Article Text

The c.859G>C variant in the SMN2 gene is associated with types II and III SMA and originates from a common ancestor
  1. S Bernal1,
  2. L Alías1,
  3. M J Barceló1,
  4. E Also-Rallo1,
  5. R Martínez-Hernández1,
  6. J Gámez2,
  7. E Guillén-Navarro3,
  8. J Rosell4,
  9. I Hernando5,
  10. F J Rodríguez-Alvarez6,
  11. S Borrego7,
  12. J M Millán8,
  13. C Hernández-Chico6,
  14. M Baiget1,
  15. P Fuentes-Prior9,
  16. E F Tizzano1
  1. 1Servei de Genètica, Hospital de la Santa Creu i Sant Pau, and CIBERER U-705, Barcelona, Spain
  2. 2Servei de Neurología, Hospital Vall d'Hebron, Barcelona, Spain
  3. 3Unidad de Genética Médica, Hospital Universitario Virgen de la Arrixaca, Murcia, Spain
  4. 4Departamento de Genética, Hospital Son Dureta, Palma de Mallorca, Spain
  5. 5Servicio de Genética, Hospital Universitario de Asturias, Oviedo, Spain
  6. 6Unidad de Genética, Hospital Ramón y Cajal, and CIBERER U-728, Madrid, Spain
  7. 7Unidad de Gestión Clínica de Genética, Reproducción y Medicina Fetal, Hospital Virgen del Rocío and CIBERER U-702, Sevilla, Spain
  8. 8Unidad de Genética, Hospital La Fe, and CIBERER, Valencia, Spain
  9. 9Unitat Bases Moleculars de les Malalties, Research Institute, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain
  1. Correspondence to Eduardo F Tizzano, Genetics, Hospital de la Santa Creu i Sant Pau, Sant Antoni Ma. Claret 167, 08025 Barcelona, Spain; etizzano{at}


Homozygous mutations of the telomeric SMN1 gene lead to degeneration of motor neurons causing spinal muscular atrophy (SMA). A highly similar centromeric gene (SMN2) can only partially compensate for SMN1 deficiency. The c.859G>C variant in SMN2 has been recently reported as a positive disease modifier. We identified the variant in 10 unrelated chronic SMA patients with a wide spectrum of phenotypes ranging from type II patients who can only sit to adult walkers. Haplotype analysis strongly suggests that the variant originated from a common ancestor. Our results confirm that the c.859G>C variant is a milder SMN2 allele and predict a direct correlation between SMN activity and phenotypic severity.

  • Clinical genetics
  • molecular genetics
  • neurology
  • motor neurone disease
  • neuromuscular disease

Statistics from

Spinal muscular atrophy (SMA) is an autosomal recessive neuromuscular disorder characterised by loss and degeneration of motor neurons of the spinal cord. The disease is caused by homozygous absence or defects of the telomeric SMN1 gene located in a complex region of chromosome 5 (5q13).1 A highly similar centromeric gene (SMN2) can only partially compensate for SMN1 deficiency. The c.859G>C variant in SMN2 has been recently reported as a positive disease modifier.2 3 We identified the variant in 10 unrelated chronic SMA patients with a wide spectrum of phenotypes ranging from type II patients who can only sit to adult walkers.

Based on age at onset and achieved milestones, SMA is classified into types I (the most severe form, characterised by generalised muscle weakness and hypotonia within the first 6 months of life); II, which manifests after the age of 6 months, affected children are able to sit but they never walk unaided; and III, in which patients eventually lose their ability to walk unassisted. Type III SMA is further divided into IIIa, with clinical symptoms appearing before the age of 3, and IIIb, with onset between 3 and 20 years.4

Although SMN1 and SMN2 encode in principle identical proteins, a C-to-T transition within exon 7 of SMN2 leads to a high proportion of transcripts that lack this exon, resulting in an unstable, non-functional protein.5 However, the expected inverse correlation between SMN2 copy number and severity of SMA disease (most type I, II and III SMA patients carry two, three, and three or four SMN2 copies, respectively) is not absolute,6 7 and some patients with two SMN2 copies show mild phenotypes, suggesting that other factors modulate disease progression.

We studied 261 unrelated Spanish SMA patients (table 1) and six pairs of SMA discordant siblings with homozygous absence of the SMN1 gene, 10 parents of patients with the c.859G>C variant and 105 unrelated healthy individuals, five of whom lack the SMN2 gene. Informed consent was obtained from all participants studied. Criteria to evaluate phenotype variability of the patients were age of onset, progression of motor milestones and age of loss of walking ability. We used LightCycler and multiplex ligation-dependent probe amplification for an accurate estimation of SMN gene copy number as described elsewhere.7 8

Table 1

Distribution of the c.859G>C variant among 261 SMA patients according to their phenotype and number of SMN2 copies. Note that the variant is positive in 8 of 16 (50%) discordant type II and III cases with two SMN2 copies

By direct sequencing using primers and methodology reported previously,9 10 we identified the SMN2 c.859G>C variant in 10 unrelated type II or III SMA patients (nine men,one woman; patients' clinical and molecular data are summarised in table 2). Five of these individuals had first clinical symptoms of the disease between 4 and 15 years of age, indicating type IIIb SMA; they all maintained walking capacity for several years. Three of these cases (patients 1–3) showed two SMN2 copies carrying the c.859G>C nucleotide change in homozygous state. The remaining two patients had three SMN2 copies and the variant in heterozygous state, with two copies of wild-type SMN2 and one that carries the c.859G>C variant. Patient 6 (type IIIa) carries two SMN2 copies and the variant in heterozygous state. He started to walk at 18 months, but muscle weakness manifested around 2 years and was generalised at 6 years old. The other four individuals with two SMN2 copies and the variant in heterozygous state (patients 7–10) developed type II disease. Altogether, 8 out of 16 patients with the chronic forms but only two SMN2 copies were positive for the variant (table 1). The analysis of ten parents confirmed that the variant has been transmitted either maternally and/or paternally (table 2).

Table 2

Clinical and molecular data of 10 Spanish SMA patients with homozygous absence of the SMN1 gene and with the c.859G>C variant in the SMN2 gene

The variant was not detected in any of the non-discordant cases: 126 type I SMA patients with two SMN2 copies, 55 type II patients with three SMN2 copies and 8 walkers with four SMN2 copies. Furthermore, the variant was neither found in the six pairs of SMA discordant sibling or individuals without SMN2 genes but was present in 2 of 105 healthy individuals analysed.

Haplotype characterisation10 showed that the 10 patients with the variant presented a common allele of 193bp or (CA)21 in the C272 multi-copy marker, which is closest to the 5′-end of the SMN genes. Analysis of the C212 marker revealed two alleles of 227bp or (CA)30 repeats (n=7) and 225bp or (CA)29 repeats (n=3); this minimal size difference could be explained by replication slippage in a previous generation (table 2 and additional figure 1). In striking contrast, these alleles were rarely associated with SMN2 in SMA patients without the variant. Thus, the linkage disequilibrium of the c.859G>C variant with these alleles strongly suggests that it originated from a common ancestor; studies in other populations are warranted to confirm this observation.

Two previous reports of the c.859G>C variant described patients with type III SMA.2 3 However, in our series, it was also found in four unrelated patients with type II disease (two SMN2 copies with the variant in heterozygous state; table 2). Underscoring the impact of SMN dosage on disease phenotype, we detected the c.859G>C variant in heterozygous form in two type IIIb cases with three SMN2 copies, while a patient also carrying the variant in heterozygous state but only two SMN2 copies suffered from the more severe type IIIa form. A homozygous type IIIa patient who inherited two copies of the c.859G>C variant from his mother in the same chromosome in cis position has been previously reported.3 The parents of one of our patients homozygous for the SMN2 variant were consanguineous, and both transmitted the variant (patient 2). Thus, the phenotype spectrum associated with the SMN2 c.859G>C variant ranges from classical type II patients to adult type IIIb walkers depending on the homozygous or heterozygous state of the variant, the copy number of SMN2 genes. Conceivably, cis and trans SMN2 copies may be differently expressed, modulating the disease phenotype.

The c.859G>C variant was not found in 126 patients with type I SMA and two SMN2 copies, similar to previous observations in 82 cases.2 3 Altogether, over 200 patients with the most severe disease form from various origins have been studied, which strongly suggests that the variant should not be expected in these patients. Instead, the c.859G>C mutation appears to be restricted to discrepant cases of milder phenotype and might be found in approximately half of the patients with only two SMN2 copies but chronic SMA. When stratifying these patients according to SMA type, the frequency reaches 80% in our type III cases (table 1). Interestingly, all but two type IIIb patients who carry the c.859G>C variant lack NAIP, which contrasts with the low frequency of NAIP deletion reported in chronic SMA cases (10–15%).

In conclusion, we confirm that the SMN2 c.859C>G variant is a milder SMA allele, which is present in a minority of patients with chronic SMA, but not in any with type I disease. Our results indicate that there is a direct correlation between SMN activity and disease severity.


We wish to thank the consenting parents and patients who made this study possible. We are indebted to all the colleagues and laboratories that referred samples for this study.


Supplementary materials


  • Funding Other funders: GENAME Project, CIBERER Intramural U-705 and FIS08-0729.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the Hospital Santa Creu i Sant Pau, Ethics Committee.

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

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.