Genetic Diagnosis of Hyperoxaluria Type 3 Patients Using Haplotype Analysis

Primary hyperoxaluria (PH) is an autosomal recessive genetic disease of the metabolism of glyoxylate, which results in excessive oxalate formation [1‒3]. PH has 3 types and the most prevalent form is PH type I, which results from a mutation in the AGXT gene (AGT) [4‒6]. Primary hyperoxaluria type 2 (PH II), caused by deficiency of the enzyme glyoxylate reductase/hydroxypyruvate reductase (GRHPR), and PH III result from mutations in the HOGA1 gene. Numbers of research indicated that PH III is more common than PH II [7]. PH patients usually reveal polyurea, dysuria; and rising urine oxalate levels may appear in childhood and adulthood [1, 8‒11].

HOGA1 gene, which is located on the 10q24.2, contains 7 exons and is translated into the mitochondrial protein of 4-hydroxy-2-oxoglutarate aldolase (HOGA1), which is also known as dihydrodipicolinate synthase like (DHDPSL). This protein consists of 327 amino acids, 17 helices, 8 B strands, and 3 turns and it also has 2 dimers; it produces a tetrameric structure [12].

There are numerous techniques to map the mutation location such as haplotype analysis [13]. Haplotype analysis uses the fact that patients who are born to consanguineous marriages or from a small geographical area probably inherit two recessive copies of a mutant allele from a common ancestor. In this study, Sanger sequencing was done followed by haplotype analysis utilizing SNP markers on individuals suspected of being affected by PH III.

Fourteen patients from 11 unrelated families who were determined to have PH III were referred to the Ali Asghar Children’s Hospital. Consent forms were obtained from all participants before sampling. This study was approved by the Iran University of Medical Sciences (IUMS) research committee (IR.IUMS.FMD.REC.1401.149). PH I was ruled out in all studied patients using Sanger sequencing of the AGXT gene. Patients’ clinical information is summarized in Table 1.

DNA was extracted from whole blood samples by salting-out approach [14]. Haplotype analysis was performed using SNPs. Four different SNPs surrounding the HOGA1 gene were selected including rs2275090, rs1124116, rs3750614, and rs2296438. ARMS-PCR primers were designed for each of our selected SNPs to genotype them. Primers for all coding exons and exon-intron boundaries of the HOGA1 gene were designed. Sanger sequencing of this gene was performed on those affected individuals who showed the homozygous haplotypes.

Different in silico tools such as Combined Annotation Dependent Depletion (CADD) [15], MutationTaster [16], Mutation Surveyor [17], Have Our Protein Explained (HOPE), Protein Variation Effect Analyzer (PROVEN) [18], Sorts Intolerant From Tolerant (SIFT) [19], Polymorphism Phenotyping (PolyPhen) [20], and Genomic Evolutionary Rate Profiling (GERP) [21] were applied for variant analysis and finally they were interpreted according to American College of Medical Genetics and Genomics (ACMG) guideline [22].

One of our studied patients (F3) had a large inbred family with lots of members affected with stone kidney. Sanger sequencing of the AGXT gene did not reveal any causative variant in the family’s proband. The haplotype analysis for the HOGA1 gene did not show homozygous haplotypes. This patient was selected to be tested by NGS. Whole exome enrichment was performed using an Agilent sure select V7 target enrichment kit and the library was sequenced on the Illumina HiSeq 6000 platform with an average depth of 100X for target regions. Nearly, all exons and flanking 10 bp were covered. Paired-end reads are aligned to the NCBI reference sequence (GRCh37). Bioinformatics analysis of the sequencing results was performed using international databases and standard bioinformatics software. The analysis was performed with emphasis on the variants within more than 3,500 genes with the phenotype-causing mutation in the “Online Mendelian Inheritance in Man” (OMIM; 2024-09-01) catalog.

Our study was performed on 14 unrelated patients who were suspicious of being affected with PH III from different regions of Iran. All of the patients were aged from 1 to 5 years. Overall, 64.2% of them are from consanguineous marriages, and the remaining are from the same ethnicity. The median age at onset of clinical symptoms was 2.93 (range: 1.5–5.5) years. Figure 1 shows the pedigrees of our studied patients.

Prior to genetic analyses like ruling out of PH1, haplotype analysis for the HOGA1, and Sanger sequencing of this gene, clinical and paraclinical evidence of the patient F11 was highly consistent with PH III: kidneys were normal in size and parenchymal echogenicity. They had a few scattered stones, but no hydronephrosis was seen. The ratio of oxalate to creatinine was high in this patient [23, 24].

The result of the haplotype analysis of 14 affected individuals along with their parents showed that only 2 patients (F4 and F11) had the homozygous haplotypes (Fig. 2). Sanger sequencing was done in the patients of both families. A homozygous mutation of c.266G>A (p.Arg89His) has been detected in F11, which was found in exon 2 of the HOGA1 gene. As shown in Figure 3, her parents were heterozygous for this variant. The result of ARMS-PCR of 4 SNPs in this family along with controls is shown in Figure 4.

We performed NGS on the proband of the family of F3. No pathogenic or likely pathogenic variant related to the patients’ phenotype was found after analysis. More analysis showed some (run of homozygosity) ROH regions in the affected individual, which are shown in Table 2. The genes whose mutations potentially cause kidney stones are also revealed in Table 2.

DANN score is a pathogenicity scoring methodology and it ranges from 0 to 1 and score 1 is predicted to be the most damaging. CADD score is a tool for scoring the deleteriousness of variants in which a score of greater than or equal to 20 is demonstrated to be the 1% most deleterious (Table 3). According to the ACMG guideline, this variant is a likely pathogenic one.

PH III is the less severe form of PH with a milder phenotype and good prognosis in most patients [25]. The frequency of each type of mutation based on the HGMD database in this gene is 77% missense/nonsense and splice site/deletions are 16% and the rest are indels [26].

The identified variant (c.266G>A[p.Arg89His]) in the present study is a missense one in the HOGA1 gene, which is the first report in the Iranian population (Fig. 5). According to the ACMG guideline, this variant is likely pathogenic because (1) it was not found in the gnomeAD and ExAC database (PM2), (2) several in silico tools such as BayesDel addAF, MetaRNN, DANN, DEOGEN2, FATHMM-XF, M-CAP, and MetaRNN revealed it could be a disease-causing variant (Table 3). Based on the HOPE [27], the mutant residue is smaller than the wild-type residue and the size difference between them puts the new residue in the incorrect position to make the hydrogen bond as the original wild-type residue did. Moreover the wild-type residue charge was positive whereas the mutant residue charge is neutral (PP3). (3) This variant is found in a homozygous state (PM3). (4) This non-synonymous variant is located in a mutational hot spot and critically well-established functional domain (PM1). (5) Missense variant is a common mechanism of this disorder and the HOGA1 gene has a low rate of benign missense variants (PP2) after all functional study is suggested for confirming the pathogenicity of this variant.

No causative variant was detected in the F4 family’s proband. This might be due to that the variant may locate in the regulatory elements/promoter regions, which were not covered by Sanger sequencing, or other genes except for HOGA1 may be responsible for the disease. It is suggested to perform whole genome sequencing to find out the causative mutation in this family.

After NGS analysis and checking the relevant genes in the different databases such as GeneCard, MalaCard, and OMIM, no gene related to the phenotypic disorder of the affected person was found. It is suggested that the mutation may be located in the non-exonic regions such as promoter or regulatory elements. Since the patient (F3) was born into a consanguineous marriage, we tried to find an ROH region to guide us to the possible causative locus in the patient. Different genes were found in these regions as shown in Table 2; the only gene that was related to this disease was the GRHPR gene. As no pathogenic/likely pathogenic variant was detected in all coding regions of this gene, we assume that the causative variant may be in the non-exonic region of the GRHPR gene; therefore, performing a whole genome sequencing for this case is suggested. To the best of our knowledge, this is the first genetic study on hyperoxaluria patients in Iran and it can provide a good context for future research.

Out of 14 patients taking part in this study, a likely pathogenic variant (HOGA1: exon 2:c266G>A) in one of them was detected. NGS analysis could be a good approach for those who did not show homozygous haplotypes.

The authors would like to thank Ali Asghar Clinical Research Development Center for search assistance.

This study protocol was reviewed and approved by the Iran University of Medical Sciences (IUMS) research committee (Approval No. IR.IUMS.FMD.REC.1401.149). For this study, written informed consent was obtained from participants’ parents. Written informed consent was obtained from their parents for publication of the details of their medical case and any accompanying images.

The authors have no conflicts of interest to disclose.

This research was funded by IUMS. Author Marzieh Mojbafan has received research support from IUMS.

Sadegh Tavakoli Ataabadi performed the experiment and collected data and analyzed and interpreted data and drafted the manuscript. Leila Behi cooperated in the project. Marzieh Mojbafan designed and supervised the study and critically revised the manuscript. Nakysa Hooman contributed to sample collection and clinical and paraclinical examination of patients.

All date generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.