This lab has previously generated alpha-crystallin gene knockout mice to study the in vivo function of these remarkable proteins. The alpha-crystallins comprise a large fraction of the soluble protein in the vertebrate lens where they were, for many years, believed to function solely as structural proteins. Lenticular alpha-crystallin is comprised of two similar subunits alphaA and alphaB, each encoded by a single gene. They are related to the small heat shock proteins, and in vitro they exhibit molecular chaperone activity, autokinase activity, and interact with, and affect the state of, several cytoskeletal components. alpha-Crystallin, especially alphaB-crystallin, has been shown to be a normal constituent of many non-lenticular tissues, and has been detected in cytoplasmic inclusion bodies found in several human pathological conditions. alphaA-Crystallin, and indeed many of the formerly """"""""lens-specific"""""""" crystallins, has also recently been shown to be expressed on non lenticular tissues. Toward understanding the major roles of alpha-crystallin in vivo, we previously generated alphaA- and alphaB-crystallin gene knockout mice and alphaA-/alphaB-crystallin gene double knockout mice (DKO). We have previously shown that the lenses of DKO mice exhibit disintegration of fiber cells surrounding the lens nucleus, and have shown that these morphological abnormalities result from elevated DEVDase and VEIDase activities in lenses lacking alpha-crystallin, suggesting involvement of the apoptosis pathway in this pathology. To analyze alpha-crystallin's possible regulation of genes in the apoptotic pathway, we employed a PCR array strategy, using lenses from alpha-crystallin-null mice, A/B-crystallin/caspase-3 triple KO mice, A/B-crystallin/caspase-6 triple KO mice, and wild type mice. These same two lines of triple KO mice were used to assess the roles of caspase-3 and caspase-6 in the lens disintegration process observed in alpha-crystallin-null mice, and attempt the rescue of this cell disintegration phenomenon by elimination of these individual caspases. PCR array analysis of apoptotic gene expression from lenses of alpha-crystallin-null mice revealed a predominant down-regulation in most functional gene groups including Anti-apoptosis, CIDE Domain, IAP, Caspase, Bcl-2, and the TNF Ligand Family. Interestingly, the CARD family of anti-apoptotic genes was down-regulated and the pro-apoptotic gene APAF1 was up-regulated. PCR array studies of lenses from triple knockout mice showed only minor, if any, difference compared to alpha-crystallin-null mice. This observation is consistent with a histological study demonstrating that elimination of either caspase-3 alone or caspase-6 alone failed to rescue the lens secondary fiber cell disintegration phenotype in alpha-crystallin-null mice. We are currently breeding mice which are deficient in alphaA-crystallin, alphaB-crystallin, caspase-3, and caspase-6 to determine whether quadruple knockouts can recue lens disintegration. Western blot analyses show age dependent APAF1 protein expression, the pattern of which correlates with caspase-3 activity in the mouse lens, suggesting a regulatory role for APAF1 and caspase-9 in secondary lens fiber cell differentiation. Western blot analyses show no difference in APAF1 expression in liver, spleen, brain, heart, muscle, and kidney in DKO and WT mice suggesting regulation of expression by alpha-crystallin selectively in lens tissue. We have hypothesized that, in secondary lens fiber cells, alpha-crystallin interferes with the apoptosis-like maturation program, halting it after organelle removal, but before cellular disintegration. Our new data, along with data from others, suggest that alpha-crystallin interferes with the caspase cascade in the apoptotic pathway. Caspase activity regulation could be at the protein-protein level by direct interaction as has recently been demonstrated by several labs, and/or at the level of gene expression regulation. Changes in apoptotic pathway gene expression patterns in lenses of alpha-crystallin-null mice, reported here, suggest a novel role for alpha-crystallin as a regulator of gene expression in the apoptotic pathway. A collaboration with Albee Messing at University of Wisconsin investigated the role of alphaB-crystallin in Alexander disease (AxD), an astrocyte disorder. In mouse models of AxD, absence of alphaB increases mortality, while astrocyte-specific overexpressionn of alphaB can reverse the increased incidence of lethal seizures, and decrease both GFAP levels and Rosenthal fibers. Although alphaB is a constituent of Rosenthal fibers in AxD, this research also demonstrated that alphaB is not required for their formation. The data are consistent with normal levels of alphaB providing a protective effect against stress induced by mutant GFAP or overexpression of normal GFAP, but being overwhelmed in AxD. Thus, increasing alphaB levels in astrocytes may be a therapeutic target for treating AxD. A mouse modeling project, investigating the non-lens functions of betaA3/A1-crystallin, is a continuing collaboration with Dr. Debasish Sinha of Johns Hopkins University, who has been on extended sabbatical (IPA) in our laboratory up through the first half of this year. Dr. Sinha and associates have shown that a naturally occurring mutation of this gene in rat causes abnormalities in astrocytes, which leads to altered development of the neural retina and retinal vasculature, with similar defects in the brain. Using an astrocyte-specific GFAP promoter, we have made transgenic mouse models in which expression of the mutant rat betaA3/A1 or the normal betaA3/A1 is restricted to astrocytes. Retinal astrocytes in transgenic mice expressing the normal protein appear similar to wild type, forming a honey comb-like network. However, in mice expressing the mutant beta A3/A1 in astrocytes, the retinal astrocyte network is irregular, exhibiting bundle-like structures and short, thickened processes. Moreover, primary vessel patterning is affected by this abnormal astrocyte arrangement. The lenses however remain transparent and intact, achieving one of the primary goals of this modeling project. The disintegrating lens in the original mutant rat had raised concerns that the retinal phenotype may not be caused by expression of the mutant crystallin in astrocytes, but a secondary effect of lens debris and associated processes. These initial studies suggest that, like a similar phenotype observed in the mutant rat brain, the retinal phenotype does indeed result from mutant betaA3/A1 expression in astrocytes, and not from lens degradation. Further analysis of these mice is ongoing. In a continuing collaboration with Dr. Lijin Dong of the NEI Genetic Engineering Facility, conditional knockout mice in which the betaA3/A1 gene is specifically disrupted in astrocytes have been generated. We have attained germline transmission of the ES cell component, and are in the process of analyzing offspring to determine if the targeted mutation has been passed through the germline. Again, this approach avoids interference or complicating secondary effects from a disintegrating lens. The approach has been to insert loxP recombinase sites flanking exons 4 and 5 of the mouse betaA3/A1-crystallin gene, and excising this portion of the gene by astrocyte-specific expression of cre recombinase, again using the GFAP promoter.
| Hagemann, Tracy L; Boelens, Wilbert C; Wawrousek, Eric F et al. (2009) Suppression of GFAP toxicity by alphaB-crystallin in mouse models of Alexander disease. Hum Mol Genet 18:1190-9 |
| Fujimoto, C; Klinman, D M; Shi, G et al. (2009) A suppressive oligodeoxynucleotide inhibits ocular inflammation. Clin Exp Immunol 156:528-34 |
| Shi, Guangpu; Cox, Catherine A; Vistica, Barbara P et al. (2008) Phenotype switching by inflammation-inducing polarized Th17 cells, but not by Th1 cells. J Immunol 181:7205-13 |
