Mutational signatures on ischemic stroke-associated genes in Saudi human genome

1 Introduction

Stroke is a heterogeneous neurological impairment and is the third leading cause of mortality (WHO 2017 Global disease burden data: 12.8 million deaths) and disability worldwide (GBD 2017 Mortality Collaborators, 2018). Although multiple factors including age, sex differences, race, diet, hypertension, smoking, sedentary lifestyle etc. makes one susceptible to stroke, genetics is the most prominent one (Boehme et al., 2017). This neurological syndrome is not a single disease and has multiple types such as Ischemic Stroke (Clots), Hemorrhagic stroke (Bleeds), Transient Ischemic Attack (TIA), Cryptogenic stroke, Brain stem stroke. While Ischemic stroke is more prevalent, Hemorrhagic stroke is highly fatal. Several genetic variation have been associated to with the risks of stroke types or subtypes of Ischemic and Hemorrhagic stroke (Lindgren, 2014). Several challenges in studies and factors that influence the incidence in the Kingdom of Saudi Arabia (KSA) have been reviewed recently (Robert and Zamzami, 2014). One study estimated the stroke incidence rate KSA is 29.8/100,000/year (Al Rajeh et al., 1998). Another study as early as 1993 estimated an annual incidence rate 43.8 per 100,000 (Al Rajeh et al., 1993). Diabetes, Hypertension are some of the major risk factors for Saudi population (El Sayed et al., 1999). Ischaemic stroke was the commonest subtype in all Arab countries which corresponds to 85–90% of all strokes worldwide (Benamer and Grosset, 2009).

Emergence of next generation sequencing methods coupled with technological advances in exome-sequencing and whole genome sequencing techniques serves as a clinical molecular diagnostic testing and prediction of risk of diseases. Genetic basis of disease are explained by the use of genome sequence and mutations such as single nucleotide polymorphisms (SNP) to identify individuals and their risk for disease traits. Genome-wide association studies and several decades of work on stroke genetics on European population created a vast amount of both DNA and RNA sequence data that could be tapped into Saudi nationals.

Mendelian and mitochondrial disorders associated with stroke are well documented (Silliman, 2002). Mutation in single gene HDAC9 could lead to a novel stroke type and this variant increases risk of Ischemic stroke by 42% (Bellenguez et al., 2012). For example, genetic variation in the flanking region of PITX2, ZFHX3 and chromosome 9 short arm region are used for classifying Ischaemic stroke. Mitochondrial genome and stroke are highly interrelated since many mitochondrial disorders are shown to cause stroke related disorders (Anderson et al., 2011). Our OMIM search for stroke shows Ischemic stroke susceptibility loci on genes such as NOS3, F5, AALOX5AP, PRKCH, NOTCH3. Stroke is both single gene and polygenic disorder and availability of comprehensive gene panels for Saudi population can not only provide new avenues for prevention but also improve potential molecular approaches, which may be suitable for the identification of stroke susceptibility genes (Saudi Mendeliome Group, 2015). Similarly frequency of mitochondrial sub haplo groups K affects the risk of IS, hence the Saudi and Arabian peninsula mitochondrial genome should be explored (Abu-Amero et al., 2008).

Despite the better health care services in KSA, a genetic approach is needed for tackling high incidence of stroke as it is both monogenic Mendelian and polygenic disorder. One of the main goals of SHGP (Saudi Human Genome project) is disease prevention. Exploration of non-coding and genic region of the Saudi human genome provides us with novel Saudi specific mutations that needs to be associated with stroke. Also, we conjecture that, by identifying and classifying these mutational signatures, a pan Arabian genetic diversity and susceptibility to stroke can be uncovered. To investigate this, a interdisciplinary approach combined with high-throughput data containing human genomes from middle eastern population to explore stroke associated variations was performed. Whole genome sequencing of humans is one of the unique strategies for identifying variants associated with disease and prediction of risk of certain genetic disorders. The objective of the current study is to provide a comprehensive description of variants in Saudi human genome, in particular mutational signatures on genes associated with ischemic stroke. Upon sequencing a cohort of prospective stroke individuals of Saudi population and comparison to the currently available variant annotation would reveal more on genetics of stroke specific to this population.

2 Methods

3 Results

We have retrieved 29 samples of two groups of Saudi individuals from 137 whole genome sequencing from Amari et al. Depth and coverage of these samples remain almost similar (Fig. 1).

Close

Fig. 1

Scatterplot showing the sequencing depth and sequencing quality of the SNPs. (a) sequencing depth and sequencing quality of the SNPs of individual from groupA population (b) sequencing depth and sequencing quality of the SNPs of individual from groupB population.

Export to PPT

3.1

3.1 Analysis of genetic variation in whole genome of Saudi population

We merged variants from two groups of individuals of Saudi population which resulted in 4,146,365 unique variants. Binning of these variants on individual chromosome and visualization of density histogram on a genome browser show (Fig. 2). These variants are evenly distributed on all chromosomes in both individuals meaning that there is genomic variation similarity (Fig. 3a and b).

Close

Fig. 2

Genome browser view of SNP distribution over 10 kb bins. (a) SNP density before filtering (b) SNP density distribution after filtering and overlapping over Refseq genes collection.

Export to PPT

Close

Fig. 3

Single nucleotide polymorphism counts over each chromosome. (a) Chromosome wise distribution of genetic variation in groupA Saudi population; (b) Chromosome wise distribution of genetic variation in groupB Saudi population; (c) Barplot showing the count of different types of nucleotide substitution observed in groupA Saudi population and (d) Type of nucleotide substitution in groupB Saudi population.

Export to PPT

3.2

3.2 Distribution of substitution type and bias in variation

Distribution of different substitution types for the variants, which passed the filter criteria shows a specific pattern. Most observed SNP conversions are G > A and C > T, which is around 35% of the total variant derived from the substitutions in both groups. Other substitutions such A > C, C > G etc are equally low abundant. A > T and T > A are the least abundant of all substitutions (Fig. 3c and d). Homozygous SNPs which have the same allele on both homologous chromosomes of the non-reference genome, and which is different from the reference genome is almost equally represented as that of heterozygous SNPs (Fig. 4a and b).

Close

Fig. 4

Stacked barplot showing the homozygous SNPs and heterozygous SNPs (a) Allelic variations in groupA Saudi population; (b) Allelic variations in group B Saudi population; (c) Barplot of count of different categories of SNPs based on whole exome based functional annotation using dbNSF4; (d) Top 10 genes harboring most genetic variation resulting from Annovar based annotation of genetic variation using RefSeq database.

Export to PPT

We determined the global nucleotide mismatch rate profile for sequencing reads in each tumor sample across all 6 nt substitution types; A·T > C·G, A·T > G·C, A·T > T·A, C·G > A·T, C·G > G·C and C·G > T·A.

3.5

3.5 Single nucleotide polymorphisms in stroke-associated genes

Both monogenic disorders and polygenic disorders involving genetic variation can lead to stroke. These variants show patterns of Mendelian inheritance. Genes associated with stroke include protein coding ones such as SUPT3H/CDC5L, TSPAN2, NOTCH3, TREX1, HTRA1, GLA, NOS3, F5, AALOX5AP, PRKCH, HDAC9, CDKN2A/CDKN2B, EDNRA, TM4SF4-TM4Sn, LINC01492, PITX2, ZFHX3, ZNF566, PDZK1IP1, RGS7, NKX2-5, HABP2, ADCY2, ANK2, FGA, LOC100505841, CDK6, PDE3A, FURIN-FES, PRPF8, ILF3-SLC44A2, ABO, MMP12, SH2B3, CASZ1, WNT2B, KCNK3, SLC22A7-ZNF318, chr9p21, SH3PXD2A, TBX3, LRCH1, SMARCA4-LDLR, PMF1-SEMA4A, FOXF2, ZCCHC14, COL4A1/COL4A2 and PDE4D. We analyzed these 49 genes for additional mutations. Among this, atleast 13 genes show mutation with amino acid changes in one to multiple positions (Fig. 5). For example, COL4A2 shows missense mutation at 517 and 718 with conversion being R517K and P718S. Similarly, in ZFHX3 two variation which are missense mutation at two different positions, one at 460 and other at 777 resulting in amino acid change of E460Q and V777A.

Close

Fig. 5

Lollipop plot of genes showing SNPs that could alter the function of the resulting protein by changing the aminoacid (missense mutation).

Export to PPT

4 Discussion

Analysis of genome sequencing data from 29 samples of two groups of Saudi individuals from 137 middle eastern populations with around 4.2 million single nucleotide polymorphisms (SNPs) shows that there are large number of genomic variations specific to middle eastern populations. Also, this implies that the available data on genomic variation could be improved by sequencing of human genome of underrepresented ancestry (Scott et al., 2016). Diversifying the genomic data of general population and increasing in cohort size is important for further exploring rare diseases in population of middle eastern ancestry (Abou Tayoun and Rehm, 2020). Substitutions such as G > A and C > T, which is around 35% of the total variant derived from the substitutions in both groups. Other substitutions such A > C, C > G etc are equally low abundant. A > T and T > A are the least abundant of all substitutions. This is observed earlier in other SNP studies reported earlier that the most common change was C > T (or G/A) which is about 70.2%. This may arise from the deamination of 5-methylcytosine at CpG dinucleotides (Dawson et al., 2001).

Genetic contribution to Ischemic stroke risk have been well studied (Rubattu et al., 1996). Ischemic stroke is a manifestation of both monogenic and polygenic disorders with variants on genes being shared among different ancestry. Candidate gene variants detected earlier which are increase in risk of stroke are shared among European and Non-european population, including Chinese population (Ariyaratnam et al., 2007). Our literature survey with resulted in atleast 49 genes involved in risk of stroke. The proportion of strokes of undetermined or rare causes is much higher for young adults and in many cases underlying causes are genetic-related (Jood et al., 2005). We observed that SNPs are extensively distributed in genes which are large and involved in multiple functions. Genome-wide studies on variation on genes also show that these genes are involved in other diseases such as parkinsonian syndrome (Lang et al., 2019).

Variation in genome explains the differences among population such as blood group and susceptibility to diseases (De Mattos, 2013). Monogenic (variation in single gene) or polygenic disorders could be associated to splice-altering variants on protein-coding gene(s) (Vuckovic et al., 2020). Genome sequencing, variant detection and annotation of genotypes are crucial steps in the evaluating, predicting, and concluding diseases in humans (Ku et al., 2010). For example, heterozygous variation in chr17 at position 44,352,876 that we identified in Saudi population is annotated in ClinVar as directly involved in temporal lobar degeneration resulting in susceptibility to Ischemic stroke. In addition, annotation and classification of stroke associated mutation profile and the clusters with similar mutation profiles would help in identifying new subtypes of stroke. Based on this specific treatment strategy differs. Our work contributes to understanding the genetics of stroke and predicts the genomic loci susceptible to stroke. Further work is required for individuals who are susceptible to stroke with a therapy-selection in the future. The result of this study on annotation of variants against clinical database is a valuable contribution as precision medicine based on genome of an individual is the future of health care in Saudi Arabia.

5 Conclusion

A fundamental aspect of this project is to explore the genomic data to enlighten the stroke research using interdisciplinary exploratory approach. Whole genome sequencing and annotation of variation of population of Gulf region including Kuwait, Qatar, and Saudi Arabia expand our understanding of human genetics and diseases in general. Integrating genome sequencing and computational approaches could provide the ability to predict the on-set of stroke and to make possible an early intervention.