Editor—The Townes-Brocks syndrome (TBS, MIM 107480) is an autosomal dominantly inherited association of imperforate anus, supernumerary/triphalangeal thumbs, and dysplastic ears. In addition to this, sensorineural or conductive hearing loss, renal malformations, cardiac defects, and mental retardation maybe present in affected subjects. TBS is caused by mutations of the putative zinc finger transcription factor geneSALL1.2 SALL1 has four double zinc finger domains which are evenly distributed over the protein.3 The majority of SALL1mutations identified to date in TBS patients are located 5′ to the first double zinc finger encoding region.4 5 Most mutations (nonsense mutations, small insertions/deletions, and one larger deletion) have been predicted to result in prematurely truncated proteins lacking all double zinc finger domains presumed to be essential for SALL1 gene function or to result in unstable transcripts, thus causing TBS via haploinsufficiency.4 5

TBS is known to overlap phenotypically with other conditions like Goldenhar syndrome, VACTERL association, or oculo-auriculo-vertrebral spectrum.1 However, a SALL1mutation has so far only been reported in one patient with a clinical picture attributable both to Goldenhar syndrome and TBS.6This patient, as well as the TBS patients in whom mutations were detected, showed at least two out of three major criteria for TBS, that is, malformations of the thumbs, ears, or anus.4 5Therefore, the phenotypic spectrum associated withSALL1 mutations seemed to be quite characteristic.

Here we report the first family in which aSALL1 mutation is associated with a phenotype which is different from TBS. The two affected daughters, their affected father, and the unaffected mother (fig 1A) were examined for SALL1 mutations after giving informed consent to the analysis. In all affected subjects, chromosome analysis before DNA studies had shown a normal karyotype. The older girl, now aged 19, was admitted to hospital as a baby because of failure to thrive which turned out to be because of renal failure. On examination, both kidneys were found to be hypoplastic. After treatment, renal function recovered but remained impaired. Besides the renal problems, mildly dysplastic ears with slight overfolding of the superior helices were seen (fig 1D). She also showed pes planus and an unusual bony fusion in one foot. Further findings include mild developmental delay with an IQ of 71 at the age of 18 (verbal 73, performance 74, assessed by WISCII), mild sensorineural hearing loss, mild hypermetropia, gastro-oesophageal reflux resulting from a small hiatus hernia, and chronic abdominal pain. She has no anal or thumb malformation (fig 1E). Her sister, now aged 13, was born with bilateral dysplastic ears (fig1F) and preauricular tags on one side. As a baby, she had bilateral grade 3 vesicoureteric reflux and bilateral hypoplastic kidneys. Her kidney function is mildly impaired. Like her sister, she has hypermetropia and gastro-oesophageal reflux. She also has dental crowding, mild developmental delay (IQ 71), and mild bilateral high frequency hearing loss, but no anal or thumb malformation (fig 1G). The mother has no health problems, whereas her husband shows impaired renal function, based on thin membrane disease with focal glomerulosclerosis, and dysplastic ears. He also has a Barrett ulcer resulting from chronic gastro-oesophageal reflux.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Pedigree of the family reported here. Both daughters (II.1 and II.2) and their affected father (I.1) show the SALL1 mutation (B, upper part), whereas the mother (I.2) shows the wild type sequence (B, lower part). The position of the mutation with respect to the SALL1 protein is shown in (C) possibly resulting in a truncated protein lacking double zinc finger domains 2-4 (zinc fingers are indicated by oval symbols). (D-G) Ears and hands of the affected girls: (D, E) II.1 aged 19, (F, G) II.2 aged 13. Note typical overfolded superior helices (D, F), but absence of thumb malformations (E, G) in both patients, which was also confirmed by x ray (not shown).

Genomic DNA was prepared from peripheral lymphocytes by routine procedures. SALL1 mutation analysis was performed by PCR amplification and direct sequencing of PCR products as described previously.5 Both affected children as well as the affected father had a heterozygous 1819delGSALL1 mutation, which is located in exon 2 between the coding regions for the first and the second double zinc finger unit (fig 1C).

Most SALL1 mutations previously reported in TBS reside in exon 2 5′ of the coding region for the first double zinc finger domain and are predicted to result inSALL1 haploinsufficiency.4 51819delG is yet another short deletion but is located 3′ of the region where most previously known mutations cluster. While this mutation could result in a prematurely terminated SALL1 protein lacking double zinc finger domains 2-4, it remains unclear if the mutated transcript and the corresponding protein are indeed expressed. Mutations at similar positions have been reported by Marlin et al.4 Here, the patients with the mutations 1565delC or 1966del10 showed at least two out of three major criteria for TBS. Both mutations should have a similar effect (translation stop at nucleotides 1624-1626 and 2074-2076) as compared to the mutation reported here (translation stop at nucleotides 1924-1926). Therefore, if proteins were translated from these mutated alleles, all would be missing the double zinc finger domains 2-4. The most likely effect, however, is that such mutations would lead to an unstable transcript and thereby to SALL1haploinsufficiency.

The clinical presentation in the family reported here is quite unusual, since all affected family members do not have thumb or anal malformations. Therefore, TBS could not be diagnosed. While the features in the patients carrying the 1565delC or 1966del10 mutation4 can be explained by haploinsufficiency forSALL1, the clinical picture in the family shown here could suggest that the 1819delG mutation has a different effect. As seen with mutations in GLI3, truncating mutations at similar positions may lead to either Pallister-Hall syndrome (PHS), different polydactylies, or Greig cephalopolysyndactyly syndrome (GCPS), suggesting that some might result in GLI3 haploinsufficiency while others lead to truncated proteins with aberrant functions.7 8

Based on these observations one could speculate that, unlike 1565delC or 1966del10, the 1819delG mutation does not result in an unstable mRNA but in a truncated SALL1 protein. On the other hand, TBS is known to vary even between families with the same mutation, thus pointing to a strong influence of modifying genes or environmental factors. Yet, while families withSALL1 mutations may differ from each other with respect to the occurrence of renal or cardiac malformations, no family has been reported to date in which all affected members have aSALL1 mutation but not TBS.

Patients like those reported here may instead carry the diagnosis of branchio-oto-renal (BOR) syndrome (MIM 113650), as was initially suspected in our patients based on the combination of dysplastic ears, hearing loss, and hypoplastic kidneys. However, BOR syndrome patients commonly present with cup shaped ears, preauricular pits, and branchial fistulae,9 none of which was seen in our patients. In addition, gastro-oesophageal reflux and borderline mental retardation are not typical of BOR syndrome.9 Therefore, our report shows that a phenotypic overlap not only exists between TBS and VACTERL association, oculo-auriculo-vertebral spectrum, or Goldenhar syndrome, but also between TBS and BOR syndrome. SALL1mutation analysis should therefore be considered for patients who present with dysplastic ears, hearing loss, and renal malformations but do not have a causative mutation in the EYA1gene.10