Clinical and Molecular Features of First Mexican Friedreich's Ataxia Patients with Compound Heterozygous
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.329555
Source of Support: None, Conflict of Interest: None
Keywords: Cardiomyopathy, compound heterozygous, diabetes, Friedreich's ataxia, FXN gene mutation, hearing loss, phenotypiv variability.
Friedreich's ataxia (FRDA) usually occurs in people <25 years; however, ≈15% have a late-onset presentation (LOFA, >26 years). Its main cause is a biallelic GAA repeat expansion within the first intron of FXN gene, but 2% to 5% of patients show one expanded allele paired with a point mutation or a large deletion in the second allele, namely, compound heterozygous individuals (CH).,
A previous study in Mexican patients with recessive ataxia reported 9.3% FXN biallelic expansions.,
Herein, we aimed to determine the frequency of FXN mutations in patients with sporadic or recessive ataxia over a 10-year period. Characteristics of the first CH Mexican patients are provided.
Between 2009 and 2019, one-hundred and eight Mexican patients (including five pairs of affected siblings) with sporadic or autosomal recessive ataxia were referred to our institution for FXN molecular analysis. Inclusion criteria were ataxic gait as a core feature of the phenotype and sporadic cases that were negative for dominant spinocerebellar ataxias types 2, 3, and 7.
This is a retrospective study that was approved by the Institutional Review Board or Independent Ethics Committee (IRB/IEC) and performed in accordance with the Helsinki Declaration (protocol_145/18). After giving written informed consent, patients were evaluated by a movement-disorder specialist following FRDA international criteria, and a geneticist interrogated them for family history (FH). Laboratory tests, neurophysiological studies, brain magnetic resonance imaging (MRI), and transthoracic echocardiograms were performed.
Blood DNA was isolated by standard procedures. Positive samples for triplet repeat-primed PCR (TP-PCR) as first screening were amplified with a second set of primers by long PCR. In CH patients, the nonexpanded allele was direct sequenced and analyzed with Mutation Surveyor Version 4.0 software, using the GenBank file NG_002601.2 as reference sequence. Then, multiplex ligation-dependent probe amplification (MLPA) was performed (APTX, SETX, FXN) for those patients with one FXN expanded allele and no sequence variants. Capillary electrophoresis for TP-PCR, sequencing, and MLPA was done on an AB3130 genetic analyzer (Life Technologies, Carlsbad, CA, USA).
Originally, 98 nonrelated patients were recruited. Because some patients referred to having affected siblings, they were also enrolled in this study. Thus, we studied 108 patients, belonging to 98 families, from 21 Mexican states. Family history (FH), consistent with autosomal recessive inheritance, was considered positive if ataxia was reported by a first- or second-degree relative. FH was positive in 52 cases (53.06%).
Participants' age at onset (AAO) ranged from 1 to 53 years (mean ± standard deviation = 17.47 ± 11.43 years) with 62% identified as males. The observed wild-type FXN alleles varied from 6 to 22 GAA repeats, with 6 to 8 repetitions being the most frequent.
TP-PCR revealed 23/98 positive cases (23.47%); 18 of them showed biallelic FXN expansions (18.37%), ranging from 75 to 1,033 GAA repeats by long-PCR (no premutation allele was observed). The remaining five positive cases had a monoallelic expansion and were suspected as CH patients [Table 1]. The AAO in both groups was similar (P = 0.5247) [Table 2].
Sequencing analysis of the nonexpanded allele in CH patients revealed two point mutations: p.Met1Ile (Patient 1) and p.Q153H (Patient 2) [Figure 1]a and [Figure 1]b, being classified as pathogenic and of uncertain significance variants, respectively, at Varsome using the American College of Medical Genetics (ACMG) rules.
MLPA analysis evidenced a heterozygous deletion of exon 3 (Ex3del_FXN) in CH Patient 3 [Figure 1]c, mean ratio = 0.6 in two-independent assays), whereas CH Patients 4 and 5 did not show structural variations.
Additionally, six benign variants were observed in CH patients (rs145006100, rs2481598, rs3829062; and three novel intronic changes: c.186+4G > A, c.284 + 24G > C, c.511 + 53G > A), and the variant of uncertain significance, rs772778611, in Patients 1, 3, 4, and 5.
CH Patient 1 referred positive FH, with a deceased brother in young adulthood, allegedly with walking disability, diabetes mellitus, and a presumptive Charcot-Marie-Tooth disease (CMT).
None of the CH patients had diabetes, hearing loss, or a severe heart disease, with a mean functional staging FARS = 2.1, and a loss of autonomous ambulation (LAA) = 32.2 ± 9.98 years. CH Patients 2 and 4 had mild cerebellar atrophy [Table 1].
We confirmed the low frequency of FXN mutations in Mexican patients with sporadic or autosomal recessive ataxias (23.47%) versus other reported populations (e.g., FRDA represents 75% of the recessive ataxias in Caucasians). Genetic counseling was offered to FRDA-confirmed cases; 18.37% of them showed biallelic FXN expanded alleles, and 5.1% were CH patients.
We observed that CH patients have phenotypic differences compared with individuals homozygous for (GAA) n expansions, in contrast with the classic FRDA phenotype, in which an earlier onset has faster progression, and nearly all patients developing cardiomyopathy at some point in their lives. Our CH patients had a slightly slower progression (time of LAA-AAO ranged from 11 to 20 years), without diabetes, severe heart disease, or hearing loss [Table 1]. However, because of our small sample size, these differences were nonsignificant [Table 2].
LOFA patients with smaller GAA expansions generally show a mild phenotype even in the absence of cardiomyopathy; they may mimic hereditary spastic paraplegias with subsequent development of ataxic symptoms. As is the case of Patient 5, who showed spastic paraparesis as the first symptom, and subsequently presented minimal ataxic gait and fine movements, without cardiac manifestation. The presence of nucleotide interruptions of the GAA tract modulating FXN methylation levels may be associated with LOFA, but this was not tested in our patient.
Sixty-seven FXN mutations have been documented in the Human Gene Mutation Database (HGMD). The mutation identified in Patient 1 (p.M1I) has been published in FRDA cases;, in fact, the initiator codon is a mutation hotspot with three documented additional substitutions p.M1T/S/L (rs142133355, rs149724959, and rs147643987, respectively) resulting in a shortened protein,,, and supposedly from a common founder. This variant is found in gnomAD database (frequency = 0.00000913) and reported in ClinVar: VCV000003983.1.
The mutation p.Q153H observed in Patient 2 is also a recurrently mutated amino acid (p.Q153H/R, rs780387020, and rs77781994, respectively), and is found in gnomAD database (frequency = 0.00000398). Q153 is clustered to the affected residues p.I154F and p.W155R located in one surface of the frataxin-causing FRDA., These mutations decrease the interaction of frataxin with the iron-sulfur biogenesis Nfs1/ISCU complex through ISD11.
Five gross exonic deletions and and one spanning the entire gene are reported in HGMD; CH patients carrying exonic deletions are considered rare and severe variants of early-onset FRDA. Interestingly, Patient 3 (AAO = 9 years), with an FXN_Ex3del, has a slow disease progression.
Because no CH patients' relatives were included, we could not check whether the expanded allele and the other FXN mutations were in cis or trans.
The two remaining CH patients, in whom no second mutation was identified, may present rare rearrangements or intronic variants, undetected by our FXN analysis, or be FRDA carriers with another Friedreich-like ataxia. The observed phenotypic variability in CH patients depends on the expression of partially functional mutant FXN, epigenetics, and the pathogenicity of mutations present in the expanded allele (i.e., number of GAA repeats, sequence interruptions) and the nonexpanded allele. Thereby, CH patients may mimic diverse phenotypes such as CMT (Patient 1), or spastic paraparesis (Patient 5), delaying their diagnosis and causing an inappropriate management of the disease (e.g., Patient 2 diagnosed after 14 years of disease progression). Our results emphasize the need to screen repeat expansions, dosage, and point mutations to rule out the FRDA molecular diagnosis avoiding additional time-consuming/expensive analyses. Recently, it has been described a CMT-like case carrying the first biallelic FXN point mutation; however, our cases negative for FXN analysis did not report that presentation.
Patient 1 also presented autoimmune hypothyroidism; this concomitant condition has been described in one pediatric FRDA patient, highlighting that in case of presenting cardiomyopathy could be exacerbated when they coincide. Although the echocardiogram was normal at the time of the patient's assessment, clinicians should be aware of this possibility.
Little is known about FXN variants in Latin Americans, whereas similar findings to ours have been reported in Brazilian patients. The underlying cause for 76.53% of our cases remains unknown and should be explored by whole-genome sequencing.
The phenotypic variability observed in CH patients could be related to the length of the expanded allele and the type of mutation in the second allele (with its corresponding amount of residual frataxin). These findings should be considered when developing future specific mutation-directed therapeutics.
This study presents the first Mexican patients with a CH FXN genotype, expanding the clinical and mutational spectrum of FRDA. Our findings have important implications for genetic counseling and management of FRDA Mexican patients.
We would like to thank all the patients who participated in this study.
Financial support and sponsorship
DGS was supported by a master's grant from CONACyT #960931.
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2]