Ankyrin G Overexpression in Hutchinson-Gilford Progeria Syndrome fibroblasts Identified Through Biological filtering of Expression Profiles
Total Page:16
File Type:pdf, Size:1020Kb
J Hum Genet (2006) 51:934–942 DOI 10.1007/s10038-006-0042-0 ORIGINAL ARTICLE Ankyrin G overexpression in Hutchinson-Gilford progeria syndrome fibroblasts identified through biological filtering of expression profiles Jian Wang Æ John F. Robinson Æ Caroline H. O’Neil Æ Jane Y. Edwards Æ Christina M. Williams Æ Murray W. Huff Æ J. Geoffrey Pickering Æ Robert A. Hegele Received: 5 March 2006 / Accepted: 23 July 2006 / Published online: 11 October 2006 Ó The Japan Society of Human Genetics and Springer-Verlag 2006 Abstract Hutchinson-Gilford progeria syndrome in culture, including increased expression of ANK3/ (HGPS; MIM 176670) is a rare disease characterized ankyrin G. Furthermore, other genes that co-clustered by accelerated aging. In this study, light and immuno- with ANK3 might provide mechanistic clues regarding fluorescence microscopy were used to assess morpho- senescence in cultured HGPS cells. logical changes, measures of cell growth kinetics and gene expression profiles in HGPS cells and normal fi- Keywords RNA microarrays Á Bioinformatics Á broblasts in culture. A filtering strategy was developed Genomics Á Cytoskeleton Á Aging based on differentially expressed transcripts seen con- sistently across three culture stages based on cell pas- sage number. This filtering strategy produced a list of Introduction 66 unique differentially expressed genes, of which ~40% were upregulated in HGPS cells compared to Hutchinson-Gilford progeria syndrome (HGPS; MIM normal fibroblasts. The increased mRNA expression in 176670) is an extremely rare disease that is character- HGPS cells that was seen for one gene defined using ized by accelerated aging and early death (Pollex and this strategy—namely ANK3— was validated using Hegele 2004). Organ systems degenerate to such an quantitative reverse-transcriptase amplification, Wes- extent that the affected subject resembles an old per- tern analysis and immunofluorescence microscopy, all son. Clinical features include short stature, microgna- of which showed significantly increased ankyrin G thia, alopecia, prominent scalp veins, prominent joints, expression. These findings demonstrate differences in hyperlipidemia and atherosclerosis, often with pre- morphology, growth kinetics and mRNA expression mature death from coronary artery disease. The ge- profiles in HGPS cells compared to normal fibroblasts netic basis for >90% of cases of HGPS is a de novo recurrent c.2036C>T splicing mutation in LMNA, encoding nuclear lamin A/C, although there is some J. Wang Á J. F. Robinson Á C. H. O’Neil Á J. Y. Edwards Á heterogeneity of mutations (Eriksson et al. 2003;De C. M. Williams Á M. W. Huff Á J. G. Pickering Sandre-Giovannoli et al. 2003; Cao and Hegele 2003; Vascular Biology Research Group, Robarts Research Institute, 100 Perth Drive, Csoka et al. 2004a). London, ON, Canada N6A 5K8 In cultured HGPS fibroblasts, others have shown differential gene expression profiles, including changes R. A. Hegele (&) in expression of genes within the ontologies of tran- Vascular Biology Research Group, Robarts Research Institute, 406–100 Perth Drive, scription factors and extracellular matrix proteins London, ON, Canada N6A 5K8 compared to normal fibroblasts (Csoka et al. 2004b). e-mail: [email protected] Since the LMNA mutation is the primary insult within the germline of HGPS fibroblasts, changes in gene M. W. Huff Á J. G. Pickering Á R. A. Hegele Department of Medicine, University of Western Ontario, expression profiles are secondary and may provide London, ON, Canada clues about dysregulated pathways that lead to 123 J Hum Genet (2006) 51:934–942 935 abnormal phenotypes in cultured cells and perhaps and total RNA was then isolated using an RNeasy kit within intact tissues and organs. Our current objectives (Qiagen, Mississauga, ON). Isolated total RNA was were to: (1) develop a biological filtering approach to quantified by ultraviolet spectroscopy, and RNA characterize cellular and molecular differences be- integrity was evaluated using the BioAnalyzer 2100 tween cultured fibroblasts from a HGPS proband with Model G2938A (Agilent Technologies, Mississauga, the dominant LMNA c.2036C>T splicing mutation and ON). fibroblasts from a normal subject; and (2) validate this approach at the protein level using Western analysis Immunofluorescence studies and immunocytochemistry to study the product of a candidate gene that showed marked differential In order to determine proliferation differences be- mRNA expression. The filtering criteria included tween normal and HGPS fibroblasts, cells were grown opposite patterns of expression between HGPS and on coverslips to 80% confluence. Cells on coverslips normal fibroblasts at three different stages of cell cul- were washed twice with PBS, fixed with 4% parafor- ture maturity defined by passage number. maldehyde on ice for 20 min, and incubated with 0.5% Triton X-100 at room temperature for 5 min. Fixed cells were next incubated with a 1:150 dilution of Materials and methods monoclonal mouse anti-human Ki-67 antibody (Dako- Cytomation, Mississauga, ON) at 37°C in a humidified Cell culture chamber for 1 h, washed three times with PBS, then incubated with a 1:100 dilution of FITC goat anti- Human normal (AG04456) and HGPS (AG03513, mouse IgG (Cedarlane Laboratories, Hornby, ON) at LMNA mutation c.2036C>T) fibroblast cell lines were room temperature in a humidified chamber for 30 min. obtained from the Coriell Cell Repository (Camden, After washing three times with PBS, cells were stained NJ). Cells were grown in Dulbecco’s Modification of with Hoechst dye (2.5 lg/ml in PBS) (Sigma-Aldrich, Eagle’s Medium (DMEM), supplemented with 15% Oakville, ON) at room temperature for 5 min. Cells non-inactivated fetal bovine serum (FBS), at 37°Cin were then washed three times with PBS and mounted 5% CO2. Cells were passaged every 3 or 4 days and on glass slides using PermaFluor Aqueous Mounting seeded at a density of 2·105 per 90 mm diameter dish. Medium (Fisher, Markham, ON). Proliferating cells Cell growth curves were obtained by measuring accu- were identified by positive staining with the Ki-67 mulated population doublings (APD), using the for- antibody. mula (logH–logS)/log 2.0, where logH and logS are the In order to localize ankyrin G within fibroblasts, logarithms of the numbers of cells harvested and see- cultures grown on cover slips were fixed with 2% ded, respectively (Bridger and Kill 2004). paraformaldehyde at room temperature for 10 min, followed by incubation with 0.5% Triton X-100 at Genomic DNA isolation and direct sequencing room temperature for 2 min. Fixed cells were incu- bated with a 1:50 dilution of monoclonal mouse Genomic DNA was isolated from cultured cells using anti-human ankyrin G antibody (Santa Cruz Biotech- the Puregene kit (Gentra Systems, Minneapolis, MN). nology, Santa Cruz, CA) at 37°C in a humidified All coding exons, intron–exon boundaries and >300 bp chamber for 1 h, washed three times with PBS, then of 5¢ and 3¢ untranslated regions of the LMNA gene incubated with a 1:200 dilution of FITC goat anti- were amplified as described (Cao and Hegele 2003) mouse IgG at room temperature in a humidified and were then sequenced on a ABI 3730 DNA Ana- chamber for 30 min. After washing three times with lyzer (Applied Biosystems, Mississauga, ON). Muta- PBS, cells were stained with Hoechst dye (2.5 lg/ml in tions were identified using Sequence Navigator PBS) at room temperature for 5 min. Cells were then software (Applied Biosystems). washed three times with PBS and mounted on glass slides using PermaFluor Aqueous Mounting Medium Total RNA isolation (Fisher, Markham, ON). RNA was isolated at the elapsed cell culture ages of Microarray studies APD 2.5 (young), APD 10 (middle-aged) for both the normal and HGPS fibroblast cell lines; and at APD 26 Standard procedures of the London Regional Ge- (old) for the normal cell line and APD 17 (old) for the nomics Centre were used (Carter et al. 2005). Total HGPS fibroblasts. Cells were washed twice with PBS RNA isolates from three different cell ages—young, 123 936 J Hum Genet (2006) 51:934–942 middle, and old—of both normal and HGPS cell lines pre-processing (Irizarry et al. 2003), a summary measure were analyzed using the GeneChip Expression Array to normalize arrays that simultaneously adjusted back- HU133A (Affymetrix, Santa Clara, CA). A 25 ng ali- ground, normalized and log-transformed perfect match quot of total RNA from each sample was processed intensity values from the .cel files. A gene list was gen- using the GeneChip Expression 3¢-Amplification Re- erated with all Affymetrix technical quality assurance agents Two Cycle cDNA Synthesis Kit (Affymetrix, probes, marked using the ‘‘Affx’’ prefix. Santa Clara, CA). Sample quality was assessed using A compound filter was used to create the final gene the BioAnalyzer 2100 Model G2938A (Agilent Tech- list: (1) remove all ‘‘Affx’’ prefix genes; (2) remove all nologies) and degradation software as described intensity values £150 units in at least three out of six (Carter et al. 2005). Only samples that registered ‘‘no conditions; (3) remove all values £ 2-fold difference alert’’ were analyzed (Carter et al. 2005). A 600 ng between normal–young and HGPS–young expression sample of cRNA was used for the second cycle of values; and (4) filter on confidence of association with cDNA and cRNA synthesis. BIOTINYLATED frag- nominal P£0.1. This filter yielded a set of 66 unique mented cRNA (15 lg) was then hybridized onto genes that were up-regulated in normal and down- HU133A chips. GeneChips were washed and scanned regulated in the HGPS cell line and vice versa (Ta- on an Affymetrix Fluidics Module Model 400 and ble 1). A similar compound filter, with the exception of GeneChip Scanner Model 3000 controlled by Gene- not filtering on confidence, yielded a list of 462 genes Chip Operating System v1.3 (GCOS) (Affymetrix).