KANK1 inhibits cell growth by inducing apoptosis through regulating CXXC5 in human malignant peripheral nerve sheath tumors

Abstract

Malignant peripheral nerve sheath tumors (MPNSTs) are a type of rare sarcomas with a poor prognosis due to its highly invasive nature and limited treatment options. Currently there is no targeted-cancer therapy for this type of malignancy. Thus, it is important to identify more cancer driver genes that may serve as targets of cancer therapy. Through comparative oncogenomics, we have found that KANK1 was a candidate tumor suppressor gene (TSG) for human MPNSTs. Although KANK1 is known as a cytoskeleton regulator, its tumorigenic function in MPNSTs remains largely unknown. In this study, we report that restoration of KANK1 in human MPNST cells inhibits cell growth both in human cell culture and xenograft mice by increasing apoptosis. Consistently, knockdown of KANK1 in neurofibroma cells promoted cell growth. Using RNA-seq analysis, we identified CXXC5 and other apoptosis-related genes, and demonstrated that CXXC5 is regulated by KANK1. Knockdown of CXXC5 was found to diminish KANK1-induced apoptosis in MPNST cells. Thus, KANK1 inhibits MPNST cell growth though CXXC5 mediated apoptosis. Our results suggest that KANK1 may function as a tumor suppressor in human MPNSTs, and thus it may be useful for targeted therapy.

Figure 1. Restoration of KANK1 in human MPNST cells suppresses cell growth in cell culture.

KANK1 gene expression was restored in lentiviral KANK1 stable STS26T (a–c) and S462 (d–f) cells using doxycycline. (a,d) Restoration of KANK1 measured by Western blots in both lines (STS26T & S462). The lentiviral constructs pLIX405-KANK1 and pSLIK-neo-KANK1 are labeled above the Western blot gels. NT, non-transfected. ACTB, beta-actin control. (b,e) MTT assays revealed that cell growth was inhibited by KANK1 restoration in both cells. Growth curves were created using the optical density values collected over five days. (c,f) Plate colony formation assay on the STS26T and S462 cells with or without KANK1 restoration. Crystal violet stained cell colonies (diameter ≥0.2 mm) were counted. (g,h) HEI-193 cells with KANK1 knocked-down using shRNAs were confirmed by QRT-PCR (g) and Western blot (h). NS, non-silencing; sh1 and sh2, KANK1-shRNAs. (i). MTT assays showed that KANK1 knockdown is able to promote cell growth. Asterisk (*) shows statistical significance, p < 0.05.

Figure 2. KANK1 inhibits MPNST growth in xenografted mice.

(a) Xenografted human MPNST STS26T cells in NOD/SCID mice with (+Dox) and without (-Dox) doxycycline feeding (200  μg/ml). (b) Changes in tumor volume over time. p < 0.001 (two-way ANOVA). (c,d) MPNST histology by hematoxylin and eosin (HE) staining. (e,f) KANK1 protein in STS26T cells measured by immunohistochemistry. (g,h) CXXC5 protein in STS26T cells measured by immunohistochemistry. (i,j) Cleaved caspase 3 measured by immunohistochemistry. Green arrow heads indicate the positively stained cells. (k,l) PCNA, a proliferation maker, measured by immunohistochemistry. (e–l) Slides were counter stained with hematoxylin. A scale bar is on the right-bottom corner for each panel (c–l).

Figure 3. KANK1 inhibits cell growth through apoptosis.

(a) Representative of three experiment replicates with STS26T. KANK1 expression cells were prepared with (+Dox) and without (−Dox) doxycycline at 1 μg/ml in culture medium. Apoptosis was analyzed with FlowJo after Fluor 647 conjugated Annexin V and PI double-staining flow cytometry. Quadrant I-IV represents dead cell population, early apoptotic cell population, late apoptotic cell population, and live cell population, respectively. (b) Statistical analysis showed significant difference between KANK1 expressing and control groups in STS26T. Asterisk (*) shows p = 0.0161. (c) KANK1 restoration induced apoptosis measured by cleaved caspase 3 expression. Both control and doxycycline-treated cells were cultured at medium with 0.5% FBS. ACTB, beta-actin control. (d) Representative of three experimental replicates with S462 cell line. Flow cytometry was formed using Fluor 488 conjugated Annexin V and PI double staining. (e) Statistical analysis showed significant difference between KANK1 expressing and control groups. Asterisk (*) shows statistical significance, p = 0.0087. (f) KANK1 restoration induced apoptosis measured by cleaved caspase 3 expression.

Figure 4. Identification of CXXC5 as a downstream gene of KANK1.

(a) Differential gene expression by RNA-seq before and after KANK1 restoration. Each dot represents the mean expression level plotted against the fold change for a given transcript. Black points are transcripts that are not statistically significant, red points are significant at FDR 10% (adjusted p-value < 0.1), and fold change is greater than 2. KANK1 and the five apoptosis-related genes are labeled with blue circles. The RP11-130C19.3 is a putative transcripts within KANK1 gene locus according to the current Ensembl assembly. (b) Quantitative RT-PCR validations for KANK1 and the five apoptosis related genes using the same RNA samples for RNA-seq. Asterisk (*) indicates statistical significance: p = 0.0004 (CXXC5), 0.0039 (BAD), and 0.0039 (SIVA1). (c,d) CXXC5 protein levels increase when KANK1 is restored in both STS26T and S462 stable cells using doxycycline. ACTB, beta-actin control. (e,f). CXXC5 is down-regulated in KANK1 knock-down HEI-193, by both QRT-PCR (e) and Western blot (f). NS, stable cell line made using non-silencing shRNA. Asterisk (*) shows statistical significance: p = 0.0065 (NS vs. sh1) and p = 0.0035 (NS vs. sh2).

Figure 5. CXXC5 knockdown diminishes KANK1 induced apoptosis.

(a) Apoptosis measured by flow cytometry using Annexin V and PI in CXXC5 shRNA stable STS26T cells, and non-silencing control shRNA cells. Quadrant I-IV represents dead cell population, early apoptotic cell population, late apoptotic cell population, and live cell population, respectively. Later apoptotic cell percentage was reduced from 67.5% to ~31% in both CXXC5 shRNAs infected cells. (b) CXXC5 knockdown rescued KANK1-induced apoptosis, which is evident by reduced expression of cleaved caspase 3. NS, non-silencing shRNA. sh1 and sh2, CXXC5 shRNAs. Dox, doxycycline (1 μg/ml).