II after the knockdown of RAB7 and ATG7 (Figure 3E and F). The results show that LAMTOR1 degrades cell surface MHC-II through endocytosis.

Figure 3.

LAMTOR1 promotes MHC-II endocytic trafficking to the lysosome through an autophagy-dependent pathway. (A) Treatment of Huh7, HepG2 and Hepa1-6 cells overexpressing LAMTOR1 with sh-RAB7 or/and sh-ATG7 to detect the effect of MHC-II, HLA-A, HLA-B and HLA-C expression levels. (B) Effect of sh-RAB7 as well as sh-ATG7 on plasma membrane MHC-II was detected by immunoblotting. (C) Flow cytometry plot showing plasma membrane MHC-II expression after knockdown of ATG7 and inhibitor 3-MA. (D) Flow cytometry plot showing plasma membrane MHC-II expression after knockdown of RAB7 and inhibitor Dynasore. (E, F) Localization of LAMTOR1 relative to MHC-II (red) in cells treated with sh-RAB7 and sh-ATG7.

LAMTOR1 degrades cell surface MHC-II to promote immune escape

Next, multiplex immunofluorescent staining and multispectral imaging were used to identify LAMTOR1, RAB7 and MHC-II coexpression in HCC cells (Figure 4A). We hypothesized that endocytic pathway-induced reduction in the surface levels of MHC-II might contribute to their immune escape from CD4⁺ T cell recognition. Therefore, we found that the expression of LAMTOR1 negative associated with CD4⁺ T cell infiltration in HCC patients from China Medical University by immunohistochemical (IHC) staining of serial sections on tissue microarrays (n = 45) (Figure 4B). To assess antigen presentation, CD8^+, and CD4⁺ T cells were cocultured with mouse Hepa1-6 cells pretreated with sh-LAMTOR1 and were found to show substantially higher tumor cell apoptosis compared with coculture with CD8⁺ T cells alone (P < 0.05) (Figure 4C). In addition, CD8⁺ and CD4⁺ T cells were cocultured with mouse Hepa1-6 cells pretreated with sh-LAMTOR1+shRAB7 and were found to show substantially higher tumor cell apoptosis compared with coculture with CD8⁺ T cells alone (P < 0.05) (Figure 4D). These studies suggest that CD4⁺ T cells play an important role in tumor-specific killing by CD8⁺ T cells. We further clarified that the surface level of MHC-II degraded via the endocytic pathway may lead to evasion of recognition by CD4⁺ T cells, which in turn inhibits the production of tumor-killing CD8⁺ T cells.

Figure 4.

LAMTOR1 trafficking to the endocytic pathway inhibits the antitumor T cell response. (A) Colocalization of LAMTOR1, RAB7 and MHC-II in HCC cell. (B) IHC staining showing positive correlation of LAMTOR1 expression with CD4 (scale bar: 200 μm; scale bar: 1.25 mm). (C) Flow cytometry analysis of cell lines cocultured with CD4⁺ T, CD8⁺ T cells to test apoptosis. (D) Flow cytometry analysis of cell lines cocultured with CD4⁺ T, CD8⁺ T cells to test apoptosis.

LAMTOR1 can switch the cell surface expression of MHC-II and promote tumor development in HCC

To verify the key role of endocytosis in tumor development, we performed tumor studies in mice and found that the change in expression of LAMTOR1 clearly influenced the degree of malignancy after administration for 2 weeks (Figure 5A). Interestingly, combination therapy completely inhibited tumor growth (P < 0.05) (Figure 5B and C). Treatment with LAMTOR1-OE+shRAB7 or LAMTOR1-OE+shATG7 increased the surface levels of MHC-II in vivo (Figure 5D and E). We examined the CD4, CD8 and MHC-II expression in vivo treatment with LAMTOR1-OE+shRAB7, LAMTOR1-OE+shATG7 or LAMTOR1-OE+shRAB7 +shATG7, the results showed that CD8⁺ T and CD4⁺ T cells higher infiltration and the expression with MHC-II was associated with LAMTOR1-OE+shRAB7+shATG7 group (P < 0.05) (Figure 5F). The total tumor area dramatically decreased in AAV-LAMTOR1 mice killed four weeks after treatment, thus demonstrating that AAV-LAMTOR1 persisted as the tumors progressed (Figure 5G). Therefore, these results confirm that LAMTOR1 downregulated the expression of surface levels of MHC-II and reduced tumor cell death.

Figure 5.

LAMTOR1 promotes tumor development through endocytic pathway in vivo. (A) Mouse Hepa1-6 cells were transplanted into homozygous (C57BL/6) mice for implantation, and expression analysis was performed using hematoxylin and eosin and IHC for LAMTOR1 staining in mouse liver cancer samples. (B, C) Mouse Hepa1-6 cells were transplanted into homozygous (C57BL/6) mice. (B) Tumor images at the endpoint. (C) Tumor weight. (D, E) Immunofluorescence staining images showing the localization of MHC-II and LAMTOR1 within sh-RAB7- or sh-ATG7-treated homozygous (C57BL/6) mice. Scale bar: 20 μm. (F) IHC staining showing positive expression of CD8⁺ T and CD4⁺ T (scale bar: 100 μm). (G) Representative images of livers of tumor-bearing mice compared with control AAV mice highlighting the tumor attenuating effect of sh-LAMTOR1.

Discussion

Endocytosis is a key regulator of immunogenicity in HCC. LAMTOR1 expression was related to the stage in HCC. Our results suggest that LAMTOR1 regulates endocytosis by interacting with the GTPase domain of DNM2 and triggers the formation of endosomes and autophagosomes, which induces the production of amphisome bodies. We further clarified that the surface level of MHC-II degraded via the endocytic pathway may lead to evasion of recognition by CD4⁺ T cells, which in turn inhibits the production of tumor-killing CD8⁺ T cells. Therefore, these results confirm that LAMTOR1 downregulated the expression of surface levels of MHC-II and reduced tumor cell death. This is consistent with a recent study that states that autophagy targets intracellular and extracellular components of lysosomal degradation, which leads to antigen processing for MHC presentation and affects the transport of MHC-I molecules (37).

CD8⁺ T cell responses are often considered to represent the most important immune cell type for controlling tumor growth, and it has been observed that tumors can immune escape by downregulating MHC class I expression (2–5,38–40). Tumor-specific CD4⁺ T cells provide the task of helper signaling, when CD4⁺ T and CD8⁺ T lymphocytes are present together, T cell responses and tumor rejection are greatly enhanced, whereas CD8⁺ T cell transfer alone is not sufficient to induce tumor remission. Consistently, tumor-specific CD4⁺ T cells help to maintain the function of tumor-directed CD8⁺ T cells. The presence of tumor-specific CD4⁺ T cells enhances the recruitment, proliferation and effector functions of CD8⁺ T cells through IFN-γ and IL-2 production (2–5,38–40). CD4⁺ T cells can further enhance the function of CD8⁺ T cells with high-affinity TCRs. Given the critical role of endocytic systems in supporting cancer metabolism and growth, our data elucidate the relevant functions of the endocytic system in the immune escape of HCC.

Endocytic receptor movement is coordinated by ubiquitinated Rab GTPases, SNARE and endosomal subgroups of regulators and effectors (41,42). RAB5, RAB7, RAB35, TSG101, VPS26A and DNM2 have now been identified to play a role in the vesicle trafficking and endocytic pathways. Interestingly, our results suggest that LAMTOR1 promotes endocytosis by regulating RAB7, RAB35 and TSG101. Pharmacologically, the DNM2 inhibitor Dynasore similarly regulates the expression of TSG101 as well as RAB35. DNM2 is a GTPase that directly interacts with caveolin to mediate membrane division and endosome formation (43) and is a key regulator mediating the independent endocytosis of caveolin (44). Importantly, our data demonstrated that LAMTOR1 is a key regulator mediating the non-dependent endocytosis of caveolin proteins.

The endocytic recruitment of microenvironmental substances to lysosomes is regulated by autophagy, a process that helps to expand the concept of autophagy-mediated organelle homeostasis. Selective degradation of the endoplasmic reticulum, mitochondria and ribosomes has been shown to be essential for various cellular processes (45,46). The localization of autophagy-related proteins in endosomes has been observed during amphisome formation (47,48). Targeting late endosomes by autophagy revealed that endosomes and autophagy disrupt the membrane-associated expression and functional role of MHC-II. These phenomena contribute to the understanding of the role of endosomal and autophagic proteins in MHC-II transport. It follows that the balanced role of endosomes and autophagy affects important physiological and pathological processes such as development, immunity and cancer (49). Moreover, in endosome-deficient tumor models, MHC-II is considered a key driver of oncogenic effects and can be an important strategy for tumor survival by avoiding immune recognition and attack. There are some shortcomings in this study, we were not able to verify the validity of the study by CRISPR-edited mice, and we were not able to explore the binding site of LAMTOR1 to DNM2. In this study, we explored the interaction of MHC-II with endocytosis, gathered mechanistic insights into how endocytosis contributes to immune escape and provide a strong rationale for developing novel therapeutic approaches for patients with HCC. Furthermore, anti-LAMTOR1 could support other approaches such as cancer vaccinations in HCC as it increases the immunogenicity by augmenting the antigen presentation in T cells.

Supplementary material

Supplementary data are available at Carcinogenesis online.

Supplementary Fig 1. Representative image of IHC staining for LAMTOR1 expression in 45 HCC patients (scale bar:1.25 mm).

Supplementary Fig 2. A, Confocal co-localization analysis of endogenous RAB5 (red) and LAMTOR1 (green) in human HCC and adjacent tissues. Scale bar:20 μm. B, C, Fluorescence microscopy and western blot images of EGFR in sh-Ctrl and sh-LAMTOR1 cells. Scale bar:20 μm. D, Dynasore-treated cells for 0.5, 1, 2, 4 and 6 h to disrupt endocytosis followed by western blotting.

Supplement Fig 3. LAMTOR1 accelerates the fusion of autophagosomes and endosomes. A, The effects of sh-Ctrl and sh-LAMTOR1 pretreated with Earle’s balanced salt solution and mammalian target of rapamycin (10 nM) on LC3 and P62 protein levels in Huh7, HepG2, and Hepa1-6 cells were analyzed using western blotting. B, Amount of autophagosomes of Huh7, HepG2, and Hepa1-6 cells with Ctrl or sh-LAMTOR1. Arrows represent autophagic vacuoles (scale bar:1 μm). C, D, Confocal analysis of endogenous LC3 (red) and TSG101 (green) with HCC tissues or cells treated with sh-Ctrl or sh-LAMTOR1. Scale bar:20 μm. E, F, Confocal analysis of endogenous LC3 (red) and CD63 (green) with HCC tissues or cells treated with sh-Ctrl or sh-LAMTOR1. Scale bar:20 μm. G, H, Confocal analysis of endogenous LAMP1 (green), LC3B (red), and CD63 (green) in HCC tissues or cells treated with sh-Ctrl or sh-LAMTOR1. Scale bar:20 μm.

Supplementary Fig 4. A, Confocal co-localization analysis of endogenous LC3 (red) and LAMP1 (green) in human HCC and adjacent tissues. Scale bar:20 μm. B, Fluorescence microscopy images of endogenous CD63 (red) and LAMP1 (green) in human HCC and adjacent tissues. Scale bar:20 μm.

Supplementary Fig 5. A-D, Endocytic pathway positively correlated with cell adhesion (CAM) and (A) immune cell, (B) extracellular vesicle, (C) and stem cell (D) expression from the TCGA data of HCC. E, Heat map showing 11 differentially expressed genes associated with LAMTOR1 of CAM based on TCGA datasets. F, Data analysis revealed that memory CD4+ T cells were associated with LAMTOR1 based on GSE140228 datasets.

Supplementary Fig 6. LAMTOR1 is involved with cell adhesion molecules (CAM). A, B, The correlation between endocytosis and CAM were validated by univariate and multivariate Cox regression analysis. C, Gene Set Enrichment analysis (GSEA) of LAMTOR1 using the TCGA dataset showing the extent of enrichment of CAM gene features in HCC cells. NES, normalized enrichment score. D, The relationship of enrichment of top 40 CAM gene sets with LAMTOR1 expression level. Signature gene sets were used to run GO analysis. E, F, EdU and cell migration assays demonstrating the effect of sh-LAMTOR1 on cell proliferation and migration (scale bar:10 μm). G, H, Localization of MHC-II (red) relative to CD63 (green) (G) and LC3B (green) (I) in HCC and adjacent tissues. Scale bar:20 μm.

Supplementary Fig. 7A-D, Treatment of Huh7, HepG2, and Hepa1-6 cells overexpressing LAMTOR1 with sh-RAB7 or/and sh-ATG7 to detect the effects of MHC-II, HLA-A, HLA-B, and HLA-C expression levels.

Glossary

Abbreviations

AAV8

Adeno-associated virus serotype 8

Ag

antigens

APCs

antigen-presenting cells

Co-IP

co-immunoprecipitation

CTL

cytotoxic T lymphocytes

EDU

5-Ethynyl-2ʹ-deoxyuridine

GO

Gene Ontology

HCC

hepatocellular carcinoma

ICT

immune checkpoint blockade therapy

IHC

immunohistochemistry

KEGG

Kyoto Encyclopedia of Genes and Genomes

LAMTOR1

late endosomal/lysosomal Adaptor, MAPK and mTOR activator 1

MHC-I

major histocompatibility complex class I

MHC-II

major histocompatibility complex class II

OS

Overall survival

TCGA

The Cancer Genome Atlas

TCR

T cell receptors

TME

tumor microenvironment

Funding

This work was supported by the National Natural Science Foundation of China (No.81901434, No.81970466), China Postal Science Foundation (No.2018M641743, No.2019M661168) and Natural Science Foundation of Liaoning Province (2019-ZD-0784).

Author contributions

BW and MX-J contributed to conception and design of the study. BW, QW, and MX-J performed the in vitro experiments with cell lines. BW-L provided expertise in mouse models. BW and BW-L carried out the TCGA data set analysis. BW provided expertise in statistical analysis. BW and QW drafted the manuscript. MX-J revised the manuscript critically for important intellectual content.

Data availability

All experimental data during this research are included in the published article and its supplementary files. The datasets and materials in this study are available on reasonable request from corresponding authors.