Abstract
Mirror movements (MM) disorder is characterized by involuntary movements on one side of the body that mirror intentional movements on the opposite side. We performed genetic characterization of a family with autosomal dominant MM and identified ARHGEF7, a RhoGEF, as a candidate MM gene. We found that Arhgef7 and its partner Git1 bind directly to Dcc. Dcc is the receptor for Netrin-1, an axon guidance cue that attracts commissural axons to the midline, promoting the midline crossing of axon tracts. We show that Arhgef7 and Git1 are required for Netrin-1–mediated axon guidance and act as a multifunctional effector complex. Arhgef7/Git1 activates Rac1 and Cdc42 and inhibits Arf1 downstream of Netrin-1. Furthermore, Arhgef7/Git1, via Arf1, mediates the Netrin-1–induced increase in cell surface Dcc. Mice heterozygous for Arhgef7 have defects in commissural axon trajectories and increased symmetrical paw placements during skilled walking, a MM-like phenotype. Thus, we have delineated how ARHGEF7 mutation causes MM.
INTRODUCTION
Congenital mirror movements (MM) is a neurological condition in which affected individuals have involuntary movement of a body part that mirrors the intentional movement of the contralateral homologous body part (1). MM predominate in the upper limbs, can be associated with pain in the upper limbs during sustained manual activities, and impair the ability of affected individuals to perform tasks requiring skilled bimanual coordination (2). MM result from functional and structural abnormalities in lateralized motor control, including abnormalities in commissural tracts that connect the left and right sides of the brain and spinal cord (3, 4).
We previously found pathogenic variants in DCC (deleted in colorectal carcinoma) in individuals with congenital MM, thus establishing DCC as the first MM gene (5). DCC encodes the receptor for Netrin-1, an axon guidance cue that attracts axons to the midline during development, promoting the midline crossing of commissural axon tracts (6–10). Pathogenic variants in DCC are the most common genetic cause of congenital MM (1, 11); however, pathogenic variants in other genes have also been found to cause MM: NTN1 (encoding Netrin-1) (12), RAD51 (13), and DNAL4 (dynein axonemal light chain 4) (14). Despite these advances, the genetic, molecular, and physiological mechanisms remain poorly understood in most individuals with congenital MM.
In this study, we performed genetic characterization of a large kindred with autosomal dominant MM (15) and identified ARHGEF7 as a candidate MM gene. Given that the known MM genes include DCC and NTN1, which are critical for Netrin-1/DCC axon guidance, and a third MM gene, RAD51, may function downstream of Netrin-1 (16), we hypothesized that ARHGEF7 is also involved in Netrin-1/DCC signaling and is important for axon guidance to the midline.
In the developing spinal cord, commissural neurons send axons that project ventrally toward and subsequently across the floor plate at the ventral midline, forming axon commissures. Netrin-1 attracts commissural axons toward the ventral midline (9, 17–20). Netrin-1 signaling activates the Rho guanosine triphosphatases (GTPases) Rac1 and Cdc42 (21–24). Rho GTPases are molecular switches that orchestrate remodeling of the actin cytoskeleton. The precise spatiotemporal control of Rho GTPases in the growth cone regulates actin cytoskeleton dynamics in response to guidance cues and drives growth cone turning (25, 26).
Rho GTPases are active when bound to guanosine triphosphate (GTP) and inactive when bound to guanosine diphosphate (GDP). Guanine nucleotide exchange factors (GEFs), promote Rho GTPase activity by exchanging GDP with GTP to promote the active GTP-bound state. GEFs also integrate upstream extracellular signals from receptors to activate Rho GTPases in a context-specific manner (27). ARHGEF7 encodes a RhoGEF, which regulates cell polarity, adhesion, and migration (28). ARHGEF7 is a GEF for Rac1 and Cdc42 (29). Therefore, we hypothesized that ARHGEF7 may act in Netrin-1/DCC signaling by activating Rac1 and Cdc42 downstream of Netrin-1 and that defects in ARHGEF7 disrupt Netrin-1/DCC signaling and cause defects in motor control lateralization. In this study, we show that ARHGEF7WT but not ARHGEF7mut, an ARHGEF7 variant we found in MM individuals, directly binds to Dcc. Moreover, the ARHGEF7 binding partner Git1, an ArfGAP (adenosine diphosphate–ribosylation factor GTPase-activating protein), also directly binds to Dcc. We show that Arhgef7 activates Rac1 and Cdc42 and inhibits Arf1 downstream of Netrin-1. We also found that Arhgef7 is required for Netrin-1–mediated commissural axon guidance in vitro and for the correct trajectory of commissural axons in vivo. Mice heterozygous for Arhgef7, a model for the MM individuals studied herein who have a loss-of-function variant in one allele of ARHGEF7, displayed abnormal symmetric movements during skilled walking, recapitulating some aspects of MM. Thus, through human genetics, we have identified ARHGEF7 as a component of Netrin-1/Dcc signaling. Disruption of ARHGEF7 impairs Netrin-1/Dcc signaling and axon guidance, resulting in MM.
RESULTS
Familial MM is associated with a variant of ARHGEF7
To discover previously unidentified MM genes, we examined a large four-generation family with nonsyndromic congenital MM (Fig. 1A) (15). Pedigree analysis suggested autosomal dominant inheritance with incomplete penetrance. Previous characterization of the affected individuals by clinical assessment, neurophysiology, and neuroimaging showed that they have a similar neurophysiological profile to individuals with MM resulting from DCC or RAD51 pathogenic variants (5, 15, 30, 31). However, sequencing of DCC, RAD51, NTN1, and DNAL4 did not reveal any pathogenic variants in the affected individuals (15), suggesting that defects in other, yet to be identified, gene(s), cause MM in this family. Therefore, we performed whole-exome sequencing to sequence all the protein coding exons and their flanking regions of six affected members (Fig. 1A, individuals II-2, II-11, III-4, III-8, III-13, and IV-6) and one obligate carrier (II-7) of this family. We found only one predicted damaging variant in the exome data that is absent in the Genome Aggregation Database (gnomAD) and present in all seven affected individuals: a frameshift variant, c.1751_1752del, p.(Asn584Thrfs*90), in ARHGEF7 (NM_001113511.1) (Fig. 1, A and B). ARHGEF7 has a probability of loss-of-function intolerance score of 1, and a loss-of-function observed/expected upper bound fraction score of 0.07 in gnomAD (v2.1.1), indicating a strong intolerance to loss-of-function mutations (32). We then screened for presence of the ARHGEF7 variant in three additional family members (III-1, III-10, and III-14). The variant segregated with the phenotype….
