Background

Hyaline fibromatosis syndrome (HFS, OMIM# 228600) is a rare heritable disorder with variable severity and frequent lethality, characterized by thickened skin with nodules, papules and plaques, often with a periorificial and perianal distribution, gingival hypertrophy, and joint contractures. Osteopenia, predisposition to respiratory infections, and diarrhea are often present.

HFS is an autosomal recessive disorder, caused by mutations in the ANTXR2 gene, also referred to as CMG2. This gene encodes ANTXR2, a type I transmembrane protein initially characterized for its role in angiogenesis and as a receptor for the anthrax toxin [1, 2]. While over 40 mutations in the ANTXR2 gene have been reported in association with HFS so far, most of them (~70%) being missense and frameshift mutations scattered along the protein, the molecular pathogenesis of this disease has only recently begun to be understood. Specifically, ANTXR2 has been shown to act as a receptor for collagen VI, promoting lysosome-mediated degradation of collagen VI in the extracellular matrix [3], consistent with the finding that patients with HFS develop an accumulation of collagen VI [4]. Thus, defective collagen VI degradation due to nonfunctional ANTXR2 likely leads to collagen accumulation in patients’ tissues, resulting in the clinical manifestations of HFS.

In this report, we examine four patients from consanguineous Iranian backgrounds diagnosed with HFS, and present their clinical and genetic findings (Table 1).

Table 1 Clinical and genetic findings in four patients with HFS

Case presentation

Case 1 presented in infancy with dusky, erythematous plaques and hyperpigmentation over the torso and joints (Fig. 1a). Painful joint contractures, recurrent diarrhea, gingival hypertrophy, and recalcitrant upper respiratory tract infections were present. The patient died from infectious complications at 6 months of age.

Fig. 1
figure 1

Clinical and genetic findings in four cases of HFS. Pedigrees displaying consanguinity, clinical findings, and Sanger sequencing of the identified ANTXR2 mutations. In each case, the upper sequence panel represents the mutant allele, as compared to reference sequence in a healthy control below. a Clinical features of Case 1 included erythematous plaques, hyperpigmentation, and joint contractures of the lower extremities. Sanger sequencing identified a novel homozygous mutation, c.969del, which is predicted to result in truncated protein product, p.Ile323Metfs*14. b First cousin consanguinity, lower extremity contractures with hyperpigmentation over the medial malleolus, and Sanger sequencing of the c.134C > T (p.Leu45Pro) mutation in Case 2. c Case 3 developed flesh-colored papules on the face, including periauricular lesions, as well as gingival hyperplasia. Sequencing revealed the recurrent mutation c.1073dup (p.Ala359Cysfs*13). d Case 4 presented with characteristic perianal lesions and perioral papules. The same mutation as in Case 3 was identified by Sanger sequencing. The green in Case 3 and 4 represents the amino acid change, p.Ala357Pro, resulting from a common benign polymorphism, c.1069G > C (rs12647691)

Case 2 presented at 1 month of age with joint contractures, skin thickening, hyperpigmentation, and perianal plaques (Fig. 1b). The patient died at 10 months of age following recurrent respiratory infections and severe diarrhea.

Case 3 presented neonatally with painful contractures of the lower extremities (Fig. 1c). The patient developed papular lesions on the forehead, nose and ears, as well as hyperpigmentation over the medial malleoli. Gingival hyperplasia was present, and the patient developed frequent respiratory infections.

Case 4 presented in infancy with painful contractures of the legs and interphalangeal joints, perianal erythematous plaques, and perioral papules (Fig. 1d). Additional features included hyperpigmentation over the medial malleoli, gingival hypertrophy, recurrent diarrhea, and respiratory infections. The patient died at 18 months of age from infectious complications. On the basis of clinical presentations, HFS was suspected in all four cases.

Genetic analysis of DNA isolated from peripheral blood of each proband was performed using PCR-based amplification using 17 pairs of primers spanning all exons and ~ 50 bp of flanking intronic sequences of the ANTXR2 gene (NM_058172; primer sequences available upon request). PCR products were bidirectionally sequenced using an automated sequencer (3730; Applied Biosystems, Foster City, CA, USA). Sequencing revealed a previously unreported homozygous frameshift deletion mutation c.969del (p.Ile323Metfs*14) in exon 12 in Case 1, and the previously reported homozygous missense mutation c.134 T > C (p.Leu45Pro) in exon 1 in Case 2 [5]. A homozygous frameshift insertion c.1073dup (p.Ala359Cysfs*13) in exon 13 was identified in Cases 3 and 4. This mutation has been identified in several previous cases of HFS, including an Iranian case reported by our laboratory [6]. The mutation occurs in a cytosine/guanosine rich region of DNA that has been suggested to be prone to mutational events including insertions and deletions [7, 8].

To determine whether this c.1073dup (p.Ala359Cysfs*13) mutation was due to independent mutational events at this hotspot, or due to a single distant founder effect mutation in the Iranian population, haplotype analysis was performed using a series of informative single nucleotide polymorphisms (SNPs) around the ANTXR2 gene (Table 2). HapMap data (https://snpinfo.niehs.nih.gov/snpinfo/snptag.html) were used to choose tag SNPs, spanning a region of 2.91 Mb encompassing the ANXTR2 gene. Selected tag SNPs were rs4692955, rs1493177, rs12509909, rs10011562, rs7685006, rs4975133, rs4975132, and rs4975131 (https://www.ncbi.nlm.nih.gov/SNP). Five intragenic SNPs in ANTXR2 were additionally included as tag SNPs: rs13140055, rs4594664, rs11730210, rs4336166, and rs12647691 (https://www.ncbi.nlm.nih.gov/SNP). Overall, 13 SNPs were genotyped by PCR (primer sequences available upon request) and bidirectionally sequenced using an automated sequencer (3730; Applied Biosystems).

Table 2 Haplotype analysis in cases with the c.1073dup mutation in ANTXR2 gene

Haplotype analysis was performed in Cases 3 and 4, as well as in our previously reported Iranian case of HFS with the same c.1073dup (p.Ala359Cysfs*13) mutation [6] (Table 2). Comparison of the haplotypes of these three Iranian patients showed a lack of haplotype conservation for SNPs in this region.

Discussion and conclusions

Collectively, these cases add to the genetic understanding of HFS. Cases with mutations in the exons encoding the cytoplasmic tail of ANTXR2 protein have been reported to be clinically less severe than those with mutations upstream in the gene, which is likely to be due to the role of the cytoplasmic tail in receptor turnover, rather than directly in ligand-binding [9]. While Cases 3 and 4 harbored truncating mutations affecting the cytoplasmic tail, the upstream frameshift mutation in Case 1 is predicted to result in a truncated protein, likely impairing receptor stability in the cellular membrane. Case 2 harbored a mutation in the von Willebrand domain, essential for ligand binding, including to collagen VI [3]. Case 1 and 2 had the most severe phenotypes and earliest lethality, supporting the notion that mutations upstream of the cytoplasmic tail domain result in increased phenotypic severity.

The consequences of specific frameshift mutations in the ANTXR2 gene have previously been examined, including the c.1073dup (p.Ala359Cysfs*13) mutation disclosed in Cases 3 and 4 [10]. Patients who are homozygous for this mutation have ANTXR2 mRNA expression levels that are half of those found in normal individuals [11]. However, the protein product of this mutated form of ANTXR2 has been shown to reach the cell membrane with partial functionality [10]. Thus, this mutation is a candidate for targeted inhibition of nonsense-mediated mRNA decay. The c.1073dup (p.Ala359Cysfs*13) mutation has been reported multiple times previously, and understanding the inheritance and epidemiology of this mutation in different populations is important in determining which HFS patients would benefit from potential therapy. While this mutation could be expected to have developed in isolated mutational events in most cases, we sought to determine the inheritance pattern of this mutation in a highly consanguineous population of Iran, in which inheritance of recurrent mutations causing recessive diseases is often due to transmission of a mutated haplotype from a common ancestor through generations, i.e., the founder effect. The haplotype analysis of our three patients showed that the haplotypes harboring this mutation are identical by state, but not by descent, which supports the notion that this mutation is not the result of a founder effect in this Iranian cohort. Consequently, this mutation may occur sporadically in different populations, and therapeutic approaches targeting this mutation would likely have benefits in any population affected by HFS.

This study reports a novel causative mutation in HFS, contributes to our understanding of genotype-phenotype correlations for this syndrome, and improves our understanding of a recurrent ANTXR2 gene mutation that may be a candidate for targeted therapy.