The majority of the mutations that have been reported are present at a heterozygous state. There are only rare cases of children inheriting two abnormal copies of the COL3A1 genes. However, this situation is more likely to occur if the parents of the child are related, as the mutation is frequent in the family (Johnson and Falls, 1949).
​
​
​
​
​
​
​
​
​
​
​
​
​
​
Heterozygous COL3A1 mutations result in a decrease in the number of normal collagen III proteins produced. As detailed in the biochemical section, each procollagen III protein consists in a trimer formed by three chains of procollagen III, each of them being encoded by one of the two inherited copies of the COL3A1 gene. The wild-type copy of the gene encodes for normal procollagen III chains, and the mutant one encodes for abnormal or even non-produced procollagen III chains. (Leeming and Karsdal, 2016). In the end, you end up with only a small proportion of the trimers of procollagen III being entirely normal. Trimers containing mutant chains have been shown to be not secreted and to be either degraded or to accumulate within cells (Mao and Bristow, 2001).
Different kinds of mutations have been found. The first mutation to be identified in a patient affected with vEDS was described in 1988 by A Superti-Furga et al (Superti-Furga et al., 1988). It consisted in an inframe 3.3 kbp heterozygous deletion in a part of the COL3A1 gene encoding for the triple helical coding domain of the procollagen III chain. The mutant procollagen III was shown to be shortened and less efficiently processed and secreted.
​
​
​
​
​
​
​
​
​
​
​
​
Two major subgroups of COL3A1 variants for vEDS can be put forward, depending on their clinical outcomes.
One first type contains all variants presenting a nonsense mutation or a frameshift one. Nonsense mutations consist in a change in the nucleotides sequence giving rise to a premature termination codon, which results in a shortened mRNA which is accordingly degraded through the nonsense-mediated mRNA decay mechanism. Frameshift mutations consist in a change in the nucleotides sequences inducing a shift in the reading frame of the mRNA, e.g. when one nucleotide is deleted and two nucleotides are inserted. mRNA arising from such variants are also rapidly degraded. In the end, no protein is actually produced from the mutant allele. Patients only produce the normal protein encoded by the wild-type COL3A1 allele.
The second type of variants include all variants showing missense mutations, which consist in a single nucleotide change resulting in a codon encoding for a different amino acid. The mutant mRNA is as long as the normal one and the protein is actually produced.
Interestingly, these two types lead to different degrees of severity in vEDS. While missense mutations appeared to generate the full phenotype early in the lifetime and to cause premature death, frameshift mutations and nonsense ones revealed to correlate with a delay in the onset of the phenotype and in death. Thus, it appears that it is actually better having only normal proteins produced, even half the amount needed, than the correct amount of both normal and abnormal proteins (Dalgleish, 2018).
​
Creating variants databases can turn to be useful in making decisions about the significance of variants found in patients. Finding the same variant across many people presenting vEDS supports evidence that this variant actually causes the disease. In this perspective, a new database has recently been launched as part of Pr. Dalgleish’s work to register variants from a thousand British and North American people affected with hypermobile EDS, a.k.a. EDS type III, whose genetic bases are still not known. Making databases can also potentially help predict natural history of a disease depending on the variant found in an individual. But this does not really apply to vEDS as this type of the syndrome is fairly uniform in its clinical manifestations, except the difference in life expectancy previously evoked. In some other diseases such as osteogenesis imperfecta, a bone disease also due to collagen defects, you can actually develop either mild or severe phenotypes, and they appear to correlate with certain types of mutations (Dalgleish, 2018 ; Dalgleish, 2014).
​
References :
​
Johnson, M.D. Falls H.F. (1949), ‘Ehlers-Danlos Syndrome. A Clinical and Genetic Study.’, Arch Derm Syphilol, 60(1):82-105.
​
Mao, J.R. Bristow, J. (2001), ‘The Ehlers-Danlos syndrome: on beyond collagens’, J Clin Invest. 107(9): 1063–1069.
​
Leeming D.J. Karsdal M.A. (2016) Biochemistry of Collagens, Laminins and Elastin : Structure, Function and Biomarkers. Academic Press.
​
Superti-Furga, A. Gugler, E. Gitzelmann, R. Steinmann, B.(1988), ‘Ehlers-Danlos syndrome type IV: a multi-exon deletion in one of the two COL3A1 alleles affecting structure, stability, and processing of type III procollagen.’, J Biol Chem. 263(13):6226-32.
​
Plancke, A. Holder-Espinasse, M. Rigau, V. Manouvrier, S. Claustres, M. Khau Van Kien, P. (2009), ‘Homozygosity for a null allele of COL3A1 results in recessive Ehlers-Danlos syndrome.’ European journal of human genetics, 17(11), 1411-6.
​
Dalgleish R. (2014). COL3A1. Available at https://eds.gene.le.ac.uk/home.php?select_db=COL3A1
​
Dalgleish, R. (2018) Interview with Claire Baudelet, 16 November.
An example of an individual carrying a homozygous COL3A1 mutant allele is shown opposite. The two mutant alleles were identical and came from consanguineous parents. Interestingly, the two parents were not affected with vEDS, which appeared to be segregated in an autosomal recessive manner in this case (Plancke et al., 2009).
In another case study, siblings were found to be compound heterozygous for COL3A1 variants, meaning that they each carried two different mutant COL3A1 alleles.
​
Biallelic sequence variants such as the ones presented lead to a non-production of any normal protein of collagen III. This was reported to be associated with a worse outcome than heterozygous variants (Jørgensen et al., 2014).
Many other variants have continued to emerge from research studies and from clinical practice. There was an important work conducted by Pr. Dalgleish at the University of Leicester and through the Ehlers Danlos Society about creating a worldwide database to report as many variants leading to EDS as possible. So far most of the data have been described in the COL3A1 gene causing vEDS. 654 unique variants of the COL3A1 gene have been registered in the database, across more than a thousand individuals. All of these variants very uniformly lead to vEDS.
As shown in the picture presented opposite, numerous variants consist in single nucleotide substitutions in the triple-helix coding domain of the COL3A1 gene, leading to a single amino-acid change in the protein.
Different variants of the gene
