Assessment of SGO1 and SGO1-AS1 contribution in breast cancer

1. Introduction

Shugoshin-like protein 1 (SGO1) participates in the control of mitosis through sheltering of sister chromatid interconnection. Such protective roles are particularly essential at the first meiotic division. In this step, cohesin is broken along the chromosome arms but centromeric cohesin should be guarded from cleavage [1]. While in higher organisms, shugoshin shields centromeric cohesin from prophase detachment, in yeast has additional functions in detecting the absence of tension at kinetochores and in producing the spindle checkpoint warnings [1]. Further evidences for contribution of SGO1 in proper mitosis have risen from studies in HeLa cells where RNAi-mediated knock down of this gene has resulted in early chromosome separation and consequent induction of apoptosis due to sustained arrest in a prometaphase-like phase [2, 3]. While its transcript levels are rather low in normal tissues, researchers have detected its over-expression in the majority of hepatocellular carcinoma (HCC) samples in association with high alpha-fetoprotein and early age of cancer occurrence. Mechanistic investigations have shown that down-regulation of this gene significantly decreases cell viability in hepatoma cell lines but not in immortalized cells. Such mitotic cell death has been shown to be accomplished through insistent stimulation of the spindle assembly checkpoint [4]. Expression level of SGO1 has been assessed in a number of human cancers. The Pancreatic Expression database as the chief source for pancreatic-originated-omics data has indicated over-expression of this gene in pancreatic cancer [5]. Expression assays in breast cancer and benign breast tissues have shown SGO1 among several differentially expressed genes between these two sets of specimens [6]. Studies in animal models have demonstrated that haploinsufficiency of SGO1 leads to chromosomal instability (CIN), development of preneoplastic as well as malignant lesions in colon [7].

A previous high throughput microarray study has resulted in identification of an antisense transcript for SGO1 which was named SGO1-AS1 [8]. This long non-coding RNA (lncRNA) has been among the utmost abnormally expressed lncRNAs in ling cancer tissues versus adjacent non-cancerous tissues (ANCTs) [8].

In the present investigation, we compared transcript quantities of SGO1 and SGO1-AS1 between breast cancer specimens and ANCTs to unravel their significance in the pathogenesis of this common female malignancy.

In the present investigation, we compared transcript quantities of SGO1 and SGO1-AS1 between breast tumor tissues and ANCTs. SGO1 has an acknowledged role in the preservation of correct mitosis in hepatoma cells and studies have proposed this gene as a possible oncotarget for HCC [4]. This evolutionarily conserved protein takes part in appropriate detachment of centromeric cohesin under the control of the spindle checkpoint in fist stages of mitosis. A splicing variant of SGO1 also regulates the spindle pole integrity. Notably, SGO1 knock-down leads to both chromosomal and centrosomal instabilities [10]. Decreased expression of SGO1 has compromised cell proliferation, triggered DNA damage response and induced senescence features in neuroblastoma cells with high levels of MYCN [11]. Moreover, knock down of SGO1 in mammary epithelial cells with high expression of E2F1, E2F2, and E2F3a has led to chromosome missegregation and centrosome amplification which indicates the role of SGO1 in preservation of genomic integrity sub-sequential to the E2F inducers [12]. Based on these evidences, we expected aberrant expression of SGO1 in breast cancer samples. However, we did not detect any remarkable difference in expression levels of this gene between tumoral tissues and ANCTs. Previous studies reported inconsistent results regarding expression pattern of this gene in malignancies. While pancreatic and breast cancer tissues have high expression of this gene [5, 6], decreased expression of SGO1 has been associated with CIN phenotype in colorectal cancer [7, 13]. Some studies suggest that over-expression of this gene in cancers is due to higher expression of truncated and malfunctioned forms [10]. As we did not assess expression of this gene at protein level, we could not deduce the functionality of expressed SGO1 in tumoral tissues and ANCTs.

We detected significant down-regulation of SGO1-AS1 in tumoral tissues compared to ANCTs. Expression of this lncRNA has been dys-regulated in lung cancer [8]. However, the mechanism and the consequences of such dys-regulation are not clear. This lncRNA might contribute in the carcinogenesis process through changing the expression of splice variants of SGO1 or through SGO1-independent mechanisms. Future mechanistic studies are required to identify the mechanism. We also detected remarkable association between expression level of this lncRNA and age of breast cancer onset in a way that expression of this lncRNA was mostly down-regulated in samples from older patients. The mechanism of this association is not understood. However, SGO1 has been among whose expression was decreased in polar bodies of aged animals compared with young animals. Moreover, such decreased expression was associated with increased levels of aneuploidy [14]. Future researches are needed to show the relevance of SGO1-AS1 down-regulation and aneuploidy in cancer.

We also reported inverse correlations between expression levels of SGO1 and SGO1-AS1 in both sets of samples. This pattern of correlation is compatible with one of the proposed roles for natural occurring antisense transcripts (NATs) in regulation of sense strands which is as following: As an RNA polymerase II complex moving on a certain DNA strand cannot move in the presence of another complex on the opposite strand, transcription of each of the transcripts will consequently intrude transcription of the other [15].

Finally, we assessed diagnostic power of SGO1-AS1 in breast cancer tissues and reported fair value for its diagnostic power. So in spite of its down-regulation in breast cancer tissues, transcript amounts of this lncRNA cannot be used as appropriate markers in this malignancy. This fact is especially prominent in its low sensitivity value.

Taken together, our study highlights dysregulation of SGO1-AS1 in breast cancer and warrants both mechanistic studies and expression assays in larger sample sizes.