New insights into the expression and function of neural connexins with transgenic mouse mutants

Abstract

Gap junctions represent direct intercellular conduits between contacting cells. The subunit proteins of these conduits are called connexins. To date, 20 and 21 connexin genes have been described in the mouse and human genome, respectively, many of them represent sequence-orthologous pairs. Targeted deletion of connexin genes in the mouse genome opened new insights into the biological function of these channel forming proteins, which, in some cases, could be correlated to phenotypic abnormalities in humans, suffering from inherited diseases caused by mutations in the corresponding orthologous connexin gene. Replacing the connexin coding DNA by an appropriate reporter gene has clarified in several cases its cell type specific expression in mouse brain. Various studies demonstrated that connexin36 is mainly expressed in interneurons of retina and brain. Targeted deletion of connexin36 evoked a loss of electrical signal transduction and interferes with synchrony which probably leads to defects in visual transmission and memory. Deletion of connexin43 in astrocytes of mouse brain resulted in increased spreading depression consistent with the notion of altered “spatial buffering” of K⁺ ions and glutamate secreted by active neurons. General connexin30-deficiency led to hearing impairment and apoptosis of hair cells, similar to that observed in mice with cochlea specific deletion of connexin26. Reporter gene expression in connexin30-deficient mice indicated that astrocytes in certain brain regions and leptomeningeal as well as ependymal cells are labelled. Reporter gene expression in connexin45- and connexin47-deficient mice was used to reassign connexin45 expression to certain CNS neurons and connexin47 expression to oligodendrocytes.

Introduction

Gap junctions are formed by docking of two hemichannels in contacting plasma membranes which results in conduits between the cytoplasmic compartments of adjacent cells. These gap junction channels allow diffusional exchange of ions, metabolites and second messenger molecules and are designated as electrical synapses when they occur between neurons where they could serve for fast transfer of electrical signals. Each hemichannel (connexon) is assembled of six protein subunits, called connexins (Cx). So far, 20 connexin genes have been described in the mouse genome, whereas 21 connexin genes have been identified in the human genome [44], [153]. Very recently, two orthologous gene sequences were identified in the mouse and human genomic data base that code for connexin like proteins of 23 kDa theoretical molecular mass, but with two instead of three cysteine residues in the predicted extracellular loops [137]. Most, but not all, mouse connexin genes described [44], [136] correspond to a human orthologous connexin gene. Targeted deletion of connexin genes in the mouse genome offers two functional insights: (a) analysis of the functional impact of this connexin gene on molecular physiology of the mouse, and (b) phenotypic comparison of the connexin defective mouse with patients suffering from mutations in the orthologous connexin gene [153].

In order to generate a new transgenic connexin-deficient mouse line, several technical prerequisites have to be fulfilled. Isogenic genomic DNA flanking the connexin gene to be targeted has to be isolated for cloning of the targeting vector. This vector should contain an appropriate selection gene for isolation of homologously recombined ES cell clones and, optionally, a reporter gene which replaces the connexin coding DNA. Homologous recombination of the targeting vector with the corresponding gene in isogenic mouse embryonic stem cells (ES) must be properly monitored before selected ES-cell clones can be injected into blastocysts. These blastocysts are implanted in pseudopregnant female foster mice which subsequently give birth to chimeric litters that have to be back-crossed, in order to identify germ line transmission. General connexin-deficient mouse lines were generated in which the connexin gene was deleted (knock-out) or replaced by either a reporter gene (LacZ, PLAP, eGFP, see Table 1) or a different connexin coding DNA (knock-in) [cf. [114], [146]. Replacement of the connexin coding DNA by a reporter gene turned out to be helpful for determining the cell type specific expression of the corresponding connexin gene. Furthermore, loxP or frt recognition sequences for Cre or FLP recombinases, respectively, allow conditional and cell type specific deletion of a connexin gene, thereby circumventing possible embryonal lethality due to ubiquitous deletion of this connexin gene [20], [36], [146]. In this work, we discuss recent results from the literature and from our own laboratory regarding the use of transgenic mouse mutants to study the cell type specific expression and function of neural connexin genes. In order to explain the implications of the new findings with transgenic mice, we have also reviewed some previous results on connexin expression obtained by other experimental approaches. When connexin antibodies are used to identify gap junctions by immunofluorescence, they react with gap junction plaques in contacting plasma membranes. The assignment of immunofluorescent signals to one of the coupled cells can be rather ambiguous, in particular when cells with long processes are being studied. In this situation, the analysis of a reporter gene that is expressed instead of the connexin gene is often easier to interpret. In addition, connexin-deficient mice can be used as unambiguous negative controls for the specificity of connexin antibodies and in situ hybridization probes.