sign in sign up
<<
Home , Calretinin, YWHAB, YWHAZ

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 580 (2006) 331–335

Microarray analysis of expression in adult retinal cells

Dmitry Ivanova,b,*, Galina Dvoriantchikovaa, Lubov Nathansonc, Stuart J. McKinnond, Valery I. Shestopalova,e,* a Bascom Palmer Eye Institute Department of Ophthalmology, University of Miami Miller School of Medicine, 163 NW 10th Avenue, Miami, FL 33136, USA b Vavilov Institute of General Genetics RAS, Moscow, Russia c Department of Molecular and Cellular Pharmacology, University of Miami Miller School of Medicine, Miami, FL, USA d Departments of Ophthalmology and Neurobiology, Duke University Medical Center, Durham, NC, USA e Department of Cell Biology and Anatomy, University of Miami Miller School of Medicine, Miami, FL, USA

Received 5 November 2005; revised 25 November 2005; accepted 6 December 2005

Available online 19 December 2005

Edited by Lukas Huber

to characterize the molecular profile of the fully differentiated Abstract Retinal ganglion cells (RGCs) transfer visual infor- mation to the brain and are known to be susceptible to selective adult RGCs and determine how it differs from other retinal degeneration in various neuropathies such as glaucoma. This cells. Microarrays have been effectively implemented in such selective vulnerability suggests that these highly specialized neu- studies. Recent work reported successful isolation of juvenile rons possess a distinct profile that becomes al- RGCs by immunopanning followed by in vitro cultivation for tered by neuropathy-associated stresses, which lead to the the purposes of EST analysis [7]. However, this approach re- RGC death. In this study, to identify expressed predomi- lied on in vitro sub-cultivation, which is unlikely to work with nantly in adult RGCs, a global transcriptional profile of purified adult due to their low survival rate in vitro [8]. Re- primary RGCs has been compared to that of the whole retina. To cent advances in RNA amplification have made it possible avoid alterations of the original gene expression profile by cell to study gene expression in ultra-small samples, and even de- culture conditions, we isolated RNA directly from adult RGCs tect differences between individual cell types within the same purified by immunopanning without prior sub-cultivation. Genes expressed predominantly in RGCs included: Nrg1, Rgn, 14-3-3 tissue [9–11]. This approach allows gene expression profiling family (Ywhah, Ywhaz, Ywhab), Nrn1, Gap43, Vsnl1, Rgs4. of a small number of purified adult RGCs without an Some of these genes may serve as novel markers for these neu- in vitro cultivation, and eliminates potential concerns over cell rons. Our analysis revealed enrichment in genes controlling the culture-induced alterations of the gene expression profile. In pro-survival pathways in RGCs as compared to other retinal this work we performed microarray analysis using amplified cells. RNA extracted both from the retina and adult RGCs purified Ó 2005 Federation of European Biochemical Societies. Published by immunopanning without in vitro sub-cultivation. The by Elsevier B.V. All rights reserved. analysis of genes enriched in RGCs allowed us to identify novel RGC markers and genes essential for survival and Keywords: Retina; Retinal ganglion cells; Immunopanning; homeostasis in these cells. RNA amplification; Gene expression; Cell markers

2. Materials and methods

2.1. Isolation of RGCs 1. Introduction All experiments were performed in compliance with the ARVO statement for use of animals in ophthalmic and vision research. Mammalian retina is a complex tissue composed of neuro- Brown Norway rats (250–300 g) were euthanized according to the IA- nal, glial, and vascular cell types among which retinal gan- CUC approved protocol. Eyes were enucleated and retinas were glion cells (RGCs) are the only type of output neurons that mechanically dissected out. RGCs were isolated according to the two-step immunopanning method reported earlier to yield 95–99% send their axons to the visual cortex of the brain [1].Itis enrichment for RGCs [12]. In order to preserve RNA and limit po- now well established that the crosstalk between RGCs and tential gene expression changes we modified the procedure to shorten different retinal cell types is critical for RGC survival [2–4]. the total treatment time to 4 h. Briefly, the whole retinas were incu- In various neuropathies like glaucoma known to progress bated in papain solution (16.5 U/ml) for 30 min. To analyze similarly with aging, RGCs die, while other retinal cells remain mostly treated samples, whole retina suspensions were sampled for RNA extraction only at this point. In the next step macrophage and endo- unaffected [5,6]. As an essential step to understanding the thelial cells were removed from the cell suspension by panning with molecular basis of such selective vulnerability, it is important the anti-macrophage antiserum (Axell Accurate Chemical Corp., Westbury, NY). RGCs were specifically bound to the panning plates containing anti-Thy1.1 antibody, and unbound retinal cells were re- *Corresponding authors: Fax: +1 305 547 3658. moved by washing with DPBS. Purified RGCs were released by tryp- E-mail addresses: [email protected] (D. Ivanov), sin incubation and immediately processed for RNA extraction. To [email protected] (V.I. Shestopalov). eliminate the effects of genetic variability and reduce the animal group size, the comparison of the RGC expression profile to that Abbreviations: GCL, ganglion cell layer; INL, inner nuclear layer; of a whole retina was performed using samples derived from the same ONL, outer nuclear layer; PRL, photoreceptor layer animal.

0014-5793/$32.00 Ó 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2005.12.017 332 D. Ivanov et al. / FEBS Letters 580 (2006) 331–335

2.2. RNA extraction 1792 in the UniGene library). Selected genes were classified according RNA from whole retina and RGCs was extracted from three differ- to category ‘‘biological process’’ using Onto-Express ent biological samples. Processed individual samples from the same (http://vortex.cs.wayne.edu/Projects.html) [15]. animal contained about 100000 cells in each. Cells were collected, trea- Ò ted with the lysis buffer supplied with the Absolutely RNA Nanoprep 2.5. qRT-PCR, in situ hybridization, immunocytochemistry kit (Stratagene, USA), and RNA was purified using this kit according For qRT-PCR, in situ hybridization and immunocytochemistry we to the manufacturer’s protocol. Prior to amplification, the RNA qual- applied standard techniques; brief protocols including the table of ity was tested in each sample by qRT-PCR for Thy1, Vim, GFAP, primers are included in Supplement 1. S100b, CD11b, CD34 and Actb (Supplement 1).

2.3. RNA amplification, labeling and array hybridization 3. Results Amplification was performed using the Amino Allyl MessageAmpe Kit (Ambion, USA), as described previously [11]. Amplified Rat Uni- versal Reference RNA was labeled with Cy-3; experimental aRNAs 3.1. RGC and retina specific gene expression from retina and RGCs were labeled with Cy-5 dye (Amersham, We compared both RGC and retina specific gene expression USA). Equivalent amounts of reference and experimental aRNAs were profiles using 2.0-fold change as a threshold and identified 326 hybridized with the Agilent Rat Oligo Microarrays (Agilent, USA) genes enriched in RGCs, and 330 genes enriched in the whole according to the manufacturer’s instructions. retina (Supplements 2 and 3; Table 1). Relative transcript enrichment detected by microarrays was confirmed by qRT- 2.4. Microarray image and data analysis PCR for 21 genes randomly selected from both the RGC and Each array was normalized for signal intensities across the whole ar- retina gene groups (Table 2). Functional classification of a sub- ray and locally, using the Lowess normalization [13]. We analyzed only those genes that passed quality control criteria described previously set of RGC-enriched transcripts is shown in Table 1. Several [11]. To compare the gene expression levels in RGCs with those in genes, like Amhr2 and Rgn, were under-represented in the ret- the retina, the ‘‘RGC/reference RNA’’ ratios were divided by the ratios ina and for this reason were not detected by microarrays with of ‘‘retina/reference RNA’’ obtained from the same animal. One class retina samples. For those genes we performed qRT-PCR and SAM (http://www-stat.stanford.edu/tibs/SAM) with FDR <1% and fold change of 2.0 was used to determine genes with consistent differ- used the data to calculate the RGC/retina ratios (included in ences in expression [14]. To verify the ‘‘RGC/reference RNA’’ micro- Table 2). Analysis of the main gene ontology (GO) categories array dataset for specificity to RGCs we compared our data with the revealed that genes vital for neurogenesis and survival were en- data from UniGene library (ID.14630) for RGCs [7]. Our dataset riched in RGCs; conversely, genes essential for visual percep- was obtained as the result of the application of one class SAM tion were enriched in the whole retina. Notably, about two (FDR <1%) to identify genes with consistent ratios of ‘‘RGC/reference RNA’’. The second dataset contained genes from the UniGene library, thirds of the differentially enriched transcripts belonged to which are present on the Agilent Rat Oligo microarray (1137 out of ESTs, some of which potentially representing novel genes.

Table 1 Functional annotation for the selected top genes enriched in RGCs Gene RGCs/retina FDR Function Nrn1; neuritin 6.18 0.088 Promotes neurite outgrowth Prph1; peripherin 1 5.35 0.088 Cytoskeleton organization Gap43; growth associated 43 5.22 0.088 Neurite outgrowth and nerve regeneration Vsnl1; visinin-like 1 5.01 0.088 Calcium dependent signalling Rgs4; regulator of G-protein signaling 4 4.85 0.287 G-protein coupled signaling Nef3; neurofilament 3, medium 4.32 0.088 Cytoskeleton organization Ywhah; 14-3-3 family, eta polypeptide 4.08 0.088 Negative regulation of Ywhaz; 14-3-3 family, zeta polypeptide 3.91 0.088 Mitochondrial import stimulation factor Sema6b; semaphorin 6B 3.82 0.238 Neurogenesis and axon guidance Nrp; neuropilin 1 3.59 0.088 Neuronal development Rnd1; Rho family GTPase 1 3.57 0.895 Neuritic process formation Plcb1; phospholipase C, beta 1 3.52 0.287 Production of the second messenger molecules Gsto1; glutathione S-transferase omega 1 3.40 0.446 NCAML1; neural L1 3.39 0.088 Neural and glial cell adhesion Nfl; neurofilament, light polypeptide 3.34 0.088 Cytoskeleton organization Ptpn5; protein tyrosine phosphatase, non-receptor type 5 3.24 0.088 Regulating the duration of ERK activation Inpp4a; inositol polyphosphate-4-phosphatase, type 1 3.19 0.088 Phosphatidylinositol-3,4-bisphosphate 4-phosphatase activity Stmn2; stathmin-like 2 3.12 0.895 Microtubule depolymerization and synaptic plasticity Ret; Ret proto-oncogene 3.07 0.355 Neurogenesis Nrg1; neuregulin 1 2.86 0.789 Neural and organ development Rgs3; regulator of G-protein signalling 3 2.76 0.355 G-protein coupled receptor signaling Map2k1; activated protein kinase 1 2.73 0.088 Pik3r1; phosphatidylinositol 3-kinase, regulatory subunit, polypeptide 1 2.66 0.088 Negative regulation of apoptosis Coch; coagulation factor C homolog, cochlin 2.41 0.789 Perception of sound Ywhab; 14-3-3 family, beta polypeptide 2.34 0.088 Signal transduction Fbxo2; F-box only protein 2 2.30 0.355 Maintaining neurons in a postmitotic state Nell2; Nel-like 2 homolog 2.22 0.287 Growth and differentiation of neural cells Calb2; 2 (Calretinin) 2.22 0.713 Calcium ion binding Ywhag; 14-3-3 family, gamma polypeptide 2.17 0.713 Signal transduction Grb2; receptor bound protein 2 2.15 0.088 Signal transduction D. Ivanov et al. / FEBS Letters 580 (2006) 331–335 333

Table 2 ina according to our results: 14-3-3 genes family, Nrn1, Stmn2, Validation of microarray data for randomly selected genes by Rgs4, Rgs3, Vsnl1 (Table 1). The set of 326 RGC-enriched quantitative RT-PCR genes identified in this study had 66 genes common with the Gene Real time RT-PCR Array Confirmed by EST gene set. The expression of 260 non-overlapping genes (expression ratio) (expression ratio) real-time PCR was a novel finding (Supplement 4). Out of 260 genes nine were Amhr2a 42.1 Only RGCs Yes confirmed by qRT-PCR (Table 2), seven others were the estab- Apoe À11.2 À12.5 Yes lished RGC (Uchl1 [19]), or neuronal markers (Prph1, Gap43, Cfh Only retina À3.41 Yes Cocha 2.60 2.41 Yes Sema6b, Nrp, NCAML1, Nrg1 see Table 1). Gap43a 4.70 5.22 Yes Map2k1 1.92 2.73 Yes 3.3. Confirmation of RGC-specific gene expression a Nrg1 2.55 2.86 Yes Location of signal distribution in our in situ hybridization Nrn1 7.00 6.18 Yes Nrpa 3.14 3.59 Yes experiments was carried out by screening for the RGCs based Pik3r1a 3.49 2.66 Yes on their characteristic appearance (large round cell bodies Pls3a 4.90 5.67 Yes filled with nuclei [20]) and location (ganglion cell layer Pros1 Only retina À6.50 Yes a (GCL) of the retina). We examined cellular localization of Prph1 6.06 5.35 Yes transcripts Ywhah (14-3-3 eta), Rgn and Nrg1 that were highly Rgna 36.5 Only RGCs Yes Rgs4 5.83 4.85 Yes enriched in RGCs according to our microarray data. Our re- Tgfb2 À3.74 À5.57 Yes sults confirmed the relative accumulation of these transcripts Thy1 3.13 1.68 Yes in the GCL of the rat retina (Fig. 1). We also examined cellular Vsnl1 3.89 5.01 Yes distribution of Nrg1 and 14-3-3 zeta/beta , which were Ywhab 2.75 2.34 Yes Ywhah 5.72 4.08 Yes markedly enriched in RGCs in both microarray and qRT-PCR Ywhaz 2.06 3.91 Yes data (Table 2). Representative images shown in Fig. 2 demon- a strate a prominent immunostaining for Ywhaz/Ywhab and –RGC-enriched genes first identified in this study. Nrg1 localizing primarily in the GCL, and to a lesser extent, in the inner nuclear layer (INL) of the retina. Specific labeling 3.2. Validation of the microarray data localized in the RGC cytoplasm and processes (Fig. 2). A total of 8069 genes (out of 20500) passed the quality con- trol criteria, and showed consistent results in all experiments with RGC-derived samples. To further verify this gene set as 4. Discussion being RGC-specific, we searched it for the established RGC markers and compared it to the published EST data generated In this study we aimed to identify genes preferentially ex- from juvenile rat RGCs [7]. The microarrays used in our exper- pressed in the adult rat RGCs. To achieve this we developed iments contained 1137 genes previously identified by the EST and successfully used a technique for microarray profiling analysis, 864 (76%) of these genes were also present in the the gene expression of RGCs purified by the immunopanning RGC-derived gene set. Notably, the overlapping genes in- technique. This technique was earlier reported to produce up cluded RGC markers Thy1, Nef3, Nfl, Nefh, Calb2 [16–18], to 99% pure isolations of both newborn and juvenile RGCs as well as genes that were enriched in RGCs compared to ret- after in vivo sub-cultivation [7,12]. However, it has never been

Fig. 1. Photomicrograph of in situ hybridization of fixed rat retina slices with the anti-sense probes for 14-3-3 (A), Rgn1 (B) and Nrg1 (C) genes; D, E and F – sense (control) probes. The bulk of specific DIG labeling (blue) localized to the ganglion cell layer (GCL) of the retina. The bar is 50 lm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 334 D. Ivanov et al. / FEBS Letters 580 (2006) 331–335

Fig. 2. Immunostaining for 14-3-3 beta/zeta (red) and Nrg1 (green) in rat retina confirmed preferential expression of these proteins in the cytoplasm of RGCs (arrows). In addition, staining for the both marker genes was evident in the RGCs processes (arrows) and the INL (14-3-3 only, yellow arrowhead). Nuclei were stained with DAPI (blue). The bar is 50 lm; inserts – 10 lm. used to isolate RNA from adult RGCs because they typically cant enrichment of RGCs in 14-3-3 proteins suggests a critical fail to survive in vitro. Here we showed that amplification of role for these proteins in preserving RGC homeostasis. Along total RNA from purified adult RGCs could be successfully with a number of genes implicated in neurogenesis, genes in- used for microarray profiling. This method eliminates the need volved in neurite outgrowth, axon guidance and regeneration to sub-cultivate adult RGCs. The reference-based experimen- were also enriched in RGCs (Table 1 and Supplement 2). tal design allowed for the comparison of gene expression data In conclusion, we demonstrated in this work that purified with future experimental data. adult RGCs could be successfully analyzed by microarrays. Our approach has been successful in identifying novel cell 4.1. RGC-enriched and marker genes markers and genes essential for the metabolism and survival Both the significant overlap (76%) with published EST data of these neurons. [7] and the presence of all major RGC marker genes among the 8069 genes detected by the microarray data confirms the good Acknowledgments: We thank Jeffrey Goldberg for the gift of T11D7e2 cells and the advice in RGCs immunopurification, Vlad Slepak, Sara quality of both RGC purification and RNA isolation. These Cleveland for the manuscript discussion and the staff of the University factors also indicate that most of the genes of interest in the of Miami Microarray Core Facility for the expert assistance. Grant microarray results are expressed in these neurons. The 66 genes support: NIH R01-EY14232 (V.S.), NIH R01-EY16516 (S.M.), NIH enriched in RGCs vs. retina common with the published EST Center Grant P30 EY014801, Fight for Sight fellowship PD05034 gene set encode proteins that could be used as novel RGCs (D.I.), Inc., a Research to Prevent Blindness (RPB) Career Develop- ment Award (V.S.) and an unrestricted grant to the University of markers (Table 1). Members of the 14-3-3 protein family Miami, and Duke University Medical Center. known for their pro-survival functions [21] were also found in this group. Our data showed two to fivefold enrichment in RGC vs. retina for four out of seven characterized 14-3-3 genes Appendix A. Supplementary data (Ywhah, Ywhaz, Ywhab, Ywhag). Moreover, among the 326 RGC-enriched genes identified in this study are several poten- Supplementary data associated with this article can be tially useful markers such as, Nrg1,Rgn, Gap43, Nrn1, Vsnl1 found, in the online version, at doi:10.1016/j.febslet.2005. and Rgs4. Immunohistochemistry for two of these markers 12.017. (Ywhaz/Ywhab and Nrg1) confirmed the microarray data and showed significant enrichment of these proteins in RGCs (Fig. 2). References

[1] Adams, S.T. (1965) Retina and optic nerve. Arch. Ophthalmol. 4.2. Functional classification of the enriched genes 73, 724–745. Being one of the most metabolically active and energy- [2] Sucher, N.J., Lipton, S.A. and Dreyer, E.B. (1997) Molecular demanding cell types in the mammalian body, RGCs rely on basis of glutamate toxicity in retinal ganglion cells. Vision Res. 37, the increased activity of pro-survival genes to sustain their nor- 3483–3493. [3] Fruttiger, M., Calver, A.R., Kruger, W.H., Mudhar, H.S., mal function [22,23]. Our analysis of RGC-enriched genes re- Michalovich, D., Takakura, N., Nishikawa, S. and Richardson, vealed among them Nrg1, Rgn and Nrn1, known for the pro- W.D. (1996) PDGF mediates a –astrocyte interaction in survival functions [24–26]; these genes may be potential novel the developing retina. Neuron 17, 1117–1131. RGC markers. A number of genes encoding components and [4] Sueishi, K., Hata, Y., Murata, T., Nakagawa, K., Ishibashi, T. and Inomata, H. (1996) Endothelial and glial cell modulators of the pro-survival PI3K/AKT signaling pathways interaction in diabetic retinopathy via the function of vascular were also found enriched in RGCs: Inpp4a [27], Pik3r1 [28], endothelial growth factor (VEGF). Pol. J. Pharmacol. 48, Grb2 [29] and Map2k1 [30]. Activated Akt has been shown 307–316. to mediate cell survival by phosphorylation of several targets [5] Quigley, H.A. (1995) Ganglion cell death in glaucoma: including Bad [31]. The sequestration of phosphorylated Bad recapitulates ontogeny. Aust. N Z J. Ophthalmol. 23, 85–91. [6] Quigley, H.A., Nickells, R.W., Kerrigan, L.A., Pease, M.E., and Bax by 14-3-3 proteins leads to release of anti-apoptotic Thibault, D.J. and Zack, D.J. (1995) Retinal ganglion cell death proteins Bcl-xL or Bcl-2, and promotes survival [21]. Signifi- D. Ivanov et al. / FEBS Letters 580 (2006) 331–335 335

in experimental glaucoma and after axotomy occurs by apoptosis. (1992) Distribution of protein gene product 9.5 (PGP 9.5) in the Invest. Ophthalmol. Vis. Sci. 36, 774–786. vertebrate retina: evidence that immunoreactivity is restricted to [7] Farkas, R.H., Qian, J., Goldberg, J.L., Quigley, H.A. and Zack, mammalian horizontal and ganglion cells. J. Comp. Neurol. 322, D.J. (2004) Gene expression profiling of purified rat retinal 35–44. ganglion cells. Invest. Ophthalmol. Vis. Sci. 45, 2503–2513. [20] Masland, R.H. (1986) The functional architecture of the retina. [8] Hu, D.N. and Ritch, R. (1997) Tissue culture of adult human Sci. Am. 255, 102–111. retinal ganglion cells. J. Glaucoma 6, 37–43. [21] Masters, S.C., Subramanian, R.R., Truong, A., Yang, H., Fujii, [9] Tsubaki, T., Arita, N., Kawakami, T., Shiratsuchi, T., Yamam- K., Zhang, H. and Fu, H. (2002) Survival-promoting functions of oto, H., Takubo, N., Yamada, K., Nakata, S., Yamamoto, S. and 14-3-3 proteins. Biochem. Soc. Trans. 30, 360–365. Nose, M. (2005) Characterization of histopathology and gene- [22] Osborne, N.N., Melena, J., Chidlow, G. and Wood, J.P. (2001) A expression profiles of synovitis in early rheumatoid arthritis using hypothesis to explain ganglion cell death caused by vascular targeted biopsy specimens. Arthritis Res. Ther. 7, 825–836. insults at the optic nerve head: possible implication for the [10] McClain, K.L., Cai, Y.H., Hicks, J., Peterson, L.E., Yan, X.T., treatment of glaucoma. Br. J. Ophthalmol. 85, 1252–1259. Che, S. and Ginsberg, S.D. (2005) Expression profiling using [23] Carelli, V., Ross-Cisneros, F.N. and Sadun, A.A. (2004) Mito- human tissues in combination with RNA amplification and chondrial dysfunction as a cause of optic neuropathies. Prog. microarray analysis: assessment of Langerhans cell histiocytosis. Retin. Eye Res. 23, 53–89. Amino Acids 28, 279–290. [24] Steinthorsdottir, V., Stefansson, H., Ghosh, S., Birgisdottir, B., [11] Ivanov, D., Dvoriantchikova, G., Pestova, A., Nathanson, L. and Bjornsdottir, S., Fasquel, A.C., Olafsson, O., Stefansson, K. and Shestopalov, V.I. (2005) Microarray analysis of fiber cell matu- Gulcher, J.R. (2004) Multiple novel initiation sites ration in the lens. FEBS Lett. 579, 1213–1219. for NRG1. Gene 342, 97–105. [12] Meyer-Franke, A., Kaplan, M.R., Pfrieger, F.W. and Barres, [25] Fukaya, Y. and Yamaguchi, M. (2005) Overexpression of B.A. (1995) Characterization of the signaling interactions that suppresses apoptotic cell death in the cloned rat promote the survival and growth of developing retinal ganglion hepatoma H4-II-E cells induced by a naturally occurring isothi- cells in culture. Neuron 15, 805–819. ocyanate sulforaphane. Int. J. Mol. Med. 15, 853–857. [13] Yang, Y.H., Dudoit, S., Luu, P., Lin, D.M., Peng, V., Ngai, J. [26] Naeve, G.S., Ramakrishnan, M., Kramer, R., Hevroni, D., Citri, and Speed, T.P. (2002) Normalization for cDNA microarray Y. and Theill, L.E. (1997) Neuritin: a gene induced by neural data: a robust composite method addressing single and multiple activity and neurotrophins that promotes neuritogenesis. Proc. slide systematic variation. Nucl. Acids Res. 30, 15–25. Natl. Acad. Sci. USA 94, 2648–2653. [14] Tusher, V.G., Tibshirani, R. and Chu, G. (2001) Significance [27] Nystuen, A., Legare, M.E., Shultz, L.D. and Frankel, W.N. analysis of microarrays applied to the ionizing radiation response. (2001) A null mutation in inositol polyphosphate 4-phosphatase Proc. Natl. Acad. Sci. USA 98, 5116–5137. type I causes selective neuronal loss in weeble mutant mice. [15] Khatri, P., Draghici, S., Ostermeier, G.C. and Krawetz, S.A. Neuron 32, 203–212. (2002) Profiling gene expression using onto-express. Genomics 79, [28] Ueki, K., Fruman, D.A., Brachmann, S.M., Tseng, Y.H., 266–336. Cantley, L.C. and Kahn, C.R. (2002) Molecular balance [16] Barnstable, C.J. and Drager, U.C. (1984) Thy-1 : a between the regulatory and catalytic subunits of phosphoino- ganglion cell specific marker in rodent retina. Neuroscience 11, sitide3-kinase regulates and survival. Mol. Cell 847–855. Biol. 22, 965–977. [17] Kong, W.C. and Cho, E.Y. (1999) Antibodies against neurofil- [29] Egan, S.E., Giddings, B.W., Brooks, M.W., Buday, L., Sizeland, ament subunits label retinal ganglion cells but not displaced A.M. and Weinberg, R.A. (1993) Association of Sos Ras amacrine cells of hamsters. Life Sci. 64, 1773–1778. exchange protein with Grb2 is implicated in tyrosine kinase [18] Volgyi, B., Pollak, E., Buzas, P. and Gabriel, R. (1997) Calretinin signal transduction and transformation. Nature 363, 45–51. in neurochemically well-defined cell populations of rabbit retina. [30] Robinson, M.J. and Cobb, M.H. (1997) Mitogen-activated Brain Res. 763, 79–86. protein kinase pathways. Curr. Opin. Cell. Biol. 9, 180–186. [19] Bonfanti, L., Candeo, P., Piccinini, M., Carmignoto, G., Comelli, [31] Thompson, J.E. and Thompson, C.B. (2004) Putting the rap on M.C., Ghidella, S., Bruno, R., Gobetto, A. and Merighi, A. Akt. J. Clin. Oncol. 22, 4217–4226.