A: Sampling and DNA extraction

The family was recruited from a rural part of Sindh province of Pakistan, after getting prior approval from Institutional Review Board (Quaid-I-Azam U IRB#1- Biomedical; IORG0002926; IRB00003532), and blood samples were taken from available affected and unaffected family members and DNA was extracted from whole blood by following the standard proteinase K/phenol/chloroform isolation method. Written informed and photography consents (Translated in local Urdu language) were obtained from the parents/guardians of patients participating in this study, which conforms to Helsinki Declaration and local legislation.

B: Clinical Assessments

Affected family members were evaluated with the help of standard questionnaire (translated and amended version of Wechsler Intelligence Scale in Urdu) for severity of disease and IQ assessments. Photographs of affected members were taken with written consent to study and publish their facial features. Two individuals from two different loops of the pedigree were selected for cranial CT scan (IV-10 and IV-14) to screen for abnormal brain anatomical features. For information on disease onset, parents were interviewed about the prenatal, perinatal and neonatal medical history of the proband.

C: Molecular Assessments

Clinical description

The family under examination consisted of six affected individuals from three different extended branches (Figure 1) and who suffer from cognitive impairment with an IQ in the range of severe ID (25-35). Prenatal, perinatal and neonatal medical histories of all affected individuals were normal. Ophthalmological and otorhinolaryngological findings were also normal. Cranial computed tomography (CT) performed on the patients from two separate loops did not reveal any brain dysmorphology or neurologic symptoms (Data not shown). No facial dysmorphism was observed except for affected individual IV-10 who had strabismus (Figure 2). In individual IV-15 cognitive impairment was also accompanied by a form of muscular dystrophy, so this individual was not treated as affected for the purposes of the analysis which was later supported by the homozygosity and CNV data. The clinical information of the affected individuals (on the basis of questionnaire) is presented in Table 1.

Molecular studies

(i) Genome-wide Homozygosity mapping and CNV analysis

Genome wide homozygosity mapping revealed homozygosity-by-descent (HBD or autozygosity) among four affected individuals (IV-11, IV-12, IV-13) on 8p23 (Figure 3) between SNPs rs6989820 and rs2237834, which delineates a critical region of 12.494 Mb {UCSC genome Browser, May 2004 (NCBI35/hg17)}. In order to confirm segregation of the identified HBD region in the entire family STS markers D8S1781, D8S262, D8S518, D8S1140, D8S277, D8S351, D8S1469, D8S1721, D8S550, D8S552, D8S1106, D8S1790, D8S1731, D8S1145 and D8S298 were genotyped. Haplotypes were generated and are presented in Figure 1, which also indicate the segregation of a minimum critical region flanked by above mentioned SNPs. Linkage analysis yielded two-point and multipoint LOD scores above 3 and 5 respectively at several markers (Table 2). This region contained a total of 96 known genes {UCSC Genome Browser, May 2004 (NCBI35/hg17)} including MCPH1 and TUSC3, which have already been reported to be involved in microcephaly with ID. Initially MCPH1 gene was sequenced to detect pathogenic mutation in this family but the sequence analysis only revealed the presence of known SNPs in exon 1 (rs2305023), 6 (rs2442513) and 8 (rs930557 and rs2920676). Subsequent CNV analysis of microarray data showed homozygous deletion of 170.673 Kb region in all affected individuals. The deletion encompassed almost the entire TUSC3 gene (minus the promoter and first exon) and its downstream region.

Markers	Genetic Position in cM (Rutgers map, build 36)	Physical Position in bp {Feb.2009 (GRCh37/hg19}	Two point LOD Score	Multipoint LOD Score
D8S1781	6.8	3678400	2.2342	4.2286
D8S262	7.13	3777275	1.0814	4.2143
D8S518	9.88	4587958	0.5706	1.456
rs6989820	-	5041417	-3.8284	-4.0196
D8S1140	11.96	5617059	3.1114	2.9368
D8S277	14.92	6616946	2.9565	5.0000
D8S351	18.94	8877155	3.0286	5.1617
D8S1469	19.38	9090104	1.523	5.1693
D8S1721	19.7	10240822	1.8241	5.1717
D8S550	20.85	10981913	- Infinity	- Infinity
D8S552	24.76	12842458	2.9823	5.1646
D8S1106	24.76	12936149	3.2113	5.1646
D8S1790	26.35	13166462	3.2843	5.1229
D8S1731	27.95	15338590	1.3825	4.5748
rs2237834	-	17479493	-3.8284	-4.4686
D8S1145	32.78	18452816	- Infinity	- Infinity
D8S298	40.11	21862145	- Infinity	- Infinity

(ii) Deletion breakpoint mapping

The CNV analysis of microarray data revealed that in all affected individuals no hybridization signals were generated for 22 SNPs on chromosome 8; from rs4258002 to rs352769 as borderline deleted SNPs. With the evidence of microarray data of four affected individuals, size of deletion fragment was mapped by PCR amplification and sequencing of junction fragment by designing primer (forward primer 5 '-TGCTCTCTGCTCTTCCTCGT-3'; reverse primer: 5'-CTTTCCTGGCAAGCTGCTAC-3') between deleted and intact SNPs (between rs10094375 and rs4258002 towards the distal end, and rs352769 and rs6530906 proximally). The BLAT search for the junction fragment sequence against human genome revealed the actual physical co-ordinates of deletion as being between 15521688 bp to 15692362 bp, and spanning a region of 170.673 Kb. The deletion was also checked in all available family members by using TUSC3 exon 3 and 5 and junction fragment primers. Both of these primer sets showed amplification in the normal individuals which represented the heterozygous segregation of this deletion except in individual IV-15. All the affected individuals of this family were homozygous for the deletion, and no homozygous or heterozygous deletion carriers were identified among 276 unrelated healthy Pakistani individuals. This deletion encompassed almost the entire TUSC3 gene (minus the promoter and first exon) and its downstream region (Figure 4). According to the database of genomics variants [21] 57 CNVs have been reported in HBD region identified in this family, but 27 involve TUSC3 gene. The detailed analysis of the CNV data indicates that all the presently identified TUSC3 involving CNVs exist in heterozygous state.

Discussion and Conclusion

Autosomal recessive NSID accounts for ~25% of genetic ID cases and may be more common than X linked cases, however, the molecular basis of AR-NSID is relatively poorly known because of clinical and genetic heterogeneity, and the absence of distinguishing clinical criteria. In non-syndromic ID, cognitive deficit is the sole clinical feature among the patients. It is suggested that ID may be caused by the disruption of the biological and molecular processes in the nervous system such as neuronal differentiation and synaptic plasticity, synaptic vesicles cycling and gene expression, regulation profiling etc [22]. There are number of co-translational and post-translational modifications required for protein stability and proper protein folding into its 3D structure, which are essential for normal protein function. One such post-translational modification, N-glycosylation, has previously reported to be involved in non-syndromic X-linked ID at a TUSC3 paralogous gene, IAP [9].

In the current study we analyzed a four generation Pakistani family with 6 affected individuals having severe psychomotor developmental delay, segregating autosomal recessive mode of inheritance. The major clinical presentations of all patients were normal except for the cognitive dysfunction. However comparison of the clinical data (Biometric and neurological data) of our family with the patients of earlier reports did not reveal any significant difference in the phenotypic expression [9, 10]. TUSC3 is the 5^th gene in autosomal recessive NSID and the 2^nd in which a 3^rd mutation has been identified, after TRAPPC9. Our study will aid in diagnostic assessment of AR-NSID individuals.