KCTD: A new gene family involved in neurodevelopmental and neuropsychiatric disorders

Abstract

The underlying molecular basis for neurodevelopmental or neuropsychiatric disorders is not known. In contrast, mechanistic understanding of other brain disorders including neurodegeneration has advanced considerably. Yet, these do not approach the knowledge accrued for many cancers with precision therapeutics acting on well-characterized targets. Although the identification of genes responsible for neurodevelopmental and neuropsychiatric disorders remains a major obstacle, the few causally associated genes are ripe for discovery by focusing efforts to dissect their mechanisms. Here, we make a case for delving into mechanisms of the poorly characterized human KCTD gene family. Varying levels of evidence support their roles in neurocognitive disorders (KCTD3), neurodevelopmental disease (KCTD7), bipolar disorder (KCTD12), autism and schizophrenia (KCTD13), movement disorders (KCTD17), cancer (KCTD11), and obesity (KCTD15). Collective knowledge about these genes adds enhanced value, and critical insights into potential disease mechanisms have come from unexpected sources. Translation of basic research on the KCTD-related yeast protein Whi2 has revealed roles in nutrient signaling to mTORC1 (KCTD11) and an autophagy-lysosome pathway affecting mitochondria (KCTD7). Recent biochemical and structure-based studies (KCTD12, KCTD13, KCTD16) reveal mechanisms of regulating membrane channel activities through modulation of distinct GTPases. We explore how these seemingly varied functions may be disease related.

Keywords: KCTD11; KCTD13; KCTD7; Neurodegeneration; Neurodevelopmental disorders.

Figure 1

The diverse human KCTD protein family and yeast Whi2. Line diagrams of the 25 human KCTD family proteins and Saccharomyces cerevisiae Whi2 are drawn to scale, grouped in color‐coded clades (A‐H), ordered as in Figure 2, and aligned with respect to their BTB domain (solid rectangles). Additional protein domains with known or inferred structures (KHA, YjbI, WD40, H1) and similarity region H2 are also represented. KCTD11L starts at an AUU start codon adding 39 N‐terminal residues (hashed box) before the first in‐frame AUG translate start. Gray line diagrams indicate proteins not discussed in detail. Scale bar indicates protein length in amino acid residues

Figure 2

Phylogenetic tree of isolated BTB domains from KCTD family homologs. Amino acid sequences of KCTD family proteins from human (Homo sapiens, HOMSA), mouse (Mus musculus, MUSMU), zebrafish (Danio rerio, DANRE), Drosophila melanogaster (DROME), Caenorhabditis elegans (CAEEL), and three yeast species (Saccharomyces cerevisiae, SACCE; Schizosaccharomyces pombe, SACPO; Candida albicans, CANAL) were collected from UniProt (release 2019_02) or after searches using the DELTA‐BLAST algorithm on the NCBI website. Sequences were aligned using MAFFT (version 7), and a neighbor‐joining (NJ) analysis was performed with 1000 bootstrap replicates. Bootstrap support values above 50 are shown at each node. The tree was rooted using Whi2p from S pombe. Yeast sequences were represented as an outgroup (red branches). The arbitrary cluster designations for groups A‐G were assigned to match those reported by Skoblov et al.1 The new H group is deduced from this analysis. Compared to Skoblov et al,1 we found that KCTD9 segregates within group E. Amino acid sequences (Table S1) and alignment results (Table S2) for this analysis are found in Supporting information

Figure 3

Proposed role for a subset of KCTD family proteins as adaptors for cullin‐3 ubiquitin ligase complexes (CRLs)

Figure 4

Altered mitochondrial morphology in KCTD7 mutant patient fibroblasts. Primary passage‐matched human fibroblasts from (A) an age‐matched control and (B) a patient with compound heterozygous R84W/D106fs mutations in KCTD7 were confirmed by Sanger sequencing and qRT‐PCR analysis as described.13 To visualize mitochondrial organelles, cells grown on round 12‐mm‐diameter glass coverslips (FisherBrand) were fixed (10 min in cold 4% paraformaldehyde), permeabilized (5 min with 0.2% Triton X‐100) and immunostained 1 h with anti‐Tom20 antibody and Alexa Fluor® secondary antibodies (Santa Cruz), mounted in Prolong Gold, and 0.5 μmol/L Z‐stack images were captured on a Nikon 90i at 40x or 60x magnification using Volocity software for deconvolution. (For quantification, mitochondria in some experiments were labeled instead with 100 nmol/L Mitotracker Red for 15 min prior to fixation.) Double‐blinded images were converted to 8‐bit grayscale, binarized and skeletonized using a custom ImageJ plug‐in, and mitochondrial structure parameters (including length, size, branching, degree of clustering, circularity) were quantified using “Analyze Skeleton 2D/3D” ImageJ plug‐in for 2‐3 independent experiments. The total mitochondrial network per cells was significantly reduced in long‐branch frequency in KCTD7 mutant fibroblast compared to control fibroblast. Individual mitochondrial subnetworks (skeletons) are rainbow colored according to total length (red longest, blue shortest). Position of the nucleus in each cell is marked by a gray circle