Structure of the amino terminus of a gap junction protein

Amino terminal glutamate residues confer spermine sensitivity and affect voltage gating and channel conductance of rat connexin40 gap junctions

Connexin40 (Cx40) contains a specific binding site for spermine (affinity approximately 100 microm) whereas connexin43 (Cx43) is unaffected by identical concentrations of intracellular spermine. Replacement of two unique glutamate residues, E9 and E13, from the cytoplasmic amino terminal domain of Cx40 with the corresponding lysine residues from Cx43 eliminated the block by 2 mm spermine, reduced the transjunctional voltage (V(j)) gating sensitivity, and reduced the unitary conductance of this Cx40E9,13K gap junction channel protein. The single point mutations, Cx40E9K and Cx40E13K, predominantly affected the residual conductance state (G(min)) and V(j) gating properties, respectively. Heterotypic pairing of Cx40E9,13K with wild-type Cx40 in murine neuro2A (N2A) cells produced a strongly rectifying gap junction reminiscent of the inward rectification properties of the Kir (e.g. Kir2.x) family of potassium channels. The reciprocal Cx43K9,13E mutant protein exhibited reduced V(j) sensitivity, but displayed much less rectification in heterotypic pairings with wtCx43, negligible changes in the unitary channel conductance, and remained insensitive to spermine block. These data indicate that the connexin40 amino terminus may form a critical cytoplasmic pore-forming domain that serves as the receptor for V(j)-dependent closure and block by intracellular polyamines. Functional reciprocity between Cx40 and Cx43 gap junctions involves other amino acid residues in addition to the E or K 9 and 13 loci located on the amino terminal domain of these two connexins.

Specific amino-acid residues in the N-terminus and TM3 implicated in channel function and oligomerization compatibility of connexin43

To identify signals that convey connexin oligomerization compatibility, we have aligned amino-acid sequences of alpha and beta group connexins (Cx) and compared the physico-chemical properties of each homologous amino-acid residue. Four positions were identified that consistently differed between alpha and beta-type connexins; two are located in the N-terminal domain (P1 and P2, corresponding to residues 12 and 13 of the Cx43 sequence), and two in the third trans-membrane-spanning domain TM3 (P3 and P4, corresponding to residues 152 and 153 of the Cx43 sequence). Replacement of each of these residues in Cx43 (an alpha-type connexin) with the corresponding residues of Cx32 (a beta-type connexin) resulted in the assembly of all variants into gap junctions; however, only the P4 variant was functional, as indicated by lucifer yellow dye transfer assays. The other three variants exerted a moderate to severe dose-dependent, dominant-negative effect on co-expressed wild-type (wt) Cx43 channel activity. Moreover, a significant dose-dependent, trans-dominant inhibition of channel activity was observed when either one of the N-terminal variants was co-expressed with wt Cx32. Assembly analyses indicated that dominant and trans-dominant inhibitory effects appeared to be based on the oligomerization of wt and variant connexins into mixed connexons. Interestingly, the identified N-terminal amino acids coincide with the position of naturally occurring, disease-causing missense mutations of several beta-connexin genes (Cx26, Cx30, Cx31, Cx32). Our results demonstrate that three of the identified discriminative amino-acid residues (positions 12, 13 and 152) are crucial for Cx43 channel function and suggest that the N-terminal amino-acid residues at position 12/13 are involved in the oligomerization compatibility of alpha and beta connexins.

Divergent effects of two sequence variants of GJB3 (G12D and R32W) on the function of connexin 31 in vitro

Recently, we identified several missense mutations of the connexin gene GJB3 encoding connexin 31 (Cx31) in erythrokeratodermia variabilis (EKV), an autosomal dominant skin disorder. These mutations include G12D, which replaces a conserved glycine residue in the amino-terminus of Cx31 and is associated with a severe EKV phenotype. In contrast, the biologic relevance of the GJB3 sequence variant R32W located in the first transmembrane domain of Cx31 is disputed. To examine the effects of these sequence variants on Cx31 biogenesis and gap junction activity we expressed wild type and mutant Cx31-Flag constructs in HeLa cells. Using immunostaining, all expression variants were detected in the cytoplasm and in a punctate pattern at the cell surface, indicating that G12D and R32W did not interfere with either protein synthesis or transport to the cell membrane. Similarly, oligomerization into hemichannels appeared not impaired when expressing either Cx31 mutant as assessed by size exclusion chromatography, immunoblotting and immunostaining. However, dye transfer experiments and monitoring of intracellular calcium levels in response to serum stimulation revealed that G12D-Cx31 did not form functional gap junction channels, probably due to incorrect assembly or altered properties of Cx31 channels. In contrast, intercellular coupling between cells expressing R32W-Cx31 was comparable to that of wtCx31, suggesting that R32W is a functionally inconsequential polymorphism of Cx31.

Genetic heterogeneity in erythrokeratodermia variabilis: novel mutations in the connexin gene GJB4 (Cx30.3) and genotype-phenotype correlations

Erythrokeratodermia variabilis is an autosomal dominant genodermatosis characterized by persistent plaque-like or generalized hyperkeratosis and transient red patches of variable size, shape, and location. The disorder maps to a cluster of connexin genes on chromosome 1p34-p35.1 and, in a subset of families, results from mutations in the gene GJB3 encoding the gap junction protein connexin-31 (Cx31). A recent report suggested the involvement of another connexin gene (GJB4) in the etiology of erythrokeratodermia variabilis. In this study, we sequenced the coding region of GJB4 in 13 unrelated erythrokeratodermia variabilis families without detectable mutations in GJB3. Mutation analysis revealed six distinct missense mutations in five families and a sporadic case of erythrokeratodermia variabilis, all of which were not found in controls. Mutation G12D, identified in an extended Dutch family, lies in the predicted amino-terminus and may interfere with the flexibility of this domain, connexin selectivity, or gating polarity of gap junction channels. Other mutations (R22H, T85P, F137L, F189Y) were located in the transmembrane domains of Cx30.3, and are predicted to hinder regulation of voltage gating or alter the kinetics of channel closure. Affected individuals of two unrelated families harbored point mutations leading to amino acid substitution F137L, which was also reported in GJB3, yet the extent and severity of hyperkeratosis was milder compared to the corresponding mutation in GJB3. Two mutations (T85P, F137L) were associated with the occurrence of rapidly changing erythematous patches with prominent, circinate, or gyrate borders in affected children but not in adults, supporting the notion that this feature is specific to Cx30.3 defects. Nevertheless, we observed highly variable intrafamilial phenotypes, suggesting the strong influence of modifying genetic and epigenetic factors. In addition to pathogenic mutations, we identified several missense mutations and a 4 bp deletion within the GJB4 coding region, which might represent either inconsequential polymorphisms or recessive mutations. In conclusion, our results demonstrate genetic heterogeneity in erythrokeratodermia variabilis, and emphasize that intercellular communication mediated by both Cx31 and Cx30.3 is crucial for epidermal differentiation.

pH-dependent intramolecular binding and structure involving Cx43 cytoplasmic domains

pH-induced closure of connexin43 (Cx43) channels involves interaction of the Cx43 carboxyl-terminal (Cx43CT) with a separate "receptor" domain. The receptor location and structure and whether the interaction is directly intramolecular are unknown. Here we show resonant mirror technology, enzyme-linked sorbent assays, and nuclear magnetic resonance (NMR) experiments demonstrating pH-dependent binding of Cx43CT to region 119-144 of Cx43 (Cx43L2), which we propose is the receptor. NMR showed that acidification induced alpha-helical order in Cx43L2, whereas only a minor modification in Cx43CT structure was detected. These data provide the first demonstration of chemically induced structural order and binding between cytoplasmic connexin domains.

Expression of a connexin31 mutation causing erythrokeratodermia variabilis is lethal for HeLa cells

The autosomal dominant skin disorder erythrokeratodermia variabilis (EKV) has been linked to mutations in the human connexin31 (hCx31) gene, which is expressed in the epidermis. We characterized and compared a pathogenic mutation resulting in replacement of amino acid glycine 12 with arginine (G12R) with wild-type hCx31 protein. HeLa cells were transfected with wild-type and mutant hCx31 cDNA, respectively, using different-constitutive and inducible-vector systems. Independent of the expression vector, wild-type and mutant hCx31 were expressed at comparative levels and localized at the plasma membranes. Mutated channels (hCx31G12R) showed higher conductance in dye coupling studies than wild type channels. Furthermore, HeLa cells died within 5 days after constitutive expression of the mutant protein. Using an inducible expression system, we demonstrated a direct correlation between survival/life span of transfected HeLa cells and expression level of the mutant protein, indicating a gain-of-function mechanism due to a defective channel closure mechanism.

Missense mutations in GJB2 encoding connexin-26 cause the ectodermal dysplasia keratitis-ichthyosis-deafness syndrome

Keratitis-ichthyosis-deafness syndrome (KID) is a rare ectodermal dysplasia characterized by vascularizing keratitis, profound sensorineural hearing loss (SNHL), and progressive erythrokeratoderma, a clinical triad that indicates a failure in development and differentiation of multiple stratifying epithelia. Here, we provide compelling evidence that KID is caused by heterozygous missense mutations in the connexin-26 gene, GJB2. In each of 10 patients with KID, we identified a point mutation leading to substitution of conserved residues in the cytoplasmic amino terminus or first extracellular domain of Cx26. One of these mutations was detected in six unrelated sporadic case subjects and also segregated in one family with vertical transmission of KID. These results indicate the presence of a common, recurrent mutation and establish its autosomal dominant nature. Cx26 and the closely related Cx30 showed differential expression in epidermal, adnexal, and corneal epithelia but were not significantly altered in lesional skin. However, mutant Cx26 was incapable of inducing intercellular coupling in vitro, which indicates its functional impairment. Our data reveal striking genotype-phenotype correlations and demonstrate that dominant GJB2 mutations can disturb the gap junction system of one or several ectodermal epithelia, thereby producing multiple phenotypes: nonsyndromic SNHL, syndromic SNHL with palmoplantar keratoderma, and KID. Decreased host defense and increased carcinogenic potential in KID illustrate that gap junction communication plays not only a crucial role in epithelial homeostasis and differentiation but also in immune response and epidermal carcinogenesis.

Expression of gap junction genes connexin32 and connexin43 mRNAs and proteins, and their role in hepatocarcinogenesis

AIM

To investigate the relationship between hepatocarcinogenesis and the expression of connexin32 (cx32), connexin43 (cx43) mRNAs and proteins in vitro.

METHODS

Gap junction genes cx32 and cx43 mRNA in hepatocellular carcinoma cell lines HHCC, SMMC-7721 and normal liver cell line QZG were detected by in situ hybridization (ISH) with digoxin-labeled cx32, and cx43 cDNA probes. Expression of Cx32 and Cx43 proteins in the cell lines was revealed by indirect immuno-fluorescence and flow cytometry (FCM).

RESULTS

Blue positive hybridization signals of cx32 and cx43 mRNAs detected by ISH with cx32 and cx43 cDNA probes respectively were located in cytoplasm of cells of HHCC, SMMC-7721 and QZG. No significant difference of either cx32 mRNA or cx43 mRNA was tested among HHCC, SMMC-7721 and QZG (P=2.673, HHCC vs QZG; P=1.375, SMMC-7721 vs QZG). FCM assay showed that the positive rates of Cx32 protein in HHCC, SMMC-7721 and QZG were 0.7%, 1.7% and 99.0%, and the positive rates of Cx43 protein in HHCC, SMMC-7721 and QZG were 7.3%, 26.5% and 99.1% respectively. Significant differences of both Cx32 and Cx43 protein expression existed between hepatocellular carcinoma cell lines and normal liver cell line (P=0.0069, HHCC vs QZG; P=0.0087, SMMC-7721 vs QZG). Moreover, the fluorescent intensities of Cx32 and Cx43 proteins in HHCC, SMMC-7721 were lower than that in QZG.

CONCLUSIONS

Hepatocellular carcinoma cell lines HHCC and SMMC-7721 exhibited lower positive rates and fluorescent intensities of Cx32, Cx43 proteins compared with that of normal liver cell line QZG. It is suggested that lower expression of both Cx32 and Cx43 proteins in hepatocellular carcinoma cells could play pivotal roles in the hepatocarcinogenesis. Besides, genetic defects of cx32 and cx43 in post-translational processing should be considered.

Reversal of the gating polarity of gap junctions by negative charge substitutions in the N-terminus of connexin 32

Intercellular channels formed by connexins (gap junctions) are sensitive to the application of transjunctional voltage (V(j)), to which they gate by the separate actions of their serially arranged hemichannels (Harris, A. L., D. C. Spray, and M. V. L. Bennett. 1981. J. Gen. Physiol. 77:95-117). Single channel studies of both intercellular and conductive hemichannels have demonstrated the existence of two separate gating mechanisms, termed "V(j)-gating" and "loop gating" (Trexler, E. B., M. V. L. Bennett, T. A. Bargiello, and V. K. Verselis. 1996. Proc. Natl. Acad. Sci. U.S.A. 93:5836-5841). In Cx32 hemichannels, V(j)-gating occurs at negative V(j) (Oh, S., J. B. Rubin, M. V. L. Bennett, V. K. Verselis, and T. A. Bargiello. 1999. J. Gen. Physiol. 114:339-364; Oh, S., C. K. Abrams, V. K. Verselis, and T. A. Bargiello. 2000. J. Gen. Physiol. 116:13-31). A negative charge substitution at the second amino acid position in the N-terminus reverses the polarity of V(j)-gating of Cx32 hemichannels (Verselis, V. K., C. S. Ginter, and T. A. Bargiello. 1994. Nature. 368:348-351;. J. Gen. Physiol. 116:13-31). We report that placement of a negative charge at the 5th, 8th, 9th, or 10th position can reverse the polarity of Cx32 hemichannel V(j)-gating. We conclude that the 1st through 10th amino acid residues lie within the transjunctional electric field and within the channel pore, as in this position they could sense changes in V(j) and be largely insensitive to changes in absolute membrane potential (V(m)). Conductive hemichannels formed by Cx32*Cx43E1 containing a negatively charged residue at either the 8th or 10th position display bi-polar V(j)-gating; that is, the open probability of hemichannels formed by these connexins is reduced at both positive and negative potentials and is maximal at intermediate voltages. In contrast, Cx32*Cx43E1 hemichannels with negative charges at either the 2nd or 5th positions are uni-polar, closing only at positive V(j). The simplest interpretation of these data is that the Cx32 hemichannel can adopt at least two different open conformations. The 1st-5th residues are located within the electric field in all open channel conformations, while the 8th and 10th residues lie within the electric field in one conformation and outside the electric field in the other conformation.