Human organoids for rapid validation of gene variants linked to cochlear malformations

Abstract

Developmental anomalies of the hearing organ, the cochlea, are diagnosed in approximately one-fourth of individuals with congenital. The majority of patients with cochlear malformations remain etiologically undiagnosed due to insufficient knowledge about underlying genes or the inability to make conclusive interpretations of identified genetic variants. We used exome sequencing for the genetic evaluation of hearing loss associated with cochlear malformations in three probands from unrelated families deafness. We subsequently generated monoclonal induced pluripotent stem cell (iPSC) lines, bearing patient-specific knockins and knockouts using CRISPR/Cas9 to assess pathogenicity of candidate variants. We detected FGF3 (p.Arg165Gly) and GREB1L (p.Cys186Arg), variants of uncertain significance in two recognized genes for deafness, and PBXIP1(p.Trp574*) in a candidate gene. Upon differentiation of iPSCs towards inner ear organoids, we observed developmental aberrations in knockout lines compared to their isogenic controls. Patient-specific single nucleotide variants (SNVs) showed similar abnormalities as the knockout lines, functionally supporting their causality in the observed phenotype. Therefore, we present human inner ear organoids as a potential tool to validate the pathogenicity of DNA variants associated with cochlear malformations.

Fig. 1

Pedigrees, variants, and confirmation of pluripotency in edited lines. (a )Graphical representation of participating families, segregating variants, and their location on respective gene and protein. Squares represent males, and circles indicate females. Filled symbols show affected individuals. Arrows point to the probands (p) of the respective families. The mutated residue is mentioned with an arrow in each chromatogram as well as in the protein schematic. (b) Graphical representation of mutated genes, proteins, and localization of variants. (c) Immunostaining of OCT4 and TRA-1-60 in monoclonal lines derived after CRISPR/Cas9 editing in GREB1L^WT, GREB1L^c.556T>C, GREB1L^KO, FGF3^WT, FGF3^c.493 A>G, FGF3^KO, PBXIP1^WT and PBXIP1^c.1722G>A, respectively. Scale bar is 50 µm. SP: Signal Peptide, MBD: Microtubule Binding Domain, CC1: Coil-coiled Domain 1, CC2: Coil-coiled Domain 2, NLS: Nuclear Localization Sequence, PID: PBX1 interacting Domain, ERID: Erα interacting Domain, NES: Nuclear Export Sequence, TM: Transmembrane

Fig. 2

Expression ofGREB1L,FGF3, and PBXIP1in WT and edited lines. (a) qRT-PCR analysis of GREB1L expression levels in GREB1L^WT, GREB1L^c.556T>C, and GREB1L^KO. (b) qRT-PCR analysis of FGF3 expression levels in FGF3^WT, FGF3^c.493 A>G and FGF3^KO. (c) qRT-PCR analysis of PBXIP1 expression levels in PBXIP1^WTand PBXIP1^c.1722G>A. (d,g) Effect of missense p.Cys186Arg variant on GREB1L in iPSCs. (e,h) Effect of missense p.Arg165Gly variant on FGF3 in iPSCs. (f,i) Effect of nonsense variant p.Trp574* showing slight change in signal for PBXIP1 (j) Western blot showing the smaller sized band in mutated PBXIP1. The results are expressed as Mean ± SEM (n = 3), and the statistical difference of p ≤ 0.05 was considered significant. The significant differences are marked with (*) whenever comparisons were made between edited lines and their respective controls. *: p < 0.05; **: p < 0.01; ***: p < 0.001

Fig. 3

Differences of early otic lineage markers and cross-sectional area in WT and edited lines. (a,b) PAX2 and PAX8 quantification in inner ear organoids (IEOs) (n = 3), respectively. The results are expressed as Mean ±SD, and the statistical difference of p ≤ 0.05 was considered significant. The significant differences are marked with (*) whenever comparisons were made between GREB1L^c.556T>C, FGF3^c.493 A>G, FGF3^KO, GREB1L^KO, PBXIP1^c.1722G>A, and their respective controls GREB1L^WT, FGF3^WT, PBXIP1^WT. (c,d) Cross-sectional area analysis of day 25 and day 35 IEO (n = 3), respectively. Area quantifications were done using ImageJ; results are expressed as Mean ± SD, and the statistical difference of p ≤ 0.05 was considered significant. The significant differences are marked with (*) whenever comparisons were made between GREB1L^c.556T>C, FGF3^c.493 A>G, FGF3^KO, GREB1L^KO, PBXIP1^c.1722G>A, and their respective controls GREB1L^WT, FGF3^WT, PBXIP1^WT. (e) Representative images with PAX2/PAX8 immunostaining in GREB1L^WT, GREB1L^c.556T>C, GREB1L^KO, FGF3^WT, FGF3^c.493 A>G, FGF3^KO, PBXIP1^WT and PBXIP1^c.1722G>A. (f) Representative H&E images of GREB1L^WT, GREB1L^c.556T>C, GREB1L^KO, FGF3^WT, FGF3^c.493 A>G, FGF3^KO, PBXIP1^WT and PBXIP1^c.1722G>A IEOs on day 25 and day35. *: p < 0.05; **: p < 0.01; ***: p < 0.001

Fig. 4

Analysis of the presence of hair cell-like populations. (a) MYO7A and SOX2 localization in GREB1L^WT, GREB1L^c.556T>C, GREB1L^KO, FGF3^WT, FGF3^c.493 A>G, FGF3^KO, PBXIP1^WT, and PBXIP1^c.1722G>A. (b-d) Quantification of MYO7A and SOX2 signals in GREB1L^WT, GREB1L^c.556T>C, GREB1L^KO, FGF3^WT, FGF3^c.493 A>G, FGF3^KO, PBXIP1^WT, and PBXIP1^c.1722G>A, respectively (n = 3). The results are expressed as Mean ± SEM, and the statistical difference of p ≤ 0.05 was considered significant. The significant differences are marked with (*) whenever comparisons were made between GREB1L^c.556T>C, FGF3^c.493 A>G, FGF3^KO, GREB1L^KO, PBXIP1^c.1722G>A, and their respective controls GREB1L^WT, FGF3^WT, PBXIP1^WT. *: p < 0.05; **: p < 0.01; ***: p < 0.001