Association of mutation and low expression of the CTCF gene with breast cancer progression.

BACKGROUND: CTCF encodes 11-zinc finger protein which is implicated in multiple tumors including the carcinoma of the breast. The Present study investigates the association of CTCF mutations and their expression in breast cancer cases.
METHODS: A total of 155 breast cancer and an equal number of adjacent normal tissue samples from 155 breast cancer patients were examined for CTCF mutation(s) by PCR-SSCP and automated DNA sequencing. Immunohistochemistry (IHC) method was used to analyze CTCF expression. Molecular findings were statistically analyzed with various clinicopathological features to identify associations of clinical relevance.
RESULTS: Of the total, 16.1% (25/155) cases exhibited mutation in the CTCF gene. Missense mutations Gln > His (G > T) in exon 1 and silent mutations Ser > Ser (C > T) in exon 4 of CTCF gene were analyzed. A significant association was observed between CTCF mutations and some clinicopathological parameters namely menopausal status (p = 0.02) tumor stage (p = 0.03) nodal status (p = 0.03) and ER expression (p = 0.04). Protein expression analysis showed 42.58% samples having low or no expression (+), 38.0% with moderate (++) expression and 19.35% having high (+++) expression for CTCF. A significant association was found between CTCF protein expression and clinicopathological parameters include histological grade (p = 0.04), tumor stage (p = 0.04), nodal status (p = 0.03) and ER status (p = 0.04).
CONCLUSIONS: The data suggest that CTCF mutations leading to its inactivation significantly contribute to the progression of breast cancer.

Introduction

Breast cancer, that represents nearly 1/4th of total cancers diagnosed in female (Ferlay et al., 2015), results from the interaction between multiple genes and environmental factors (Kaaks et al., 2005) (Xie et al., 2006).

CTCF (CCCTC-binding factor) regulates gene expression through activation/repression of the promoter, chromatin lining and imprinting of genomes (Ong and Corces, 2014), along with involvement in the establishment of the 3D structure of the genome and genomic segments (Holwerda and de Laat, 2013). It also has a tumor suppressor function and is commonly deleted or mutated in breast cancer cell lines and breast tumors (Ji et al., 2016, Kaiser et al., 2016, Sabarinathan et al., 2016, Umer et al., 2016).

11-zinc fingers enable the protein to bind diverse gene sequences making it a universal transcription factor (Kim et al., 2007, Nakahashi et al., 2013). CTCF is a tumor suppressor suspects linked with familial breast cancer (Ohlsson et al., 2001). Genomic platforms and integration of sequence data suggest that CTCF may harbor driver mutation in breast cancer (Cancer Genome Alas, 2012, Nik-Zainal et al., 2016). CTCF protein levels are elevated in many breast tumors as well as cancer cell lines (Docquier et al., 2005). The epigenetic control of the BAX gene by CTCF helps the cancer cells to evade apoptosis in addition to influencing genome imprinting, intronic transcription, inactivation of X-chromosome, and post-transcription processing (Docquier et al., 2005, Mendez-Catala et al., 2013). It further downregulates the transcription of the c-Myc oncogene and impacts the expression of maternal H19 allele (Holmgren et al., 2001), as well as establishes chromatin boundaries and mediating long-range chromatin interactions (Phillips and Corces, 2009).

16q22-24 region, the location of CTCF, commonly exhibits loss of heterozygosity (LOH) in breast cancer (Lindblom et al., 1993, Cleton-Jansen et al., 1994). Interestingly, this phenomenon is associated with longer survival and delayed metastasis (Lindblom et al., 1993, Hansen et al., 1998).

CTCF can be deregulated in multiple ways, including germline and somatic (missense and nonsense) mutations resulting in the onset and progression of genetic disorders like human cancers (Lupianez et al., 2015, Filippova et al., 1998, Rubio-Perez et al., 2015). The CTCF gene mutation(s) leading to many human cancers strongly suggest the critical loss of function as important as tumor suppression (Marshall et al., 2017, Ohlsson et al., 2001, Filippova et al., 1998).

An early in vitro report indicated the anti-proliferative activity of CTCF by showing that it repressed the cell proliferation (Lutz et al., 2000). However, the precise role of CTCF in the onset or progression of cancer remains to be elucidated. The efforts have been made in many human tumors, but breast cancer is scarcely studied (Takai et al., 2001, Filippova et al., 2002, Yeh et al., 2002, Aulmann et al., 2003, Ulaner et al., 2003).

The current study aimed to find mutations in various hot spot exons of CTCF by polymerase chain reaction-single stranded conformation polymorphism (PCR-SSCP) followed by sequencing of DNA isolated from cancerous and adjacent normal breast tissues to identify mutational hot spots of CTCF contributing to carcinogenesis. Additionally, the expression of CTCF protein was also analyzed to show the relationship between CTCF mutation and its expression with various clinico-pathological parameters.

Materials and Methods

Biological samples

The subjects were informed, and consent forms were collected from all the participants after the ethical clearance from the institutional ethics committee of All India Institute of Medical Sciences (AIIMS), New Delhi, India. 155 female breast cancer tissue samples and an equal number of adjacent normal tissues (measuring 5–10 mm), were obtained and stored in PBS and formalin between 2010 and 2014 from the Department of Surgical Oncology, AIIMS, New Delhi India. Breast cancer stages were determined using the TNM staging system or American Joint Committee on Cancer (AJCC). The clinicopathological variables are age, histological type, tumor size, histological grade, menopausal status, tumor stage, nodal status, ER expression, PR expression and Her2/Neu.

DNA isolation

DNA was isolated from the cases and healthy control tissue by using the standard phenol/chloroform method as described previously (Sambrook et al., 1989). DNA concentration and purity were determined using electrophoresis and UV spectrophotometry before storing it in TE buffer.

PCR-SSCP analysis

Hot spot exon 1 and 4 of CTCF gene were probed for the presence of mutation using the tumor and normal control DNA with the primers (table 1). PCR product showed the presence of 206 bp and 246 bp amplicons (Fig. 1) as identifed by using Quantity One Software (Bio-Rad Laboratories, CA, USA). The purified product was assayed for any alteration in the electrophoretic mobility as described previously (Orita et al., 1989). Comparison of single-stranded DNA bands of the tumor and normal control to the identification of SSCP positive samples (Fig. 2).

Table 1

Oligonucleotide primer sequences used for amplification of different exons.

Gene	Consensus sequence	Annealing temperature (oc)	Amplicon size (bp)
CTCF Exon 1 FP	5′- GGTGATGATGGAACAGCTGG -3′	60	206
CTCF Exon 1 RP	5′- TGGTAGCAACAGGTACAGTC -3′
CTCF Exon 4 FP	5′-TCACATTCGCTCTCATACTGG-3′	58	246
CTCF Exon 4 FP	5′-CGGAGAAGCATTATCAATTC-3′

Fig. 1

Amplified exon products (206 and 246 bp) of CTCF gene. Lane L: Molecular marker of 100 bp, Lanes 1–7: Amplicons from the Breast cancer tissues samples.

Fig. 2

SSCP (non-radio active) analysis of CTCF gene shows a variation/ shift in band pattern lane 2, 3, 5, 6, 7 and 8 of the breast tumor samples when compared with normal sample in lane 1.

DNA sequencing

The SSCP positive samples were re-amplified and purified before sequencing twice to prevent the formation of any artifacts (ABI PRISM310 dye Terminator Cycle Sequencing Ready reaction Kit) and analyzed using Sequencing Analysis Software 3.4.1 (Fig. 3).

Fig. 3

Representative partial electropherograms of Mutant (B) (shown by arrow) with Normal (A) adjacent forms of CTCF gene showing substitutin G>T.

IHC analysis

Immunohistochemical staining was done to assess CTCF protein expression using the anti-human CTCF antibody (GeneScript, USA, Catalog # A01529) (Barbareschi et al., 1996). Briefly, breast cancer tissue samples cut into 2 to 4 μm section embedded in poly-L- lysine coated slides were treated with xylene, alcohol, and heat to retrieve the antigen. The slides were finally incubated with anti-CTCF antibody and developed using straptavidin Horse-reddish peroxidase detection kit (GeneScript USA). Slides scoring as Low or no expression (+), Moderate (++) and High expression (+++) representation of number and distribution of cells among cases and controls was done using Olympus BX 50, Tokyo.

Statistical analysis

Chi-square test (χ2) was used to assess the association of CTCF mutation and its expression with various clinicopathological parameters using GraphPad Prism 6.0. The P-value ≤ 0.05 was considered significant.

Results

Mutation(s) in CTCF and clinicopathological parameters

A total of 25 (16.12%) breast cancer cases showed mutations in the exon 1 and exon 4 including a missense mutation (Gln > His, G > T) in seventeen cases and silent mutation (Ser > Ser, C > T) in eight cases (Table 2) (Fig. 6).

Table 2

Details of CTCF gene Mutation(s) in Female Breast Cancer Cases from India.

Affected Codon	Base Position	Base Change	Amino Acid Change	Mutation Effect	No. of Patients
72	216	CAG > CAT (G > T)	Glutamine > Histidine (Gln > His)	Missense	17
388	1455	TCC > TCT (C > T)	Serine > Serine (Ser > Ser)	Silent	08

Fig. 6

Graphical abstract.

No mutation was observed in normal control samples. The relationship between CTCF mutations and clinicopathological parameters showed a significant correlation with patients menopausal status (p = 0.02), tumor stage (p = 0.03), nodal status (p = 0.03) and estrogen receptor (ER) expression (p = 0.04 (Table 3). However, the association with age, histological type, tumor size, histological grade, progesterone receptor (PR) expression, and Her2/Neu failed to reach statistical significance.

Table 3

Correlation between mutations of human CTCF gene with clinicopathological parameters.

Number of patients	155
Parameters	No. of cases (n = 155)	Mutations	Mutation rate (%)	χ2 value	P value
Age
>50	80	16	20.00	1.831	0.176
≤50	75	09	12.00
Menopausal status
Pre	70	06	8.57	5.390	0.020**
Post	85	19	22.35
Histological Type
Invasive ductal Carcinoma	150	25	16.67	0.993	0.318
(IDC)
Invasive lobular carcinoma	00	0.00	05
(ILC)
Tumor Size
≤2cm	65	07	10.77	2.377	0.123
˃2cm	90	18	20.00
Histological Grade
Poorly differentiated (PD)	40	05	12.50	1.596	0.450
Moderately differentiated (MD)	69	14	20.29
Well differentiated (WD)	46	06	13.04
Tumor Stage
Stage II (a+b)	73	07	09.59	4.363	0.036**
Stage III (a+b) + IV	82	18	21.95
Nodal Status
Positive	81	18	22.22	4.656	0.030**
Negative	74	07	09.46
Estrogen Receptor (ER) Expression
Positive	72	07	09.72	4.080	0.043**
Negative	83	18	21.68
Progesterone Receptor (PR) Status
Positive	66	08	12.12	1.365	0.242
Negative	89	17	19.10
Her2/Neu
Positive	69	07	10.14	3.292	0.069
Negative	86	18	20.93

P-value < 0.05** was considered significant.

CTCF expression and clinicopathological features

Of all 155 cases, 66 (42.59%) showed low/ no expression (+), 59 cases (38.06%) with moderate (++) expression and 30 cases (19.35%) had high (+++) expression for CTCF nuclear staining (Table 4, Fig. 5). A significant correlation was detected between CTCF protein expression and histological grade (p = 0.04), nodal status (p = 0.03), tumor stage (p = 0.04), and ER status (p = 0.04) (Table 5). However, the association with age (p = 0.29), menopausal status (p = 0.84), histological type (p = 0.56), tumor size (p = 0.464), PR status (p = 0.10) and Her2/Neu (p = 0.49) failed to reach significance (Table 5).

Table 4

Profile of CTCF protein expression.

CTCF gene expression
Low	66/155	42. 59%
Moderate	59/155	38. 06%
High	30/155	19. 35%

Fig. 5

Representative Immunohistochemical slides (seen at 10X) showing (a) Low expression (+), (b) Moderate expression (++), (c) High expression (+++) of CTCF protein in Indian female breast cancer cases and (d) CTCF protein expression in normal control.

Table 5

Correlation between the expression of CTCF protein and Clinico-pathological parameters of breast carcinoma.

Parameters	n=155	Low	Normal	High	χ2	p value
Age
˃50	80	35	14	31	2.447	0.294
≤50	75	29	21	25
Menopausal status
Pre	70	32	14	24	0.328	0.848
Post	85	35	18	32
Histological Type
Invasive Ductal Carcinoma (IDC)	150	60	39	51	1.158	0.560
Invasive Lobular Carcinoma (ILC)	05	02	02	01
Tumour Size
≤2cm	65	23	18	24	1.532	0.464
˃2cm	90	40	24	26
Histological Grade
First	40	20	09	11	09.791	0.044**
Second	69	23	25	21
hird	46	12	11	23
Tumor stage
Stage2 (a+b)	73	23	18	32	6.079	0.047**
Stage3 (a+b) +4	82	38	23	21
Nodal status
Positive	81	38	21	22	6.824	0.033**
Negative	74	24	15	35
ER status
Positive	72	22	17	33	6.008	0.049**
Negative	83	40	19	24
PR Status
Positive (+ve)	66	36	12	18	4.475	0.106
Negative (−ve)	89	34	26	29
Her2/Neu
Positive (+ve)	69	24	19	26	1.413	0.493
Negative (−ve)	86	38	20	28

Correlation between mutation(s) and expression of CTCF cases

The mutation(s) found in the breast cancer patients was analyzed along with CTCF expression to elucidate the potential role of CTCF in breast cancer. The relationship of CTCF mutation with its expression was observed significantly in the case of low level (+) protein expression (p = 0.03). However, the link in cases of moderate (++) (p = 0.11) and high level (+++) (p = 0.43) protein expression was found not significant (Table 6).

Table 6

Correlation between mutations and protein expression of CTCF gene in breast cancer patients

Level of expression	No. of cases (n = 155)	Mutations (25/155 = 16.12%)	χ2	p-value
Low	69/155 (44.52%)	16/155 (10.32%)	4.581	0.032**
Moderate	59/155 (38.06%)	06/155 (3.87%)	2.501	0.113
High	27/155 (17.42%)	03/155 (1.93%)	0.608	0.435

Discussion

Aberrant CTCF is linked with several diseases/disorders, including cancer (Aulmann et al., 2003, Prawitt et al., 2005, Herold et al., 2012, Gregor et al., 2013, Bastaki et al., 2017). The tumor suppressor function of CTCF is speculated based upon its impact on critical genes like p53, Myc, BRCA1, p19/ARF involved in cancer onset and progression (Bell and Felsenfeld, 2000, Klenova et al., 2002, Qi et al., 2003, Ohlsson et al., 2001).

Recent studies suggest that overexpression of CTCF contributes to tumor development in breast cancer by downregulating HOXA10 and H3K27me3 expressions (Mustafa et al., 2015, Lee et al., 2017). Interestingly, the repression of CTCF leads to the overexpression of BAX and eventual apoptosis (Docquier et al., 2005).

Mutations have been detected in CTCF chromatin binding sites (CBS) in multiple cancers, especially mutations of A-T base pairs (Katainen et al., 2015). The loss of CTCF poly ADP-ribosyltion in breast cancer leading to the expression of both 180-kDa and 130-kDa in comparison to only 180-kDa CTCF in normal breast tissues likely contributes to the progression of breast cancer (Docquier et al., 2009).

Although CTCF-130-kDa or Rb2/p130 can be used as the biomarker for the cancer progression, the utility of CTCF-130-kDa as the disease prognosis biomarkers remain to be investigated (Long et al., 2018, Shi et al., 2018, Wang et al., 2017, Wu et al., 2018, Kawamura et al., 2018). Recent studies show that CTCF regulates changes of the 3D genome organization (Wang, 2018, Singh and Shrivastava, 2017, Liu and Wu, 2018) in the pathogenesis of disease (Szalaj and Plewczynski, 2018, Ma et al., 2018, Terabayashi and Hanada, 2018).

In the present study, we screened the hotspot coding regions of CTCF gene for the mutation(s) by PCR-SSCP in 155 cases of female breast carcinoma along with corresponding adjacent healthy control. We found 25 (16.1%) missense and silent mutations in female breast cancer tissues as shown in the table 2. The mutation(s) were identified at codon 72 leading to Gln > His (G > T), and at codon 1455 leading to Ser > Ser(C > T). The missense codon mutations in CTCF zinc finger domain 3, observed in the present study may impact the CTCF binding to the promoters of genes related to cellular proliferation like MYC, PLK, PIM-1, p19ARF, and Igf2/H19 (Fig. 4) (Filippova et al., 2002). The mutation(s) may also have resulted in repression of the wild-type allele and low/ no protein expression (Aulmann et al., 2003). Moreover, our data exhibited the altered expression profiles of CTCF which may be due to the result of potential mutation(s) in the CTCF exonic region and thus can contribute in the progression of breast cancer as shown in an early study (Tiffen et al., 2013).

Fig. 4

Impact of mutation on CTCF protein.

Importantly, we observed that CTCF mutations were only found in breast cancer tissue and not in healthy tissues. The analysis of potential relationship with the patient’s ages, menopausal status, histological types, tumor sizes, histological grades, tumor stages, lymph node metastases, steroid receptors (ER & PR) and Her2/neu amplifications to elucidate the role of mutation(s) in CTCF gene in the progression of breast cancer revealed a significant relationship between CTCF mutations and patients’ menopausal status (p = 0.02), tumor stages (p = 0.03), lymph node metastases (p = 0.03) and ER (p = 0.04). The significant association with the clinical parameters further emphasizes the link between CTCF and breast cancer progression. No significant association was found with patient ages, histological types, tumor sizes, histological grades, PR and Her2/neu amplifications.

CTCF protein expression analysis was performed to explore the possible role of CTCF in female breast cancer cases and also to define the biomarker property.

An early study showed moderate to strong nuclear staining of CTCF protein (Aulmann et al., 2003). More than 80% of the cases included in our study showed a low or moderate CTCF expression which was significantly linked with histological grades (p = 0.04), tumor stages (p = 0.04), lymph node involvements (p = 0.03) and ER (p = 0.04). No significant association was observed with parameters such as age, menopausal status, histological types, tumor sizes, PR status and Her2/neu amplification of breast cancer progression (Table 5) suggesting the involvement of detected mutations in expression and altered formations of the protein contributing to the onset and progression of breast cancer which is in agreement with the findings of an early study (Tiffen et al., 2013).

A significant association found between CTCF mutation and protein expression in the low level (+) category of protein expression (p = 0.03), whereas no significant association with the moderate (++) and high level (+++) of protein expression suggest that the progression of breast cancer may be related to the lower level of CTCF protein expression indicating towards the anti-proliferative activity of CTCF, and a potent breast cancer susceptible genes. The altered expression profiles of CTCF showing mainly nuclear expression in the current study are in agreement with the earlier studies using immunofluorescence to without any cytoplasmic staining (Zhang et al., 2004). In the present study, we found an association between CTCF expression and histological grade in the high percentage of low-grade tumors having less proliferative activity showing positive nuclear expression, whereas high-grade tumors primarily showed low or no expression. Our results are concordant with early reports indicating the ability of CTCF to inhibit cell growth and proliferation (Rasko et al., 2001).

Conclusions

Our study showed mutations (missense Gln > His, G > T and silent Ser > Ser, C > T) of the CTCF gene in Indian female breast cancer cases. The detected mutations showed that 16 (64%) mutations had a statistically significant association with low or no expression for CTCF when analyzed with IHC data. The findings suggest that CTCF may be a tumor suppressor gene and its inactivation may play an essential role in the progression of breast carcinoma. However, further clinical studies with larger sample size are needed to elucidate the fundamental role of CTCF gene in the breast cancer onset and progression in Indian population.