Hypermethylation of mitochondrial DNA facilitates bone metastasis of renal cell carcinoma

Abstract

Kidney cancers including clear cell carcinoma (RCC) are identified with very vulnerable mitochondria DNA (mtDNA) and frequent epigenetic aberrations. Bone metastasis from RCC is prevalent and destructive. Bone marrow contains a quite hypoxic microenvironment that usually insitigate 50% of hypermethylation events in conferring a selective advantage for tumor growth. We hypothesized that hypermethylation of mtDNA in RCC cells would significantly contribute to bone metastatic tumor progression. Methylation-specific polymerase chain reaction assay (MSP) was adopted to measure the methylation status of D-loop region of mtDNA in 15 pairs of bone metastatic and primary RCC as well as tumor adjescent normal kidney tissues. mtDNA copy number was examined by the real-time quantitative polymerase chain reaction (qPCR). Western blotting analysis was used to measure the accumulation of several DNA methyltransferases (DNMTs) in the mitochondria and nucleus fractions of bone metastatic RCC cells. mRNA expression of mitochondria encoded genes was examined by RT-PCR. Reactive oxygen species (ROS), mitochondrial membrane potential and ATP content were measured using in vitro cells treated with de-methylation drug 5-Azacytidine (5-Aza). Non-invasive bioluminescent imaging was performed to monitor tumor occurrence in skeleton in mice. Our results showed that the D-loop region in bone metastatic tumor cells was markedly hypermethylated than those in primary RCC tumor cells, that is associated with a decreased mtDNA copy number and accumulation of DNMT1 in the mitochondria. The bone-tropism tumor colonization and progression of RCC cells was significantly suppressed by demethylating the D-loop region of mtDNA and reducing the intracellular level of ROS and ATP by 5-Aza treatment. In conclusion, our study provided a direct association between hypermethylation of mtDNA in RCC with bone metastastic tumor growth.

Keywords: 5-Azacytidine; Bone metastasis; clear cell carcinoma; hypermethylation; mitochondria DNA.

Figure 1

Methylation status of D-loop region in bone metastasis (BO) vs primary (P) RCC tumor and normal kidney tissue (Norm) specimens and cells. A. Representative MSP results of D-loop methylation in 3 pairs of BO and P tumor specimens. M: methylated, U: unmethylated. B. Statistical analysis of methylation ratios of D-loop region in all BO, P and Norm specimens (n=15). C. Methylation status of D-loop region in ACHN-P vs ACHN-BO and OS-RC-P vs OS-RC-BO cells. D. Statistical analysis of methylation ratios of D-loop region in tumor cell lines and normal human bone cell line (HOB) and human kidney cell line (HEK293) (n=3). *P < 0.05, by student's t-test. Error bars show ± SD.

Figure 2

mtDNA copy number and mitochondria DNMT1 expressions in BO and P RCC tumor specimens and cells. A. Relative mtDNA copy number in 15 pairs of BO and P specimens. B. Comparison of the average relative mtDNA copy number between the 15 cases. * P<0.05. C. Representative blots of DNMT1, DNMT3A, and DNMT3B in nuclear (Nucl) and mitochondrial (Mito) fractions of ACHN and OS-CR-2 cells. D. Semi-quantification of DNMT1 and DNMT3A in C. Expression was normalized to the respective internal reference markers (VDAC1 or Histone 3), protein expression in P cells was set as 1. * P<0.05.

Figure 3

The mRNA expression levels of mtDNA genes in BO vs P RCC cells detected by quantitative RT-PCR (n = 5). * P<0.05.

Figure 4

In vitro effects of 5-Aza on D-loop methylation (A), cell proliferation (B) and apoptosis (C), as well as mitochondria function (D-G) in ACHN-BO and OS-RC-2-BO cells. *P < 0.05 by student's t-test. Error bars show ± SD.

Figure 5

Treatment with 5-Aza on bone metastatic tumor development in xenograft animal models. A. Experimental design for the assessment of in vivo bone metastatic tumor growth after epigenetic methylation reprogramming. ACHN-BO and OS-RC-2-BO cells were pretreated with 100nM 5-AZA for 1 week in vitro. The reprogrammed and control cells were injected into the left cardiac ventricle of athymic nude mice. Tumor growth in skeleton (spine or limbs) were monitored twice a week for 8 weeks. B. Representative bioluminescent images of bone metastasis in 3 mice in the control and 5-Aza treatment groups. C. Kaplan-Meier curves for the bone metastasis-free survival of both ACHN-BO and OS-RC-2-BO xenograft animal models. The p value was calculated with a log-rank test (N = 14). D. Representative images of HIF1α immunohistochemistry staining and quantification (H scores) of the HIF1α immuno-reactivities in the bone tissue sections from the mice who developed bone metastasis. Images were taken under 10x objective.