Article Text

Download PDFPDF
Mitochondrial molecular genetic results in a South African cohort: divergent mitochondrial and nuclear DNA findings
  1. Surita Meldau1,2,
  2. Elizabeth Patricia Owen1,2,
  3. Kashief Khan2,
  4. Gillian Tracy Riordan3,4
  1. 1 Division of Chemical Pathology, Department of Pathology, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa
  2. 2 Chemical Pathology, National Health Laboratory Services, Groote Schuur Hospital, Cape Town, South Africa
  3. 3 Division of Paediatric Neurology, Department of Paediatrics and Child Health, University of Cape Town, Cape Town, Western Cape, South Africa
  4. 4 Red Cross War Memorial Children's Hospital, Cape Town, South Africa
  1. Correspondence to Surita Meldau, Division of Chemical Pathology, Department of Pathology, University of Cape Town Faculty of Health Sciences, Cape Town 7925, South Africa; surita.meldau{at}


Aims Mitochondrial diseases form one of the largest groups of inborn errors of metabolism. The birth prevalence is approximately 1/5000 in well-studied populations, but little has been reported from Sub-Saharan Africa. The aim of this study was to describe the genetics underlying mitochondrial disease in South Africa.

Methods An audit was performed on all mitochondrial disease genetic testing performed in Cape Town, South Africa.

Results Of 1614 samples tested for mitochondrial DNA (mtDNA) or nuclear DNA (nDNA) variants in South Africa between 1994 and 2019, there were 155 (9.6 %) positive results. Pathogenic mtDNA variants accounted for 113 (73%)/155, from 96 families. Mitochondrial encephalopathy with lactic acidosis and stroke-like episodes, 37 (33%)/113, Leber’s hereditary optic neuropathy, 26 (23%)/113, and single large mtDNA deletions, 22 (20%)/113, accounted for 76%. Thirty eight of 42 nDNA-positive results were homozygous for the MPV17 pathogenic variant c.106C>T (p.[Gln36Ter, Ser25Profs*49]) causing infantile neurohepatopathy, one of the largest homozygous groups reported in the literature. The other nDNA variants were in TAZ1, CPT2, BOLA3 and SERAC1. None were identified in SURF1, POLG or PDHA1.

Conclusions Finding a large group with a homozygous nuclear pathogenic variant emphasises the importance of looking for possible founder effects. The absence of other widely described pathogenic nDNA variants in this cohort may be due to reduced prevalence or insufficient testing. As advances in therapeutics develop, it is critical to develop diagnostic platforms on the African subcontinent so that population-specific genetic variations can be identified.

  • DNA
  • medical laboratory science
  • genetic diseases
  • inborn

Data availability statement

Data are available upon reasonable request. Our data are kept on a secure server as part of a diagnostic-linked patient database. It contains sensitive patient information only viewed by laboratory diagnostic staff under strict patient confidentiality agreements. In order to share the data used, it would first need to be deidentified and cleaned up prior to sharing. This can be done upon request, pending additional ethics clearance.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Mitochondrial respiratory chain disorders are recognised collectively as one of the largest groups of inborn errors of metabolism and one of the most common causes of inherited neurological disease at any age. They are caused by pathogenic variants in either mitochondrial DNA (mtDNA), or nuclear DNA (nDNA) genes. The double stranded 16 569 bp mtDNA genome, situated in multiple copies inside the mitochondrial matrix, encodes 13 subunits of the mitochondrial respiratory chain as well as 22 transfer RNAs (tRNAs) and two mitochondrial ribosomal RNAs. However, up to 1900 proteins,1 2 are involved in human mitochondrial function, the majority of which are encoded by nDNA. Pathogenic variants in more than 400 genes spanning mtDNA and nDNA genomes have been implicated in mitochondrial disease. New disease associations have been frequently identified since the advent of whole exome sequencing (WES) and whole genome sequencing.

Overall prevalence of mitochondrial disease in well-studied populations, of predominantly European descent, is estimated to be as high as 1/5000.3–5 There is a paucity of prevalence data from the African continent. Reasons for this include the complexity and cost of clinical and laboratory diagnosis; difficulty accessing complex diagnostic platforms, such as respiratory chain enzymology and next generation sequencing (NGS); and the preventable infectious disease burden on the continent which has been justifiably prioritised by healthcare systems.

The National Health Laboratory Services (NHLS) Inherited Metabolic Disease (IMD) Laboratory at Groote Schuur Hospital in Cape Town, South Africa, has been the primary diagnostic referral laboratory for mitochondrial genetics since 1991.


To conduct a retrospective audit of all mitochondrial disease-related molecular diagnostic findings over the past 26 years (1994–2019), in order to better define the genetic background underlying this group of disorders in the South African population.


The following data fields from all mitochondrial disease genetic testing requests were retrieved from the UCT/NHLS Mitochondrial disease database (UCT HREC/REF: R010/2017) and analysed: age at presentation, gender, referral site, clinical data provided, sample type, genetic diagnostic result (variant information, where identified), final diagnosis (if available), other supporting lab results, family history of note, variants of unknown significance detected and mitochondrial haplogroup (where available, post 2015).

Prior to 2015, mitochondrial molecular testing included targeted analysis for common pathogenic mtDNA variants and long-range PCR-based deletion screening on unscreened referrals, as well as limited Sanger sequencing on a handful of nDNA genes (including POLG, PDHA1, TAZ1, CPT2 and SURF1). Respiratory chain enzyme analysis was not performed, as it was unavailable in the state health sector. More appropriate tissue sampling and newer genetic techniques were introduced from 2013. NGS of the mtDNA genome was implemented and the first WES runs were performed in selected cases.

The diagnostic rate prior to and after the introduction of more extensive testing strategies and improved referral practices in 2015 was assessed to measure efficiency and utility of these measures.

Regional as well as state versus private referral efficiency was assessed by measuring the positive genetic diagnostic rate obtained for each referral site.


The IMD laboratory processed a total of 1614 requests for mitochondrial genetics between 1994 and 2019, excluding carrier and prenatal testing requests. Disease was confirmed in 155 of these cases, resulting in a diagnostic rate of 9.6% over the entire period (7.7% excluding MPV17 requests). The positive diagnostic rate prior to 2015 was 6.2%, while the diagnostic rate between 2015 and 2019 was 16.7% (11.6% excluding MPV17 requests and results). Yearly diagnostic rates are presented in figure 1.

Figure 1

Graph representing the diagnostically confirmed cases each year since 1994. Blue bars are the total number of diagnostic confirmations. Orange bars represent diagnostic confirmations excluding MPV17 cases.

All identified pathogenic variants are depicted in tables 1 and 2. The most common genetic cause of disease identified was the c.106C>T (p.[Gln36Ter, Ser25Profs*49]) variant in exon 3 of the nDNA encoded MPV17 gene, with 38 patients homozygous for this variant. Apart from a single case of Barth syndrome (TAZ1), one of carnitine palmitoyl transferase 2 deficiency (CPT2), and one each of the previously reported multiple mitochondrial dysfunctions syndrome type 2 (MDDS2, BOLA3)6 and 3-methylglutaconic aciduria with deafness, encephalopathy and Leigh-like syndrome (MEGDEL, SERAC1),7 8 no other pathogenic nDNA variants have been identified in this setting to date. The m.3243A>G (MT-TL1): variant, primarily causing mitochondrial encephalopathy, lactic acidosis and stroke-Like episodes (MELAS), or maternally inherited diabetes and deafness, was the most commonly occurring mtDNA variant in our cohort, accounting for 33% of all primary pathogenic mtDNA variants identified and 24% of total mitochondrial disease diagnoses. The common m.11778G>A (MT-ND4) and m.14484 (MT-ND6) LHON-associated variants, with single large mtDNA deletions accounted for most of the remaining cases, documented in table 2. At least 4 MELAS cases, from separate families (4/30), 12 LHON cases (11/25) and 9 patients with single large mtDNA deletions (9/22) had traditional black African surnames. No information regarding self-identification of race or haplogroup is available.

Table 1

Pathogenic nuclear DNA (nDNA) variants identified in the South African cohort

Table 2

Pathogenic mtDNA variants identified in the South African mitochondrial disease cohort

Supplemental material

Between 2013 and 2019, most referrals were received from tertiary referral hospitals in the Western Cape (37.3%) and KwaZulu Natal (13.7%) regions, or through one of three private pathology services nationwide (29.6%). A complete breakdown of national referral figures and diagnostic rates are presented in figure 2.

Figure 2

(A) Graph representing the distribution of referrals between state and private sectors in South Africa, as well as the diagnostic confirmation rates in cases referred from each of these sectors. (B) Geographic distribution of referral origins for mitochondrial genetic testing from state healthcare services in South Africa. (C) Geographical map showing genetic diagnostic rates per province in South Africa. EC, Eastern Cape; FS, Free State; GP, Gauteng; KZN, Kwazulu-Natal; NW, North West; WC, Western Cape.

Only 8.2% of samples submitted were muscle biopsies, the historical gold standard for mtDNA investigations, because of the higher degree of heteroplasmy found in muscle. Only 23 (32%) out of 73 requests for mtDNA deletion screening in patients with progressive external ophthalmoplegia or features suggestive of Kearns-Sayre syndrome were muscle biopsies. Of these, 15/23 were positive for either single or multiple deletions (65%), compared with only 3 (6%) out of the remaining 50 where no muscle was available. Overall, of the 132 muscle biopsies submitted to the laboratory during this time, for any mitochondrial diagnostic testing, 28 (21%) resulted in confirmation of mitochondrial disease.

DNA from 21, 13 and 9 patients, respectively, were subjected to full gene Sanger sequencing of the POLG, SURF1 and PDHA1 genes, respectively. No pathogenic variants were found in any of these in the South African cohort.

Haplogroup information was recorded for all positive cases where full mtDNA sequencing was performed (since 2015, see tables 1 and 2).


The results of this study are dominated by the high number of cases of the South African variant of MPV17 infantile neurohepatopathy, due to homozygosity for the common MPV17:c.106C>T(p.[Gln36Ter, Ser25Profs*49]) variant. Clinical phenotyping, suggestive liver histology and focused WES led to the identification of this pathogenic variant in a cohort of 24 South African patients.9 A further 14 patients have been diagnosed since the original publication. The c.106C>T variant has a predicted carrier frequency of 1/68 (95% CI 1/122 to 1/38) in the black South African population in the Western Cape and an estimated newborn incidence of 1/18 622 births (95% CI 1/59 536 to 1/5776).9 This is one of the largest cohorts of patients with mitochondrial disease with a specific clinical phenotype, homozygous for a single nDNA variant, described in the literature to date. With an annual national birth rate of more than 1 000 000, the anticipated number of infants born yearly with this severe condition could be over 50. Actual numbers remain much lower, emphasising the diagnostic gap anticipated in less well-resourced countries, where access to tertiary healthcare and genetic testing is limited or unevenly distributed. The low diagnostic rate in this group, with a known estimated prevalence and a very severe clinical phenotype, suggests that other milder disease phenotypes may be underdiagnosed. Pathogenic variants have been identified in other metabolic disorders in South Africa, including galactosaemia, primary hyperoxaluria, cystinosis, glutaric aciduria, familial hypercholesterolaemia and variegate porphyria.10–15 This emphasises the need for better genetic characterisation of inherited disease in migratory populations.

There was a relatively low number of requests for mitochondrial genetic studies in general. With a worldwide estimated prevalence of about 1 in 5000 to 1 in 8000, a population of ~58.8 million should have as many as 11 700 affected patients. Only just over 1600 cases were referred for testing over 26 years, with an additional 212 patients forming part of a research cohort originating primarily from the northern provinces of South Africa16–19 and another cohort harbouring RRM2B variants as a suspected cause of renal dysfunction, rod-cone dystrophy and sensorineural hearing loss in the Afrikaner subpopulation.20 Very few of the positive cases in our cohort led to further investigation of family members, highlighting a possible lack of capacity to follow-up on affected families in the clinical setting. Clinics with good interdisciplinary liaison and which provide genetic counselling, as well as social support, are needed.

Referral patterns varied across provinces and between state and private sectors. Most state referrals were received from the Western Cape, as two of the major academic centres in the province are closely situated to the laboratory. Thirty per cent of referrals between 2013 and 2019 were from private pathology services, which is high relative to the proportion of the population served. Eighty per cent of the South African population rely on state healthcare services. Only 4.7% of private referrals resulted in confirmed diagnoses, compared with 16.7% in the state healthcare sector for this period. State referrals have largely been from academic hospitals with collective experience of managing rare diseases.

Close to 10% of referrals resulted in confirmed diagnoses over the entire 26-year period, an increase from the previously described rate of about 6%.16 17 This increase can largely be ascribed to the introduction of more advanced testing strategies and improved referral practices introduced after 2015. Improved clinical screening, clinician scientist liaison, more appropriate tissue sampling and newer genetic techniques were introduced from 2013. NGS of the mtDNA genome was implemented and the first WES runs were performed in 2015 in selected cases on a research basis. These measures doubled the diagnostic confirmations (from 6.2% to 11.6%, excluding MPV17 cases) with an upward trend year on year. The identification of the likely founder MPV17 variant in a single population subgroup led to a spike in diagnostic confirmations, resulting in a total positive diagnostic rate of 16.7% between 2015 and 2020.

As expected, there was a higher percentage of positive findings from muscle biopsies. These are invasive, especially in children who may be at risk from general anaesthesia, as well as costly. There is a move internationally towards whole exome or genome sequencing in cases of suspected nDNA disease such as mtDNA maintenance syndromes. This requires good clinical profiling combined with strong bioinformatics infrastructure and laboratory support to achieve success. Without this, complex genetic screening strategies are likely to generate data that cannot be appropriately validated.

The absence of pathogenic variants in common nDNA mitochondrial genes (such as POLG, SURF1 and PDHA1) in this study group puts emphasis on the need for broader genetic testing. WES is currently being done in a research capacity, but more widely available implementation of this technology is necessary if we are to elucidate the nDNA genetics underlying disease in Sub-Saharan Africa.

One very important gap remains in our data. The South African population is genetically and culturally very diverse. Most cases where common pathogenic mtDNA variants (such as the MELAS m.3243A>G and LHON m.11778G>A variant) were identified were diagnosed through testing for a small group of select variants and therefore did not generate any information about the haplogroup context in which they occur. Further work should define these variants in the context of their diverse genetic landscape and address the paucity of information surrounding the occurrence of these variants in African (L) haplogroups. Recent work by O’Keefe et al has indicated that haplogroup context could potentially affect disease phenotypes, particularly in mt-tRNA genes.21–23 African data addressing this phenomenon are severely lacking,17 contributing to the speculation that some populations may be protected, leading to absent or milder phenotypes.21 The clinical presentations of the four m.3243A>G positive patients with traditional African surnames in this cohort were compatible with clinical findings described with MELAS, without unusual or specific features. Similar findings were seen for the LHON and mtDNA deletion cases. This contrasts with previous findings describing a high predominance of muscle phenotypes in South African paediatric patients from the northern provinces. In our group in whom a genetic diagnosis was not made, there were children with Leigh syndrome or features of suspected mitochondrial encephalopathy, who came from families speaking African home languages.18 19 To be certain of genetic variance in mitochondrial disease in African populations, ongoing research and increased awareness is needed.

With evolving reproductive and therapeutic interventions to support families living with mitochondrial disorders, improving genetic diagnosis becomes imperative. Research into mitochondrial disease in Sub-Saharan Africa is necessary to enable patients to benefit from advances in the field.

The data reported here represent the largest and most diverse cohort of patients with mitochondrial disease reported to date from Africa. It highlights the fact that common pathogenic mtDNA variants do occur in Sub-Saharan African populations, although the haplogroups in which they occur remain to be elucidated, while nDNA variants that are common elsewhere have not been identified thus far. Health planning and allocation of resources depend on accurate representation of disease in local populations and therefore improved molecular diagnostic accuracy is vital to ensuring future management is equitable and directed.

Take home messages

  • Mitochondrial diseases occur in Sub-Saharan Africa.

  • The mtDNA findings in this cohort resemble those of well-described populations, while nDNA findings do not.

  • MPV17 accounts for the majority of nDNA inherited mitochondrial disease known in Sub-Saharan Africa.

Data availability statement

Data are available upon reasonable request. Our data are kept on a secure server as part of a diagnostic-linked patient database. It contains sensitive patient information only viewed by laboratory diagnostic staff under strict patient confidentiality agreements. In order to share the data used, it would first need to be deidentified and cleaned up prior to sharing. This can be done upon request, pending additional ethics clearance.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by the University of Cape Town Human Research Ethics Committee with ref: HREC/REF:024/2018.


The authors thank Professor Komala Pillay for histopathology support; Associate Professor George van der Watt for biochemistry support; Dr Ronalda De Lacy for clinical work leading to MPV17 variant identification; patients, families and referring clinicians of the cases described in this audit; NHLS for funding support.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Tahir S Pillay.

  • Contributors SM helped in conceptualisation, planning, conduct and reporting. EPO helped in conceptualisation and conduct. KK helped in conducting the study. GTR helped in conceptualisation, planning and reporting.

  • Funding This study was funded by National Health Laboratory Service.

  • Map disclaimer The depiction of boundaries on the map(s) in this article does not imply the expression of any opinion whatsoever on the part of BMJ (or any member of its group) concerning the legal status of any country, territory, jurisdiction or area or of its authorities. The map(s) are provided without any warranty of any kind, either express or implied.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.