A Gja8 (Cx50) point mutation causes an alteration of α3 connexin (Cx46) in semi-dominant cataracts of Lop10 mice

Human Molecular Genetics, Mar 2002

Mutations of connexin α8 (GJA8 or Cx50) and connexin α3 (GJA3 or Cx46) in humans have been reported to cause cataracts with semi-dominant inheritance patterns. Targeted null mutations in Gja8 and Gja3 in mice cause cataracts with recessive inheritance. The molecular bases for these differences in inheritance patterns and the mechanism for cataractogenesis in these mutants are poorly understood. We recently mapped an autosomal semi-dominant cataract [lens opacity 10 (Lop10)] mutation to mouse chromosome 3 and identified a missense mutation (G→C) in the Gja8 gene, which causes glycine at codon 22 to be replaced with arginine (G22R). Moreover, we demonstrated that the α8 G22R isoform is a loss-of-function mutant for α8, as well as a dominant mutation for reducing the phosphorylated forms of α3 connexin in vivo. To test the hypothesis that the alteration of endogenous α3 connexin in Lop10 mice led to a unique lens phenotype, we generated double mutant offspring between Lop10 and the Gja3tm1 (α3–/–) mice. The double homozygous mutant mice (Lop10/Lop10 α3–/–) showed relatively normal lens cortical fibers compared to the Lop10 mice. A functional impairment of endogenous α3 connexin is therefore partly responsible for cellular phenotypes in the Lop10 mice. This study has provided some novel molecular insights into mouse and human cataractogenesis caused by α8 and α3 mutations. These mouse models will be useful for investigating the mechanistic relationship between gap junction impairment and cataract formation.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://hmg.oxfordjournals.org/content/11/5/507.full.pdf

A Gja8 (Cx50) point mutation causes an alteration of α3 connexin (Cx46) in semi-dominant cataracts of Lop10 mice

Bo Chang 1 3 Xin Wang 1 2 3 Norman L. Hawes 1 3 Ryan Ojakian 1 2 3 Muriel T. Davisson 1 3 Woo-Kuen Lo 0 1 3 Xiaohua Gong 1 2 3 0 Morehouse School of Medicine , Atlanta, GA, USA 1 10550 North Torrey Pines Road, La Jolla, CA 92037, USA 2 Department of Cell Biology, The Scripps Research Institute , IMM10 3 The Jackson Laboratory , Bar Harbor, ME, USA Mutations of connexin 8 (GJA8 or Cx50) and connexin 3 (GJA3 or Cx46) in humans have been reported to cause cataracts with semi-dominant inheritance patterns. Targeted null mutations in Gja8 and Gja3 in mice cause cataracts with recessive inheritance. The molecular bases for these differences in inheritance patterns and the mechanism for cataractogenesis in these mutants are poorly understood. We recently mapped an autosomal semi-dominant cataract [lens opacity 10 (Lop10)] mutation to mouse chromosome 3 and identified a missense mutation (GC) in the Gja8 gene, which causes glycine at codon 22 to be replaced with arginine (G22R). Moreover, we demonstrated that the 8 G22R isoform is a loss-of-function mutant for 8, as well as a dominant mutation for reducing the phosphorylated forms of 3 connexin in vivo. To test the hypothesis that the alteration of endogenous 3 connexin in Lop10 mice led to a unique lens phenotype, we generated double mutant offspring between Lop10 and the Gja3tm1 (3-/-) mice. The double homozygous mutant mice (Lop10/Lop10 3-/-) showed relatively normal lens cortical fibers compared to the Lop10 mice. A functional impairment of endogenous 3 connexin is therefore partly responsible for cellular phenotypes in the Lop10 mice. This study has provided some novel molecular insights into mouse and human cataractogenesis caused by 8 and 3 mutations. These mouse models will be useful for investigating the mechanistic relationship between gap junction impairment and cataract formation. - The mouse lens and eye are remarkably similar in structure to the human lens and eye, and both species have many similar ocular disorders (1). Mouse models for inherited ocular disease allow for rapid genetic analysis, pathophysiological characterization and the understanding of disease processes (2,3). The ocular lens has three unique features: (i) avascular organ structure, (ii) transparency and (iii) continued growth throughout life. The lens is composed of a monolayer of cuboidal epithelial cells covering the anterior surface of a buck of elongated fibers. At an early embryonic stage, the posterior cells of the lens vesicle differentiate to form the primary lens fibers. The growth of the lens is dependent upon the production of secondary fiber cells in the equator from the anterior epithelial cells. The newly differentiated fibers lay on top of the previously differentiated fibers which eventually lose all their intracellular organelles, such as the nucleus, mitochondrion, endoplasmic reticulum, Golgi apparatus, etc., to become mature fibers in the interior region of the lens (4). More than 90% of the soluble proteins in the lens are crystallins. In order to maintain the solubility properties of crystallins and to ensure transparency, the lens must have sophisticated mechanisms to control ionic balance and water content. Hence, an extensive network of gap junctions provides low-resistance pathways for both electrical and metabolic coupling of lens epithelialepithelial, epithelialfiber and fiberfiber cells (5). The development and function of mammalian lenses are associated with the utilization of gap junction channels formed by the products of at least three connexin genes: 1 (Cx43) (6), 3 (Cx46) (7) and 8 (Cx50) (8). Connexin 1 is expressed mainly in the lens epithelial cells. The co-existence of 3 and 8 connexins in the lens fibers has been verified in different vertebrates, including the chick (9), sheep (10), cow (11), mouse (12), rat (8), primate (13) and human (14). Genetic studies of 3 and 8 connexins in humans and mice are summarized below. (i) Three mutations in the 3 connexin and two mutations in the 8 connexin have been reported to cause autosomal dominant cataracts in humans (1518). (ii) Genetically engineered mice with an absence of either the 3 or 8 connexin developed distinctive cataractous phenotypes. Connexin 3 homozygous knockout (/) mice develop variable expressions of nuclear cataracts, based upon the genetic background of the strain (12,19). The connexin 8 homozygous knockout (/) mice have microphthalmia with zonular pulverulent nuclear cataracts (20,21). Both 3 and 8 connexin knockout mice showed recessive lens phenotypes, since their heterozygous knockout (+/) mice have relatively normal and transparent lenses. (iii) No2 mice (8 point mutation A47D) have been reported to develop semi-dominant cataracts (22). This genetic evidence verified an essential role of 3 and 8 connexins in the lenses of humans and mice. However, two essential questions were not investigated in the previous studies. (i) What are the molecular bases for the differences in inheritance patterns? Point mutations caused dominant phenotypes, whereas their null mutations produced recessive phenotypes. (ii) What is the distinctive molecular pathology that is associated with dominant cataracts in connexin point mutants versus the recessive phenotypes in null mutants? Lop10 is an autosomal semi-dominant mutation causing cataracts, which was discovered among progeny of a cross between BALB/cJ and AKR/J mice (23). Mice homozygous for the Lop10 mutation developed microphthalmia with dense cataracts, first observed when the mice opened their eyes at day 12. Heterozygous mice have relatively normal sized eyes with a variable expression of cataracts, ranging from no expression with clear lenses to partial expression with snowflake opacities, and to full expression with dense fetal nuclear opacities (23). In this study, we have identified that Lop10 is a point mutation of 8 connexin and further investigated morphological and biochemical alterations in the dominant cataractous phenotype of Lop10 mice in comparison with 8 and 3 knockout mice. Lop10 segregated as an autosomal semi-dominant mutation. These data were confirmed by the (BALB/cJ-Lop10/ Lop10CAST/Ei)F1BALB/cJ backcross that produced 42 Lop10/+ mice with cataract and 47 wild-type mice with normal eye. We used a genome-wide screen to map the Lop10 mutation to mouse chromosome 3 between D3Mit11 and D3Mit288 (Fig. 1), where Gja8 is located. DNA sequence analysis showed that Lop10 is a missense mutation (GC) at codon 22 of the Gja8 gene that results in glycine being replaced by arginine (G22R) (Fig. 1). Therefore, the gene symbol for Lop10 should be Gja8Lop10. The 8 null mutation (hereafter 8/) is recessive, causing microphthalmia with mild cataracts in homozygous mice (20,21), while Lop10 acts semi-dominantly to produce cataracts. We tested for allelism between Lop10 and the Gja8 targeted mutation. Lop10/8 F1 mice that were generated from a cross between homozygous Lop10 and 8/ mice developed microphthalmia with dense cataracts (Fig. 2B). In contrast, Lop10/+ F1 mice, which were generated from a cross between homozygous Lop10 and wild-type mice, had normal-sized eyes (Fig. 2A) with a variable expression of cataracts, depending on the mouse strain background (data not shown). Heterozygous 8 mutant (+/) mice developed normal eyes with transparent lenses (20,21). We therefore concluded that 8 G22R in Lop10 mice not only causes a loss-of-function of its own protein product, but is also a direct demonstration that the Lop10 and 8 targeting mutation are in the same gene. The morphological differences between the lenses of Lop10 and 8/ mutants were obvious in standard histological sections (Fig. 3). The lens histology of 1-day-old homozygous Lop10 mice was similar to that of the 8/ mice (Fig. 3A and B). The homozygous Lop10 (Gja8Lop10) mice eventually developed severe cataracts associated with a posterior lens rupture, characterized by a number of vacuoles and/or enlarged intercellular spaces (Fig. 3D and F), but the 8/ mice did not (Fig. 3C and E). Interestingly, the differentiation of the secondary fibers in the homozygous Lop10 lenses proceeded in a relatively normal fashion in the bow region (Fig. 3F), but severely disrupted fibers were observed in the deeper regions. These data suggest that Lop10 is a loss-of-function mutation for 8 connexin itself, and also affects the function of other lens proteins. Double immunohistochemical labeling was used to detect 8 G22R and 3 connexins in the frozen sections of Lop10 lenses by using a mouse monoclonal antibody against 8 and a rabbit polyclonal antibody against 3. Diffuse signals (much like a high level fluorescent background) with a few punctuated spots were detected by the anti-8 antibody, whereas small punctuated fluorescent spots were detected by the anti-3 antibody (Fig. 4). This anti-3 antibody detected small punctuated fluorescent spots in 8/ lenses compared to those of wild-type lenses (21). Moreover, only small gap junction plaques were observed in the Lop10 lenses by thin-section electron microscopic analysis (data not shown). Thus, we concluded that the 8 G22R mutant protein had lost its ability to form normal gap junctions. Surprisingly, an obvious reduction in the phosphorylated forms of 3 connexins was found in the lens homogenates of Lop10 mutant lenses. It is known that 3 connexin is mainly phosphorylated in postnatal lenses (24,25). The mutant 8 G22R protein was detected in lens homogenates of Lop10/Lop10 and Lop10/ 8 mice, but not in 8/ mice (Fig. 5A, left). We have found that there was at least a 4-fold reduction (n = 3), measured by densitometric analysis, in absolute levels of the phosphorylated forms of 3 connexin in lens homogenates of 1-week-old Lop10/Lop10 mice, as compared to the levels in 8/ sibs (Fig. 5A, indicated by an arrow in the right-hand panel). We did not find any obvious changes in the phosphorylated 3 connexin in 8/ lenses when compared to wild-type lenses small lens with posterior rupture small lens with posterior rupture small lens with mild nuclear cataract normal size lens with nuclear cataract small lens with dense cataract or posterior rupture cortical vacuoles with severe disrupted fibers cortical vacuoles with several disrupted fibers no vacuoles with relative normal fibers no vacuoles with relative normal fibers no cortical vacuoles with relative normal cortical fibers and disrupted nuclear fibers (Fig. 5B, indicated by an arrow in the right-hand panel). To test the hypothesis that an altered 3 connexin mediated unique lens phenotypes in the Lop10 mutant, we mated the Lop10 mutant mice with the 3 null targeted mutant mice to produce double mutant Lop10/Lop10 3/ mice. Histological sections indicated that vacuoles and/or enlarged intercellular spaces in the cortical regions of the homozygous Lop10/Lop10 lenses were absent in the double homozygous Lop10/Lop10 3/ lenses (data not shown). Yet, some of the double mutant Lop10/Lop10 3/ mice still developed lenses with posterior ruptures, and others showed severely disrupted fibers in the center regions of the lens without posterior ruptures that were similar to the lens phenotypes in the double homozygous targeted mutant 8/ 3/ mice (data not shown). A summary of the ocular phenotypes in 3 knockout, 8 knockout, Lop10 and Lop10 3 double mutants is listed in Table 1. Moreover, according to ultra-structural scanning electron graphics, the enlarged balls-and-sockets and globulization (Fig. 6, right) present in the cortical fibers of the homozygous Lop10/Lop10 lenses were absent in the double homozygous Lop10/Lop10 3/ lenses (Fig. 6, bottom left). We therefore believe that an altered function of the endogenous wild-type 3 connexin in the Lop10 mutants mediates at least a part of the unique lens phenotypes, and the functional impairment of other unknown lens protein(s) might cause the posterior ruptures in the 8 G22R mutant. Lop10, 8 connexin knockout and 3 connexin knockout mice develop distinctive lens phenotypes (Table 1). We have identified that the dominant cataracts in Lop10 mice are caused by a missense mutation (GC) at codon 22 of the Gja8, which results in glycine being replaced by arginine (G22R). Furthermore, we have demonstrated that: (i) the Lop10 mutant develops distinctive morphological and biochemical alterations in comparison with those of the 8 null targeted mutant mice; (ii) the presence of the 8 G22R mutant in Lop10 mouse lenses leads to a reduction in the phosphorylated forms of endogenous 3 connexin; and (iii) the functional impairment of endogenous 3 connexin mediated a part of the dominant cataractous phenotypes in the Lop10 mice. Thus, the study of the 8 G22R mutant provided some novel and important insights, not only for the human and mouse cataractogenesis that is caused by 8 and 3 mutations (23,26,27), but also for other human and mouse diseases that are associated with mutations of different connexin isoforms (2831). Both 3 and 8 connexins were detected in the same gap junctional plaques of fiber cells by indirect methods, including immunohistochemical staining of frozen sections (12) and immuno-gold labeling with an EM replica (32). Heteromeric connexons have been demonstrated biochemically in the lens (33,34), so 3 connexin is likely to be one of the in vivo candidates that directly interact with the 8 G22R mutant protein. The alteration of the phosphorylated 3 connexin in the Lop10 mouse lenses provided direct in vivo evidence for the regulation and interactions between 8 and 3 connexins. With this study, however, we are unable to determine whether the reduction in the phosphorylated 3 connexin in Lop10 lenses is due to a direct physical interaction between 8 G22R and 3 connexin via heteromeric or heterotypic interactions to form the gap junction channels between the two subunits, or whether a downstream effect of cataract formation resulted in an altered function of 8 G22R. An elucidation of the molecular mechanism(s) by which the 8 G22R mutant protein affects the phosphorylated forms of 3 connexin will help us to further understand the functional role of the interactions between the two connexin isoforms in normal lenses, as well as in cataractous lenses. More importantly, identification of the mechanisms for the interactions among different connexin isoforms is essential for one to understand the regulation of gap junction channels, since in any given cell type in vivo at least two different connexin isoforms are expressed. Glycine 22 residue in the N-terminal domain of 8 connexin, which was replaced with arginine in Lop10, is an evolutionarily conserved residue in all the members of the connexin gene superfamily, crossing different species (35). The N-terminal domain has been reported to form a part of the transjunctional voltage sensor of gap junction channels and to play a fundamental role in controlling ion permeation (36,37). Therefore, it will be interesting to investigate whether 8 G22R mutant will be able to alter the ion permeation in gap junction channels formed by mixing wild-type 3 and/or 8 G22R mutant forms. A different dominant cataractous mouse mutant No2 was reported to be caused by a point mutation of 8 connexin within codon 47, a normally encoded aspartic acid (D) residue replaced with alanine (A) (22). It is not known whether the cataracts that appeared in the No2 mice were similar to the cataracts in Lop10 mice, since no histological and biochemical data of No2 cataracts have been shown. The studies of the 8 D47A mutant protein, by using Xenopus oocytes, indicated that the D47A mutation acts as a loss-of-function mutation without dominant inhibition for wild-type 8 or 3 connexin (38). Therefore, the 8 D47A mutation might act differently from the 8 G22R mutation, since 8 G22R caused a reduction of phosphorylated 3 connexin in the Lop10 lens. The question remains that if 8 D47A acts only as a loss-of-function mutant, why do the No2 mice show a dominant phenotype rather than a recessive phenotype, like that found in 8 null mutant (knockout) mice? It is possible that 8 D47A also impairs the function of other lens protein(s) to cause a dominant cataract formation. Further studies are necessary to investigate the in vivo and in vitro functional similarities and differences between 8 D47A and G22R point mutations. It is surprising that distinctive morphological alterations, such as vacuoles and posterior lens ruptures, were observed only in the dominant cataractous lenses of the Lop10 mice and not in the 8/ mice. Therefore, different mutations in 8 connexin can lead to distinctive cataract pathology. It is not clear how the 8 G22R mutant leads to vacuole formation and/or globulization in the Lop10 lenses. However, the impairment of the endogenous 3 connexin must contribute to vacuole formation in the cortical fibers, as the loss of 3 connexin eliminated this phenotype (Fig. 6). The vacuole formation in the cortical fibers has been observed in other mouse models that were related to alterations of extracellular or intracellular signaling molecules or transporters, such as MEK1(E) and SMIT transgenic mice (39,40) and SPARC knockout mice (41,42). Gap junction communication is likely to be involved in transmitting cellcell signals such as IP3, calcium or other small molecules such as secondary messages, but a direct link between gap junctional communication and other signaling pathways in the lens has not been discovered. Therefore, the 8 connexin mutant mice will be extremely valuable for studying the signaling mechanism for cellcell communication via gap junctions in the lens. MATERIALS AND METHODS Gene mapping and sequence To determine the chromosomal location of the Lop10 gene, we mated BALB/cJ-Lop10/Lop10 mice with CAST/Ei mice. The F1 mice that exhibited cataracts were backcrossed to wild-type BALB/cJ mice for the initial linkage study. Tail DNA was isolated for the linkage analysis and genotyping. Two pools of DNA with equal contributions from 10 mice with cataracts or 10 mice with normal lenses were genotyped for 57 simple sequence length polymorphism (SSLP) markers (three markers per chromosome) that were polymorphic between the two parental strains. Loci showing significant skewing of the alleles were identified and additional markers around them were selected in order to genotype individual mice and further confirm the linkage. We designed two pairs of PCR primers (Gja8-R with Gja8-F, Gja8-2F and Gja8-2R) based on the DNA sequence from mouse 8 connexin sequence in GenBank (accession no. M91243) covering the whole coding region. The sequences for oligos used for characterizing the 8 coding region in the Lop10 mutant were Gja8-f, 5-GGCACTTGATAGAAGCTGTTGG-3 (125146); Gja8-r, 5-GGTGGACCAAGGGACACATAAG-3 (16901669); Gja8-2r, 5-TTCTCAGCAATCTCCCCAGTG-3 (10361015); and Gja8-2f, 5-GGCCATTACTTCCTGTATGG-3 (789808). PCR methods for genotyping different mutant mice Different pairs of primers were used to genotype the mutant and wild-type alleles of Gja8 and Gja3. A 300 bp fragment was expected for wild-type Gja3 (3) by using the primer pair 5-CCCAGGCTCTACCTCAGGTT-3 (sense) and 5-CTTTGCCGATGACTGTAGAG-3 (antisense); a 500 bp fragment for knockout Gja3 with the primer pair 5-CCCAGGCTCTACCTCAGGTT-3 (sense) and 5-CAGGGTTTTCCCAGTCACGAC-3 (antisense); a 320 bp fragment for either G22R or wild-type allele of Gja8 (8) with the primer pair 5-GGATCCTTTCAAACAAC-3 (sense) and 5-GCCGATGACAGTGGAGTGCTC-3 (antisense); a 450 bp fragment for the targeted mutant (null) allele of Gja8 with the primer pair 5-GGATCCTTTCAAACAAC-3 (sense) and 5-CAGGGTTTTCCCAGTCACGAC-3 (antisense). Histological, immunohistochemical and biochemical analyses Standard histological methods were used for analyzing lenses of both mutant and normal mice. Frozen lens sections were prepared and used for the antibody staining, following the procedure described in our previous paper (12). A laser-confocal microscope (Bio-Rad, Model 1024) was used to collect the fluorescence from the indirect immunostaining. We used both a mouse monoclonal antibody against 8 (6-4-B2-C6, provided by Dr J.Kistler, University of Auckland, Auckland, New Zealand) (43) and a rabbit polyclonal antibody against 3 connexin. Western blotting was carried out by using a standard protocol described previously (12). A rabbit polyclonal antibody against the C-terminal region of 8 connexin (generously provided by Dr M.J.Wolosin, Mt Sinai School of Medicine, New York) was used. The total lens crude membrane fractions were insoluble pellets after whole lenses were homogenized with 20 mM NaOH in 1 mM Na2CO3 following centrifugation at 16 000 g for 15 min. The pellets were dissolved into the buffer (10 mM Tris pH 8.0, 1 mM MgCl2, 0.2% SDS, 0.2% -mercaptoethanol). Dephosphorylation was performed by incubating the crude membrane fraction (5 g) with 3 U calf intestinal alkaline phosphatase (Boehringer-Mannheim Corp.) for 3 h at 37C. In parallel, equal aliquots were incubated either in buffer without phosphatase, to control for endogenous enzyme activities, or with phosphatase inhibitors including 5 mM sodium pyrophosphate, 5 mM sodium fluoride and 1 mM sodium orthovanadate. The reactions were terminated by adding SDS-containing gel loading buffer. Ultra-structural examination using a scanning electron microscope and transmission electron microscope Mouse lenses were fixed in 2.5% gluteraldehyde with 0.075 M of cacodylate buffer pH 7.2 for 5 days at 4C, and the fixative was changed daily. Lenses were dissected or fractured and further processed in 1% OsO4 and dehydrated through a standard ethanol series. Lens samples were critical point dried, mounted on aluminum stubs and coated with a Hummer II gold-palladium sputter coater, then viewed with a Hitachi S-2700 SEM. The authors wish to thank Jordan McMullin for her help in editing this manuscript, and Dr Beverly Richards-Smith for her suggestions about the genetic nomenclature. This work was supported in part by National Eye Institute grants EY13849, EY12808, EY07758 and grant CA34196 from the National Cancer Institute. REFERENCES


This is a preview of a remote PDF: https://hmg.oxfordjournals.org/content/11/5/507.full.pdf

Bo Chang, Xin Wang, Norman L. Hawes, Ryan Ojakian, Muriel T. Davisson, Woo-Kuen Lo, Xiaohua Gong. A Gja8 (Cx50) point mutation causes an alteration of α3 connexin (Cx46) in semi-dominant cataracts of Lop10 mice, Human Molecular Genetics, 2002, 507-513, DOI: 10.1093/hmg/11.5.507