Analysis of specialized DNA polymerases expression in human gliomas: association with prognostic significance

Abstract

Aberrant activation of the translesion DNA synthesis (TLS) pathway has been suggested to play a role in tumorigenesis by promoting genetic mutations. We therefore examined glioma specimens for the expression of specialized DNA polymerases involved in TLS and assessed their prognostic significance. The expression levels of DNA polymerase κ (Pol κ), Pol ι, and Pol η were assessed in 40 primary glioma samples and 10 normal brain samples using quantitative real-time PCR and Western blot analysis. Their prognostic significance was evaluated using a population-based tissue microarray derived from a cohort of 104 glioma patients. Overexpression of Pol κ and Pol ι was observed in 57.5% (23-40) and 27.5% (11-40) of patients, respectively, whereas no significant expression of Pol η was seen in the specimens. Immunohistochemical studies revealed positive Pol κ and Pol ι staining in 72 (69.2%) and 33 (31.7%) of the 104 glioma specimens, respectively. Pol κ expression was associated with advanced stages of the disease. Both Pol κ- and Pol ι-positive staining were associated with shorter survival in glioma patients (P < .001 and P = .014, respectively). A multivariate survival analysis identified Pol κ as an independent prognostic factor for glioma patients (P < .001). These findings demonstrate, for the first time, that the expression of Pol κ and Pol ι is deregulated in gliomas, and upregulation of Pol κ is associated with poorer prognosis in glioma patients.

Fig. 1.

Expression analysis of specialized DNA Pol κ, Pol ι, and Pol η mRNA and protein in primary glioma tissues (G) and normal brain tissues (N). (A) Quantitative real-time PCR analysis of Pol κ, Pol ι, and Pol η mRNA expression in glioma and normal brain samples. GADPH was used as loading control. (B) Representative images of Western blot analysis of Pol κ, Pol ι, and Pol η protein expression in glioma and normal brain samples. Expression of β-actin was served as a quantity control. (C) Average sample ratios of Pol κ (left) and Pol ι (right) protein expression quantified using ImageJ software. Expression levels were normalized with β-actin.

Fig. 2.

Immunohistochemical analysis of Pol κ, Pol ι, and Pol η expression in primary glioma tissues (G) and normal brain tissues (N). Thick sections of paraffin-embedded TMA derived from a total of 104 primary glioma specimens including WHO grades I–IV were stained using an anti-Pol κ, anti-Pol ι, and anti-Pol η antibody, respectively. Representative images of Pol κ, Pol ι, and Pol η expression: Aa, Ab, Ba, Bb, Ca, and Cb, normal brain tissue; Ac, Ad, Bc, Bd, Cc, and Cd, pilocytic astrocytoma (WHO grade I); Ae, Af, Be, Bf, Ce, and Cf, diffuse astrocytoma (WHO grade II); Ag, Ah, Bg, Bh, Cg, and Ch, anaplastic astrocytomas (WHO grade III); Ai, Aj, Bi, Bj, Ci, and Cj, glioblastoma multiforme (WHO grade IV); magnification: ×100 (Aa, Ac, Ae, Ag, Ai, Ba, Bc, Be, Bg, Bi, Ca, Cc, Ce, Cg, and Ci) and ×400 (Ab, Ad, Af, Ah, Aj, Bb, Bd, Bf, Bh, Bj, Cb, Cd, Cf, Ch, and Cj).

Fig. 3.

Association of Pol κ and Pol ι protein expression with survival in glioma patients. Kaplan–Meier survival curves comparing Pol κ and Pol ι expression status (positive or negative) are shown. The analysis of Pol κ or Pol ι included (A and D) all patients with the log-rank test (P < .001 or P = .014), (B) anaplastic astrocytoma (WHO grade III) between Pol κ–positive and Pol κ–negative glioma patients with the log-rank test (P = .014), (C) glioblastoma multiforme (WHO grade IV) between Pol κ–positive and Pol κ–negative glioma patients with the log-rank test (P = .048), (E) anaplastic astrocytoma (WHO grade III) between Pol ι–positive and Pol ι–negative glioma patients with the log-rank test (P = .089), (F) glioblastoma multiforme (WHO grade IV) between Pol ι–positive and Pol ι–negative glioma patients with the log-rank test (P = .528).