Deficiency in the endocytic adaptor proteins PHETA1/2 impairs renal and craniofacial development

ABSTRACT A critical barrier in the treatment of endosomal and lysosomal diseases is the lack of understanding of the in vivo functions of the putative causative genes. We addressed this by investigating a key pair of endocytic adaptor proteins, PH domain-containing endocytic trafficking adaptor 1 and 2 (PHETA1/2; also known as FAM109A/B, Ses1/2, IPIP27A/B), which interact with the protein product of OCRL, the causative gene for Lowe syndrome. Here, we conducted the first study of PHETA1/2 in vivo, utilizing the zebrafish system. We found that impairment of both zebrafish orthologs, pheta1 and pheta2, disrupted endocytosis and ciliogenesis in renal tissues. In addition, pheta1/2 mutant animals exhibited reduced jaw size and delayed chondrocyte differentiation, indicating a role in craniofacial development. Deficiency of pheta1/2 resulted in dysregulation of cathepsin K, which led to an increased abundance of type II collagen in craniofacial cartilages, a marker of immature cartilage extracellular matrix. Cathepsin K inhibition rescued the craniofacial phenotypes in the pheta1/2 double mutants. The abnormal renal and craniofacial phenotypes in the pheta1/2 mutant animals were consistent with the clinical presentation of a patient with a de novo arginine (R) to cysteine (C) variant (R6C) of PHETA1. Expressing the patient-specific variant in zebrafish exacerbated craniofacial deficits, suggesting that the R6C allele acts in a dominant-negative manner. Together, these results provide insights into the in vivo roles of PHETA1/2 and suggest that the R6C variant is contributory to the pathogenesis of disease in the patient. This article has an associated First Person interview with the first author of the paper.


Figure S1 .
Figure S1.pheta1 and pheta2 transcripts are maternally inherited and expressed during development.(A) RT-PCR for pheta1, pheta2, and rpl4 (a housekeeping reference gene) was performed from total RNA extracted from wild-type 512-cell, 1 dpf, and 3 dpf zebrafish.pheta1/2 expression was detected in all stages, with the amount of transcript being the highest at the 512-cell stage, indicating high levels of maternal expression.DNA ladders are shown in the left-most lanes.DNA band sizes are as indicated by arrowheads.(B-C'') Whole-mount in situ hybridization for pheta1 at the 1-cell and 1 dpf stage.Hybridization with the anti-sense probe (B-B'') showed that pheta1 is expressed maternally (1-cell) and during development (1 dpf).The sense probe serves as negative control (C-C'').Boxed regions in B' and C' are shown in higher magnification in B'' and C'', respectively.Scale=100 µm.

Figure S2 .
Figure S2.Loss of pheta1/2 has no effect on ciliogenesis in the inner ear, olfactory placode, lateral line, and photoreceptors.(A) Representative confocal images of cilia in WT and dKO in the inner ear, the olfactory placode, and the lateral line.Cilia are labeled with anti-acetylated α-tubulin (green), basal bodies labeled with anti-γ tubulin (red), and nuclei labeled with DAPI (blue).Scale bar=25 µm.(B)Zpr1 (green) and zpr3 (green) antibodies were used to label cones and rods, respectively.Scale bar=50 µm.

Figure S3 .
Figure S3.Loss of pheta1/2 has no effect on OKR.(A) A zebrafish larva is placed in an arena with moving black and white gratings.An infrared camera records the position and speed of each eye during the visual stimulation.(B) Velocity of tracking movements in response to moving gradients at various contrasts at 5-6 dpf.(C) Correlation in angle and velocity between left and right eye.Error=SEM.

Figure S4 .
Figure S4.Ceratohyal ossification and chondrogenesis marker analysis.(A) Alizarin Red stained larvae, viewed from the ventral side.Bilateral bone collars in WT and pheta1-/-animals are indicated by arrowheads.The majority of pheta2-/-and dKO animals lack bilateral bone collars at this stage.The number of animals imaged with the displayed phenotype is shown in the upper right corner of each image.(B-E) RNA-seq normalized transcript counts for acana (B), dcn (C), col2a1a (D), and sox9a (E) at 5 dpf.**:p<0.01.(F) Fluorescent in situ hybridization for sox9a.Images show confocal maximum intensity projections, viewed from the ventral side.The number of animals imaged with the displayed phenotype is shown in the lower-left corner of each image.Error bar=100 µm.Abbreviations: ch, ceratohyal cartilage; m, Meckel's cartilage.

Figure S6 .
Figure S6.The effects of Tg(R6C) on cranial distance and ceratohyal length.Cranial distance (A) and ceratohyal length (B) in pheta1 +/-and pheta1 -/- backgrounds at 6 dpf, with and without the Tg(R6C) transgene.Measured structures are shown in the schematics on the left.t-test p values are as shown above graphs.

Figure S7 .
Figure S7.Splice site analysis of the PHF6 variant in the UDP patient.Sequence chromatograms showing the normal allele (upper panel) and mutant allele (lower panel) of the UDP patient (i.e., proband).There was no splice defect except the in frame deletion of Leu244 at the exon 7 and 8 boundary (marked in red).