Altered Expression of Genes of the Bmp/Smad and Wnt/Calcium Signaling Pathways in the Cone-only Nrl-/- Mouse Retina, Revealed by Gene Profiling Using Custom cDNA Microarrays
2004; Elsevier BV; Volume: 279; Issue: 40 Linguagem: Inglês
10.1074/jbc.m408223200
ISSN1083-351X
AutoresJindan Yu, Shirley He, James S. Friedman, Masayuki Akimoto, Debashis Ghosh, Alan J. Mears, David Hicks, Anand Swaroop,
Tópico(s)Photoreceptor and optogenetics research
ResumoMany mammalian retinas are rod-dominant, and hence our knowledge of cone photoreceptor biology is relatively limited. To gain insights into the molecular differences between rods and cones, we compared the gene expression profile of the rod-dominated retina of wild type mouse with that of the cone-only retina of Nrl-/- (Neural retina leucine zipper knockout) mouse. Our analysis, using custom microarrays of eye-expressed genes, provided equivalent data using either direct or reference-based experimental designs, confirmed differential expression of rod- and cone-specific genes in the Nrl-/- retina and identified novel genes that could serve as candidates for retinopathies or for functional studies. In addition, we detected altered expression of several genes that encode cell signaling or structural proteins. Prompted by these findings, additional real-time PCR analysis revealed that genes belonging to the Bmp/Smad and Wnt/Ca2+ signaling pathways are expressed in the mature wild type retina and that their expression is significantly altered in the Nrl-/- retina. Chromatin immunoprecipitation analysis of adult retina identified Bmp4 and Smad4, which are down-regulated in the Nrl-/- retina, as possible direct transcriptional targets of Nrl. Consistent with these studies, Bmp4 and Smad4 are expressed in the mature rod photoreceptors of mouse retina. Modulation of Bmp4 and/or Smad4 by Nrl may provide a mechanism for integrating diverse cell signaling networks in rods. We hypothesize that Bmp/Smad and Wnt/Ca2+ pathways participate in cell-cell communication in the mature retina, and expression changes observed in the Nrl-/- retina reflect their biased utilization in rod versus cone homeostasis. Many mammalian retinas are rod-dominant, and hence our knowledge of cone photoreceptor biology is relatively limited. To gain insights into the molecular differences between rods and cones, we compared the gene expression profile of the rod-dominated retina of wild type mouse with that of the cone-only retina of Nrl-/- (Neural retina leucine zipper knockout) mouse. Our analysis, using custom microarrays of eye-expressed genes, provided equivalent data using either direct or reference-based experimental designs, confirmed differential expression of rod- and cone-specific genes in the Nrl-/- retina and identified novel genes that could serve as candidates for retinopathies or for functional studies. In addition, we detected altered expression of several genes that encode cell signaling or structural proteins. Prompted by these findings, additional real-time PCR analysis revealed that genes belonging to the Bmp/Smad and Wnt/Ca2+ signaling pathways are expressed in the mature wild type retina and that their expression is significantly altered in the Nrl-/- retina. Chromatin immunoprecipitation analysis of adult retina identified Bmp4 and Smad4, which are down-regulated in the Nrl-/- retina, as possible direct transcriptional targets of Nrl. Consistent with these studies, Bmp4 and Smad4 are expressed in the mature rod photoreceptors of mouse retina. Modulation of Bmp4 and/or Smad4 by Nrl may provide a mechanism for integrating diverse cell signaling networks in rods. We hypothesize that Bmp/Smad and Wnt/Ca2+ pathways participate in cell-cell communication in the mature retina, and expression changes observed in the Nrl-/- retina reflect their biased utilization in rod versus cone homeostasis. In mammals, vision is initiated in the retina, which is a highly structured part of the brain consisting of over 50 types of neurons that are organized in three distinct layers (1Dowling J.E. The Retina: an Approachable Part of the Brain. Harvard University Press, Cambridge, MA1987Google Scholar, 2Masland R.H. Curr. Opin. Neurobiol. 2001; 11: 431-436Crossref PubMed Scopus (243) Google Scholar). The outer nuclear layer consists exclusively of two types of photoreceptors, rods and cones, responsible for detection and transduction of light energy. Rods function under low ambient light and form the major photoreceptor population of many mammals, including humans (95%) and mice (97%). Cones are responsible for phototransduction in bright light, providing high acuity and color vision (1Dowling J.E. The Retina: an Approachable Part of the Brain. Harvard University Press, Cambridge, MA1987Google Scholar). Cones are needed for maintaining central vision, and their survival is vital for preserving visual function in retinal and macular diseases (3Hicks D. Sahel J. Invest. Ophthalmol. Vis. Sci. 1999; 40: 3071-3074PubMed Google Scholar, 4Curcio C.A. Owsley C. Jackson G.R. Invest. Ophthalmol. Vis. Sci. 2000; 41: 2015-2018PubMed Google Scholar). Rods and cones possess distinct subsets of proteins involved in the phototransduction cascade (5Nathans J. Neuron. 1999; 24: 299-312Abstract Full Text Full Text PDF PubMed Scopus (255) Google Scholar, 6Molday R.S. Invest. Ophthalmol. Vis. Sci. 1998; 39: 2491-2513PubMed Google Scholar), but despite significant neuroanatomical and physiological advances (2Masland R.H. Curr. Opin. Neurobiol. 2001; 11: 431-436Crossref PubMed Scopus (243) Google Scholar, 7Leskov I.B. Klenchin V.A. Handy J.W. Whitlock G.G. Govardovskii V.I. Bownds M.D. Lamb T.D. Pugh Jr., E.N. Arshavsky V.Y. Neuron. 2000; 27: 525-537Abstract Full Text Full Text PDF PubMed Scopus (159) Google Scholar, 8Marc R.E. Jones B.W. Mol. Neurobiol. 2003; 28: 139-147Crossref PubMed Scopus (204) Google Scholar), little progress has been made toward delineating the molecular mechanisms that underlie functional distinctions between the two photoreceptor types, their communication with other neurons, and their maintenance, survival, or remodeling in response to extrinsic or intrinsic insults. One approach to systematically dissect the regulatory networks and molecules associated with rod or cone photoreceptor function is to take advantage of animal models that exhibit preferential utilization of one or the other photoreceptor sub-type. Unfortunately, many species with cone-rich retinas (e.g. ground squirrel and chick) present difficulties with respect to experimentation and are less amenable to genetic manipulations. Because of a large number of naturally occurring and experimentally generated mutants (available at jaxmice.jax.org/info/index.html), the mouse offers a unique opportunity to examine this complex question. The Nrl 1The abbreviations used are: Nrl, neural retina leucine zipper; ChIP, chromatin immunoprecipitation; NRE, Nrl response element; Bmp, bone morphogenetic protein; EST, expressed sequence tag; P, postnatal day; E, embryonic day; qRT-PCR, quantitative reverse transcription-PCR; Hprt, hypoxanthine guanine phosphoribosyl transferase; PBS, phosphate-buffered saline; PNA, peanut agglutinin; GFP, green fluorescence protein; EGFP, enhanced GFP; ONL, outer nuclear layer; FACS, fluorescence-activated cell sorting; IS, inner segment.-knockout (Nrl-/-) mouse, recently generated in our laboratory, exhibits a unique rod-less and cone-only retinal phenotype (9Mears A.J. Kondo M. Swain P.K. Takada Y. Bush R.A. Saunders T.L. Sieving P.A. Swaroop A. Nat. Genet. 2001; 29: 447-452Crossref PubMed Scopus (717) Google Scholar). Nrl was originally identified from a subtracted human retina library and shown to be expressed, by Northern analysis, specifically in the retina and retinoblastoma cell lines (10Swaroop A. Xu J.Z. Pawar H. Jackson A. Skolnick C. Agarwal N. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 266-270Crossref PubMed Scopus (269) Google Scholar). In situ hybridization analysis identified Nrl transcripts in developing mouse brain and lens, although the expression became restricted to the retina after birth (11Liu Q. Ji X. Breitman M.L. Hitchcock P.F. Swaroop A. Oncogene. 1996; 12: 207-211PubMed Google Scholar). Later studies, however, demonstrated that Nrl is specifically and highly expressed in the rod (and not cone) photoreceptors (9Mears A.J. Kondo M. Swain P.K. Takada Y. Bush R.A. Saunders T.L. Sieving P.A. Swaroop A. Nat. Genet. 2001; 29: 447-452Crossref PubMed Scopus (717) Google Scholar, 12Swain P.K. Hicks D. Mears A.J. Apel I.J. Smith J.E. John S.K. Hendrickson A. Milam A.H. Swaroop A. J. Biol. Chem. 2001; 276: 36824-36830Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar) and pineal gland 2A. J. Mears and A. Swaroop, unpublished data. and that the transcripts in developing brain and lens probably represented cross-hybridization with p45, L-Maf, or another homologous sequence in the mouse genome (12Swain P.K. Hicks D. Mears A.J. Apel I.J. Smith J.E. John S.K. Hendrickson A. Milam A.H. Swaroop A. J. Biol. Chem. 2001; 276: 36824-36830Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar, 13Ogino H. Yasuda K. Science. 1998; 280: 115-118Crossref PubMed Scopus (234) Google Scholar). Nrl is shown to interact with other transcription factors, such as the homeodomain protein Crx, zinc finger protein Fiz1, and orphan nuclear receptor Nr2e3, and regulate (either alone or synergistically) the expression of several rod-specific genes (14Chen S. Wang Q.L. Nie Z. Sun H. Lennon G. Copeland N.G. Gilbert D.J. Jenkins N.A. Zack D.J. Neuron. 1997; 19: 1017-1030Abstract Full Text Full Text PDF PubMed Scopus (577) Google Scholar, 15Lerner L.E. Gribanova Y.E. Ji M. Knox B.E. Farber D.B. J. Biol. Chem. 2001; 276: 34999-35007Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar, 16Mitton K.P. Swain P.K. Chen S. Xu S. Zack D.J. Swaroop A. J. Biol. Chem. 2000; 275: 29794-29799Abstract Full Text Full Text PDF PubMed Scopus (175) Google Scholar, 17Mitton K.P. Swain P.K. Khanna H. Dowd M. Apel I.J. Swaroop A. Hum. Mol. Genet. 2003; 12: 365-373Crossref PubMed Scopus (36) Google Scholar, 18Rehemtulla A. Warwar R. Kumar R. Ji X. Zack D.J. Swaroop A. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 191-195Crossref PubMed Scopus (186) Google Scholar, 19Pittler S.J. Zhang Y. Chen S. Mears A.J. Zack D.J. Ren Z. Swain P.K. Yao S. Swaroop A. White J.B. J. Biol. Chem. 2004; 279: 19800-19807Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar, 20Cheng H. Khanna H. Oh E.C. Hicks D. Mitton K.P. Swaroop A. Hum. Mol. Genet. 2004; 13: 1563-1575Crossref PubMed Scopus (197) Google Scholar). Missense mutations in the human NRL gene are associated with autosomal dominant retinitis pigmentosa (21Bessant D.A. Payne A.M. Mitton K.P. Wang Q.L. Swain P.K. Plant C. Bird A.C. Zack D.J. Swaroop A. Bhattacharya S.S. Nat. Genet. 1999; 21: 355-356Crossref PubMed Scopus (152) Google Scholar, 22DeAngelis M.M. Grimsby J.L. Sandberg M.A. Berson E.L. Dryja T.P. Arch. Ophthalmol. 2002; 120: 369-375Crossref PubMed Scopus (48) Google Scholar). Consistent with these findings, the targeted deletion of Nrl (Nrl-/-) in mouse resulted in a retina with no rod photoreceptors; instead, a concomitant increase in functional S-opsin expressing cones was observed (9Mears A.J. Kondo M. Swain P.K. Takada Y. Bush R.A. Saunders T.L. Sieving P.A. Swaroop A. Nat. Genet. 2001; 29: 447-452Crossref PubMed Scopus (717) Google Scholar). This apparent functional switching of photoreceptor sub-types (from rods to cones) has been validated by histology, electrophysiology, and biochemical and molecular analysis (9Mears A.J. Kondo M. Swain P.K. Takada Y. Bush R.A. Saunders T.L. Sieving P.A. Swaroop A. Nat. Genet. 2001; 29: 447-452Crossref PubMed Scopus (717) Google Scholar, 23Zhu X. Brown B. Li A. Mears A.J. Swaroop A. Craft C.M. J. Neurosci. 2003; 23: 6152-6160Crossref PubMed Google Scholar, 24Yoshida S. Mears A.J. Friedman J.S. Carter T. He S. Oh E. Jing Y. Farjo R. Fleury G. Barlow C. Hero A.O. Swaroop A. Hum. Mol. Genet. 2004; 13: 1487-1503Crossref PubMed Scopus (134) Google Scholar). 3S. S. Nikonov, L. Daniele, C. Lillo, A. J. Mears, A. Swaroop, D. Williams, and E. N. Pugh, Jr., submitted for publication. Microarray-based global profiling of gene expression, in combination with bioinformatic tools, can yield valuable insights into cell- or tissue-specific functions. Expression profiling of tissues from mice deficient in a transcription factor gene can point to downstream regulatory targets, provide candidates for functional studies, and facilitate positional cloning of human disease loci (25Livesey F.J. Furukawa T. Steffen M.A. Church G.M. Cepko C.L. Curr. Biol. 2000; 10: 301-310Abstract Full Text Full Text PDF PubMed Scopus (236) Google Scholar, 26DeRyckere D. DeGregori J. Methods. 2002; 26: 57-75Crossref PubMed Scopus (10) Google Scholar, 27Horton J.D. Shah N.A. Warrington J.A. Anderson N.N. Park S.W. Brown M.S. Goldstein J.L. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 12027-12032Crossref PubMed Scopus (1097) Google Scholar, 28Mu X. Beremand P.D. Zhao S. Pershad R. Sun H. Scarpa A. Liang S. Thomas T.L. Klein W.H. Development. 2004; 131: 1197-1210Crossref PubMed Scopus (101) Google Scholar). Analysis of the retinal transcriptome during development and aging and in mouse models of retinal dysfunction has been the subject of intense investigation (28Mu X. Beremand P.D. Zhao S. Pershad R. Sun H. Scarpa A. Liang S. Thomas T.L. Klein W.H. Development. 2004; 131: 1197-1210Crossref PubMed Scopus (101) Google Scholar, 29Blackshaw S. Fraioli R.E. Furukawa T. Cepko C.L. Cell. 2001; 107: 579-589Abstract Full Text Full Text PDF PubMed Scopus (254) Google Scholar, 30Swaroop A. Zack D.J. Genome Biol. 2002; (http://genomebiology.com/2002/3/8/reviews/1022)PubMed Google Scholar, 31Yoshida S. Yashar B.M. Hiriyanna S. Swaroop A. Invest. Ophthalmol. Vis. Sci. 2002; 43: 2554-2560PubMed Google Scholar). We have used the Nrl-/- mouse model to identify molecular differences between rods and cones (24Yoshida S. Mears A.J. Friedman J.S. Carter T. He S. Oh E. Jing Y. Farjo R. Fleury G. Barlow C. Hero A.O. Swaroop A. Hum. Mol. Genet. 2004; 13: 1487-1503Crossref PubMed Scopus (134) Google Scholar). However, commercially available microarrays do not have adequate representation of genes transcribed in developing and mature eye/retina; hence, several groups have produced custom slide microarrays of eye/retina-expressed genes (28Mu X. Beremand P.D. Zhao S. Pershad R. Sun H. Scarpa A. Liang S. Thomas T.L. Klein W.H. Development. 2004; 131: 1197-1210Crossref PubMed Scopus (101) Google Scholar, 30Swaroop A. Zack D.J. Genome Biol. 2002; (http://genomebiology.com/2002/3/8/reviews/1022)PubMed Google Scholar, 32Chowers I. Gunatilaka T.L. Farkas R.H. Qian J. Hackam A.S. Duh E. Kageyama M. Wang C. Vora A. Campochiaro P.A. Zack D.J. Invest. Ophthalmol. Vis. Sci. 2003; 44: 3732-3741Crossref PubMed Scopus (56) Google Scholar). For gene profiling, we have isolated and sequenced cDNAs from mouse eye/retina libraries, annotated over 10,000 expressed sequence tags (ESTs), and produced cDNA microarrays (called I-gene microarrays) (33Yu J. Farjo R. MacNee S.P. Baehr W. Stambolian D.E. Swaroop A. Genome Biol. 2003; (http://genomebiology.com/2003/4/10/R65)Google Scholar, 34Farjo R. Yu J. Othman M.I. Yoshida S. Sheth S. Glaser T. Baehr W. Swaroop A. Vision Res. 2002; 42: 463-470Crossref PubMed Scopus (40) Google Scholar). Here, we report the expression profile of the mature retina from the rod-less (and cone-only) Nrl-/- mice using I-gene microarrays and compare it to the gene profile of the rod-dominated wild type mouse retina. We demonstrate differential expression of several genes, encoding phototransduction, structural, and signaling proteins, in the Nrl-/- retina. Our data reveal novel differentially expressed genes for future functional studies. Of particular importance are the findings that genes encoding components of Bmp/Smad and Wnt/Ca2+ signaling pathways are expressed in the mature retina and their expression is altered in the Nrl-/- retina. Expression analysis by real-time PCR and chromatin immunoprecipitation studies suggest that the activity of the Smad-mediated Bmp signaling pathway is modulated by Nrl in the mature retina. In support of this, we show that Bmp4 and Smad4 are expressed in the rod photoreceptors of the mature mouse retina. We also propose that rods and cones exhibit selective bias in the utilization of different signaling pathways for cell-cell communication and controlling intracellular functions. Tissue Preparation, RNA Isolation, and Northern Analysis—All procedures involving mice were approved by the University Committee on Use and Care of Animals of the University of Michigan. Retinas were dissected from the wild type and Nrl-/- mice at postnatal day (P) 21 and snap-frozen on dry ice. Total RNA was isolated using TRIzol reagent (Invitrogen) and purified by using an RNeasy kit (Qiagen). RNA integrity was verified by denaturing formaldehyde-agarose gels. Total RNA samples with a 260- to 280-nm absorbance ratio of greater than 1.9 were used for studies. Northern analysis was performed as described previously (10Swaroop A. Xu J.Z. Pawar H. Jackson A. Skolnick C. Agarwal N. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 266-270Crossref PubMed Scopus (269) Google Scholar). Reference RNA for Retinal Gene Profiling—To generate the reference RNA for microarray hybridizations, total RNA was pooled from the following tissues and cell lines: mouse eye or retina at embryonic day (E) 14–16, P2–3, P10–12, and adult (5.5 mg); mouse adult brain (1.5 mg); P19 embryonic carcinoma cells (35McBurney M.W. Jones-Villeneuve E.M. Edwards M.K. Anderson P.J. Nature. 1982; 299: 165-167Crossref PubMed Scopus (577) Google Scholar) (3 mg); retinoic acid-induced P19 cells that differentiated into neuronal and glial-like cells (3 mg); and neuroblastoma cell line N1E-115 (36Amano T. Richelson E. Nirenberg M. Proc. Natl. Acad. Sci. U. S. A. 1972; 69: 258-263Crossref PubMed Scopus (609) Google Scholar) (4 mg). The pooled RNA was divided into aliquots (at a concentration of 1.28 μg/μl) and stored at -80 °C until use. Target Labeling and Microarray Hybridization—Mouse I-gene cDNA microarray slides (34Farjo R. Yu J. Othman M.I. Yoshida S. Sheth S. Glaser T. Baehr W. Swaroop A. Vision Res. 2002; 42: 463-470Crossref PubMed Scopus (40) Google Scholar) contained PCR-amplified products from over 6,500 cDNAs, printed in duplicate; cDNAs were isolated from libraries constructed from E15.5 eyes, P2 eyes, and adult retinas, sequenced, and annotated (33Yu J. Farjo R. MacNee S.P. Baehr W. Stambolian D.E. Swaroop A. Genome Biol. 2003; (http://genomebiology.com/2003/4/10/R65)Google Scholar). Target RNA (10 μg of total RNA) was labeled using a 3DNA Submicro Expression Array Detection kit (Genisphere, Hatfield, PA) and hybridized to microarray slides, as described (37Yu J. Othman M.I. Farjo R. Zareparsi S. MacNee S.P. Yoshida S. Swaroop A. Mol. Vis. 2002; 8: 130-137PubMed Google Scholar). The slides were scanned using an Affymetrix 428 scanner (Affymetrix, Santa Clara, CA) to obtain the highest intensity of signal, without reaching saturation for a maximum of 10 out of the 13,440 spots. Image Processing and Data Analysis—Scanned array images without major defects (such as scratches or blobs) were analyzed using AnalyzerDG (MolecularWare Inc., Cambridge, MA) in a batch mode. A "contour shape" was utilized to detect spots for intensity calculations, whereas a "cell method" was set to calculate background on an individual basis in a local square region centered on the spot. A data file containing spots' intensities and annotations was exported for each array in the tab-delimited text format and then imported into the statistical package, R (available at www.r-project.org/). Intensities of Cy3 and Cy5 channels for each spot were calculated separately by subtracting corresponding median background from the mean signal intensity. Genes with negative background-corrected intensities in either channel were filtered out. To reduce systematic variation caused by experimental procedures (such as dye effects), a data-driven normalization was applied to individual datasets using a cluster of least-altered genes on the array identified by a rank-based algorithm. Briefly, let (Rj, Gj) denote the measurements for the jth gene in the red (Cy5) and green (Cy3) channels, respectively, j = 1,..., m. We calculate the ranks in each channel separately, take the difference in ranks between the two channels, and fit a three-component normal mixture model to the difference in ranks for the m genes. There will be three classes of genes to consider: those for which the rank in the red channel is significantly higher than that in the green channel, those for which the rank in the green channel is significantly higher than that in the red channel, and those for which the ranks do not substantially change between the two channels. The genes whose ranks did not change between the two channels were used to perform a slide-dependent normalization based on a locally weighted linear squares procedure (38Cleveland W.S. J. Am. Stat. Assoc. 1979; 74: 829-836Crossref Scopus (7474) Google Scholar). This allowed us to normalize genes and redefine a new horizontal zero axis that was used to compute log ratios of intensities in two channels. For indirect comparisons, log ratios of the wild type slide were subtracted from the Nrl-/- slide to obtain reference-corrected values for each gene. An Empirical Bayes method was then applied to the replicated arrays to obtain a B statistic for each gene (39Lonnestedt I. Speed T. Statistica Sinica. 2002; 12: 31-46Google Scholar); the B statistic estimates the posterior log odds of differential expression. Quantitative Reverse Transcription-PCR—To validate gene expression changes, qRT-PCR analysis was performed as described (33Yu J. Farjo R. MacNee S.P. Baehr W. Stambolian D.E. Swaroop A. Genome Biol. 2003; (http://genomebiology.com/2003/4/10/R65)Google Scholar). Briefly, total RNA (2.5 μg) treated with RQ1 RNase-free DNase (Promega, Madison, WI) was subjected to reverse transcription using oligod(T) primers and with (+RT sample) or without (-RT) SuperScript II (Invitrogen). For each gene, real-time PCR reactions from wild type and Nrl-/- samples were performed in triplicate with SybrGreen I (Molecular Probes, Eugene, OR) and analyzed with the iCycler IQ real-time PCR detection system (Bio-Rad). The average threshold cycle (Ct) differences between the two samples were normalized against Hprt (control) in the corresponding cDNA preparation. Immunohistochemistry—Anti-Gnb3 polyclonal antibodies were generated in rabbit against the peptide (ADITLAELVSGLEVV) and affinity-purified (Invitrogen). Anti-Smad4 rabbit polyclonal antibody (H-552) was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Due to the low number of cones in wild type mouse retina (3% of photoreceptors), we also performed immunohistochemical analyses on frozen sections of adult pig retina, which contains ∼15–20% cones (40Hendrickson A. Hicks D. Exp. Eye Res. 2002; 74: 435-444Crossref PubMed Scopus (141) Google Scholar). Cryostat sections of mouse and pig retina were permeabilized with 0.1% Triton X-100 (5 min), and then preincubated in blocking buffer (PBS supplemented with 0.2% bovine serum albumin, 0.1% Tween 20, 5% normal rabbit serum, and 0.1% NaN3) for 30 min. The sections were then incubated overnight at 4 °C in anti-Gnb3 and anti-Smad4 polyclonal antibodies (diluted 1:200 in blocking buffer), combined either with rho-4D2 anti-rod opsin monoclonal antibody (1 μg/ml) (41Hicks D. Molday R.S. Exp. Eye Res. 1986; 42: 55-71Crossref PubMed Scopus (192) Google Scholar) or with biotinylated peanut agglutinin (PNA, Vector Laboratories, 10 μg/ml). After extensive washing in PBS, sections were incubated in a mixture of goat anti-rabbit IgG-Alexa488 or -Alexa594 (Molecular Probes; 2 μg/ml), and either rabbit anti-mouse IgG- or streptavidin-Alexa594 or -Alexa488 (each 2 μg/ml) in blocking buffer for 2 h. Slides were washed extensively, mounted, and viewed by fluorescence microscopy (Nikon Optiphot 2) or by laser scanning confocal microscopy (Zeiss LSM 510 version 2.5 scanning device with Zeiss Axiovert 100 inverted microscope). Control experiments were performed by omitting the primary antibody. Chromatin Immunoprecipitation—A commercially available assay kit (Upstate Biotechnologies, Charlottesville, VA) was used for ChIP studies. Briefly, four snap-frozen retinas from wild type mice were cross-linked for 15 min at 37 °C with 1% formaldehyde in PBS containing proteinase inhibitors. The retinas were washed four times in ice-cold PBS with proteinase inhibitors and then incubated on ice for 15 min. The tissue was then sonicated on ice eight times using 20-s pulses. The remaining steps were essentially performed as described by the manufacturer, using anti-NRL polyclonal antibody (12Swain P.K. Hicks D. Mears A.J. Apel I.J. Smith J.E. John S.K. Hendrickson A. Milam A.H. Swaroop A. J. Biol. Chem. 2001; 276: 36824-36830Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar). Putative promoter regions (5′ upstream of the transcription start site) for Rho, Bmp4, Smad4, and Bmpr1a were determined in silico (www.ncbi.nlm.nih.gov/mapview). Each DNA sequence was analyzed using MatInspector (www.genomatix.de/index.html). PCR primers were designed to flank the putative Nrl binding sites (Nrl response element, NRE) (18Rehemtulla A. Warwar R. Kumar R. Ji X. Zack D.J. Swaroop A. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 191-195Crossref PubMed Scopus (186) Google Scholar) either predicted manually or by MatInspector (Table I). The sequence closest to the transcription start site was chosen. ChIP DNAs from two independent experiments were used for PCR using equal amounts for input, with antibody, and no antibody reactions.Table IPutative NRE and the PCR primers used for ChIP enrichment assaysGenePutative NRE (base pairs upstream of the transcription start site)ChIP PCR primersBmp4GACAGTGACGCAGGGAATCAA (—1495)Forward 5′-AACCTGCTATGGGAGCACAG-3′Reverse 5′-GGAATGTCAGGTTGGAAGGA-3′Smad4CAATTTTGATGACAATGAAGAATTG (—1299)Forward 5′-GCATCCAAAGGGTCATGAGT-3′Reverse 5′-GGGTAAGCCAAAGGGACAAT-3′Bmpr1aATGTATGACTGTGCATCACAT (—1684)Forward 5′-GGTGGATATGAGGGAATGGA-3′Reverse 5′-TGGTGGTCCACTACCATCTG-3′RhodopsinGGATGCTGAATCAGCCTCT (—70)Forward 5′-GATGGGATAGGTGAGTTCAGGA-3′Reverse 5′-GAGAAGGGCACATAAAAATTGG-3′ Open table in a new tab In Situ Hybridization—The 35S-labeled antisense and sense cRNA transcripts were synthesized from the mouse Bmp4 and Smad4 cDNAs cloned in pSPORT1 vector, according to the manufacturer's instructions (Ambion). The labeled transcripts were subjected to alkali hydrolysis (to give an average size of 70 nucleotides) and then hybridized overnight to 5- to 7-μm serial sections of adult mouse retina using standard protocols (Phylogeny Inc., Columbus, OH). After stringent washing, the slides were dipped in Kodak NTB-2 nuclear track emulsion and exposed for 1–2 weeks. Slides were developed in Kodak D-19, counterstained with toluidine blue, and analyzed by light and dark-field optics. Expression Analysis Using Purified Rod Photoreceptors—To obtain a highly enriched population of rod photoreceptors, we used Nrl-GFP transgenic mice that express enhanced green fluorescent protein (EGFP) under the control of Nrl promoter specifically in rods. 4M. Akimoto and A. Swaroop, unpublished data. Retinas of Nrl-GFP mice at P28 were dissected and dissociated using the Papain Dissociation System (Worthington). Cells that were positive and negative for EGFP were flow-sorted by FACSAria (BD Biosciences). Total RNA was used for RT-PCR using gene-specific primers, as follows. Bmpr1a: forward, AAGGAATGGGTGGGATTAGC; reverse, TGGCGATTGCCAACTAGATA; Grm6: forward, CAAGTAGCAAGGTTGAGTGT; reverse, GGAAGAATGCTGGAAGCAAG; Hprt: forward, CAAACTTTGCTTTCCCTGGT; reverse, CAAGGGCATATCCAACAACA; Nrl: forward, GCTGCATTTTCACCGAATCT; reverse, GGTGGTTTGGGTTGTGGTAG; Rho: forward, CTTCCTGATCTGCTGGCTTC; reverse, ACAGTCTCTGGCCAGGCTTA; Smad4: forward, ACCCGCGTATGCCGCCCCATCC; reverse, ACAGCGTCGCCAGGTGCTCGGC; Thy1: forward, AACTCTTGGCACCATGAACC; reverse, AGGCTGAACTCATGCTGGAT. We chose to generate gene expression profiles of P21 retinas from rod-dominated wild type and rodless (cone-only) Nrl-/- mice. At this stage, the differentiation and laminar organization of retinal neurons are complete, and phototransduction pathways in the retina (from photoreceptors to ganglion cells) are fully functional. High Concordance between Direct and Indirect Microarray Comparisons—We examined two different experimental designs, direct and indirect comparison (42Yang Y.H. Speed T. Nat. Rev. Genet. 2002; 3: 579-588Crossref PubMed Scopus (603) Google Scholar), for their ability to identify differentially expressed genes. Two direct comparison experiments (two slides), in which Cy3-labeled wild type and Cy5-labeled Nrl-/- retinal RNA targets were hybridized simultaneously to the same slide, were performed. Five indirect comparisons were carried out with the reference RNA labeled by Cy3 and hybridized in conjunction with Cy5-labeled either wild type or Nrl-/- retinal RNA (total of 10 slides). Of the 13,440 spots on the I-gene microarray slides, 97.4% (13,092 spots) showed higher spot intensity than the background (i.e. were considered detected) in both direct comparison slides, whereas 91.6% (12,307 spots) spots were detected in all 10 slides with indirect comparisons. Scatter-plot analysis of signal intensities of the wild type and Nrl-/- retinas revealed high similarity between direct and indirect comparisons (Fig. 1). A majority of genes have log2 ratios centered at 0, indicating no change between the two tested samples. Only a few spots displayed over 4-fold change in both methods, although direct comparison showed a tighter scatter within the 4-fold lines. Both methods were able to successfully identify the duplicate spots of S-opsin and rhodopsin as the most up- or down-regulated genes in the Nrl-/- retina, respectively. To compare the power of direct versus indirect methods in identifying differential gene expressi
Referência(s)