Melanoma is a highly metastatic cancer, and its incidence has increased over the past several decades. The prognosis of metastatic melanoma is grim, with a median survival rate of 9 months. Once a distant metastatic clone is established, the 5-year survival rate decreases to 10% despite the availability of advanced therapeutics . Besides their higher degree of ‘stemness’ and efficiency in evading attack by host immune cells, melanoma cells are capable of initiating vascularization, leading to metastasis . Although the extent of vascularization, as measured by intratumoural microvessel density (MVD), is not a prognostic marker in cutaneous melanoma , increased angiogenesis and lymphangiogenesis in metastatic sentinel lymph nodes are associated with distant metastasis in patients with melanoma . MVD has been associated with many clinicopathological parameters that affect prognosis, including overall survival and relapse of melanoma [5,6]. Among the angiogenic and antiangiogenic factors, vascular endothelial growth factor receptor 2 (VEGFR2) is a key receptor that is activated by vascular endothelial growth factor (VEGF) to initiate downstream molecules and recruit endothelial progenitor cells to vascularization sites . Combined application of anti-VEGFR1 and anti-VEGFR2 antibodies is sufficient to inhibit melanoma growth and metastasis , indicating that vascularization is important for melanoma metastasis.
NLK is a serine/threonine kinase that belongs to the extracellular signal-regulated kinases/microtubule-associated protein kinase families . Phosphorylation of lymphoid enhancer-binding factor 1 (LEF1) by nemo-like kinase (NLK) plays dual and opposite roles. In neural progenitor cells, NLK phosphorylates LEF1 at Thr155–Pro156 and Ser166–Pro167 sequences, resulting in the dissociation of LEF1 from histone deacetylase 1 and ultimately promoting Wnt signalling pathways . In HeLa and HEK293 cells, NLK phosphorylates LEF1 and inhibits its DNA-binding activity . Growing research supports the role of NLK as either a tumour suppressor or oncogene in various cancer types. For example, NLK expression in hepatocellular carcinoma is associated with a poor prognosis , and upregulation of NLK is an independent prognostic factor in colorectal cancer . In contrast, NLK appears to be a tumour suppressor in prostate cancer, glioma, breast cancer, ovarian cancer and lung cancer [14–19]. However, the degree of NLK expression in melanoma and its clinical significance are unknown.
In a previous study, we found that levels of the phosphorylated form of Lef1 are reduced in the lung tissue of NLK knockout animals, whereas both VEGF and VEGFR2 expressions are elevated in the lung epithelial cells of these mice . Furthermore, we showed that NLK phosphorylates histone deacetylase 1 and inhibits Wnt target gene expression in primary mouse embryonic fibroblasts . In the present study, we hypothesized that the loss of NLK expression may be one of the causes of melanoma vascularization. Hence, the aim of the present study was to investigate the potential associations between NLK expression, vascularization and metastasis in melanoma.
Patients and methods
Patients and specimens
The study was approved by the Ethics Committee of the General Hospital of Ningxia Medical University. Tissue microarrays of melanoma were purchased from Xian Alenabio Biotechnology (Xian, China) and informed consent was obtained from every patient. All patients were ethnically Asian. The collection time of all samples was from 2001 to 2014. No radiotherapy or chemotherapy was performed before surgery. Samples included 117 primary tumours, 58 lymph node metastases and 16 normal skin tissues. Melanoma specimens from the skin were assessed as stage 0, I, II, III or IV according to the tumour, node and metastasis (TNM) melanoma staging system of the American Joint Committee on Cancer. The clinicopathological features of tissue specimens are shown in Table 1.Tissues were fixed in 4% formalin and embedded in paraffin before the samples were prepared for tissue microarrays. At least two pathologists made the final diagnosis of pathological tissue on the basis of haematoxylin and eosin staining and immunohistological staining for melanoma-specific markers such as HMB45 and S100.
The streptavidin–peroxidase method was applied for immunohistochemical (IHC) staining. Slides were heated in an oven at 60°C for 1 h, and then deparaffinized with xylene and hydrated with graded ethanol. Next, sections were boiled for 13 min in citrate buffer (pH 6.0) to retrieve the antigen. After washing in PBS, endogenous peroxidase was inactivated by covering the tissue with 3% hydrogen peroxide for 10 min at room temperature. The samples were blocked with normal goat serum (ZLI-9022; ZSGB Biotechnology, Beijing, China) for 1 h at 37°C. The primary antibody was diluted according to the recommendation and incubated with the sample overnight at 4°C. The following primary antibodies were used separately: anti-NLK mouse monoclonal antibody (diluted 1: 50, sc-4836; Santa Cruz Biotechnology, Santa Cruz, California, USA) and anti-VEGFR2 rabbit monoclonal antibody (diluted 1: 500, cat. no. 9698; Cell Signaling Technology Beverly, Massachusetts, USA). The next day, sections were washed and incubated with biotin-labelled goat antibodies against rabbit/mouse IgG for 20 min at 37°C, and then with the streptavidin–peroxidase complex (KIT-9710; MXB Biotechnology, Fuzhou, China). Subsequently, the sections were washed three times in PBS, incubated in 3-amino-9-ethyl-1-carbazole chromogen for 15 min and then counter-stained with haematoxylin–eosin solution. PBS instead of the primary antibody served as the negative control.
Two pathologists who were blinded to the study carried out the IHC analysis. Immunohistochemical staining of NLK was assessed using a semiquantitative scoring system according to a method described previously by Weichert et al. . The intensity of immunostaining (category A) was classified as 0 (no staining), 1 (weak staining), 2 (moderate staining) and 3 (strong staining). The percentage of positive tumour cells (category B) was considered 0 (none), 1 ( ≤ 10%), 2 (11–50%), 3 (51–80%) and 4 ( > 80%). The final IHC score was 0–12, which was calculated by multiplying category A and B. Staining of tumour cells with the final staining scores of 0, 1–4, 5–8 and 9–12 was assigned as negative ( − ), slightly positive ( + ), moderately positive ( + + ) and strongly positive ( + + + ), respectively. For the statistical analysis, a score of 0–4 was assigned to the low-expression group and 5–12 to the high-expression group. At least three areas were counted for every sample under × 400 magnification using Aperio ImageScope software (Leica, Wetzlar, Germany) and an Olympus BX51 microscope (Tokyo, Japan). For the quantification of MVD, the hot-spot method reported by Weidner  was used to assess each specimen. We first identified the areas with the highest number of microvessels at a low magnification (× 10) as the hot spot, and then magnified the hot spot to a higher magnification (× 20) and counted the number of microvessels. Any single endothelial cell or cell mass that was stained by anti-VEGFR2 was considered a countable microvessel as long as it was clearly separated from the surrounding cells, blood vessels or other tissues. Three different areas were counted for every slide and the arithmetic mean was considered the MVD of the sample.
To study the expression of NLK in normal melanocytes, immunofluorescence assays were used to detect both NLK and HMB45 antigens in skin tissue. The processing of tissues and staining protocols were the same as described previously . Anti-HMB45 mouse antibody (Ab787; Abcam, Cambridge, UK) was used at a dilution of 1: 100; AlexaFluor 488 goat anti-mouse secondary antibody (Ab150117; Abcam) and AlexaFluor 647 goat anti-rabbit secondary antibody (Ab150079; Abcam) were used at a dilution of 1: 400. Nuclei were counter-stained with DAPI. Pictures were taken using an Olympus BX51 (Tokyo, Japan) microscopic imaging system.
Statistical analyses were carried out using IBM SPSS statistics (version 22.0, Chicago, Illinois, USA) and GraphPad Prism (version 6.0, La Jolla, California, USA). The associations between the expression of NLK and the clinicopathological parameters in melanoma patients were assessed using the Mann–Whitney test. The Mann–Whitney test was also used to determine the correlation between MVD and NLK. P values less than 0.05 were considered statistically significant.
Clinicopathological features of tissue specimens
A total of 175 patients with malignant melanoma included 84 (48.0%) patients with cutaneous melanoma, 33 (18.9%) patients with mucosal melanoma and 58 (33.1%) patients with metastatic melanoma (lymph node metastases). There were a total of 83 males and 92 females, ranging in age from 7 to 88 years (Table 1). According to the American Joint Committee on Cancer melanoma staging system, 11 (13.1%), 64 (76.2%), eight (9.5%) and one (1.2%) patients were, respectively, diagnosed with stage I, II, III and IV cutaneous melanoma. In addition, normal skin tissues from 16 healthy patients were examined as a control in this study. The clinicopathological features of the tissue specimens are shown in Table 1.
Expression of NLK is downregulated in metastatic melanoma
Immunohistochemical analysis using NLK antibodies showed that NLK was highly expressed in the cytoplasm of basal layer cells of the normal skin epidermis (Fig. 1a–e). On investigating the expression of NLK in melanoma, it was found that 129 out of 175 patient samples were NLK positive (Table 1); 26.3% of the samples showed a complete absence of NLK expression. The overall positive rate was 73.7%. However, the majority of patients (65.1%) were NLK negative or weakly positive (1 + ) and 34.9% of the total patient samples were NLK strongly positive (2 + or 3 + ). In these samples, NLK was mainly detected in the cytoplasm and perinuclear region of tumour cells (Fig. 1h–j). Furthermore, 16 (72.7%) out of 22 thin (T1, T2; < 2 mm in diameter) melanoma expressed NLK, whereas 55 (88.7%) of 62 thick (T3, T4; > 2 mm in diameter) melanoma were NLK positive. In addition, NLK expression was higher in thick melanoma than that in thin melanoma, but the difference was not statistically significant (P = 0.0539 by the Mann–Whitney U-test) (Table 1). In contrast, NLK expression in most of the metastatic melanoma was either low or absent (Fig. 1k–o). On comparing the intensity of NLK in metastatic melanoma, it was found that 95 (81.2%) cases out of 117 with primary malignant melanoma expressed NLK by immunohistochemistry, whereas 34 (58.6%) of 58 metastatic malignant melanoma cases were positive for NLK. When evaluated by scoring the expression levels, the expression of NLK was significantly higher (P = 0.0022 by the Mann–Whitney U-test) in primary malignant melanoma than in metastatic malignant melanoma (Fig. 1p). In addition, NLK expression was not correlated with age, sex, TNM stage or the presence of lymph node metastasis (Table 1). These results suggest that the expression of NLK is downregulated primarily in metastatic melanoma.
To further confirm that NLK is expressed in melanocytes, co-staining of NLK and HMB45 in normal skin slides was performed. Immunofluorescent co-localization staining showed that NLK was expressed in HMB45-positive cells in the basal layer (Fig. 1q), suggesting that NLK is expressed in normal melanocytes.
Expression of NLK inversely correlates with VEGFR2-positive microvascular formation
Previously, we showed that NLK interferes with VEGFR signalling . To determine the possible effects of reduced NLK expression in metastatic melanoma, we evaluated VEGFR2-related microvessel formation by counting the number of vessels stained by an anti-VEGFR2 antibody . As shown by pictures from eight representative patients, melanomas with a higher degree of NLK expression showed a reduced MVD (Fig. 2a, patients 1–4), whereas tumours with lower NLK scores showed a higher MVD (Fig. 2b, patients 5–8), suggesting more intratumoural vascular formation. Although the expression of VEGFR2 in tumour cells was very low in most tumour samples, some tissues with a higher density of microvessels showed moderate expression of VEGFR2 within tumour cells (Fig. 2b, lower panels). Statistical analysis of all patients with cutaneous melanoma indicated that NLK expression was correlated inversely with MVD (Fig. 2c; n = 84; P = 0.0052). This result suggests that reduced expression of NLK in tumour tissue may favour microvessel growth in melanoma.
In the present study, we showed that NLK expression was reduced in primary melanoma and completely lost in most cases of metastatic melanoma. Higher expression of NLK correlated with lower potential to metastasize and the extent of NLK expression was correlated inversely with MVD, as indicated by VEGFR2 staining, a marker of vascularization. We failed to find any association between NLK expression and tumour stage or tumour invasion. Despite lacking information on survival, the data suggest that NLK expression in melanoma may be one of the critical factors affecting vascularization and metastasis.
Vascularization is one of the most important steps affecting cancer metastasis. Angiogenesis within proliferating tumour tissue is required for a metastatic cell to survive hypoxia and establish a clone in a new niche after the malignant cell moves in. NLK negatively regulates VEGF/VEGFR2 expression in early lung development , which is similar to our present findings in tumour tissue, in that loss of NLK may enhance VEGF/VEGFR2 signalling. The correlation between NLK and MVD suggests that NLK plays a role in intratumoural vascularization. It is possible that, in metastatic melanomas, persistent Wnt signalling caused by insufficient NLK expression leads to both the proliferation of tumour cells and the recruitment of endothelial cells to sites of developing blood vessels. Interestingly, MVD has been found to be related to lymphovascular density and lymph node metastasis . Still, the mechanism by which reduced NLK expression promotes melanoma metastasis is so far unclear.
Melanoma metastasis is driven by genetic mutations in melanoma and affected by epigenetic changes. NLK expression is modulated by a number of factors. For example, NLK expression can be inhibited by miRNA-181a and miRNA-181b in hepatic cancer cell lines and natural killer cells during development [25,26]. The expression level of miRNA-181 has been shown to be elevated in uveal melanoma and is associated with metastasis . Thus, our findings that NLK is downregulated in metastatic melanoma may shed light on how metastatic cancer spreads.
The authors thank Professor Xien Gui for his support.
This work is supported by Zhongnan Hospital, Wuhan University, and funding from the National Science and Technology Department (81572902) and the West China Top Class Discipline Project in Basic Medical Sciences, Ningxia Medical University, to Hengning Ke.
Conflicts of interest
There are no conflicts of interest.
1. Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Atkins MB, Byrd DR, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009; 27:6199–6206.
2. Kumar D, Kumar S, Gorain M, Tomar D, Patil HS, Radharani NNV, et al. Notch1-MAPK signaling axis regulates CD133(+) cancer stem cell-mediated melanoma growth and angiogenesis. J Invest Dermatol. 2016; 136:2462–2474.
3. Hillen F, van de Winkel A, Creytens D, Vermeulen AH, Griffioen AW. Proliferating endothelial cells, but not microvessel density, are a prognostic parameter in human cutaneous melanoma. Melanoma Res. 2006; 16:453–457.
4. Pastushenko I, Van den Eynden GG, Vicente-Arregui S, Prieto-Torres L, Alvarez-Alegret R, Querol I, et al. Increased angiogenesis and lymphangiogenesis in metastatic sentinel lymph nodes is associated with nonsentinel lymph node involvement and distant metastasis in patients with melanoma. Am J Dermatopathol. 2016; 38:338–346.
5. Kashani-Sabet M, Sagebiel RW, Ferreira CM, Nosrati M, Miller JR III. Tumor vascularity in the prognostic assessment of primary cutaneous melanoma. J Clin Oncol. 2002; 20:1826–1831.
6. Aung PP, Leone D, Feller JK, Yang S, Hernandez M, Yaar R, et al. Microvessel density, lymphovascular density, and lymphovascular invasion in primary cutaneous melanoma-correlation with histopathologic prognosticators and BRAF status. Hum Pathol. 2015; 46:304–312.
7. Tugues S, Koch S, Gualandi L, Li X, Claesson-Welsh L. Vascular endothelial growth factors and receptors: anti-angiogenic therapy in the treatment of cancer. Mol Aspects Med. 2011; 32:88–111.
8. Gille J, Heidenreich R, Pinter A, Schmitz J, Boehme B, Hicklin DJ, et al. Simultaneous blockade of VEGFR-1 and VEGFR-2 activation is necessary to efficiently inhibit experimental melanoma growth and metastasis formation. Int J Cancer. 2007; 120:1899–1908.
9. Ishitani T, Ishitani S. Nemo-like kinase
, a multifaceted cell signaling regulator. Cell Signal. 2013; 25:190–197.
10. Ota S, Ishitani S, Shimizu N, Matsumoto K, Itoh M, Ishitani T. NLK positively regulates Wnt/beta-catenin signalling by phosphorylating LEF1 in neural progenitor cells. EMBO J. 2012; 31:1904–1915.
11. Ishitani T, Ninomiya-Tsuji J, Nagai S, Nishita M, Meneghini M, Barker N, et al. The TAK1–NLK–MAPK-related pathway antagonizes signalling between beta-catenin and transcription factor TCF. Nature. 1999; 399:798–802.
12. Chen HW, Qiao HY, Li HC, Li ZF, Zhang HJ, Pei L, et al. Prognostic significance of Nemo-like kinase
expression in patients with hepatocellular carcinoma. Tumour Biol. 2015; 36:8447–8453.
13. Zhang W, He J, Du Y, Gao XH, Liu Y, Liu QZ, et al. Upregulation of nemo-like kinase
is an independent prognostic factor in colorectal cancer. World J Gastroenterol. 2015; 21:8836–8847.
14. Emami KH, Brown LG, Pitts TE, Sun X, Vessella RL, Corey E. Nemo-like kinase
induces apoptosis and inhibits androgen receptor signaling in prostate cancer cells. Prostate. 2009; 69:1481–1492.
15. Cui G, Li Z, Shao B, Zhao L, Zhou Y, Lu T, et al. Clinical and biological significance of nemo-like kinase
expression in glioma. J Clin Neurosci. 2011; 18:271–275.
16. Huang Y, Jiang Y, Lu W, Zhang Y. Nemo-like kinase
associated with proliferation and apoptosis by c-Myb degradation in breast cancer. PLoS One. 2013; 8:e69148.
17. Shaw-Hallgren G, Chmielarska Masoumi K, Zarrizi R, Hellman U, Karlsson P, Helou K, et al. Association of nuclear-localized Nemo-like kinase
with heat-shock protein 27 inhibits apoptosis in human breast cancer cells. PLoS One. 2014; 9:e96506.
18. Zhang Y, Peng C, Wu G, Wang Y, Liu R, Yang S, et al. Expression of NLK and its potential effect in ovarian cancer chemotherapy. Int J Gynecol Cancer. 2011; 21:1380–1387.
19. Lv L, Wan C, Chen B, Li M, Liu Y, Ni T, et al. Nemo-like kinase
(NLK) inhibits the progression of NSCLC via negatively modulating WNT signaling pathway. J Cell Biochem. 2014; 115:81–92.
20. Ke H, Masoumi KC, Ahlqvist K, Seckl MJ, Rydell-Tormanen K, Massoumi R. Nemo-like kinase
regulates the expression of vascular endothelial growth factor (VEGF) in alveolar epithelial cells. Sci Rep. 2016; 6:23987.
21. Masoumi KC, Daams R, Sime W, Siino V, Ke H, Levander F, et al. NLK-mediated phosphorylation of HDAC1 negatively regulates Wnt signaling. Mol Biol Cell. 2017; 28:346–355.
22. Weichert F, Groh A, Shamaa A, Richards T, Awd S, Linder R, et al. Analysis of histological datasets by signal processing methods. Pathologe. 2008; 29(Suppl 2)129–134.
23. Weidner N. Intratumor microvessel density as a prognostic factor in cancer. Am J Pathol. 1995; 147:9–19.
24. Shayan R, Karnezis T, Murali R, Wilmott JS, Ashton MW, Taylor GI, et al. Lymphatic vessel density in primary melanomas predicts sentinel lymph node status and risk of metastasis. Histopathology. 2012; 61:702–710.
25. Cichocki F, Felices M, McCullar V, Presnell SR, Al-Attar A, Lutz CT, et al. Cutting edge: microRNA-181 promotes human NK cell development by regulating Notch signaling. J Immunol. 2011; 187:6171–6175.
26. Ji J, Yamashita T, Budhu A, Forgues M, Jia HL, Li C, et al. Identification of microRNA-181 by genome-wide screening as a critical player in EpCAM-positive hepatic cancer stem cells. Hepatology. 2009; 50:472–480.
27. Zhang L, He X, Li F, Pan H, Huang X, Wen X, et al. The miR-181 family promotes cell cycle by targeting CTDSPL, a phosphatase-like tumor suppressor in uveal melanoma. J Exp Clin Cancer Res. 2018; 37:15.