Cancer Malignancy Is Correlated with Upregulation of PCYT2-Mediated Glycerol Phosphate Modification of α-Dystroglycan

Abstract

The dystrophin-glycoprotein complex connects the cytoskeleton with base membrane components such as laminin through unique O-glycans displayed on α-dystroglycan (α-DG). Genetic impairment of elongation of these glycans causes congenital muscular dystrophies. We previously identified that glycerol phosphate (GroP) can cap the core part of the α-DG O-glycans and terminate their further elongation. This study examined the possible roles of the GroP modification in cancer malignancy, focusing on colorectal cancer. We found that the GroP modification critically depends on PCYT2, which serves as cytidine 5'-diphosphate-glycerol (CDP-Gro) synthase. Furthermore, we identified a significant positive correlation between cancer progression and GroP modification, which also correlated positively with PCYT2 expression. Moreover, we demonstrate that GroP modification promotes the migration of cancer cells. Based on these findings, we propose that the GroP modification by PCYT2 disrupts the glycan-mediated cell adhesion to the extracellular matrix and thereby enhances cancer metastasis. Thus, the present study suggests the possibility of novel approaches for cancer treatment by targeting the PCYT2-mediated GroP modification.

Keywords: CDP-glycerol; PCYT2; cancer malignancy; glycerol phosphate modification; matriglycan; α-dystroglycan.

Figure 1

Immunohistochemical analysis for glycerol phosphate (GroP) modification in human colorectal cancer tissues. (A) Representative images of staining with DG2. The left and right panels represent negative and positive DG2 staining, respectively, in human colorectal cancer tissues (×100). (B) GroP expression according to disease stage. The p-value in each stage was calculated by comparison with the positive rate in stage IV.

Figure 2

Effects of enhancement of the GroP modification on cancer cell behaviors. (A) TagD overexpression reduced the expression of matriglycans. Cell lysates containing equal amounts of total proteins prepared from HCT116 cells transfected with TagD-Flag-expressing vector or control vector were subjected to immunoblot analysis using IIH6 and AF6868 along with anti-β-actin and anti-Flag antibodies. (B) Proliferation assay of HCT116 cells transfected with TagD-expressing vector or control vector at 0, 24, 48, and 72 h. Error bars represent the standard error of the mean (SEM) (n = 3). (C) Transwell assay (scale bar = 600 μm) and (D) wound-healing assay (scale bar = 200 μm) for HCT116 cells at 24 h after transfection with TagD-expressing or control vector. (E) Wound-healing index of TagD-overexpressing cells and control cells. Individual data points are represented by black dots on the bar graph. Error bars represent the SEM (n = 3). Significant differences (*) were calculated compared with WT index using two-tailed unpaired Student’s t-test (p  <  0.05).

Figure 3

Characterization of enzymatic activities of PCYT2. (A) HPLC profiles of the reaction mixtures in the presence and absence of the candidate enzymes listed in Table 1 and positive (TagD) and negative controls. The peak corresponding to CDP-Gro is indicated with the arrow. (B) The enzymatic activities of PCYT2α and PCYT2β with various glycerol-3-phosphate or ethanolamine phosphate and CTP concentrations are plotted. CDP-Gro/CDP-Etn synthase activities of PCYT2α (▼/●) and PCYT2β (Ø◆/▲) are represented, respectively. The data were fitted to the Michaelis–Menten equation by nonlinear regression (GraphPad Prism 9; https://www.graphpad.com/scientific-software/prism/). The apparent Michaelis constant (K_m) and the maximal velocity (V_max) were calculated.

Figure 4

LC-ESI-MS/MS analysis of CDP-Gro and CDP-Etn from HCT116 cells. (A) Representative extraction ion chromatograms of CDP-Gro (upper panel) and CDP-Etn (lower panel) derived from HCT116. (B) HCT116 wild-type cells, TagD-overexpressing cells (TagD-OE), PCYT2-KO (ΔPCYT2), and PCYT2-KO rescue cells (ΔPCYT2-PCYT2-OE) were analyzed, and the data were normalized to units of pmol per milligram of protein. Individual data points are represented by black dots on the bar graph. Error bars represent the standard error of the mean (SEM) (n = 3).

Figure 5

Glycan modification on α-DG373(T322R)-Fc expressed in HCT116 wild-type and PCYT2-KO cells. The extracted ion chromatograms of selected ³¹³pyrQIHATPTPVR³²² on α-DG carrying phosphorylated core M3 glycoforms, HexNAc₄Hex₂P₂GroP₁ (left) and HexNAc₄Hex₂P₂ (right), in biological triplicate of WT (upper 3 panels) and ΔPCYT2 (lower 3 panels) derived from HCT116. The HexNAc₄Hex₂P₂GroP₁ glycoforms were only detected in WT, but not ΔPCYT2, in contrast with the detections of phosphorylated core M3 structures, HexNAc₄Hex₂P₂GroP₁.

Figure 6

The expression of matriglycans was reduced by PCYT2-overexpression. HCT116 clone stably transfected with the doxycycline (Dox)-inducible PCYT2 construct; −Dox, uninduced cells; + Dox, induced cells. Cell lysates containing equal amount of total proteins prepared from HCT116 cells with (+) or without (−) Dox were subjected to immunoblot analysis using anti-Flag antibody and anti-β-actin, or to WGA enrichment followed by immunoblot analysis using IIH6.

Figure 7

Correlation between GroP modification and PCYT2 expression levels in colorectal cancer tissues. (A) Representative images of PCYT2 staining. The left and right panels represent negative PCYT2 and positive PCYT2 staining, respectively, in human colorectal cancer tissues (×100). (B) Mapping of correlation between GroP modification and PCYT2 expression levels based on the staining scores. Staining scores of both GroP and PCYT2 were calculated at stage IV colorectal cancer tissues, and data were analyzed using the Spearman rank correlation. Dot size corresponds to the number of colorectal cancer tissue samples.