Charcot-Marie-Tooth (CMT) disease is common, affecting ∼1:2500 people, and CMTX1 is the second most common form (Fridman et al., 2015), representing 7–18% of CMT. (Kleopa, 2011) Mutations in GJB1, the gene encoding the gap-junction forming protein connexin 32 (Cx32), were first reported to cause CMTX1 in 1993. (Bergoffen et al., 1993) Shortly thereafter, Cx32 was identified in both Schwann cells and oligodendrocytes (Scherer et al., 1995), suggesting that loss of function mutations cause cell autonomous effects in myelinating Schwann cells leading to neuropathy. This was subsequently confirmed by expressing the human GJB1 gene with a Schwann cells specific promoter in Gjb1-null/Cx32 knockout (Cx32KO) mice; this transgene prevented demyelination (Scherer et al., 2005). Hundreds of different GJB1 mutations have been identified. (Scherer and Wrabetz, 2008) We (Abrams et al., 2003; Abrams et al., 2017; Abrams et al., 2013; Deschênes et al., 1997; Kleopa et al., 2002; Kleopa et al., 2006; Yum et al., 2002) and others (Martin et al., 2000; Matsuyama et al., 2001) have shown that most Cx32 mutants show abnormal trafficking when expressed in mammalian cell lines.

Recent experiments in mice have provided proof of principal that gene replacement therapy could be a treatment for CMTX1 (Kagiava et al., 2021; Kagiava et al., 2018; Kagiava et al., 2019; Kagiava et al., 2016; Sargiannidou et al., 2015) Most of these experiments have used Cx32KO mice, first described in 1996 (Nelles et al., 1996b), and subsequently characterized as an authentic model of CMTX1. (Anzini et al., 1997a; Scherer et al., 1998a) The similar PNS phenotype of patients with a variety of GJB1 mutations (including deletion of the entire open reading frame (Ainsworth et al., 1998; Lin et al., 1999; Nakagawa et al., 2001) and promoter mutations (Flagiello et al., 1998; Ionasescu et al., 1996b), indicates that disease-associated mutations cause loss of Cx32 function in the PNS. (Shy et al., 2007) However, as treatment paradigms for CMTX1 progress, it is important to have additional models which are more representative of the mutations that have more complicated effects - abnormal trafficking and intracellular accumulation, as well as other functional alterations - since these processes may impact on the efficacy of therapy. Furthermore, careful study of models representing the broader spectrum of the molecular effects of mutations on the function of Cx32 may provide insight into the roles of Cx32 in the biology of normal Schwann cells and oligodendrocytes.

Earlier models of CMTX1 used a random insertion transgenic approach followed by backcrossing into the Cx32KO mouse to produce a model expressing only the mutant form of Cx32, with Schwann cell expression driven by an Mpz or Cnp promotor. These include p.T55I (Sargiannidou et al., 2009), p.R75W (Sargiannidou et al., 2009), p.R142W (Jeng et al., 2006), p.N175D (Papaneophytou et al., 2018), and p.N175fs*68. (Abel et al., 1999) However, these models have the disadvantage of not reflecting the endogenous patterns of gene regulation of Gjb1.

In this study we set out to develop “knock-in” models of mutations that are more representative of the spectrum of cellular dysfunctions seen with GJB1 mutations, using CRISPR/Cas9 gene editing. The primary amino acid sequence of mouse and human Cx32 are almost identical, enabling us to create mouse Gjb1 mutations that correspond to the human GJB1 mutations. A knock-in ensures that the endogenous promotor regulates the expression of the mutation, thereby producing the most authentic models for pre-clinical therapeutic trials and investigations of pathogenic mechanisms. We chose two mutations producing mutant proteins that are localized to intracellular compartments: p.T55I, which is localized to the ER (Kleopa et al., 2002), and p.R75W, which is localized to the Golgi. (Yum et al., 2002) Both mutants fail to form gap junction plaques when expressed exogenously. (Kleopa et al., 2002; Yum et al., 2002) The phenotypes of patients with these mutations have been well described in the literature; (Latour et al., 1997; Panas et al., 2001; Panas et al., 1998a; Silander et al., 1997; Taylor et al., 2003) they have PNS manifestations typical of CMTX1 and some patients with these mutations also have phenotypes that owe to CNS manifestations. (Panas et al., 1998b; Parissis et al., 2017; Taylor et al., 2003).