Detection of Somatic Mutations and Germline Polymorphisms in Mitochondrial DNA of Uterine Fibroids Patients

Abstract

To identify the role of mitochondrial DNA (mtDNA) mutations in uterine fibroids patients, genomic DNA isolated from paired myometrium and fibroid tissues was screened for mutations. The present study represents the first investigation to report that 10.4% of uterine fibroids cases had either mtDNA mutations or polymorphisms or both. Among the 14 mitochondrial sequence variants identified, seven are somatic mutations (A3327C, G3352A, G3376A, G3380A, G3421A, T15312G, and C15493G) and the remaining (G3316A, C3342A, C3442T, T10205A, A10188G, A10229C, and A10301T) are gene polymorphisms. Somatic mutations were both homo- and heteroplasmic in nature. Of the seven somatic mutations located in the MTND1 and MTCYB genes, five (71.42%) are nonsynonymous in nature, whereas four (57.14%) of the polymorphisms located in MTND1 and MTND3 genes are found to be nonsynonymous. Sequence variants such as G3380A, G3421A, T15312G, G3376A, and G3316A have been earlier described in different human pathologies, but the remaining are novel ones. Mitochondrial somatic mutations and polymorphisms may predispose women to an earlier onset of degenerative cellular processes, which impair oxidative phosphorylation capacity and thereby promote tumorigenesis in uterine smooth muscle cells. Detection of mtDNA sequence variations in fibroid patients raises the need for larger case-control studies to screen the whole mitochondrial genome and evaluate as a future diagnostic biomarker in fibroid patients.

Introduction

Leiomyomas (fibroids) are benign growths of the uterine tract; these constitute the major causes of gynecological morbidity and hysterectomy in reproductive age group women (Vollenhoven et al., 1994). Even though the chance of malignant transformation of uterine fibroids is <1%, they often account for menorrhagia, pelvic pain, reproductive dysfunction, infertility, loss of pregnancy, and placental abruption in a large number of women (Parker, 2007). Genetic studies have indicated a higher susceptibility rate of fibroids in mono- over dizygotic twins (Treloar et al., 1992; Ligon and Morton, 2000; Luoto et al., 2000). Family history and polymorphisms in hormone and cytokine receptors, cell signaling, and DNA repair genes, in addition to molecular and environmental factors, are also described to contribute to the etiopathogenesis of these tumors (Flake et al., 2003; Jeon et al. 2005; Hsieh et al., 2007; Govindan et al., 2009; Lakshmi et al., 2009; Shaik et al., 2009). Despite extensive studies on the involvement of nuclear genes, the role of mitochondrial genes in the pathophysiology of fibroids has not been examined.

It is apparent that estrogen acts as a promoter of fibroid growth and mitochondria constitute an important target of action (Felty and Roy, 2005). Estrogen has been reported to alter the reactive oxygen species (ROS) levels inside the mitochondrial matrix region (Liehr and Roy, 1998; Felty et al., 2005). Mitochondrial DNA (mtDNA) is more vulnerable to oxidative damage and undergoes a higher rate of mutation than does the nuclear genome (Jeon et al., 2005); hence, it is conceivable that estrogen-stimulated ROS levels could lead to mitochondrial genetic instability in tumor cells. Several studies have identified the existence of mtDNA alterations, such as missense mutations, frame shift mutations, and multiple gene insertions/deletions in breast, ovarian, cervical, and uterine endometrial neoplasms and cell lines (Copeland et al., 2002; Chatterjee et al., 2006; Czarnecka et al., 2006). Mutations in mtDNA could in turn elevate ROS production, leading to additional mutations and oxidative stress; this could provide replication advantage to transformed cells to evade apoptosis and drive cellular proliferation (Rustin, 2002). Given the critical role of estrogen in the tumorigenesis of fibroids and in mitochondria, the role of mtDNA mutations could be hypothesized in the pathogenesis of uterine fibroids. Therefore, mutation screening was done at different gene regions of mtDNA isolated from fibroid tumors and matched myometrium tissues.

Materials and Methods

Enrollment of subjects and clinical specimen collection

Sixty-nine uterine fibroid patients from government maternity hospitals in Hyderabad were enrolled in this study. Case diagnosis was based on the gynecological and ultrasonography examinations. Clinical details including age at menarche, associated symptoms, and length and duration of menstrual cycle were recorded in case sheets. A total of 138 tissue samples, comprised of uterine fibroids and adjacent normal myometrium of the same patient, were collected after vaginal or abdominal hysterectomy. Histological characterization of tissues was performed by an experienced pathologist. The present study protocol was approved by the hospital ethics committee and informed consent was obtained from all the patients.

DNA isolation, polymerase chain reaction, and mutation detection

Genomic DNA was extracted from tissue samples by following the method described by Alluri et al. (2005). Individual polymerase chain reactions (PCRs) were performed to amplify the 2993-3493, 5909-6301, 7471-8566, 8211-8411, 10076-10357, and 14975-15744 gene regions using the sequence-specific primers (Table 1). Specific conditions for PCR including the annealing temperature and its duration, in addition to the number of cycles, are mentioned in Table 2. Single-stranded conformational polymorphism (SSCP) and restriction enzyme digestion methods were followed to screen the mutations, based on the size of PCR amplicons (Dong and Zhu, 2005). SSCP with PCR amplicons of the 2993-3493 region was carried out after initial digestion with HaeII enzyme. PCR amplicons that showed mobility shift on SSCP analysis or altered restriction pattern by restriction enzyme digestion analysis were initially purified using the QIAquick (Qiagen) gel extraction kit, followed by a bidirectional sequencing reaction using an ABI PRISM 310 Genetic Analyzer (Perkin-Elmer).

Table 1.

Primer Details, Polymerase Chain Reaction Conditions, and the Gene Regions

S. no.	mtDNA region	Primer sequence	Annealing temperatures (°C/seconds/number of PCR cycles)	Amplicon spanning gene regions	Detection method
1	2993-3020: F	5′-TTGGATCAGGACATCCCGATGGTGCAG-3′	64/45/35	MTND1+MTTL1 + MTRNR2	HaeII+SSCP
	3466-3493: R	3′-GTTTTAGGGGCTCTTTGGTGAAGAGTT-5′
2	5909-5929: F	5′-TCG CCG ACC GTT GAC TAT TC-3′	51/45/35	MTCO1	HaeIII enzyme
	6281-6301: R	3′-TAA GGG AGG GTA GAC TGT TC-5′
3	7471-7492: F	5′-CAA AGC TGG TTT CAA GCC AAC C-3′	62/45/35	MTCO2	HaeIII+MSP 1
	8545-8566: R	3′-TTG TGG GGC AAT GAA TGA AGC-5′
4	8211-8231: F	5′-TCG TCC TAG AAT TAA TTC CC-3′	50/30/33	MTCO2+MTATP8 + MTTK	SSCP
	8392-8411: R	3′-GGG GGT AAT TAT GGT GGG CC-5′
5	10076-10095: F	5′-TCA ACA CCC TCC TAG CCT TA-3′	59/30/32	MTND3	SSCP
	10337-10357: R	3′-CAG ACT TAG GGC TAG GAT GA-5′
6	14975-14998: F	5′-GAA TCA TCC GCT ACC TTC ACG CC-3′	61/45/35	MTCYB	StyI+Eam 1104I
	15744-15722: R	3′-AGG AGG TCT GCG GCT AGG AGT CA-5′

Table 2.

Clinical and Mutation Details of Fibroid Cases

										Affected tissue
Case no.	Age	Type of fibroids	Number of fibroids	Parity status	Family history	Gene region	Nucleotide sequence variation	Genetic code change	Amino acid change	Fibroids	Myometrium
F5	42	Intramural	Multiple	G2P2	No	MTND1	G3380A	CGA→CAA	Arg→Gln	Yes	No
F19	42	”	”	0	No	”	G3376A	GAA→AAA	Glu→Lys	”	”
F37	35	”	Single	G3P2	No	”	G3421A	GTT→ATT	Val→Ile	”	”
F50	44	Subserosal	Multiple	G1P1	No	”	A3327C	CTA→CTC	Leu→Leu	”	”
”	”	”	”	G2P2	”	”	G3352A	GCA→ACA	Ala→Thr	”	”
F55	47	Intramural	Single	G2P2	Yes	”	G3316A	GCC→ACC	Ala→Thr	”	Yes
”	”	”	”	G1P1	”	”	C3442T	CTA→TTA	Leu→Leu	”	”
”	”	”	”	G3P2	”	”	C3342A	CCC→CCA	Pro→Pro	”	”
F40	38	Subserosal	Multiple	G2P2	No	MTND3	T10204A	GTC→GAC	Val→Asp	”	”
”	”	”	”	G2P2	”	”	A10301T	ACA→ACT	Thr→Thr	”	”
”	”	”	”	G1P1	”	MTCYB	C15493G	CTC→CTG	Leu→Leu	”	No
”	”	”	”	G4P2	”	”	T15312G	ATT→AGT	Ile→Ser	”	”
F13	42	Intramural	”	G3P1	No	MTND3	A10188G	ATA→GTA	Met→Val	”	Yes
”	”	”	”	G2P2	”	”	A10229C	TTA→TTC	Leu→Phe	”	”

Analysis of the mtDNA sequences

The results of the mtDNA sequence analysis were compared with the published NCBI sequences, using the BLAST search tool for mutation identification (www.ncbi.com/blast). Further, sequence variants were compared against mitochondrial databanks, such as the MITOMAP and mtDBase banks to identify whether those variants were reported earlier (www.mitomap.org/cgibin/tbl14gen.pl and www.genpat.uu.se/mtDB/index.html). Somatic mtDNA mutations are defined as those present in the fibroids but not in the myometrium tissues samples from the patient concerned, whereas sequence variants, which are present in both the fibroids and matched myometrium tissues, were scored as germline polymorphisms. Somatic mutations or polymorphisms that are not found in MITOMAP database are considered as novel.

Results

Clinical analysis

Of the cases included in the study, 19 were diagnosed to have single and 50 to have multiple fibroids, which varied in size between 1.0×1.0 and 9.0×6.0 cm, with a mean area of 20.0 cm. The mean age at fibroids detection was 40±1.02 years and the age at menarche was 13.8±0.72 years. Based on three-generation pedigree analysis, 9 of 69 cases (13.04%) reported positive family history of fibroids in their first-degree relatives. Nulliparity was reported in 3 of 69 (4.34%) fibroid cases.

Somatic mutations and polymorphisms identification

On sequencing analysis of PCR amplicons, which showed altered mobility on SSCP gels (six cases) or altered banding pattern on agarose gels (one case), a total of 14 mutations were detected in seven fibroid cases (10.14%). Among these, seven are mitochondrial somatic variations (A3327C, G3352A, G3376A, G3380A, G3421A, T15312G, and C15493G) and the remaining seven (G3316A, C3342A, C3442T, T10205A, A10188G, A10229C, and A10301T) are germline polymorphisms (Table 2). Except T15312G and C15493G, all other sequence variations were seen in homoplasmic form. Of the 14 different sequence variants identified, 8 mutations were located in the MTND1 gene and the remaining 6 mutations were in the MTND3 and MTCYB genes. Of the seven somatic mutations of the MTND1 and MTCYB genes, five (71.42%) are nonsynonymous, that is, resulting in an amino acid change, and the remaining two (28.57%) are synonymous, whereas among the seven gene polymorphisms, four (57.14%) were found to be nonsynonymous and three (42.85%) are synonymous. Nine of the 14 sequence variants, A3327C, G3352A, C15493G, A10188G, T10204A, A10229C, A10301T, C3342A, and C3442T, were not recorded in the mtDBase database and MITOMAP databases; hence, these are novel mtDNA sequence variations.

Discussion

The application of mDNA mutation and/or polymorphism patterns as a biomarker is rapidly expanding in disciplines ranging from rare metabolic diseases and aging to tumorigenesis. Mitochondrial mutations are associated with breast (15.6%-93%), ovarian (26%), and cervical (27%-95%) neoplasms, which indicates the rising significance of mitochondria in estrogen-sensitive pathologies (Bianchi et al., 1995; Parrella et al., 2001; Tan et al., 2002; Zhu et al., 2004; Sharma et al., 2005; Tseng et al., 2006; Wang et al., 2007; Chen and Zhan, 2009; Shen et al., 2010). The present study represents the first investigation reporting that 10.4% of uterine fibroids cases had either mtDNA mutations or polymorphisms or both.

Upon comparison of all the identified somatic mitochondrial mutations in mitochondrial genome databases, G3380A was previously found to be detected in colon tumors (Habano et al., 1999). Interestingly, this same mutation was described in a heteroplasmic fashion in muscle fibers of mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS) patients as well (Horvath et al., 2008). On the other hand, the G3421A mutation was reported in pancreatic cancer patients (Jones et al., 2001) and also in a family with maternally inherited diabetes and deafness but in a homoplasmic fashion (Chen et al., 2006). The G3376A mutation has been implicated as pathogenic for Leber's hereditary optic neuropathy (LHON) and MELAS patients (Blakely et al., 2005). The nonsynonymous nature of all these transitions could affect the functionality of the highly conserved region of mitochondrial NADH dehydrogenase subunit 1 (MTND1) protein. Another nonsynonymous mutation, T15312G, located in mitochondrial cytochrome b (MTCYB), was previously reported in thyroid tumors (Chatterjee et al., 2006). The higher percentage of homoplasmic somatic mutations or polymorphisms observed in fibroid cases could have been driven by random genetic drift mechanism, by which preexisting heteroplasmic sequence variants were outgrown in successive proliferation of fibroid progenitor cells (Coller et al., 2001; Jones et al., 2001). On the basis of these findings, it is suggested that the observed somatic mutations are pathogenic and are functionally selected during the neoplastic initiation and growth. None of the sequence variations showed specific clinopathological correlation with tumor types, nulliparity, and family history. Differential occurrence of mtDNA mutations in various tumors is partly due to the prominence of different repair mechanisms. A majority of fibroid cases are asymptomatic and go undetected and this could be attributed to the low frequency of family history among the mutant cases.

Among the gene polymorphisms identified, G3316A, which is seen in one fibroid case, was previously reported in patients with gestational diabetes and LHON (Odawara et al., 1996; Yu et al., 2004; Phasukkijwatana et al., 2006; Tang et al., 2006). Interestingly, the C3342A transversion, seen in the same patient, was earlier reported as C to T in a healthy Caucasian population (Snejina et al., 2006). Polymorphisms, such as G3316A, T10204A, A10301T, A10188G, and A10229C, can cause subtle differences in the encoded protein structure and function. No functional significance is attributed to sequence variations such as A3327C, C3342A, C3442T, A10301T, and C15493G, as they are synonymous. Gene polymorphisms of mtDNA predispose individuals to an earlier onset of degenerative cellular processes and decline in OXPHOS capacity (Trounce and Pinkert, 2007). These finding indicates that mitochondrial polymorphisms are indeed frequent in fibroid patients and raises the need to conduct larger case-control studies to investigate them as risk markers to fibroid susceptibility.

MTND1, MTND3, and MTCYB genes encode integral polypeptide subunits of mitochondrial respiratory chain complexes I, II, and III, respectively. Nonsynonymous sequence variations in these genes could alter the sequence in the encoded proteins, which in turn affects the stability and functioning of mitochondrial respiration, resulting in a change in mitochondrial ROS levels (Ugalde et al., 2003; Hinttala et al., 2006; Malfatti et al., 2007). Elevated ROS levels damage mtDNA and also result in membrane fluidity loss and defective apoptotic effects. The mechanisms of the differences in phenotype associated with mtDNA mutations and polymorphisms are not fully understood, but differences in organ-specific thresholds for energy demand and/or the presence of additional nuclear gene mutations may also modulate mitochondrial pathologies. In view of the high energy demand that exists in uterine smooth muscle tissues, disruption in cellular energy production could promote fibroid tumorigenesis.

Although none of the sequence variations are observed in the MTCO1, MTCO2, MTRNR2, MTTL1, and MTATP8 gene regions, the possibility of mutations at other sequences cannot be ruled out. Hence, comprehensive screening of the whole mtDNA genome in fibroid tumors could identify more mitochondrial sequence variants. We were unable to observe any trend between mtDNA mutations and clinical and pathological characteristics, or family history, possibly because of our relatively small sample size and detection bias.

In conclusion, the present study demonstrates the existence of mtDNA somatic mutations and polymorphisms in uterine fibroid patients. Additional evaluation of the molecular consequences of somatic mutations in fibroids may elucidate their role in energy metabolism of fibroid tumors. In a clinical context, application of PCR-coupled restriction enzyme and/or SSCP methods for rapid multisampling analysis of fibroid patients may potentiate the detection of mtDNA sequence variants with minimum cost.

Footnotes

Acknowledgments

The authors acknowledge DBT, Government of India (BT/PR/5516/MED/14/646/2004), for their financial support. The authors greatly thank the management and staff of Bhagwan Mahaveer Medical Research Centre, Hyderabad, Govt. Maternity Hospitals (Nayapool and Kothi), Hyderabad, and Bhagwan Mahaveer Medical Research Centre, Hyderabad, for providing clinical specimens and data. The authors sincerely thank Miss. Maria Arland for helping in sample collection and S. Movva for technical help.

Disclosure Statement

No competing financial interests exist.

References

Alluri

, Mohan

, Komandur

et al. 2005. MTHFR C677T gene mutation a risk factor for arterial stroke: a hospital based study. Eur J Neurol, 12:40-44.

Bianchi

, Bianchi

, Bailliet

. 1995. Mitochondrial DNA mutations in normal and tumor tissues from breast cancer patients. Cytogenet Cell Genet, 71:99-103.

Blakely

, de Silva

, King

et al. 2005. LHON/MELAS overlap syndrome associated with a mitochondrial MTND1 gene mutation. Eur J Hum Genet, 13:623-627.

Chatterjee

, Mambo

, Sidransky

. 2006. Mitochondrial DNA mutations in human cancer. Oncogene, 25:4663-4674.

Chen

, Zhan

. 2009. Study on the mutations in the D-loop region of mitochondrial DNA in cervical carcinoma. J Cancer Res Clin Oncol, 135:649-651.

Chen

, Liu

, Song

et al. 2006. A novel mitochondrial DNA missense mutation at G3421A in a family with maternally inherited diabetes and deafness. Mutat Res, 602:26-33.

Coller

, Khrapko

, Bodyak

et al. 2001. High frequency of homoplasmic mitochondrial DNA mutations in human tumors can be explained without selection. Nat Genet, 28:147-150.

Copeland

, Wachsman

, Johnson

et al. 2002. Mitochondrial DNA alterations in cancer. Cancer Investig, 20:557-569.

Czarnecka

, Golik

, Bartnik

. 2006. Mitochondrial DNA mutations in human neoplasia. J Appl Genet, 47:67-78.

10.

Dong

, Zhu

. 2005. Single-strand conformational polymorphism analysis: basic principles and routine practice. Methods Mol Med, 108:149-157.

11.

Felty

, Roy

. 2005. Estrogen, mitochondria, and growth of cancer and non-cancer cells. J Carcinog, 4:1.

12.

Felty

, Xiong

, Sun

et al. 2005. Estrogen-induced mitochondrial reactive oxygen species as signal-transducing messengers. Biochemistry, 44:6900-6909.

13.

Flake

, Andersen

, Dixon

. 2003. Etiology and pathogenesis of uterine leiomyomas: a review. Environ Health Perspect, 111:1037-1054.

14.

Govindan

, Ahmad

, Vedicherla

et al. 2009. Estrogen receptor-α gene (T/C) Pvu II polymorphism in endometriosis and uterine fibroids. Dis Markers, 26:149-154.

15.

Habano

, Sugai

, Yoshida

et al. 1999. Mutations in the mitochondrial DNA D-Loop region occur frequently in adenocarcinoma in Barrett's esophagus. Int J Cancer, 83:625-629.

16.

Hinttala

, Smeets

, Moilanen

et al. 2006. Analysis of mitochondrial DNA sequences in patients with isolated or combined oxidative phosphorylation system deficiency. J Med Genet, 43:881-886.

17.

Horvath

, Reilmann

, Holinski-Feder

et al. 2008. The role of complex I genes in MELAS: a novel heteroplasmic mutation 3380G>A in ND1 of mtDNA. Neuromuscul Disord, 18:553-556.

18.

Hsieh

, Chang

, Tsai

et al. 2007. Interleukin (IL)-12 receptor beta1 codon 378 G homozygote and allele, but not IL-1 (beta-511 promoter, 3953 exon 5, receptor antagonist), IL-2 114, IL-4-590 intron 3, IL-8 3′-UTR 2767, and IL-18 105, are associated with higher susceptibility to leiomyoma. Fertil Steril, 87:886-895.

19.

Jeon

, Kim

, Park

et al. 2005. DNA repair gene XRCC1 Arg399Gln polymorphism is associated with increased risk of uterine leiomyoma. Hum Reprod, 20:1586-1589.

20.

Jones

, Song

, Hempen

et al. 2001. Detection of Mitochondrial DNA Mutations in Pancreatic Cancer Offers a “Mass”-ive Advantage over Detection of Nuclear DNA Mutations. Cancer Res, 61:1299-1304.

21.

Lakshmi

, Shetty

, Vottam

et al. 2009. Tumor necrosis factor alpha -C850T polymorphism is significantly associated with endometriosis in Asian Indian women. Fertil Steril, 94:453-456.

22.

Liehr

, Roy

. 1998. Pro-oxidant and antioxidant effects of estrogens. Methods Mol Biol, 108:425-435.

23.

Ligon

, Morton

. 2000. Genetics of uterine leiomyomata. Genes Chromosomes Cancer, 28:235-245.

24.

Luoto

, Kaprio

, Rutanen

et al. 2000. Heritability and risk factors of uterine fibroids the Finnish Twin Cohort study. Maturitas, 37:15-26.

25.

Malfatti

, Bugiani

, Invernizzi

et al. 2007. Novel mutations of ND genes in complex I deficiency associated with mitochondrial encephalopathy. Brain, 130:1894-1904.

26.

Odawara

, Sasaki

, Yamashita

. 1996. A G to A substitution at nucleotide position 3316 in mitochondrial DNA is associated with Japanese noninsulin-dependent diabetes mellitus. Biochem Biophys Res Commun, 227:147-151.

27.

Parrella

, Xiao

, Fliss

et al. 2001. Detection of mitochondrial DNA mutations in primary breast cancer and fine needle aspirates. Cancer Res, 61:7623-7626.

28.

Parker

. 2007. Etiology, symptomatology and diagnosis of uterine myomas. Fertil Steril, 87:725-736.

29.

Phasukkijwatana

, Chuenkongkaew

, Suphavilai

et al. 2006. The unique characteristics of Thai Leber hereditary optic neuropathy: analysis of 30 G11778A pedigrees. J Hum Genet, 51:298-304.

30.

Rustin

. 2002. Mitochondria from cell death to proliferation. Nat Genet, 30:352-353.

31.

Shaik

, Govindan

, Kodati

et al. 2009. Polymorphic (CAG)n repeats in the androgen receptor gene: a risk marker for endometriosis and uterine leiomyomas. Hematol Oncol Stem Cell Ther, 2:289-293.

32.

Shen

, Wei

, Chen

et al. 2011. Evaluating mitochondrial DNA in patients with breast cancer and benign breast disease. J Cancer Res Clin Oncol, 137:669-675.

33.

Snejina

, Naumova , Elissaveta . 2006. MITOMAP mtDNA Sequence Data. www.mitomap.org/cgi-bin/mitomap/tbl15gen.pl#20060518004.

34.

Sharma

, Singh

, Sharma

et al. 2005. Mutation in the mitochondrial DNA D-loop region is frequent in cervical cancer. Cancer Cell Int, 5:34.

35.

Tan

, Bai

, Wong

. 2002. Comprehensive scanning of somatic mitochondrial DNA mutations in breast cancer. Cancer Res, 62:972-976.

36.

Tang

, Zhou

, Li

et al. 2006. Variation of mitochondrial gene and the association with type 2 diabetes mellitus in a Chinese population. Diabetes Res Clin Pract, 73:77-82.

37.

Tseng

, Yin

, Chi

et al. 2006. Mitochondrial DNA mutations and mitochondrial DNA depletion in breast cancer. Genes Chromosomes Cancer, 45:629-638.

38.

Trounce

, Pinkert

. 2007. Cybrid models of mtDNA disease and transmission, from cells to mice. Curr Top Dev Biol, 77:157-183.

39.

Treloar

, Martin

, Dennerstein

et al. 1992. Pathways to hysterectomy: insights from longitudinal twin research. Am J Obstet Gynecol, 167:82-88.

40.

Ugalde

, Triepels

, Coenen

et al. 2003. Impaired complex I assembly in a Leigh syndrome patient with a novel missense mutation in the ND6 gene. Ann Neurol, 54:665-669.

41.

Vollenhoven

, Pearce

, Herington

et al. 1994. Steroid receptor binding and messenger RNA expression in fibroids from untreated and gonadotrophin-releasing hormone agonist pretreated women. Clin Endocrinol, 40:537-544.

42.

Wang

, Wang

, Yao

et al. 2007. Somatic mutations of mitochondrial genome in early stage breast cancer. Int J Cancer, 121:1253-1256.

43.

, Yu

, Liu

et al. 2004. Relationship between mutations of mitochondrial DNA ND1 gene and type 2 diabetes. Chin Med J, 117:985-989.

44.

Zhu

, Qin

, Sauter

. 2004. Large-scale mitochondrial DNA deletion mutations and nuclear genome instability in human breast cancer. Cancer Detect Prev, 28:119-126.