Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Both QKI and TMEFF2 genes are considered putative tumor suppressor genes (TSGs). In gastric (GC) and colorectal (CRC) cancers, downregulation of their expressions is known to be frequent. However, QKI and TMEFF2 mutations that could potentially inactivate their functions are not reported in cancers.

METHODS:

In a genome database, we observed that both QKI and TMEFF2 harbor mononucleotide repeats, which could be mutated in cancers with high microsatellite instability (MSI-H). For this, we studied 79 GCs and 124 CRCs for the mutations and their intratumoral heterogeneity (ITH).

RESULTS:

Six of 34 GCs (17.6%) and 10 of 79 CRCs (12.7%) with MSI-H exhibited QKI frameshift mutations while five of 79 CRCs (6.3%) with high MSI (MSI-H) exhibited TMEFF2 frameshift mutations. However, we found no such mutation in microsatellite stable/low MSI (MSS/MSI-L) cancers within the mononucleotide repeats. We also studied ITH for the detected frameshift mutations in 16 cases of CRCs and detected that QKI and TMEFF2 frameshift mutations showed regional ITH in 2 (12.5%) and 1 (6.3%) cases, respectively.

CONCLUSIONS:

Our data show that candidate TSG genes QKI and TMEFF2 harbor mutational ITH as well as the frameshift mutations in GC and CRC with MSI-H. From this observation, frameshift mutations of QKI and TMEFF2 may play a role in tumorigenesis through their TSG inactivation in GC and CRC.

Keywords

QK1 TMEFF2 tumor suppressor gene methylation frameshift mutation cancer microsatellite instability

1. Introduction

QKI gene encodes an RNA-binding protein that regulates diverse mRNA functions such as splicing, export and stability of mRNA [1]. QKI regulates target mRNAs such as p27, beta-catenin and c-fos that are related to cancer development [2, 3, 4]. Aberrant expression of QKI has been reported in many tumors [5, 6, 7]. For example, QKI gene is silenced in gastric (GC) and colorectal (CRC) cancers [3, 7]. QKI overexpression inhibits proliferation of GC and CRC cells [3, 7]. These reports suggest that QKI has tumor suppressor gene (TSG) activity in multiple cancers. TMEFF2, also known as HPP1, encodes the tomoregulin family of transmembrane proteins [8]. Loss of TMEFF2 expression by promoter hypermethylation has been reported in many tumors, including GC, CRC and breast cancers [8, 9, 10]. TMEFF2 expression promotes apoptosis and induces cell cycle arrest [11, 12], indicating that TMEFF2 gene is a candidate TSG as well.

About one third of GC and CRC are classified as high microsatellite instability (MSI-H) cancers [13]. Many TSGs harbor frameshift mutations at monocleotide repeats in MSI-H cancers [13]. In the human genome database, we observed that both QKI and TMEFF2 genes possess nucleotide repeats in coding sequences that might be mutated in MSI-H cancers. Intratumoral heterogeneity (ITH) is a common phenomenon in cancers, which may result in cancer evolution and influence on clinical outcomes [14, 15, 16]. Thus, identification of genetic ITH is important in understanding biological and clinicopathologic features of the cancers. The current study aimed to find whether QKI and TMEFF2 frameshift mutations are common and harbor ITH in MSI-H GC and CRC.

Table 1
Summary of pathologic features of gastric and colorectal cancers

Feature	MSI-H	MSS/MSI-L
Gastric carcinomas
Total cases	34	45
TNM stage
I	13	15
II	13	18
III	7	11
IV	1	1
Lauren’s subtype
Diffuse	4	21
Intestinal	20	15
Mixed	3	6
Indeterminate	7	3
EGC vs. AGC
EGC	3	4
AGC	31	41
Colorectal carcinomas
Total cases	79	45
TNM stage
I	15	6
II	29	20
III	32	16
IV	3	3
Location
Cecum	16	0
Ascending colon	46	3
Transverse colon	12	2
Descending and sigmoid colon	4	17
Rectum	1	23

EGC: early gastric cancer, AGC: advanced gastric cancer, TNM: tumor, lymph node, metastasis, MSI-H: high microsatellite instability, MSI-L: low microsatellite instability, MSS: stable microsatellite instability.

2. Materials and methods

2.1 Tissue samples and microdissection

In this study, we used 79 GCs and 124 CRCs. The GCs were 34 GCs with MSI-H, 45 GCs with microsatellite stable/low MSI (MSS/MSI-L), 79 CRCs with MSI-H and 45 CRCs with MSS/MSI-L (Table 1). These samples overrepresent MSI-H cases, because we collected MSI-H and MSS/MSI-L cases separately in different times. We adopted an MSI evaluation system using five mononucleotide repeats [17]. In brief, we used multiplex PCR for five quasi-monomorphic mononucleotide repeats (BAT25, BAT26, NR21, NR24, and NR27). Cancers with no allelic size variations in fewer than two of the microsatellites were considered MSS/MSI-L, while those with size variations in more than two microsatellite markers were considered MSI-H. For 16 MSI-H CRCs, we picked 4–7 tumor areas and one normal area per CRC for the ITH analysis. Each ITH area was histologically confirmed under light microscope. These ITH areas were studied for detecting mutational ITH of QKI and TMEFF2. Pathologic features of the cancers were evaluated in all blocks of all cases by a pathologist and are summarized in Table 1. Tumor and normal cells were microdissected as described previously [18, 19]. Approval of this study was obtained from the institutional review board of the Catholic University of Korea.

Figure 1.

Representative SSCP and DNA sequencings of repeats in QKI and TMEFF2 in colon and gastric carcinomas. SSCP (A) and DNA sequencing analyses (B) of the A7 repeat in QKI (left) and the A8 repeat in TMEFF2 (right) from tumor (Lane T) and normal tissues (Lane N). A. SSCPs of the PCR products show aberrant bands (arrows in lane T) as compared to SSCP from normal tissues (N).B. Direct DNA sequencing analyses show a heterozygous A deletion (left) and a heterozygous A duplication (right) in the tumor tissues as compared to normal tissues.

Table 2

Summary of QKI and TMEFF2 mutations in gastric and colorectal cancers

Gene	Location	Wild type	Mutation	MSI status of the	Incidence in	Nucleotide change
				mutation cases ( $n$ )	MSI-H cancers (%)	(predicted amino acid change)
QKI	Exon 3	A8	A7	MSI-H (15)	Colorectal: 9/79 (11.4)	c.401delA (p.Lys134Argfsx14)
					Gastric: 6/34 (17.6)
		A8	A6	MSI-H (1)	Colorectal: 1/79 (1.3)	c.401delA (p.Lys134Glyfsx5)
TMEFF2	Exon 9	A7	A6	MSI-H (4)	Colorectal: 4/79 (5.1)	c.908delA (p.Lys303Argfsx19)

MSI-H: high microsatellite instability.

Table 3

Intratumoral heterogeneity of QKI and TMEFF2 frameshift mutations in colorectal cancers with MSI-H

Case	Regional biopsy sites							Mutation status	ITH status
	#1	#2	#3	#4	#5	#6	#7
CRC3	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	n.d.	Wild type	Non-ITH
CRC15	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Non-ITH
CRC26	Wild type	TMEFF2 c.908delA	n.d.	Wild type	Wild type	Wild type	Wild type	TMEFF2 mutation	ITH
CRC27	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Non-ITH
CRC34	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Non-ITH
CRC35	Wild type	Wild type	n.d.	n.d.	n.d.	Wild type	Wild type	Wild type	Non-ITH
CRC39	Wild type	Wild type	Wild type	Wild type	n.d.	Wild type	Wild type	Wild type	Non-ITH
CRC41	Wild type	n.d.	Wild type	Wild type	n.d.	Wild type	Wild type	Wild type	Non-ITH
CRC43	Wild type	Wild type	Wild type	n.d.	n.d.	Wild type	n.d.	Wild type	Non-ITH
CRC45	QKI c.401delA	QKI c.401delA	QKI c.401delA	QKI c.401delA	Wild type	QKI c.401delA	Wild type	QKI mutation	ITH
CRC47	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Wild type	Non-ITH
CRC48	WT	n.d.	n.d.	WT	WT	WT	WT	Wild type	Non-ITH
CRC49	n.d.	WT	WT	QKI c.401delA	WT	WT	WT	QKI mutation	ITH
CRC51	WT	WT	WT	WT	WT	WT	WT	Wild type	Non-ITH
CRC53	WT	WT	WT	WT	WT	WT	WT	Wild type	Non-ITH
CRC55	WT	WT	n.d.	n.d.	WT	WT	WT	Wild type	Non-ITH

ITH: intratumoral heterogeneity, MSI-H: high microsatellite instability, n.d.: not done.

2.2 Single strand conformation polymorphism (SSCP) analysis

We analyzed DNA sequences in QKI (A8 in exon 3) and TMEFF2 (A7 in exon 9). Genomic DNA was amplified using polymerase chain reaction (PCR). [ ${}^{32}$ P]dCTP was incorporated to the PCR products for visualization in autoradiogram. We determined aberrant gel motility in the SSCP (FMC Mutation Detection Enhancement system; Intermountain Scientific, Kaysville, UT, USA) using visual inspection, which subsequently sequenced by Sanger DNA sequencing (3730 DNA Analyzer, Applied Biosystem, Carlsbad, CA, USA). Other procedures in detail were described in our earlier studies [18, 19, 20].

3. Results

3.1 Mutational analysis

Genomic DNAs of the 79 GCs and 124 CRCs were studied to detect frameshift mutations in QKI (A8 in exon 3) and TMEFF2 (A7 in exon 9) mononucleotide repeats. In SSCP, we observed aberrant bands in 20 cases among the 79 GCs and 124 CRCs (Table 2 and Fig. 1). Normal tissues of the corresponding patients revealed no aberrant migration in SSCP, which indicated that the aberrant bands had risen somatically (Fig. 1). Sanger sequencing of the cancer tissues revealed that they were frameshift mutations of QKI and TMEFF2 (Fig. 1). All detected mutations of QKI and TMEFF2 were considered heterozygous mutations by SSCP and direct sequencing patterns (Fig. 1). These mutations were deletion mutations, which would lead to premature termination of QKI or TMEFF2 translation (Table 2). Ten of 79 CRCs (12.7%) and 6 of 34 GCs (17.6%) with MSI-H exhibited QKI mutations while 5 CRCs (6.3%) with MSI-H exhibited TMEFF2 mutations. Any cancers in this study did not harbor both QKI and TMEFF2 frameshift mutations simultaneously.

Figure 2.

Intratumoral heterogeneity of QKI and TMEFF2 frameshift mutations in colon cancers. A: Direct DNA sequencing analyses show QKI c.401delA mutation (MT) in 5 regional areas (45-1, -2, -3, -4 and -6) and wild-type (WT) in the other 2 areas (45-5 and -7). B: Direct DNA sequencing analyses show TMEFF2 c.908delA mutation (MT) in one regional area (26-2) and wild-type (WT) in the other 5 areas (26-1, -4, -5, -6 and -7).

QKI and TMEFF2 frameshift mutations were found in MSI-H cancers, but there was none in MSS/MSI-L cancers (Table 2) (Fisher’s exact test, $p<$ 0.001). There was no significant difference in the mutation prevalence between CRCs (15/79) and GCs (6/34) ( $p=$ 0.547). Disease free survival (DFS) rates of our QKI-mutated and TMEFF2-mutated cases at 5-years were 62.5% and 60.0%, respectively. Overall survival (OS) rates of them at 5-years were 75.0% and 80.0%, respectively. There was no significant difference either in DFS or OS between QKI-mutated and QKI-wild-type MSI-H cancers ( $p>$ 0.05). Also, there was no significant difference of them between TMEFF-mutated and TMEFF wild-type MSI-H cancers ( $p>$ 0.05). We also analyzed the relationship of each immunostaining with various pathologic parameters (differentiation, depth of invasion and TNM stage), but there was no significant difference ( $p>$ 0.05).

3.2 Intratumoral heterogeneity of QKI and TMEFF2 frameshift mutations

Ninety-six areas from 16 CRCs with MSI-H were analyzed to find the ITH in QKI and TMEFF2 frameshift mutations. In two (#45 and 49) of the 16 CRCs (12.5%), the A8 repeat in QKI exhibited one base-pair deletion mutation in one or more regional areas (Table 3 and Fig. 2). In one (#26) of the 16 CRCs (6.3%), the A7 repeat in TMEFF2 exhibited one base-pair deletion (A7 to A6) mutation in one of the 6 regional fragments. There was no significant histological difference between mutation ( $+$ ) and mutation ( $-$ ) ITH areas.

4. Discussion

Inactivation of a TSG by somatic mutations is known to lead the cell to cancer development in combination with other genetic changes. Earlier studies identified that both QKI and TMEFF2 possess TSG activities in cells [3, 7, 11, 12]. Both QKI and TMEFF2 are known to lose their activities by expression losses through promoter hypermethylation in GC and CRC [3, 7, 8, 9]. In the COSMIC database that encompasses overall mutation incidences to date, the QKI recurrently has hotspot p.K134 frameshift mutations in GC ( $n=$ 5) and CRC ( $n=$ 23). The TMEFF2 has the hotspot p.K303 frameshift mutation in CRC ( $n=$ 3), but none in GC. Such validation in the public data may indicate that the recurrent frameshift mutations of QKI and TMEFF2 may be present irrespective of regions and ethnicities of the cancer patients. QKI are recurrently A8 of PLA2R1 and the A7 of SRPK1 harbored deletion mutations in 3 (one GC, one CRC and one melanoma) and one (one CRC) cases, respectively. However, there has been no report on inactivating mutations of QKI and TMEFF2 TSGs in GC and CRC. This study by us identified that both MSI-H GC and CRC exhibited QKI (17.6% of GC and 12.7% of CRC) and TMEFF2 (6.3% of CRC) frameshift mutations. These results indicate that QKI and TMEFF2 TSGs are altered not only by protein losses as shown previously [3, 7, 8, 9] but also by frameshift mutations identified in this study and suggest that QKI and TMEFF2 inactivating mutations might inhibit their TSG functions in MSI-H GC/CRC as well. Thus, QKI and TMEFF2 frameshift mutations might play a role in the pathogenesis of MSI-H GC/CRC by inhibiting their TSG activities.

Two opposite activities (TSG and oncogenic) of TMEFF2 have been identified [8, 9, 21]. In prostate cancers, TMEFF2 promotes cell proliferation in an ADAM17-depencdent manner [22]. By contrast, in GC and CRC, TMEFF2 expression is decreased compared to the normal cells [8, 9]. In addition, loss of TMEFF2 expression is related to the tumor stages and poor clinical outcomes of GC patients [22]. TMEFF2-mediated tumor suppression requires activation of STAT1/2 pathways that induces G0/G1 phase cell cycle arrest [11, 12]. These earlier data may indicate that TMEFF2 function either oncogenic or TSG is context- and cell type-dependent. Together, TMEFF2 may behave as a TSG that is frequently altered by frameshift mutations as well as expression loss in GC and CRC. The QKI and TMEFF2 frameshift mutations may truncate and disable QKI and TMEFF2 proteins and hence be considered loss-of-function mutations. Expression of QKI increases the enterocyte differentiation markers in CRC cells along with p27 protein level and beta-catenin, and inhibits cell proliferation [3]. The truncated mutants of QKI predicted by our data may possibly inactivate their TSG function (maintenance of cell differentiation and inhibition of proliferation), but the cellular consequences as well as clinical implications remain to be further clarified.

In this study, we identified mutational ITH of QKI or TMEFF2 in 3 of 16 CRCs (19%), which is consistent with previous studies that had reported frequent mutational ITH in CRCs with MSI-H [23]. It is well known that cancer ITH of driver mutations could result in poor clinical outcomes in patients. For instance, single or a group of mutations with a metastasis potential might redirect clinical outcomes since such rare clones may accomplish dominance during tumor progression [14, 15, 16]. Based on the TSG properties of QKI and TMEFF2 in CRC, our results suggest that loss of the TSG properties may have regional heterogeneity that could be further selected and influence the clinical outcomes. However, it should be further studied how the ITH change during the progression and the ITH play a causal role in tumor pathogenesis.

TMEFF2 has EGF-like domain that is known to bind with ERBB4 [24]. Given the importance of oncogenic erbB proteins in cancer therapies, further understanding of TMEFF2 alterations in our data (mutational inactivation in MSI-H GC and CRC) here may provide useful information for its potential clinical application. In these points, our data may reveal information on biological mechanisms as well as therapeutic strategies in cancer cells.

Footnotes

Acknowledgments

This study was supported by a grant from National Research Foundation of Korea (2012R1A5A2047939).

Conflict of interest

The authors declare no competing interests.

References

Darbelli

and Richard

, Emerging functions of the Quaking RNA-binding proteins and link to human diseases, Wiley Interdiscip Rev RNA 7 (2016), 399–412.

Deschênes

Vézina

Beaulieu

J.F.

et al., Role of p27(Kip1) in human intestinal cell differentiation, Gastroenterology 120 (2001), 423–438.

Yang

Zhang

et al., RNA-binding protein quaking, a critical regulator of colon epithelial differentiation and a suppressor of colon cancer, Gastroenterology 138 (2010), 231–240.

Yang

Wang

et al., E2F1 and RNA binding protein QKI comprise a negative feedback in the cell cycle regulation, Cell Cycle 10 (2011), 2703–2713.

Mulholland

P.J.

Fiegler

Mazzanti

et al., Genomic profiling identifies discrete deletions associated with translocations in glioblastoma multiforme, Cell Cycle 5 (2006), 783–791.

Ichimura

Mungall

A.J.

Fiegler

et al., Small regions of overlapping deletions on 6q26 in human astrocytic tumours identified using chromosome 6 tile path array-CGH, Oncogene 25 (2006), 1261–1271.

Bian

Wang

et al., Downregulation of tumor suppressor QKI in gastric cancer and its implication in cancer prognosis, Biochem Biophys Res Commun 422 (2012), 187–193.

Young

Biden

K.G.

Simms

L.A.

et al., HPP1: A transmembrane protein-encoding gene commonly methylated in colorectal polyps and cancers, Proc Natl Acad Sci USA 98 (2001), 265–270.

Shibata

D.M.

Sato

Mori

et al., Hypermethylation of HPP1 is associated with hMLH1 hypermethylation in gastric adenocarcinomas, Cancer Res 62 (2002), 5637–5640.

10.

Suzuki

Sunaga

Shames

D.S.

et al., RNA interference-mediated knockdown of DNA methyltransferase 1 leads to promoter demethylation and gene re-expression in human lung and breast cancer cells, Cancer Res 64 (2004), 3137–3143.

11.

Elahi

Zhang

Yeatman

T.J.

et al., HPP1-mediated tumor suppression requires activation of STAT1 pathways, Int J Cancer 122 (2008), 1567–1572.

12.

Hernandez

J.M.

Elahi

Clark

et al., The tumor suppressive effects of HPP1 are mediated through JAK-STAT-interferon signaling pathways, DNA Cell Biol 34 (2015), 541–549.

13.

Imai

and Yamamoto

, Carcinogenesis and microsatellite instability: The interrelationship between genetics and epigenetics, Carcinogenesis 29 (2008), 673–680.

14.

Marusyk

Almendro

and Polyak

, Intra-tumour heterogeneity: A looking glass for cancer? Nature Reviews Cancer 12 (2012), 323–334.

15.

McGranahan

and Swanton

, Biological and therapeutic impact of intratumor heterogeneity in cancer evolution, Cancer Cell 27 (2015), 15–26.

16.

Kim

T.M.

Jung

S.H.

Baek

I.P.

et al., Regional biases in mutation screening due to intratumoural heterogeneity of prostate cancer, J Pathol 233 (2014), 425–435.

17.

Murphy

Zhang

Geiger

et al., Comparison of the microsatellite instability analysis system and the Bethesda panel for the determination of microsatellite instability in colorectal cancers, J Mol Diagn 8 (2006), 305–311.

18.

Kim

M.S.

Yoo

N.J.

and Lee

S.H.

, Preferential occurrence of RHOA mutation in gastric and colorectal cancers, Pathology 47 (2015), 598–599.

19.

H.R.

C.H.

Yoo

N.J.

et al., Frameshift mutations of TAF1C gene, a core component for transcription by RNA polymerase I, and its regional heterogeneity in gastric and colorectal cancers, Pathology 47 (2015), 101–104.

20.

Y.S.

Kim

M.S.

Yoo

N.J.

et al., NSD1 encoding a histone methyltransferase exhibits frameshift mutations in colorectal cancers, Pathology 48 (2016), 284–286.

21.

Ali

and Knaüper

, Phorbol ester-induced shedding of the prostate cancer marker transmembrane protein with epidermal growth factor and two follistatin motifs 2 is mediated by the disintegrin and metalloproteinase-17, J Biol Chem 282 (2007), 37378–37388.

22.

Sun

Xiong

et al., TMEFF2 deregulation contributes to gastric carcinogenesis and indicates poor survival outcome, Clin Cancer Res 20 (2014), 4689–4704.

23.

Calin

G.A.

Gafà

, Genetic progression in microsatellite instability high (MSI-H) colon cancers correlates with clinico-pathological parameters: A study of the TGRbetaRII, BAX, hMSH3, hMSH6, IGFIIR and BLM genes, Int J Cancer 89 (2000), 230–235.

24.

Uchida

Wada

Akamatsu

et al., A novel epidermal growth factor-like molecule containing two follistatin modules stimulates tyrosine phosphorylation of erbB-4 in MKN28 gastric cancer cells, Biochem Biophys Res Commun 266 (1999), 593–602.

Frameshift mutation of candidate tumor suppressor genes QK1 and TMEFF2 in gastric and colorectal cancers

Abstract

BACKGROUND:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

Table 1 Summary of pathologic features of gastric and colorectal cancers

2.1 Tissue samples and microdissection

3. Results

3.1 Mutational analysis

4. Discussion

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Summary of pathologic features of gastric and colorectal cancers