PLOD2 Is a Prognostic Marker in Glioblastoma That Modulates the Immune Microenvironment and Tumor Progression

Abstract

This study aimed to investigate the role of Procollagen-Lysine, 2-Oxoglutarate 5-Dioxygenase 2 (PLOD2) in glioblastoma (GBM) pathophysiology. To this end, PLOD2 protein expression was assessed by immunohistochemistry in two independent cohorts of patients with primary GBM (n1 = 204 and n2 = 203, respectively). Association with the outcome was tested by Kaplan−Meier, log-rank and multivariate Cox regression analysis in patients with confirmed IDH wild-type status. The biological effects and downstream mechanisms of PLOD2 were assessed in stable PLOD2 knock-down GBM cell lines. High levels of PLOD2 significantly associated with (p1 = 0.020; p2< 0.001; log-rank) and predicted (cohort 1: HR = 1.401, CI [95%] = 1.009−1.946, p1 = 0.044; cohort 2: HR = 1.493; CI [95%] = 1.042−2.140, p2 = 0.029; Cox regression) the poor overall survival of GBM patients. PLOD2 knock-down inhibited tumor proliferation, invasion and anchorage-independent growth. MT1-MMP, CD44, CD99, Catenin D1 and MMP2 were downstream of PLOD2 in GBM cells. GBM cells produced soluble factors via PLOD2, which subsequently induced neutrophils to acquire a pro-tumor phenotype characterized by prolonged survival and the release of MMP9. Importantly, GBM patients with synchronous high levels of PLOD2 and neutrophil infiltration had significantly worse overall survival (p < 0.001; log-rank) compared to the other groups of GBM patients. These findings indicate that PLOD2 promotes GBM progression and might be a useful therapeutic target in this type of cancer.

Keywords: PLOD2; glioblastoma; neutrophils; prognostic biomarkers; tumor progression.

Figure 1

PLOD2 in GBM patients: univariate analysis of survival. PLOD2 expression was dichotomized into PLOD2^high and PLOD2^low according to the median-split method. Kaplan–Meier curves were plotted for the (A) 5-year overall survival (OS) in the Hannover cohort, (B) 3-year OS in the Magdeburg cohort and (C,D) 1-year progression-free survival (PFS) in both cohorts. The log-rank test was used for statistical analysis and the p-values are indicated in the upper-right corner of each plot.

Figure 2

PLOD2 in GBM patients: multivariate analysis of survival. Multivariate Cox regression analysis model for the (A) overall survival (OS) and (B) progression-free survival (PFS) of patients with high versus low levels of PLOD2. HR: hazard ratio; CI [95%]: 95% confidence interval.

Figure 3

PLOD2 and GBM invasion. (A) Representative micrographs of invasion assays in sh-control versus sh-PLOD2 H4 cells. The upper panels show the pre-invasion status at 0 h, the lower panels show the post-invasion status at 72 h. The red line marks the closure of the “gap”, indicating the degree of tumor invasion. (B) PLOD2 knock-down significantly reduces the invasiveness of H4 cells. The data are presented as percentage to sh-control. (C) Representative image of a gelatin zymography gel showing that both sh-PLOD2 and sh-control released MMP2, but only negligible levels of MMP9. (D) PLOD2 knock-down significantly reduces the release of MMP2 in H4 cells. For quantification, the integrated density of the bands was determined using the ImageJ software. The data are presented as percentage to medium only. Shown are the means + S.D. of 3 independent experiments. Statistical analysis was performed with the paired t-test.

Figure 4

PLOD2 and GBM proliferation. PLOD2 knock-down reduced the metabolic activity of H4 GBM cells, as indicated by the MTT assay using (A) 2000 cells and (B) 4000 cells. (C) Representative micrographs of colonies generated by sh-control and sh-PLOD2 cells after 10 days of culture in low-gelling agarose. (D) PLOD2 knock-down significantly inhibited the anchorage-independent growth of H4 cells. Shown are the means + S.D. of at least three independent experiments per assay. In all studies, statistical analysis was performed with the paired t-test.

Figure 5

Molecular mechanisms of PLOD2 in GBM. (A) Representative western blots of Catenin D1, CD44, CD99, MT1-MMP and PLOD2 levels in sh-PLOD2 versus sh-control GBM cells. Beta-actin was used as a loading control. PLOD2 knock-down significantly decreased the levels of (B) Catenin D1, (C) CD44, (D) CD99 and (E) MT1-MMP in GBM cells. Shown are the means + S.D. of at least three independent experiments. Statistical analysis was performed with the paired t-test.

Figure 6

PLOD2 and neutrophils in GBM. (A) Neutrophils were incubated with supernatants (SN) derived from sh-control or sh-PLOD2 GBM cells. Culture medium was used as control. The survival of neutrophils was assessed at 24 h and the release of MMP9 at 1 h post-stimulation. (B) Sh-control SN prolonged the survival of neutrophils in both cell lines. This effect was significantly lower upon stimulation with PLOD2 knock-down SN. (C) Representative image of a gelatin zymography gel showing that neutrophils release MMP9 upon stimulation with GBM SN. (D) Neutrophils stimulated with sh-PLOD2 SN released significantly lower levels of MMP9 compared with their sh-control stimulated counterparts. For quantification, the integrated density of the bands was determined using the ImageJ software. The data are presented as percentage to medium only. Shown are the means + S.D. of three independent experiments. Statistical analysis was performed with the paired t-test. (E) GBM patients were divided into four groups according to the combined expression of CD66b and PLOD2. The Kaplan–Meier curves were plotted for the 3-year overall survival and the statistical significance was determined with the log-rank test. The p-value is indicated in the upper right corner of the plot. (F) Multivariate Cox regression analysis model for the overall survival of the four groups of patients. HR: hazard ratio; CI [95%]: 95% confidence interval.

Figure 7

Schematic representation of the main findings integrated in the biology of GBM.