Histone Lactylation-Driven Upregulation of VRK1 Expression Promotes Stemness and Proliferation of Glioma Stem Cells

Abstract

Glioblastoma (GBM) is the most malignant primary brain tumor in adults, with glioma stem cells (GSCs) as a subpopulation contributing to treatment resistance and recurrence. This study investigates the role of lactate and its regulatory gene, vaccinia-related kinase 1 (VRK1), in the regulation of GSC stemness. Utilizing multiple glioma-related databases and patient-derived GSCs, it is discovered that lactate enhances the stemness and proliferation of GSCs via VRK1. Mechanistically, lactate promotes histone lactylation (H3K18la) at the VRK1 promoter in GSCs, thereby upregulating VRK1 expression. VRK1 enhances Y-box binding protein 1 (YBX1) protein stability by inhibiting its ubiquitination and degradation, and phosphorylates YBX1 to promote its nuclear translocation, thereby regulating GSC stemness and proliferation via the YBX1/SOX2 pathway. Additionally, the VRK1-targeted nanoliposome A/TMZ-siVRK1 effectively suppresses the stemness and proliferation of GSCs, demonstrating its therapeutic potential. In conclusion, lactate regulates the stemness and proliferation of GSCs via the H3K18la/VRK1/YBX1/SOX2 pathway. This study elucidates the role of histone lactylation in stem cell regulation and suggests that VRK1 is a potential therapeutic target for GBM.

Keywords: SOX2; VRK1; glioma stem cell; histone lactylation; nanoliposome.

Figure 1

VRK1 regulates the stemness and proliferation of GSCs in vitro. A) Screening flowchart (authorized by BioRender) for identifying key genes involved in lactate‐mediated regulation of GSC stemness and proliferation in GBM. B) Stable transfected cell lines with VRK1 knockdown were constructed using lentiviral shVRK1‐1 and shVRK1‐2 sequences in GSC4 and GSC63 cells, and the knockdown efficiency was verified by western blot. C) Neurosphere proliferation assay showed that VRK1 knockdown reduced the size of neurospheres formed by GSCs (n = 3, Student's t‐test). D,E) ELDA showed that VRK1 knockdown inhibited the self‐renewal capability of GSCs (n = 3). F) Neurosphere formation assay showed that VRK1 knockdown decreased the number of neurospheres formed by GSCs (n = 3, Student's t‐test). G) Western blot verified the overexpression efficiency of VRK1 in GSC21 and GSC40 cells. H) Neurosphere proliferation assay showed that VRK1 overexpression increased neurosphere size (n = 3, Student's t‐test). I,J) ELDA showed that VRK1 overexpression enhanced the self‐renewal capacity of GSCs. K) Neurosphere formation assay showed that VRK1 overexpression increased the number of neurospheres formed by GSCs (n = 3, Student's t‐test). Data were presented as mean ± SD, ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 2

VRK1 regulates the stemness and proliferation of GSCs in vivo. A,B) An orthotopic xenograft tumor model was established in nude mice using GSC63 cells with VRK1 knockdown. Fluorescence imaging (A) and H&E staining (B) showed that VRK1 knockdown reduced intracranial tumor volume (n = 5). C) Kaplan‐Meier survival analysis indicated that VRK1 knockdown prolonged survival (n = 5, Log‐Rank test). D) IHC staining of VRK1 and Ki67 in tumor tissues. E) In vivo limiting dilution assays revealed decreased tumorigenic potential of GSC63 cells following VRK1 knockdown (n = 5). ^* p < 0.05, ^** p < 0.01.

Figure 3

Lactate regulates the stemness and proliferation of GSCs via VRK1. A) Correlation analysis of VRK1 with stemness markers in the CGGA dataset: NES (r = 0.1200), OLIG2 (r = 0.2529), POU3F2 (r = 0.3106), PROM1 (r = 0.1061), SOX2 (r = 0.4248), and SOX9 (r = 0.3319) (n = 137, Pearson correlation). B) RT‐qPCR analysis revealed that VRK1 knockdown significantly decreased the mRNA level of SOX2 (n = 3, Student's t‐test). C) Western blot showed that VRK1 knockdown reduced the protein level of SOX2. D) Western blot revealed that lactate failed to upregulate SOX2 expression in GSCs following VRK1 knockdown. E,F) ELDA showed that lactate failed to enhance the self‐renewal capacity in GSCs following VRK1 knockdown (n = 3). G) Neurosphere proliferation assay indicated that lactate failed to increase neurosphere size in GSCs following VRK1 knockdown (n = 3, One‐way ANOVA test). H) Neurosphere formation assay indicated that lactate failed to promote neurosphere formation in GSCs following VRK1 knockdown (n = 3, One‐way ANOVA test). Data were presented as mean ± SD, ns p > 0.05, ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 4

Lactate regulates VRK1 expression by inducing H3K18la in GSCs. A) Western blot revealed increased levels of overall Kla, VRK1, and SOX2 in GSCs treated with lactate (sodium lactate, 10 mm, 24 h). B) IP indicated that H3 histone lactylation was the most significant enhancement compared with other histones following lactate treatment in GSC4 cells. C) Western blot showed the expression levels of H3K9la, H3K14la, H3K18la, and H3K56la in GSCs treated with lactate or PBS. D) CUT&Tag indicated that the H3K18la antibody bound to the VRK1 promoter in GSC63 cells. E,F) ChIP confirmed that H3K18la antibody bound to the VRK1 promoter in GSC4 and GSC63 cells (n = 3, Student's t‐test). G) Western blot showed that C646 (100 µm, 24 h) significantly reduced H3K18la levels and downregulated VRK1 and SOX2 expression, whereas TSA (1 µm, 24 h) increased H3K18la levels and upregulated VRK1 and SOX2 expression. H) MCT1/2 inhibitor AR‐C155858 inhibited the upregulation of H3K18la, VRK1, and SOX2 induced by exogenous lactate in GSCs (AR‐C155858, 1 mm, 24 h). I) LDHA inhibitor (R)‐GNE‐140 reduced H3K18la levels and downregulated VRK1 and SOX2 in GSCs ((R)‐GNE‐140, 1 µm, 24 h). Data were presented as mean ± SD, ^** p < 0.01, ^**** p < 0.0001.

Figure 5

VRK1 directly binds to YBX1 and increases its protein stability. A) A Venn diagram showed the overlapping genes among the VRK1‐specific binding molecules, the glioma intrinsic expression gene set, and the transcription factor gene set. B) Correlation analysis between SOX2 and 12 transcription factors in GSCs using sequencing data from the internal glioma database. C) Co‐IP revealed direct binding between VRK1 and YBX1 in GSC4 and GSC63 cells. D,E) RT‐qPCR analysis revealed that the mRNA level of YBX1 was not significantly altered following VRK1 knockdown or overexpression in GSCs (n = 3, Student's t‐test). F) Western blot revealed that VRK1 knockdown decreased YBX1 protein expression, and VRK1 overexpression increased YBX1 protein expression. G) The protein degradation of YBX1 significantly slowed down over time after adding CHX in GSCs with VRK1 overexpression compared to the controls. H) Western blot showed that MG132 (10 µm, 6 h) reduced YBX1 degradation in GSCs with VRK1 knockdown, while 3‐MA (10 µm, 6 h) had no effect. I) Western blot demonstrated that YBX1 expression was significantly upregulated in GSCs with normal VRK1 expression following MG132 treatment, which was consistent with YBX1 expression in VRK1‐overexpressing GSCs treated with DMSO. J) IP demonstrated that the ubiquitination level of YBX1 significantly increased after VRK1 knockdown and decreased after VRK1 overexpression in GSCs treated with MG132. K) GST pull‐down assay validated the direct binding between VRK1 and YBX1. L) GST pull‐down assay revealed that YBX1 bound to wild‐type VRK1 but not to the K71‐mutant VRK1. Data were presented as mean ± SD, ns p > 0.05.

Figure 6

VRK1 phosphorylates YBX1 to promote its nuclear translocation. A) In vitro kinase assay showed that VRK1 protein phosphorylated YBX1 in a dose‐dependent manner. B) In vitro kinase assay showed that the wild‐type VRK1 protein, but not the K71‐mutant VRK1 protein, phosphorylated YBX1 in a dose‐dependent manner. C) Western blot showed that VRK1 knockdown decreased the protein levels of YBX1, phosphorylated YBX1 (Ser102), and SOX2, whereas VRK1 overexpression increased their protein levels. D) Immunofluorescence revealed that VRK1 knockdown reduced the nuclear expression of YBX1 in GSC4 and GSC63 cells. E) Western blot showed that VRK1 knockdown or overexpression correspondingly decreased or increased YBX1, phosphorylated YBX1 (Ser102), and SOX2 protein levels in both the cytoplasm and nucleus of GSCs. F) YBX1 and phosphorylated YBX1 (Ser102) were increased in both the cytoplasm and nucleus of GSCs overexpressing wild‐type VRK1, whereas the K71‐mutant VRK1 showed no such effect. G) Nuclear‐cytoplasmic fractionation assays revealed that GSCs overexpressing wild‐type YBX1 increased YBX1 and phosphorylated YBX1 levels in the nucleus, whereas overexpressing YBX1 S102A did not increase YBX1 and phosphorylated YBX1 levels in the nucleus. H) GSCs overexpressing wild‐type YBX1 increased SOX2 expression, while GSCs overexpressing YBX1 S102A did not increase SOX2 expression.

Figure 7

VRK1 regulates the stemness and proliferation of GSCs via YBX1. A) Ch‐IP revealed that YBX1 bound to the SOX2 promoter region (n = 3, Student's t‐test). B) Western blot demonstrated that the upregulation of SOX2 expression caused by VRK1 overexpression in GSCs was inhibited by YBX1 knockdown. C) Neurosphere proliferation assay indicated that the promotion of neurosphere proliferation caused by VRK1 overexpression in GSCs was inhibited by YBX1 knockdown (n = 3, One‐way ANOVA test). D,E) ELDA showed that the promotion of self‐renewal caused by VRK1 overexpression in GSCs was inhibited by YBX1 knockdown (n = 3). F) Neurosphere formation assay demonstrated that the promotion of neurosphere formation induced by VRK1 overexpression in GSCs was inhibited by YBX1 knockdown (n = 3, One‐way ANOVA test). Data were presented as mean ± SD, ns p > 0.05, ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 8

VRK1, YBX1, and SOX2 are highly expressed and exhibit a positive correlation in GBM. A) IHC staining showed that VRK1 and YBX1 were localized in both the nucleus and cytoplasm, while SOX2 was predominantly localized in the nucleus (Scale bars: 200, 50 µm). VRK1, YBX1, and SOX2 were weakly expressed in non‐tumor tissues but were significantly overexpressed in GBM, with expression levels increasing with glioma grade. B) IRS values of VRK1, YBX1, and SOX2 (Non‐tumor tissue: n = 3, WHO II glioma: n = 14, WHO III glioma: n = 12, GBM: n = 15; one‐way ANOVA test). C) Correlation analysis indicated that the expression levels of VRK1, YBX1, and SOX2 were positively correlated in glioma tissues (VRK1 versus YBX1, r = 0.8289, p < 0.0001; VRK1 versus SOX2, r = 0.8616, p < 0.0001; YBX1 versus SOX2, r = 0.8807, p < 0.0001; Pearson correlation). D) The expression levels of VRK1, YBX1, and SOX2 were positively correlated in GBM tissues (VRK1 versus YBX1, r = 0.5174, p = 0.0482; VRK1 versus SOX2, r = 0.6939, p = 0.0041; YBX1 versus SOX2, r = 0.5510, p = 0.0332; Pearson correlation). Data were presented as mean ± SD, ns p > 0.05, ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 9

Synthesis and characterization of A/TMZ‐siVRK1. A) Size distribution of TMZ‐siVRK1, A/TMZ‐siNC, A/PLGA‐siVRK1, and A/TMZ‐siVRK1. B) TEM images of TMZ‐siVRK1, A/TMZ‐siNC, A/PLGA‐siVRK1, and A/TMZ‐siVRK1. C) Zeta potential of TMZ‐siVRK1, A/TMZ‐siNC, A/PLGA‐siVRK1, and A/TMZ‐siVRK1. D) Agarose gel electrophoresis experiments demonstrated that when the N/P ratio was 10:1, siVRK1 was completely adsorbed by DOTAP. E) Serum stability of TMZ‐siVRK1, A/TMZ‐siNC, A/PLGA‐siVRK1, and A/TMZ‐siVRK1 (n = 3). F,G) In vitro TMZ (F) or siVRK1 (G) drug release profiles under pH 6.5 or 7.4 PBS conditions (n = 3). H) The CLSM images were performed to test the ability of TMZ‐siVRK1‐FITC and A/TMZ‐siVRK1‐FITC to penetrate the BBB and target GSC63 (n = 3, Student's t‐test). Red: Dil‐labeled BBB layer cell membrane (HBMEC). Blue: DAPI‐labeled nuclei; green: TMZ‐siVRK1‐FITC or A/TMZ‐siVRK1‐FITC (FITC signal). Scale bar = 25 µm. I) CLSM images showing endosomal escape of A/TMZ‐siVRK1 in GSC63. Cells were incubated with A/TMZ‐siVRK1 at 37 °C for 1, 4, and 8 h. FITC‐labeled A/TMZ‐siVRK1 (green signal) was used. Nuclei were stained by DAPI (blue signal), and endosomes were stained with LysoTracker (red signal). Scale bars = 10 µm. J–L) The corresponding colocalization fluorescence intensity profiles of FITC (green) and endosomes (red) in GSC63 after FITC‐labeled A/TMZ‐siVRK1 treatment for 1 h (J), 4 h (K), and 8 h (L). M) The corresponding colocalization ratios of FITC (green) and endosomes (red) (n = 3, one‐way ANOVA test). N) Fluorescent image of TMZ‐siVRK1‐Cy5 and A/TMZ‐siVRK1‐Cy5 brain distribution in vivo was observed with a small animal imaging system at post‐injection 6, 12, and 24 h (left). Statistical analysis of fluorescence content of A/TMZ‐siVRK1‐Cy5 in mice brains with imaging in vivo (right) (n = 3, Student's t test). Data were presented as mean ± SD, ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 10

A/TMZ‐siVRK1 effectively inhibits glioma growth in vitro and in vivo. A) Neurosphere formation assay showed that the number of neurospheres formed by GSC63 cells treated with A/TMZ‐siVRK1 was decreased compared with PBS, TMZ‐siVRK1, A/TMZ‐siNC, and A/PLGA‐siVRK1 groups (n = 3, one‐way ANOVA test). B) Flow cytometry showed that the apoptosis rate of GSC63 cells treated with A/TMZ‐siVRK1 was increased compared with PBS, TMZ‐siVRK1, A/TMZ‐siNC, and A/PLGA‐siVRK1 groups (n = 3, one‐way ANOVA test). C) Experimental schedule of patient‐derived GSC63 inoculation and drug administration of PBS, TMZ‐siVRK1, A/TMZ‐siNC, A/PLGA‐siVRK1, and A/TMZ‐siVRK1. D) Kaplan–Meier survival curves of mice implanted with GSC63 cells (n = 5, Log‐rank test). E,F) In vivo fluorescence imaging (E) and H&E‐staining (F) of orthotopic xenograft tumor model. G) Immunohistochemical staining of VRK1, YBX1, SOX2, and Ki67 in glioma tissues of nude mice (Scale bar: 50 µm). Data were presented as mean ± SD, ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 11

Schematic illustration (authorized by BioRender) of lactate regulates the stemness and proliferation of glioma stem cells via H3K18la/VRK1/YBX1/SOX2.