IgSF11–RAP1 signaling promotes cell migration and invasion of cutaneous melanoma

Abstract

Background

Aberrant cell adhesion signaling is known to either accelerate or inhibit cancer progression, but the underlying molecular basis has yet to be established. The immunoglobulin superfamily 11 (IgSF11) functions as a cell adhesion protein and is overexpressed in several types of cancer, including high-grade glioma. However, it remains unknown whether and how IgSF11 stimulates malignant phenotypes.

Methods

Using The Cancer Genome Atlas (TCGA), we first examined the expression of IgSF11 gene in various types of cancer tissues. Next, we developed an anti-hIgSF11 monoclonal antibody (mAb) and evaluated the clinicopathological significance of high IgSF11 expression in 187 cutaneous melanoma patients via immunohistochemistry using this selective mAb. We also generated human melanoma cell lines A375 and 888mel expressing IgSF11, as well as 888mel:IgSF11^KO and 888mel:IgSF11^KO:IgSF11 cells, and compared their phenotypes with those of control cells both in vitro and in vivo. Immunoprecipitation-mass spectrometry was applied to identify an IgSF11-interacting protein, followed by validation of its association with IgSF11 and of the specific IgSF11 region responsible for the complex formation and promoting melanoma cell migration.

Results

IgSF11 mRNA was highly expressed in glioblastoma tissues and skin cutaneous melanoma tissues, but not in other malignant tumors. High IgSF11 expression was observed in 57 out of the 187 melanoma cases (30.5%) and was significantly correlated with Clark’s level and high budding, both of which are parameters of melanoma invasion. Using a series of established cell lines, we demonstrated that IgSF11 promotes melanoma cell migration and invasion, as well as the enrichment of a gene set associated with epithelial-mesenchymal transition (EMT). Importantly, we identified that IgSF11 forms a complex with RAS-associated protein 1 (RAP1). Furthermore, the L372–R378 region of IgSF11 was required for recruiting RAP1 and driving melanoma cell migration.

Conclusions

We found that IgSF11–RAP1 signaling facilitates the migration and invasion of melanoma cells. The identification of IgSF11–RAP1 machinery highlights a novel link between cell adhesion and signaling molecules in promoting the malignant phenotypes of melanoma and may serve as a promising therapeutic target for this malignancy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-025-02245-5.

Introduction

In 2022, the global incidence of cutaneous melanoma exceeded 330,000 cases, and nearly 60,000 people died of the disease [1]. The proven risk factors for cutaneous melanoma include family history, ultraviolet exposure, and multiple benign or dysplastic nevi [2, 3]. These factors cause alterations in various genes and signaling pathways, contributing to the pathogenesis of cutaneous melanoma [4–9]. However, the genetic and molecular classification of cutaneous melanoma has not yet been successful so far in predicting disease progression, recurrence, and drug response. Therefore, the identification of novel prognostic biomarkers and therapeutic targets for cutaneous melanoma is crucial.

Cell adhesion molecules, such as cadherins and claudins (CLDNs), maintain tissue architecture and homeostasis. They also propagate intercellular signaling that organizes a broad range of physiological cellular processes, including cell growth, survival, differentiation, polarity, and migration [10–12]. Moreover, accumulated evidence has indicated that dysregulated CLDN signaling accelerates or inhibits cancer progression [13–16]. For instance, we previously reported that aberrant CLDN6 signaling drives endometrial cancer progression by hijacking the CLDN6–ERα (estrogen receptor α) axis [17, 18]. We also demonstrated that excessive CLDN4–LXRβ (liver X receptor β) signaling accelerates breast cancer metabolism and advancement [19]. Furthermore, we have recently found that CLDN10 cooperates with LAT1 (L-type amino acid transporter 1) and activates intracellular signaling, resulting in the progression of clear cell renal cell carcinoma [20].

The immunoglobulin superfamily 11 (IgSF11; also known as brain- and testis-specific Ig superfamily protein [BT-IgSF] and V-set and immunoglobulin domain containing 3 [VSIG3]) functions as a cell-adhesive protein that belongs to the CTX (cortical thymocyte marker in Xenopus) family within the IgSF superfamily [21–23]. The members of this family consist of two extracellular IgSF domains (a membrane-distal V-type and a membrane-proximal C2-type Ig domains), a single transmembrane region, and a C-terminal domain [24]. IgSF11 shares homology with the coxsackie virus and adenovirus receptor (CAR), the coxsackievirus and adenovirus receptor-like membrane protein (CLMP), and the endothelial cell-selective adhesion molecule (ESAM) [25] among junctional adhesion molecule (JAM)-family proteins [26]. It is strongly expressed in brain and testis tissues [23] and plays a major role not only in synaptic connectivity and plasticity [27, 28] but also in male fertility by maintaining functional blood-testis barrier [29]. Furthermore, it has been reported to be aberrantly expressed in gastric and hepatocellular carcinomas [30], high-grade glioma [31], and invasive ductal carcinoma of the breast [32]; however, the underlying molecular basis remains unknown.

In the present study, we showed that the IgSF11 gene is aberrantly expressed not only in glioma but also in cutaneous melanoma tissues. Using immunohistochemistry with a newly established anti-hIgSF11 monoclonal antibody (mAb), we further demonstrated that approximately 30% of melanoma cases exhibited high IgSF11 protein expression. Notably, tumor invasion in cutaneous melanoma tissues was significantly enhanced in the IgSF11-high group compared with the IgSF11-low group. We also revealed that IgSF11 promotes the migration and invasion of melanoma cells in vitro and in vivo. Moreover, we disclosed that IgSF11 cooperates with RAS-associated protein 1 (RAP1) to accelerate melanoma cell migration.

Results

IgSF11 gene is expressed in skin cutaneous melanoma tissues

We first examined the expression of the IgSF11 gene in various types of cancer using The Cancer Genome Atlas (TCGA). As shown in Supplementary Figure S1A, IgSF11 mRNA was highly expressed not only in glioblastoma but also in skin cutaneous melanoma tissues, whereas it was barely detectable in other cancer tissues. Additionally, the Human Protein Atlas analysis showed that the IgSF11 gene was also expressed at least in a range of cutaneous melanoma cell lines (Supplementary Fig. S2). Furthermore, data from the Genotype-Tissue Expression (GTEx) database revealed that IgSF11 transcripts were strongly detected in the testis and moderately observed in the ovary, brain, and adrenal gland tissues (Supplementary Fig. S1B).

Establishment and characterization of an anti-human IgSF11 mAb

We subsequently generated anti-hIgSF11 mAbs using a polypeptide corresponding to the C-terminal amino acid (AA) region 379–393 as an antigen (Fig. 1 A–C; Supplementary Fig. S3). Upon screening by enzyme-linked immunosorbent assay (ELISA), 40 of 375 hybridomas showed high reactivity to the immunized peptide (Supplementary Fig. S4A), and clone #3 exhibited the highest reactivity among the candidate clones (Supplementary Fig. S4B). Next, to verify the specificity of the anti-hIgSF11 mAb (clone #3) and the commercially available rabbit anti-IgSF11 polyclonal antibody (pAb; HPA046377, Sigma-Aldrich), 293T cells were transiently transfected with either mIgSF11 or hIgSF11 expression vector. On Western blot analysis, while the anti-IgSF11 pAb reacted with both mIgSF11 and hIgSF11, the anti-hIgSF11 mAb selectively recognized hIgSF11 but not mIgSF11 (Fig. 1D). The anti-hIgSF11 mAb also detected positive signals by immunohistochemistry of formalin-fixed paraffin-embedded (FFPE) tissues using cell blocks of 293T + hIGSF11 cells (Fig. 1E). In addition, the positive IgSF11 signals disappeared upon antigen absorption (Fig. 1F). Moreover, immunofluorescence staining using the anti-hIgSF11 mAb revealed the expression of endogenous IgSF11 in the human testis tissues, further supporting the selectivity of the established anti-hIgSF11 mAb (Fig. 1G).

Fig. 1.

Establishment and characterization of an anti-human IgSF11 mAb. A Domain structure of hIgSF11 protein. The antigen region is indicated in red. B, C Amino acid sequences of the C-terminal cytoplasmic domains of hIgSF11, mouse IgSF11 (mIgSF11), and the corresponding regions of human JAM family members. Conserved amino acids and the antigen region are shown in light blue and red, respectively. D-F Western blot and immunohistochemical analyses showing the specificity of the anti-hIgSF11 mAb (clone #3). 293T cells were transfected with the indicated expression vectors and analyzed using the revealed anti-IgSF11 Abs (D) and anti-hIgSF11 mAb (E, F) with (F) or without (E, F) antigen absorption. Arrowhead indicates a specific signal for IgSF11. Scale bars, 50 µm. G Confocal images of IgSF11 in human testis tissues. Scale bar, 100 µm

High IgSF11 expression correlates with cell invasion of melanoma tissues

We subsequently performed immunohistochemical analysis to access the expression of IgSF11 protein in cutaneous melanoma tissues resected from 187 patients (Supplementary Table 1). Positive signals for IgSF11 were observed in melanoma tissues with distinct signal intensity (SI) (Fig. 2A). Based on the semi-quantification of IgSF11 expression, 57 of the 187 melanoma cases (30.5%) showed high IgSF11 expression, while the remaining 130 cases (69.5%) exhibited low IgSF11 expression (Supplementary Fig. S5 A and B). Importantly, tumor budding, defined as small clusters of invaded cancer cells, was significantly increased at Clark’s levels 3 and 3/4/5 in the IgSF11-high group compared with the IgSF11-low group (Fig. 2B and C).

Fig. 2.

High IgSF11 expression correlates with melanoma cell invasion. A Representative immunohistochemical images showing weak, moderate, and strong signal intensities for IgSF11 expression in melanoma tissues. HE, hematoxylin–eosin. B Representative IgSF11 and HE images revealing invasion in the indicated melanoma tissues. Tumor cells are indicated in green in the rightmost columns. C Quantification of budding of IgSF11-high and IgSF11-low melanoma tissues. The budding number is shown in the violin plots (solid line, median; dashed lines, first and third quartiles). Representative budding image is revealed in the left panel. Scale bars, 100 µm (A); 200 µm (B); 50 µm (C)

Kaplan–Meier plots showed that the overall and recurrence-free survival rates in the IgSF11-high group were lower than those in the IgSF11-low group, although the differences were not significant (Supplementary Fig. S5 C and D) possibly due to the high number of censored cases during the follow-up period. The 10-year overall survival rates were 43.3% and 30.0% in the IgSF11-high and IgSF11-low expression groups, respectively.

Among the clinicopathological factors, high IgSF11 expression was significantly associated with Clark’s level (p < 0.001), tumor budding (p = 0.020), and TNM stage III/IV (p = 0.030), but not with older age (p = 0.890), gender (p = 0.288), tumor size (p = 0.565), spindle cells (p = 0.095), lymph node metastasis (p = 0.088), distant metastasis (p = 0.137), BRAFV600E gene mutation (p = 0.642) or recurrence (p = 0.082) (Table 1).

Table 1.

Relationship between the IgSF11 signal and clinicopathological factors in patients with melanoma (n = 187)

Parameter		Total		Low IgSF11		High IgSF11		p-value
				(n = 130)		(n = 57)
Age								0.890
	< 70	90	(48.1)	63	(48.4)	27	(44.8)
	≥ 70	97	(51.8)	67	(51.5)	30	(55.2)
Gender								0.288
	Male	94	(50.2)	62	(47.6)	32	(56.1)
	Female	93	(49.7)	68	(52.3)	25	(43.8)
Tumor size (cm)								0.565
	< 10	12	(6.4)	7	(5.3)	5	(8.7)
	< 20	84	(44.9)	59	(45.3)	25	(43.8)
	< 30	28	(14.9)	18	(13.8)	10	(17.5)
	< 40	32	(17.1)	26	(20.0)	6	(10.5)
	≥ 40	22	(11.7)	13	(10.0)	9	(15.7)
	Unknown	9	(4.8)	7	(5.3)	2	(3.5)
Clark’s level								< 0.001
	I	33	(17.6)	33	(25.3)	0	(0.0)
	II	28	(14.9)	24	(18.4)	4	(7.0)
	III	30	(16.0)	16	(12.3)	14	(24.5)
	IV	51	(27.2)	29	(22.3)	22	(38.5)
	V	37	(19.7)	25	(19.2)	12	(21.0)
	Unknown	8	(4.2)	3	(2.3)	5	(8.7)
Budding								0.020
	< 5	33	(17.6)	25	(19.2)	8	(14.0)
	≥ 5	75	(25.1)	39	(30.0)	36	(63.1)
	Unknown	18	(9.6)	9	(6.9)	9	(15.7)
	Clark’s level I/II	61	(32.6)	57	(43.8)	4	(7.01)
Spindle cells								0.095
	(–)	157	(83.9)	113	(86.9)	44	(77.1)
	(+)	30	(16.0)	17	(13.0)	13	(22.8)
Lymph node metastasis								0.088
	(–)	128	(68.4)	93	(71.5)	35	(61.4)
	(+)	41	(21.9)	24	(18.4)	17	(29.8)
	Unknown	18	(9.6)	13	(10.0)	5	(8.7)
Distant metastasis								0.137
	(–)	164	(87.7)	114	(87.6)	50	(87.7)
	(+)	7	(3.7)	3	(2.3)	4	(7.0)
	Unknown	16	(8.5)	13	(10.0)	3	(5.2)
TNM stage								0.030
	0/I/II	123	(65.7)	91	(70.0)	32	(56.1)
	III/IV	49	(26.2)	28	(21.5)	21	(36.8)
	Unknown	15	(8.0)	11	(8.4)	4	(7.0)
BRAFV600E gene mutation								0.642
	(–)	23	(12.2)	16	(12.3)	7	(12.2)
	(+)	9	(4.8)	7	(5.3)	2	(3.5)
	Unknown	155	(82.8)	107	(82.3)	48	(84.2)
Recurrence								0.082
	(–)	119	(63.6)	88	(67.6)	31	(54.3)
	(+)	68	(36.3)	42	(32.3)	26	(45.6)

IgSF11 accelerates melanoma cell migration in vitro

We then determined the expression profile of IgSF11 in three representative human melanoma cell lines A375, 624mel, and 888mel, via RT-qPCR and Western blot analyses. As shown in Supplementary Figure S6 A and B, the expression levels of both IgSF11 mRNA and IgSF11 protein in 888mel cells were higher than those in A375 and 624mel cells. Immunofluorescence staining revealed that IgSF11 was at least partially concentrated on the cell membranes of 624mel and 888mel cells, but was hardly detected along the cell boundaries of A375 cells (Supplementary Fig. S6C).

Next, we generated A375 and 888mel cells expressing IgSF11 using a lentiviral vector system (Fig. 3A). The expression of IgSF11 protein in A375:IgSF11 cells (clones #1/2/3) and 888mel:IgSF11 cells (clones #1/2/3) was confirmed by Western blot analysis (Fig. 3B; Supplementary Fig. S7A), and IgSF11 was observed along the cell membranes by immunofluorescence staining (Fig. 3C; Supplementary Fig. S7B). Overexpression of IgSF11 did not affect the morphological appearances, cell viability, cell proliferation or cell cycle profiles of these cells (Supplementary Fig. S7 C–E; Supplementary Fig. S8 A–E). On the other hand, wound healing and double chamber assays revealed that cell migration was significantly increased in two clones of A375:IgSF11 cells compared with control A375:EGFP cells (Fig. 3 D–G; Supplementary Fig. S8F). Stimulated cell migration was also observed in 888mel:IgSF11 cells (clones #1 and 2; Supplementary Fig. S7F).

Fig. 3.

IgSF11 promotes cell migration of the melanoma cell line A375. A The construct of the hIgSF11 expression vector. EF-1α, elongation factor-1α; F, flag; IRES, internal ribosome entry site. B Western blot for the indicated proteins in the revealed A375 cells. C Confocal images of IgSF11 in the indicated A375 cells. Arrowheads indicate IgSF11 expression along the cell membranes. (D–G) Representative and quantitative wound healing (D, E) and double chamber (F, G) assays for the revealed A375 cells. The wound closure rates and migration indices are plotted and presented in histograms (mean ± SEM; n = 3 and 4 for E and G, respectively). H GSEA showing the enrichment of gene sets associated with epithelial-mesenchymal transition (EMT) in A375:IgSF11 cells. Scale bars, 50 µm (C, D); 100 µm (F)

We subsequently established 888mel:IgSF11^KO and 888mel:IgSF11^KO:IgSF11 cells, and compared their phenotypes with those in control 888mel cells. Knockout of the IgSF11 gene was verified by DNA sequencing (Supplementary Fig. S9), and the absence and rescued expression of IgSF11 protein in these cells were confirmed using Western blot analysis (Fig. 4A). As expected, the loss of IgSF11 protein in 888mel cells did not alter cell proliferation but significantly decreased cell migration (Fig. 4B and C; Supplementary Fig. S7G). In addition, re-expression of IgSF11 in 888mel:IgSF11^KO cells led to a significant increase in cell migration without morphological changes (Fig. 4 A, D and E; Supplementary Fig. S7H).

Fig. 4.

IgSF11 accelerates cell migration of the melanoma cell line 888mel. A Western blot showing the absence and presence of IgSF11 protein in the indicated 888mel cells. B–E Representative and quantitative double chamber assay for the indicated 888mel cells. The migration index is plotted and shown in the histograms (mean ± SEM; n = 3). Scale bars, 100 µm. F GSEA revealing the enrichment of gene sets associated with epithelial-mesenchymal transition (EMT) in 888mel cells

Moreover, RNA sequencing and subsequent gene set enrichment analysis (GSEA) revealed that gene sets of KRAS and IL2–STAT5 signaling, as well as epithelial-mesenchymal transition (EMT), were significantly enriched in A375:IgSF11 cells compared with control A375:EGFP cells (Fig. 3H; Supplementary Fig. S10A). These gene sets were also significantly augmented in wild-type (WT) 888mel cells compared to 888mel:IgSF11^KO cells (Fig. 4F; Supplementary Fig. S10B). IgSF11 altered expression levels of a variety of EMT-related genes in A375 and 888mel cell (Supplementary fig. S10C). Additionally, gene sets related to MAPK, ITGA/B–FAC–RAC, and ITGA/B–RHOG–RAC signaling were also significantly activated in A375 cells in an IgSF11-dependent manner (Supplementary fig. S10A).

IgSF11 promotes tumor invasion of melanoma cells in vivo

Unexpectedly, two weeks after inoculation into SCID (severe combined immunodeficiency) mice, the tumor growth of A375:IgSF11 xenografts was significantly decreased compared with that of A375:EGFP xenografts (Fig. 5A and B). IgSF11 did not affect the cell proliferation of A375 xenografts (Fig. 5 C and D), which is consistent with the results obtained from the in vitro experiments. On the other hand, IgSF11 increased apoptosis in A375 xenografts (Fig. 5 E and F), suggesting that the reduced growth of A375:IgSF11 tumor was at least partially attributed to enhanced apoptosis. Next, to gain insight into reduced growth of A375:IgSF11 xenografts, we evaluated RNA sequencing data for the tumor tissues composed of human A375:IgSF11 and mouse stromal cells. As sown in Supplementary figure S11A, the expression of marker genes for mouse dendric cells and natural killer (NK) cells, e.g., Cd86, Itgax, Gzmb, Ncam1, and Prf1, was highly upregulated in A375:IgSF11 tissues compared with that in A375:EGFP tissues. A variety of chemokine-related genes were also enriched in A375:IgSF11 tissues compared with that in A375:EGFP ones (Supplementary fig. S11B and C). Of note, the expression of mouse Cxcl9/10/14/16 and human CXCL10/11 genes, all of which regulate recruitment, migration, differentiation and activation of dendric cells and/or NK cells [33–35], were significantly augmented in A375:IgSF11 xenografts compared to those in A375:EGFP ones.

Fig. 5.

IgSF11 enhances cell invasion of the melanoma cell line A375 in vivo. A, B Gross appearance and weight of the indicated xenografts at 14 d after inoculation. Tumor weights are plotted and shown in the histograms (mean ± SEM; n = 5). C, D Representative images and quantification of Ki-67 staining of the indicated xenografts. Ki-67 index is plotted and shown in the histograms (mean ± SEM; n = 5). E, F Representative apoptotic bodies and their quantification of the indicated xenografts. Apoptotic index is plotted and shown in the histograms (mean ± SEM; n = 5). G Representative HE (hematoxylin–eosin) and GFP images of the indicated xenografts. Tumor cells are indicated in green in the middle columns. H Quantification of budding in the indicated xenografts. The number of tumor buds is plotted and shown in histograms (man ± SD; n = 10). Scale bars, 1 cm (A); 50 µm (C and E); 100 µm (G)

IgSF11 forms a complex with RAP1

To identify IgSF11-interacting proteins, we performed immunoprecipitation-mass spectrometry (IP-MS) analysis. Various proteins enriched in IgSF11-IP of A375:IgSF11 and 888mel:IgSF11 cells, but not in their IgG-IP, were isolated as candidates that form a complex with IgSF11 (Supplementary Fig. S12S). Among them, we focused on RAP1A/B because they were detected in the IgSF11-IP of both A375:IgSF11 and 888mel:IgSF11 cells and are known to accelerate tumor cell migration and invasion [36–39] similar to IgSF11. As expected, the expression of RAP1A was detected in both A375 and 888mel:IgSF11^KO cells, and the levels were not altered by overexpression of IgSF11 (Supplementary Fig. S13). Interestingly, three-dimensional structural prediction using AlphaFold strongly suggested that the C-terminal AA region L372–R378 binds to Y40–E45 of RAP1A/B (Fig. 6A and B). By IP/immunoblot (IB) assay, both RAP1A and RAP1B were associated with IgSF11 in 293T cells (Fig. 6C and D). Additionally, immunofluorescence analysis showed that endogenous RAP1A was at least in part colocalized with IgSF11 in A375:IgSF11 and 888mel:IgSF11^KO:IgSF11 cells (Supplementary Fig. S14A and B).

Fig. 6.

IgSF11 forms a complex with RAP1 in 293T cells. A Three-dimensional structural analysis using PyMOL indicating that L372–R378 of IgSF11 binds to Y40–E45 of RAP1A. B Predicted association between IgSF11L372–R378 and RAP1Y40–E45. C, D IP-IB analysis showing the IgSF11/RAP1A and IgSF11/RAP1B complexes in 293T cells. E IP-IB analysis revealing that the association between IgSF11∆L372–R378 and RAP1A is reduced in 293T cells. IP, immunoprecipitation; IB, immunoblot. F, G Quantitative wound healing assay for the revealed A375 and 888mel:IgSF11^KO cells that transiently transfected with either IgSF11 or IgSF11∆L372–R378. The wound closure rates are plotted and presented in histograms (mean ± SEM; n = 3)

L372–R378 of IgSF11 is required for recruiting RAP1 and promoting melanoma cell migration

To validate the biological relevance of L372–R378 in IgSF11, we subsequently constructed the IgSF11∆L372–R378 expression vector. As shown in Fig. 6E, the formation of the IgSF11/RAP1A complex was hindered when IgSF11∆L372–R378 and RAP1A were transfected into 293T cells. Furthermore, the migration of A375 + IgSF11∆L372–R378 and 888mel:IgSF11^KO + IgSF11∆L372–R378 cells was significantly reduced compared with that of A375 + IgSF1 and 888mel:IgSF11^KO + IgSF11 cells, respectively (Fig. 6F). Taken collectively, these results indicate that L372–R378 of IgSF11 is responsible for recruiting RAP1 and accelerating melanoma cell migration.

Discussion

The expression of the IgSF11 gene and IgSF11 protein is known to be upregulated in high grade glioma tissues [31]. In the present study, using the TCGA database, we showed that IgSF11 mRNA is highly expressed not only in glioblastoma tissues but also in skin cutaneous melanoma tissues, in good agreement with the result obtained from another database [40]. While IgSF11 is reported to be upregulated in gastric and hepatocellular carcinomas [30], as well as invasive ductal carcinoma of the breast [32], TCGA analysis revealed that IgSF11 transcripts are hardly detected in malignant tumors other than high grade glioma and melanoma tissues. Additionally, via immunohistochemical analysis using the newly developed mAb that selectively recognizes hIgSF11, we found that IgSF11 protein is expressed in cutaneous melanoma tissues with varied intensity and proportion scores, and that about 30% of melanoma cases exhibited high IgSF11 expression.

We also demonstrated in the present work that high IgSF11 expression significantly corelates with two indicators for melanoma invasion, namely Clark’s level [41] and high budding [42]. In addition, the 10-year overall and recurrence-free survival rate in the IgSF11-high group of melanoma patients was lower than that in the IgSF11-low group, though large-scale clinical analysis would be required to draw more solid conclusions. In contrast, Shekari et al. have very recently reported that high IgSF11 expression is not associated with any clinicopathological parameters, as well as overall and recurrence-free survival in melanoma patients [40]. The reason for these discrepancies could be explained by methodological differences between two analyses. We determined the expression of IgSF11 protein within cutaneous melanoma tissues by immunostaining using the selective anti-IgSF11 mAb and verified the clinicopathological relevance of high IgSF11 expression in melanoma patients. On the other hand, Shekari et al. used gene expression database in whole cancer tissues, which include not only cancer cells but also diverse non-cancer cells, and performed bioinformatic study. In this regard, we should mention our previous reports showing that high CLDN10 and low CLDN12 expression are poor prognostic markers of clear cell renal cell carcinoma and uterine cervical cancer, respectively, which are in contrast to the results obtained from the TCGA database [20, 43].

Another conclusion of this study is that IgSF11 promotes the migration and invasion of melanoma cells in vitro and in vivo. This conclusion was drawn from the following findings: 1) the migration of A375:IgSF11 and 888mel:IgSF11 cells was significantly increased compared with that of A375:EGFP and 888mel:EGFP cells, respectively, via wound healing and double chamber assays; 2) KO of the IgSF11 gene significantly reduced migration in 888mel cells, and re-expression of IgSF11 in 888mel:IgSF11^KO cells reversed the decreased cell migration; and 3) tumor budding was increased in A375:IgSF11 xenografts compared with A375:IgSF11 xenografts. The knockdown of IgSF11 expression by using siRNA decreases the migration of the A2058 melanoma cell line [40], further supporting our conclusion. Taken together with previous studies reporting that IgSF11 increases the migration of melanophores in Zebrafish [44], which correspond to melanocytes in homeotherms, aberrant IgSF11 expression enhances the migration and invasion of melanoma cells, likely by hijacking the physiological function of IgSF11 in melanocytes. In addition, since IgSF11 is preferentially expressed in normal brain and testis tissues, it may also be involved in the migration of neuronal cells and germ cells in such tissues.

The most important finding of the present study is that IgSF11 cooperates with RAP1 to accelerate melanoma cell migration and invasion. This was evidenced by the following results: 1) IP-MS analysis showed a possible association between IgSF11 and RAP1A/1B in both A375:IgSF11 and 888mel:IgSF11 cells; 2) IgSF11 promoted melanoma cell migration and invasion as described above, while RAP1 is also known to stimulate tumor cell migration and invasion [36–39]; 3) GSEA indicated that IgSF11, like RAP1 [45–48], activates the expression of EMT-related genes in A375:IgSF11 and 888mel:IgSF11 cells; 4) three-dimensional structural prediction suggested that the C-terminal AA region L372–R378 of IgSF11 binds to Y40–E45 of RAP1A; 5) IP/IB assay revealed that the IgSF11/RAP1A and IgSF11/RAP1B complexes are formed in 293T cells; and 6) RAP1A was at least in part colocalized with IgSF11 in A375:IgSF11 and 888mel:IgSF11^KO:IgSF11 cells. Moreover, since RAP1 activates various intracellular signals such as MAPK and SRC/PI3K/AKT [46, 49] and enhances melanoma cell migration [36, 37, 50], it is rational that IgSF11–RAP1 signaling facilitates cell migration and invasion of cutaneous melanoma.

We also found that L372–R378 of IgSF11 is required for recruiting RAP1A and driving melanoma cell migration. This is evident because the formation of the IgSF11/RAP1A complex was suppressed in 293T cells in the absence of L372–R378 of IgSF11. Additionally, the migration of A375 + IgSF11∆L372–R378 and 888mel:IgSF11^KO + IgSF11∆L372–R378 cells was significantly prevented compared with that of A375 + IgSF1 and 888mel:IgSF11^KO + IgSF11 cells, respectively, supporting the biological significance of IgSF11L372–R378 in melanoma cells. Taking these results together with our AlphaFold prediction, L372–R378 of IgSF11 likely binds to RAP1A via Y40–E45 to stimulate malignant phenotypes of melanoma cells. In this context, it should also be noted that JAM-A, a member of the JAM family, is indirectly associated with RAP1A via afadin/AF6 in the human colon carcinoma cell line SK-CO15, promoting cell migration [51, 52]. Additionally, JAM-A relates to afadin/AF6 and activates RAP1 in breast cancer cells to stimulate cell migration [53]. Thus, IgSF11 and JAM-A appear to recruit and activate RAP1, thereby driving cancer cell migration, through direct and indirect mechanisms, respectively.

Another issue that should be discussed is the mechanisms behind the reduced growth of A375:IgSF11 xenografts. Our RNA sequencing analysis revealed that IgSF11 enhances infiltration of both NK cells and dendric cells into A375:IgSF11 tumor tissues in SCID mice, as well as the expression of various chemokine-related genes that are involved in recruitment, migration, differentiation and activation of these innate immune cells [33–35]. Thus, we speculate that excessive NK cells trigger cell death of A375:IgSF11 melanoma cells, resulting in the suppression of the tumor growth in SCID mice. In addition, it is known that immune tolerance is established via the interaction between IgSF11 and VISTA (V-domain immunoglobulin suppressor of T cell activation) [54, 55]. Hence, even though outstanding dendric cell infiltration within A375:IgSF11 tissues, the loss of acquired immune cells in SCID mice and the consequent failure of immune tolerance may also contribute to the inhibition of tumor growth. Further studies are required that employ approaches such as autologous transplantation.

Conclusions

We developed a specific anti-hIgSF11 mAb and demonstrated that tumor invasion was significantly increased in the IgSF11-high group of cutaneous melanoma cases. By generating a series of human melanoma cell lines and comparing their phenotypes, we also demonstrated that IgSF11 accelerates melanoma cell migration and invasion in vitro and in vivo. Importantly, we uncovered that IgSF11 forms a complex with RAP1, and IgSF11–RAP1 signaling facilitates melanoma cell migration and invasion. IgSF11–RAP1 signaling could provide novel insights into the molecular mechanisms of cutaneous melanoma and may be a promising therapeutic target for this malignancy. To this end, potential adverse effects of targeting the IgSF11–RAP1 axis on normal IgSF11-positive cells should be also determined.

Methods

Gene expression data

Comprehensive gene expression data for cancer tissues [56] were obtained from cBioPortal (https://www.cbioportal.lorg/). The expression data of normal human adult tissues were obtained from the GTEx Portal on March 1, 2024 [57].

Antibodies

The antibodies used in this study are listed in Supplementary Table S2.

Rat mAbs against hIgSF11 were generated using the iliac lymph node method [58]. A polypeptide (C)-GSSPQVMSRSNGSVS, corresponding to the AA region 210–224 of hIgSF11, was conjugated to keyhole limpet hemocyanin (Imject Maleimide-Activated mcKLH, 77605,Thermo Fisher Scientific, Waltham, MA, USA) before immunization. The details of the procedure and materials were described in the previous report [20].

Cell culture

The human melanoma cell lines A375 (CVCL_0132), 624mel (CVCL_8054) and 888mel (CVCL_4632) were kindly provided by the Department of Pathology, Sapporo Medical University School of Medicine, Sapporo, Japan. These cell lines were cultivated in Dulbecco's modified Eagle medium (DMEM; 044–29765; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan), with the addition of 10% fetal bovine serum (FBS; 175012; Nichirei, Tokyo, Japan) and 1% penicillin–streptomycin (161–23181; FUJIFILM Wako Pure Chemical Corporation).

Expression vectors, transfection, and establishment of stable cell lines

The protein coding regions of hIgSF11, mIgSF11, hRAP1A, and hRAP1B were amplified by RT-PCR using cDNA of 293T as templates. The PCR products were subsequently cloned into the NotI/BamHI site of the CSII-EF-MCS-IRES2-Venus plasmid (RDB04384, Riken, Wako, Japan). The generation of hIgSF11ΔL372–R378 was achieved by using a conventional site-directed mutagenesis protocol with the mutagenic primers 5'-GGGGTCAACATATGACTGGGTCATCACCACAGGTGAT-3'and 5'-CCTGTGGTGATGACCCAGTCATATGTGACCCGGGA-3'.

The overexpression cell lines were established by lentiviral transduction using the packaging plasimids psPAX2 (#12260; Addgene, Watertown, MA, USA) and pCMV-VSV-G (#8454; Addgene), as previously described [20]. The cells were single-cell cloned by limiting dilution in 96-well cell culture plates.

Genome editing

IgSF11^KO cell lines were generated using CRISPR/Cas9 technology. Two single guide RNAs (sgRNAs), corresponding to the second exon and the subsequent intron of the human IgSF11 gene (Fig. 4A), were cloned into the lentiCRISPR v2 plasmid (#52961, Addgene). Although lentiCRISPR v2 was originally designed for lentiviral transduction, the plasmids were directly and transiently introduced into the parental cells using Lipofectamine 3000 (15292465, Thermo Fisher Scientific). Twenty-four hours after transfection, the cells were exposed to 1 µg/mL puromycin for 24 h to select positive clones. The next day, the cells were isolated by limiting dilution in 96-well plates, followed by genomic PCR screening and subsequent verification by Western blotting.

Cell blocks

293T cells were transiently transfected with vectors expressing IgSF11 or EGFP using polyethyleneimine (PEI Max; #24765–1; Polysciences, PA, USA). Two to three days after transfection, they were subjected to centrifugation and fixation with 10% formalin for 12 h at 4 °C. The fixed cell pellets were mixed with 1% sodium alginate, followed by 1 M calcium chloride, and then embedded in paraffin using Tissue-Tek VIP 5 Jr (Sakura Finetek Japan, Tokyo, Japan).

Tissue collection and immunohistochemistry

FFPE tissues were collected from 187 patients with cutaneous melanoma (ages 21–100 years) who underwent tumorectomy between 2006 and 2018 at Fukushima Medical University Hospital. A comprehensive set of clinical and pathological data, including age, gender, tumor size, Clark's level, budding, spindle cells, TNM stage, lymph node metastasis, and recurrence was obtained (see Supplementary Table S1). The presence of distant metastases was determined by imaging.

For immunohistochemical staining of IgSF11, rehydrated sections were subjected to antigen retrieval using Immunosaver (333; Nisshin EM, Tokyo, Japan), following the manufacturer's protocol. Melanin removal was achieved by applying 20% hydrogen peroxide (H₂O₂) to the tissue sections at room temperature overnight. Following treatment with an avidin/biotin blocking kit (415041; Nichirei) and 0.5% casein (218680; Merck Millipore, Darmstadt, Germany), the sections were then incubated overnight at 4 °C with 10 µg/ml of rat anti-IgSF11 mAb (clone #3). The secondary antibody reaction was performed by using the Histofine mouse PO-Rat secondary antibody (414311; Nichirei). Signal intensity was then amplified by sequential incubation of the sections with 1.5 nM of biotinyl tyramide (SML2135; Merck Millipore) and horseradish peroxidase (HRP). The color reaction was developed using Histofine DAB substrate kit (425011; Nichirei).

Immunostaining results were interpreted independently by two pathologists (Y.K. and K.S.) using a semi-quantitative scoring system (immunoreactive score; IRS) [59]. The immunostaining reactions were evaluated based on a signal intensity (SI: 0, no stain; 1, weak; 2, moderate; 3, strong) and the percentage of positive cells (PP: 0, < 1%; 1, 1–10%; 2, 11–30%; 3, 31–50%; and 4, > 50%). The SI and PP scores were then multiplied to generate the IRS for each case, and the lowest IRS among the two evaluators was used for analysis. Based on the receiver operating characteristic (ROC) analysis, we divided the samples into two groups: IgSF11-low (average IRS < 8) and IgSF11-high (average IRS ≥ 8). Tumor budding at the invasive front, which has been reported as a poor prognostic factor [42], was evaluated and compared with IgSF11 expression levels.

For the analysis of xenografted tumors, thin-sliced FFPE samples were subjected to antigen retrieval by boiling citrate buffer (pH 6.0). Following blocking with 0.5% casein (218680; Merck Millipore), the specimens were incubated overnight at 4 °C with primary antibodies. For the secondary antibody and color development, Histofine Simple Stain SAB-PO (MULTI; 424041; Nichirei), Histofine Simple Stain mouse MAX-PO(Rat; 414311; Nichirei), and Histofine DAB substrate kit (425011; Nichirei) were used.

Immunofluorescence and imaging

Melanoma and normal testis specimens were embedded in OCT compound and thin sliced to 10 µm at –20 °C. The cells were grown on a glass-based dish (3910–035; IWAKI, Shizuoka, Japan) that had been coated with Cellmatrix Type I-A (Nitta Gelatin, Osaka, Japan). The samples were then fixed in acetone-ethanol (50% acetone, 50% ethanol) for 10 min at –20 °C. Following thorough rinsing with phosphate-buffered saline (PBS), the samples were then preincubated in 0.5% casein (218680; Merck Millipore) for 10 min. They were then incubated overnight at 4 °C with primary antibodies diluted in Signal Booster Immunostain F (BCL-ISF; Beacle, Kyoto, Japan). Next, they were rinsed again with PBS, followed by a reaction for 1 h with appropriate secondary antibodies. All specimens were examined using a laser-scanning confocal microscope (FV1000, Olympus, Tokyo, Japan).

Antigen absorption

One µg of the peptide antigen (C)-GSSPQVMSRSNGSVS was mixed with 2 µg of the rat anti-IgSF11 mAb in Solution F, Signal Booster Immunostain (BCL-ISF; Funakoshi, Tokyo, Japan) and incubated overnight at 4 °C under gentle rotation to allow antigen–antibody interaction. After centrifugation at 15,000 rpm for 30 min, the supernatant was used as the primary antibody.

RNA extraction, RT-qPCR, and RNA sequencing

Total RNA was isolated from cells using TRIzol RNA Isolation Reagents (15596018; Thermo Fisher Scientific). For RT-qPCR, reverse transcription was performed using the SensiFAST cDNA Synthesis Kit (BIO-65054; Meridian Bioscience, Cincinnati, OH, USA), and target genes were quantified using THUNDERBIRD SYBR qPCR Mix (QPS-201, Toyobo, Osaka, Japan) and the Step One Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) using the primers listed in Supplementary Table S3. The expression levels of the target genes were normalized to GAPDH expression.

RNA sequencing was performed by Rhelixa (Tokyo, Japan). A total of 6 GB of 150-base paired-end data was acquired using the Illumina NovaSeq 6000 system. Trim galore was used to trim the sequenced reads. Then the trimmed FASTQ data were mapped to the human reference genome sequence (hg38) by Bowtie 2. The mapped data was quantified and visualized by SeqMonk and its standard RNA sequencing quantitation pipeline.

Cell viability, proliferation, cell cycle, and migration assays

Cell viability was analyzed by the CellTiter 96 AQueous One Solution Cell Proliferation Assay (G3582; Promega, Madison, WI, USA) according to the manufacturer’s instructions.

The proliferation index of cells was determined by evaluating the incorporation of BrdU (19–960, Sigma-Aldrich, St. Louis, MO, USA). Approximately 24–48 h after cell passage, cells were exposed to BrdU for 30 min. The specimens were then fixed with 4% paraformaldehyde and 0.1% Triton X-100, followed by immunostaining with an anti-BrdU antibody according to the standard protocol.

Cell cycle analysis was performed using flow cytometry (FACS). The cell suspension was fixed with 95% ethanol for 30 min at 4 °C. For staining, cells were incubated with 50 µg/mL propidium iodide and 100 µg/mL RNase A in PBS for 30 min at 37 °C. Subsequently, the cells were analyzed using a FACS CantoII system (BD Biosciences, San Jose, CA, USA) and FlowJo software v10.9.0 (BD Biosciences).

Cell migration was assessed by wound healing and double-chamber assays. The wound healing assay was performed as previously described [20]. For the double-chamber assay, cells were seeded in inserts (353097; Corning, NY, USA) of 24-well double-chamber culture plates and grown in DMEM supplemented with 0.1% fetal bovine serum (FBS; 175012; Nichirei) and 1% penicillin–streptomycin for 24 h. They were fixed with 4% paraformaldehyde and stained with 0.04% crystal violet solution. The number of cells that traversed the pores was meticulously counted through histological analysis.

Xenograft model

Cell suspensions containing 2 × 10⁶ of A375:EGFP or A375:IgSF11 cells were prepared in 200 µL of Type 1 A collagen (Nitta Gelatin) and stored on ice. The cells were then subcutaneously injected into the dorsal flank of the 8-week-old SCID mice (C.B-17/Icr-scid/scidJcl, CLEA Japan). Two weeks after transplantation, the mice were euthanized, and the tumors were excised. Apoptotic index was histologically assessed by a pathologist.

Immunoprecipitation and immunoblot

Total cell extracts were collected using IP Lysis Buffer (87787; Thermo Fisher Scientific) supplemented with cOmplete Protease Inhibitor Cocktail (04693116001; Merck, Darmstadt, Germany). The samples were sonicated with three or four bursts of 10 s. Immunoprecipitation was performed using the Immunoprecipitation Kit Protein G (11719386001; Sigma-Aldrich), according to the manufacturer's protocol. Normal rat IgG (147–09521; FUJIFILM Wako Pure Chemical Corporation) was used as a negative control. Whole cell lysates or immunoprecipitated samples were mixed with sample loading buffer containing 2-mercaptoethanol, and subsequently boiled for 10 min. The samples were then subjected to one-dimensional SDS-PAGE, followed by electrophoresis-based transfer onto a polyvinylidene difluoride (PVDF). The membranes were saturated with PVDF Blocking Reagent for Can Get Signal (NYPBR01; Toyobo) for 30 min. After rinsing in TBS2T (11.5 mM Tris-base, 38 mM Tris-hydrochloride, 300 mM sodium chloride with 0.1% Tween 20), the membranes were incubated overnight with a primary antibody solution at 4ºC, followed by a one-hour incubation with HRP-conjugated secondary antibodies.

Immunoprecipitation-mass spectrometry (IP-MS)

IP-MS was used to identify IgSF11-binding proteins. First, whole cell extracts of A375, A375:IgSF11, 888mel, and 888mel:IgSF11 were immunoprecipitated with a rabbit anti-FLAG mAb. Protein purification and digestion were performed using the sample preparation (SP3) method [60]. The samples were analyzed using ultra-high-performance liquid chromatography connected to a Q-Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific) through a nano electrospray ion source (AMR Inc., Tokyo, Japan). DIA-MS data files were analyzed using DIA-NN (version 1.8.1, https://github.com/vdemichev/DiaNN) by matching them against an in silico human spectral library [61].

Protein structure prediction

The prediction of protein structure was conducted using the AlphaFold2 program (v.2.3.2; DeepMind Technologies, London, UK), and the resulting visualizations were rendered using PyMol v.2.4.2. (Schrödinger, Inc., New York, NY, USA). The AA sequences of RAP1A, RAP1B, and IgSF11 were retrieved from the UniProt database (identifiers: Q5DX21 for IgSF11, P62834 for RAP1A, and P61224 for RAP1B).

Statistical analyses

Statistical significance of differences in cell viability and cell proliferation was analyzed using the Mann–Whitney U test, while differences in cell migration were analyzed using Welch’s t-test. Gene Set Enrichment Analysis (GSEA) was performed using GSEA v.4.2.3 software and hallmark gene sets, which are publicly available from the Broad Institute. Survival analyses were conducted using the Kaplan–Meier method, with differences between groups evaluated using the log-rank test. A p-value of < 0.05 was considered statistically significant for all comparisons. All statistical analyses were conducted using GraphPad Prism 9 (GraphPad Software) and SPSS v.29 (IBM). Quantitative data were presented as scatter bar charts with error bars indicating the standard error of the mean (SEM). In the MTS assay, each plot represents the mean value from six replicates for each clone.

Abbreviations

AA

Amino acid

BSA

Bovine serum albumin

BrdU

5-Bromo-2’-deoxyuridine

CLDN

Claudin

DMEM

Dulbecco's modified Eagle medium

ELISA

Enzyme-linked immunosorbent assay

FBS

Fetal bovine serum

FFPE

Formalin-fixed paraffin-embedded

GSEA

Gene set enrichment analysis

HRP

Horseradish peroxidase

IB

Immunoblot

IP

Immunoprecipitation

IgSF11

Immunoglobulin superfamily 11

IRS

Immunoreactive score

JAM

Junctional adhesion molecule

mAb

Monoclonal antibody

MS

Mass spectrometry

PBS

Phosphate-buffered saline

PCR

Polymerase chain reaction

pAb

Polyclonal antibody

PVDF

Polyvinylidene fluoride

TBS

Tris-buffered saline

Authors’ contributions

Investigation, Y.K., K.S., M.I., M.K, and S.K.; formal analysis, Y.K., K.S., M.I., M.K, S.I., S.K., T.Y., Y.H., and H.C. resources, S.I. and T.Y.; funding acquisition, Y.K.; visualization, Y.K. and K.S.; writing–original draft, Y.K., K.S., and H.C.; conceptualization, writing–review & editing, supervision, K.S. and H.C. All authors have read and approved the final manuscript.

Funding

This work was supported by JSPS KAKENHI (grant number: 23K07748).

Data availability

RNA-seq data were deposited in the DNA Data Bank of Japan (https://ddbj.nig.ac.jp/gea) at the accession number GSE285994. Further inquiries can be directed to the corresponding author.

Declarations

Ethics approval and consent to participate

All animal experiments conformed to the National Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Experiments Committee of Fukushima Medical University (approval code, 2024072; approval date, 7 Jun, 2024). The human studies were approved by the Research Ethics Committee of Fukushima Medical University (approval code, 2022-115; approval date, Sep 13, 2022) and were conducted in accordance with the 1964 Helsinki Declaration or comparable standards. Informed consent (broad consent) was obtained from all of the participants in this study. Since it was conducted as a retrospective study using cases with a follow-up period of more than ten years, the patients had already died or stopped visiting the hospital. The experimental protocol has been disclosed on the website, and the patients or their representatives were able to decline to participate in the survey if they wanted.

 Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.