A novel human pain insensitivity disorder caused by a point mutation in ZFHX2

Abstract

Chronic pain is a major global public health issue causing a severe impact on both the quality of life for sufferers and the wider economy. Despite the significant clinical burden, little progress has been made in terms of therapeutic development. A unique approach to identifying new human-validated analgesic drug targets is to study rare families with inherited pain insensitivity. Here we have analysed an otherwise normal family where six affected individuals display a pain insensitive phenotype that is characterized by hyposensitivity to noxious heat and painless bone fractures. This autosomal dominant disorder is found in three generations and is not associated with a peripheral neuropathy. A novel point mutation in ZFHX2, encoding a putative transcription factor expressed in small diameter sensory neurons, was identified by whole exome sequencing that segregates with the pain insensitivity. The mutation is predicted to change an evolutionarily highly conserved arginine residue 1913 to a lysine within a homeodomain. Bacterial artificial chromosome (BAC) transgenic mice bearing the orthologous murine p.R1907K mutation, as well as Zfhx2 null mutant mice, have significant deficits in pain sensitivity. Gene expression analyses in dorsal root ganglia from mutant and wild-type mice show altered expression of genes implicated in peripheral pain mechanisms. The ZFHX2 variant and downstream regulated genes associated with a human pain-insensitive phenotype are therefore potential novel targets for the development of new analgesic drugs.awx326media15680039660001.

Keywords: Mendelian; dorsal root ganglia; pain insensitivity; transcription factor.

Figure 1

Missense mutation in ZFHX2 identified in pain insensitive family. (A) Marsili syndrome pedigree showing autosomal dominant inheritance pattern of the pain insensitive phenotype. (B) Microphotograph of a skin punch biopsy of the proband (Individual II-4) carrying the mutation. PGP9.5 positive profiles (in red; shown by white arrows) were counted as they crossed the dermal-epidermal junction (dotted line). Using the immunofluorescence method, 11.3 fibres/mm were counted and 6.4 fibres/mm using DAB (not shown). The intraepidermal nerve fibre density (IENFD) count was normal for age and gender. Scale bar = 50 μm. (C) Sanger sequencing trace showing the ZFHX2 mutation (NM_033400:c.G5738A) that was PCR-amplified from genomic DNA isolated from the proband. All affected individuals are heterozygous for this mutation. (D) Schematic representation of ZFHX2 protein domain structure with the zinc finger motifs, PRDM9-like domain (aa 1033–1277) and three homeodomains (aa 1595–1657, 1857–1919 and 2065–2127) annotated. The location of the p.R1913K mutation within the second homeodomain is also indicated. See Supplementary Fig. 1 for more information. (E) Sequence alignment around mutation site for orthologous Tetrapoda ZFHX2 proteins.

Figure 2

Zfhx2 is highly expressed in DRG within peripherin-positive neurons. (A) Real-time qPCR assay measuring the expression level of Zfhx2 in specific tissues from adult wild-type mice (n = 3). (B) Immunocytochemical analysis of PFA fixed dissociated DRG derived from an adult C57BL/6 wild-type mouse. DAPI staining (in blue), rabbit polyclonal anti-ZFHX2 (in red), mouse monoclonal anti-peripherin (in green). ZFHX2 localizes to the nucleus of peripherin-positive small diameter DRG neurons. Scale bar = 10 μm.

Figure 3

Zfhx2 knockout mice have altered acute mechanical and thermal pain thresholds. (A) Randall Selitto test measuring withdrawal thresholds in the tail to noxious mechanical stimuli in knockout (KO; n = 15) and wild-type (WT) littermates (n = 13). Knockout mice are significantly hyposensitive compared to controls (P = 0.0017). (B) Hot plate test measuring withdrawal latency to noxious thermal (50°C) stimuli in knockout (n = 31) and wild-type littermate (n = 31) controls. Knockout mice are significantly hypersensitive compared to controls (P = 0.000008). (C) ZFHX2 knockout deep dorsal horn WDR neurons show a deficit in noxious mechanical coding compared to littermate controls but no differences in response to dynamic low threshold stimuli. Brush (wild-type n = 44, knockout n = 61, P = 0.176); Prod 100 g/cm² (wild-type n = 43, knockout n = 61, P = 0.01); Prod 150 g/cm² (wild-type n = 43, knockout n = 61, P = 0.044). All data analysed by t-test. Results are presented as mean ± standard error of the mean (SEM); ns (not significant) P > 0.05; *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001.

Figure 4

ZFHX2 p.R1907K BAC transgenic mice are hyposensitive to noxious heat and DRG neurons have reduced responses to capsaicin. (A) Hargreaves’ test measuring paw withdrawal latency to noxious thermal stimuli. Wild-type (WT) n = 16, mutant (MUT) n = 16 (genomic BAC copy number 1–5), P = 0.0048. (B) Mice with the highest genomic BAC copy number (four to five copies) have a high pain threshold to noxious heat. Wild-type n = 16, mutant n = 5, P = 0.000058. (C) Hot-plate test measuring withdrawal latency to noxious thermal (50°C) stimuli in BAC transgenic mice (mutant n = 16) and wild-type littermate (n = 16) controls. P = 0.058. (D) Similar to the Hargreaves’ test, mice with the highest genomic BAC copy number (four to five copies) have a high pain threshold to noxious heat. Wild-type n = 16, mutant n = 5, P = 0.00095. (E) Example confocal images from cultured wild-type and mutant mouse DRG neurons before and after the application of capsaicin (1 µM). Scale bar = 50 µm. (F) Averaged response of all wild-type (n = 129) and mutant (n = 138) mouse DRG neurons to capsaicin (1 µM) application. (G) Maximal relative fluorescence from baseline for wild-type and mutant mouse DRG neurons following capsaicin (1 µM) application. P = 0.0000002. All data analysed by t-test. Results are presented as mean ± SEM; ns P > 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001.

Figure 5

Potential transcriptional targets of ZFHX2. (A) The consensus AG-rich motif derived from promoter regions of deregulated genes in the microarray screen. The motif has been constructed from five common motifs highly represented in 1000-bp upstream regions (Supplementary Fig. 6). (B) The top six GO terms for genes that contain the AG-rich motif. For more information see Supplementary Table 4. The genes with the biggest change in their expression levels (P < 0.01). Zfhx2 expression level is shown as a positive control as it is expected to rise after introducing extra copies of the Zfhx2^mut gene. Some of these genes have shown an enrichment of ZFHX2 binding at their regulatory upstream regions (as indicated in right column) as analysed by ChIP-seq analysis.