Differential Gene Expression in Mouse Sclera during Ocular Development

Abstract

purpose. Ocular development involves changes in extracellular matrix components of the scleral wall as it expands. This study was conducted to determine scleral gene expression profiles during mouse ocular development to identify genes involved in normal scleral growth.

methods. Sample sets of pooled sclerae of 3- and 8-week-old mice were microdissected, and total RNA was isolated. After reverse transcription, the cDNA was in vitro transcribed to produce biotin-labeled cRNA. The purified biotin-labeled cRNA samples were hybridized to microarray chips (GeneChip Mouse Genome 430 2.0; Affymetrix, Santa Clara, CA). Gene transcript expression profiles were determined, and eight differentially expressed genes between the two age groups underwent further confirmation by real-time PCR analysis.

results. Differential regulation of 4884 gene transcripts in mouse sclera with less than 5% false-discovery rate (FDR) was identified. The top 1000 with the lowest FDR among the 4884 probe sets were filtered for threefold changes between the two age groups, and 718 gene transcripts were identified. Among these 718 gene transcripts, 210 were upregulated and 508 downregulated in adult relative to juvenile mouse sclera. TGF-β1 and several collagen genes were significantly downregulated. Microarray differential expression by real-time PCR validation of eight extracellular matrix–associated gene transcripts was confirmed.

conclusions. This is the first study to demonstrate gene expression profiles in mouse sclera during ocular growth. These findings support the role of TGFβ1 as a signaling molecule in modulating extracellular matrix during ocular development. This endeavor may be helpful in furthering understanding of how scleral remodeling is regulated during eye growth.

The ocular refractive components undergo precisely coordinated physical alterations during ocular growth, to attain and maintain normal emmetropic visual acuity, so that the image focus falls on the retinal plane. ¹ Any discordance between the axial length and other optical refractive components, such as corneal and lenticular curvatures would result in ametropia and blurred visual acuity, in which the focus of the image falls either in front of (myopia) or behind (hyperopia) the retina. It is thought that biological visual cues guide ocular growth, and determinants have been actively studied in experimental myopia in various animal species. ^{1 2 3 4 5 6} The models invariably alter the axial length of the eye with noted changes in scleral wall physiology. To date, the molecular events that synchronize the changes remain unclear. 

The sclera plays a key role in maintaining the axial length and the vitreous depth, a major factor contributing to the progression of myopia. ⁷ Studies of progressed human myopia have found significant reduction in collagen contents with corresponding reduced scleral thickness. ^{8 9 10 11} Compared with normal sclera, highly myopic eyes have smaller collagen fibril diameters with an altered lamellar interwoven fibril bundle arrangement, ¹⁰ suggesting potential changes in collagen components. 

Changes in scleral collagen fibrils were examined in animal models of induced myopia, where reduced scleral thickness has been observed. ^{12 13 14 15 16} In the tree shrew, regulation of collagen content occurs at both the mRNA and protein levels. There is noted significant reduction in the expression of type I collagen, whereas the expression of type III and V collagens remains relatively steady. ¹⁷ Increased degradation of collagen fibrils and the extracellular matrix (ECM) also occur during scleral remodeling. ^{18 19 20}  

The molecular mechanisms of scleral remodeling remain to be elucidated. Several signaling molecules have been examined for their roles in influencing scleral enlargement. ^{15 21 22 23 24 25 26} The family of multifunctional transforming growth factors (TGFs) regulate cellular processes, including ECM structure. ²⁷ Increased TGF-β levels result in increased collagen gene expression and protein synthesis in fibroblast cells. ²¹ Reduced TGF-β expression has been observed in the sclera of experimentally induced myopic animals, suggesting an important role in regulating scleral remodeling during the development of myopia. ²¹ Other genes and pathways, such as paired box gene 6 (PAX6), ^{28 29} members of the Notch signaling pathway, ^{30 31} sonic hedgehog (shh) signaling molecules, ³² and bone morphogenetic proteins (BMPs), ³³ may also play a role in scleral remodeling. 

We hypothesize that gene signaling during scleral expansion and remodeling during myopic development may have similarities to scleral enlargement signals noted during normal ocular growth toward emmetropia. Both processes may use common pathways to achieve increases in scleral surface area and eye size. To work toward testing this hypothesis and to identify genes and pathways responsible for scleral expansion during ocular development, we performed gene expression profiling using mouse sclera at two stages of development. It has been shown that eyes of mice that undergo visual form deprivation have significant myopia due to an increase in the axial length, ^{34 35} thus resembling other well-characterized mammalian and nonmammalian models of induced myopia. ^{12 13 14 15 16} This information may be useful for further studies of mouse sclera, especially in induced-myopia experiments. 

Methods

Animals

This study protocol was approved by the Institutional Animal Care and Use Committee of the Children’s Hospital of Philadelphia and was in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Eighteen each of C57BL/6J mice at 3- and 8- weeks of age were killed. Mouse eyes were microdissected to obtain sclera. Mouse sclera is a fibrous tissue containing small amounts of RNA. To obtain a sufficient amount for conducting oligonucleotide microarray analysis, sclerae from six mice in the same age group were pooled. Three independent pooled replicates from each age group were processed in parallel for all experimental procedures. 

RNA Extraction and cDNA Microarray

Scleral RNA was isolated (TRIzol) according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). RNA yields were assessed by absorbance at 260 nm, and the quality was confirmed (model 2100 bioanalyzer; Agilent Technologies, Inc., Palo Alto, CA). After reverse transcription of total RNA (Superscript II cDNA synthesis kit; Invitrogen), the cDNA was in vitro transcribed to produce biotin-labeled cRNA (Enzo Life Sciences, Inc, Farmingdale, NY). The purified biotin-labeled cRNA was fragmented into 50- to 200-bp products and hybridized to a gene microarray (GeneChip Mouse Genome 430 2.0 Array chips; Affymetrix, Inc., Santa Clara, CA) at 45°C for 16 hours. Posthybridization wash and streptavidin-phycoerythrin staining protocols were performed (Fluidics Station 450; Affymetrix, Inc.), and the microarray chips were scanned (GeneChip Scanner 3000; Affymetrix, Inc.). Six array chips (six pairs of mouse eyes per chip), representing three independent and pooled replicates for each age group were used in the study. 

Data Analysis

Data collection and probe level intensity analysis were performed (GeneChip Operating Software version 1.2; Affymetrix, Inc.). Probe intensities were normalized, and expression signals of all genes (probe sets) were calculated using GCRMA (GC robust multiarray analysis, as implemented in GeneSpring ver. 7.1; Silicon Genetics, Redwood City, CA). Principal components analysis (PCA) was used to assess variability of all samples. Data from one 3-week-old mouse array chip (chip no. 3) showed large variation and was excluded from further array statistical analysis. Although it was not included in the analysis, data from that chip showed a similar trend of gene expression relative to the two other 3-week-old mouse array chips. We are unable to repeat the experiment with an additional chip. 

GCRMA expression levels were recalculated from the five remaining chips (two from 3-week-old and three from 8-week-old mice), and differentially expressed genes were found using the local pooled error (LPE) statistical test (S+ ArrayAnalyzer ver. 2; Insightful Corp., Seattle, WA). The resampling technique was used to control the false-discovery rate (FDR). The LPE test is effective in identifying differentially expressed genes in experiments with limited numbers of replicates. ³⁶ Analysis of 45,038 probe sets of the Mouse Genome 430 2.0 Array chip using LPE testing with default parameters identified 4884 probe sets with an FDR of less than 5%. Among the 4884 probe sets, the top 1000 (as ranked by statistical significance) with an FDR of less than 5% were further filtered for changes. Differentially expressed gene transcripts at more than threefold were used to generate relevant interaction networks using Ingenuity Pathways Analysis (IPA; www.ingenuity.com/ Ingenuity Systems, Redwood City, CA). This is a Web-based application that enables the discovery, visualization, and exploration of interaction networks significant to cDNA microarray data sets. A data set containing gene identifiers and corresponding multiples of change (x-fold) in expression were uploaded as a tab-delimited text file. Each gene identifier was mapped to its corresponding gene object in the Ingenuity Pathways Knowledge Base (IPKB; Ingenuity Systems). A change cutoff of 3-fold was set to identify genes with expression that was significantly differentially regulated. These genes, called “focus genes,” were then used as the starting point for generating biological networks. A group of extracellular matrix gene–gene interactions in mouse sclera during ocular growth was identified. 

Real-Time PCR

Eight genes that showed significant differential expression profiles with cDNA microarray analysis were confirmed for expression by real-time PCR analysis (SYBR Green PCR Master Mix; Applied Biosystems [ABI], Foster City, CA). Primers were designed on computer (Primer Express software; ABI) to amplify 60- to 150-bp cDNA fragments across intron–exon boundaries (Table 1) . The mouse β-actin gene was used as an internal control based on its constant level of expression across different age groups in the present cDNA microarray analysis. Standard curves were generated for each gene. The dye (SYBR Green I; ABI) binds to the minor groove of double-stranded (ds)DNA; thus, the fluorescent quantification could contain signals from self-complementary primer–dimer pairs. To determine the specificity of DNA quantification, a melting curve was generated for each gene. 

The same scleral RNA from both pooled samples used for microarray analysis (18 each at 3- and 8- weeks of age including the 3-week-old mouse subset sample with large variations), and samples isolated from one 3- and one 8-week-old mouse were used for real-time PCR analysis. After reverse transcription of total RNA (Superscript II cDNA synthesis kit; Invitrogen), the cDNA was in vitro transcribed to produce cRNA (Epicenter, Madison, WI). The cRNA was then used as a template for the second cycle of cDNA synthesis. Real-time PCR was performed using 0.5 μL of cDNA in a 10-μL reaction for 40 cycles under the following conditions: 60°C for 1 minute, 95°C for 15 seconds and 300 nM of each primer. All experiments were performed in triplicate (model 7900HT Sequence Detection System; ABI). 

Results

The cDNA Microarray Analysis of Expressed Genes

A mouse whole-genome expression array containing representations of more than 39,000 gene transcripts (45,038 probe sets) was used in the study. Three independent pooled replicates for each age group of 3- and 8 week-old mice were prepared and hybridized to six microarray chips (GeneChip Mouse Genome 430 2.0 Array; Affymetrix). The percentage of gene transcripts detected as present on array chips ranged from 46% to 52%. Analysis of the 45,038 probe sets identified 4884 transcripts with a significant FDR of less than 5% (P < 0.05). The top 1000 transcripts with the lowest FDR among the 4884 probe sets were filtered for threefold changes between different age groups, and 718 were retained (Fig. 1) . Among these 718 gene transcripts, 210 were upregulated and 508 were downregulated in adult relative to juvenile mouse sclera. The top 40 up- and downregulated (ranked by fold-change) gene transcripts are listed in Tables 2 and 3 , respectively. 

Hierarchical clustering of gene expression data from each of the 3- and 8-week-old mouse sclera chips showed a higher degree of similarity within each age group than between different groups (Fig. 2) . The biological replicates for each age group were clearly clustered together. 

The 718 gene transcripts represent 10 functional categories based on literature review and the use of the Entrez Gene database (http://www.ncbi.nlm.nih.gov/entrez; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). The functional categories listed include cell adhesion (2.5%), extracellular matrix (11.6%), cytoskeleton (5.4%), signal transduction (17.8%), organogenesis (9.4%), apoptosis (1.2%), transcription (9.2%), transporter (8.6%), metabolism (9.2%), and EST (expressed sequence tag; 26.8%). 

Validation of Gene Expression with Real-Time PCR

Eight genes displaying a high ratio of differential expression in microarray analysis or functionally related to extracellular matrix composition were chosen for further validation by real-time PCR analysis. Four upregulated and four downregulated in adult relative to juvenile mouse sclera were selected. Upregulated genes included kinesin family member 5A (KIF5A), extracellular proteinase inhibitor (EXPI), vitronectin (VTN), and neurogenic differentiation 1 (NEUROD1). The downregulated genes included procollagen Vα1 (COL5A1), procollagen type XIα2 (COL11A2), elastin (ELN), and transforming growth factor β1 (TGFB1). Their differential expression profiles were verified by quantitative real-time PCR (SYBR Green; ABI) using total RNA from both pooled samples, the same sample used for microarray analysis, and samples isolated from one 3- and one 8-week-old mouse sclera. A melting curve was generated for each gene, and a single specific PCR product melted at the expected temperature was confirmed for all eight genes. The housekeeping gene β-actin was used as an internal control to normalize gene expression quantity, as its expression remained constant between different age groups based on our cDNA microarray analysis. 

As shown in Figure 3 , the real-time PCR results confirmed the microarray data, and the up- and downregulated genes in microarray experiments showed corresponding increased or decreased expressions with real-time PCR analysis. However, the multiples of change (x-fold) determined by the two techniques varied. For example, for collagens 5A1 and 11A2, microarray analysis showed downregulation of 12- and 10-fold after normalization by the housekeeping gene β-actin, respectively. The real-time PCR analysis showed only six- and sevenfold changes, respectively. 

Downregulation of TGF-β1 and Collagen Genes in Adult Sclera

The 4884 expressed gene transcripts with less than 5% FDR includes several major collagen types, such as types I to VI, VIII, X to XII, XIV to XV, XVIII to XIX, and XXIV. Among them, collagen types I, III to VI, and XI showed significant downregulation ranging from 3- to 24-fold in adult sclera. In addition, the extracellular matrix–associated genes elastin, integrin A5, and integrin B1, were significantly downregulated at 10-, 3- and 3-fold, respectively. Integrin A5 and B1 are components of the fibronectin receptor, which mediates cell–matrix adhesion. ³⁷ Analysis of expressed genes found TGF-β, -β1, -β2, and -β3 present in mouse sclera (<5% FDR). However, only TGF-β1 showed differential expression at more than threefold. We used the IPA program to highlight specific extracellular matrix genes thought to be involved in gene–gene interactions in mouse sclera during ocular growth, with the change notations shown in Figure 4 . 

Discussion

To our knowledge, this is the first study to investigate differential gene expression in mouse sclera. The cDNA microarray analysis in the present study identified the expression of approximate 4884 transcripts in mouse sclera with less than 5% FDR. There were 718 differentially expressed genes at more than threefold— 210 upregulated and 508 downregulated—in the adult mouse sclera. Among differentially expressed genes, TGF-β1 and several collagen genes were significantly downregulated (Fig. 4) . The broad and significant downregulation of collagen genes suggests that cellular signaling molecules mediate changes in mouse sclera remodeling. This is particularly important, considering that significant reductions in collagen contents and a corresponding reduced scleral thickness have been reported in both human myopic eyes and animal models of myopia. ^{8 9 10 11} Results of this study support the hypothesis that scleral expansion and remodeling during myopic development and scleral enlargement during normal ocular growth could share a common signaling pathway to achieve an increase in scleral surface area and anterior–posterior globe size. 

The TGFs are secreted cytokines which control extracellular matrix synthesis. ²¹ Scleral expression has been confirmed in various species. ^{15 21 38 39 40 41 42 43} It has been reported that the family of TGFs play important roles in controlling collagen gene expression. ^{15 21 41} Our cDNA microarray study found TGF-β, -β1, -β2, and -β3 present in both juvenile and adult mouse sclera. Only TGF-β1 showed significant differential expression at more than threefold. The other isoforms showed marginal changes at their expression levels. The role of TGF-β1 in scleral remodeling has been studied in detail in myopic tree shrew. ²¹ TGF-β1 expression was shown to be downregulated by 32%, 1 day after myopic development, with persistent lower levels of expression thereafter. 

Information on the role of TGF-β1 in normal ocular growth has been limited. Subconjunctival injections of TGF-β1 had no effect on refractive error in control eyes, and effects on the myopic eyes were inconsistent due to large individual variation in a chick experimental myopia eye model study. ²³ However, the presence of TGF-β1 reduced the rescue efforts by basic fibroblast growth factor (bFGF) of the myopic eye by 50%, as measured by axial length, suggesting that bFGF may act partially through the TGF-β1 pathway. Further characterization of TGF-β1 and bFGF pathways in mouse sclera may help to delineate mechanisms of ocular growth. 

Some early developmental genes may participate in late ocular growth. The homeobox gene Pax6 is essential for early ocular development, ⁴⁴ and it remains expressed in the adult eyes. ⁴⁵ In this study, we did not detect a significant change in Pax6 gene expression, nor did we find changes in the bone morphogenetic proteins (BMPs). Further data analysis and protein–protein interacting network exploration may help to uncover connections among differentially expressed genes. 

Studies of human myopic eyes show a significant decrease in scleral thickness and changes in collagen morphology. ^{8 9 10 11} Similar morphologic changes are also observed in animal models of myopia. ^{12 13 14 15 16} In myopic tree shrew sclerae, there was a 20% to 34% decrease in collagen type I expression and a 15% reduction in [³H] proline incorporation, suggesting reduced collagen synthesis. ^{17 20} The gene transcript expression profiles of collagen types III and V was unchanged. Our study also found significant downregulation in collagen type I gene expression. Unlike the myopic tree shrew, our study found decreased expression in several collagen genes, including types I, III to VI, and XI (Fig. 4) . The differences between the previous studies and the present study may be due to different animal models or different techniques. Additional information from other animal models may be useful. 

The scleral fibrils are assembled by heterogeneous collagen types, and the expressions of subtypes are dependent on various tissues and structures. ^{46 47} This study found 16 collagen subtypes present in mouse sclera. All collagen types have been recognized in sclera in previous animal studies. ^{17 47 48 49} Among these 16 subtypes, only types I, III to VI, and XI showed significant downregulation in adult sclera (Fig. 4) . Differential expression of collagen gene subtypes in juvenile relative to adult mice may suggest different scleral fibril composition and subsequent alteration in scleral rigidity and mechanical properties. 

Studies of scleral remodeling during normal and pathologic ocular growth have focused on known genes and factors historically. The cDNA microarray technology allows for gene expression study on a large scale. It is particularly important in ocular growth studies, as the molecular events occurring during development are complex, involving multiple factors and pathways. A detailed study of expressed scleral genes and a rational pathway during ocular development may help to identify signaling molecules that guide ocular growth. 

 Supported by National Eye Institute Grants R01 EY014685 and 2PEY01583, Research To Prevent Blindness, Inc., and the Mabel E. Leslie Endowment Fund.

 Submitted for publication June 16, 2005; revised November 26, 2005; accepted February 24, 2006.

 Disclosure: J. Zhou, None; E.F. Rappaport, None; J.W. Tobias, None; T.L. Young, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Real-Time PCR Gene Primer Pairs Used

Gene Name	RefSeq ID	Forward Primers (5′–3′)	Reverse Primers (5′–3′)
KIF5A	NM_008447	AGAACAACCTGGAACAGCTTACAA	CGAAGTCGTTTTTCCAATTTAGGA
EXPI	NM_007969	CTGGTAGCTTTGATTTTCATGACAA	GACAAGCGCCAGGTTTTTCT
VTN	NM_011707	CCATTCAGAGCGTCTATTTCTTCTC	TCCACTCGCCGGGTTCT
NEUROD1	NM_010894	CCCGAGGCTCCAGGGTTAT	CCCGCTCTCGCTGTATGATT
COL5A1	NM_015734	TGAATTCAAGCGTGGGAAACT	CCGCAGGAAGGTCATTTGTAC
COL11A2	NM_009926	TGGCACTCCTGGTCCAGAAG	GCCGGGCTTTCCTGCTA
ELN	NM_007925	GGCTTTGGACTTTCTCCCATT	CCGGCCACAGGATTTCC
TGFB1	NM_011577	CACCGGAGAGCCCTGGATA	TGCCGCACACAGCAGTTC
ACTB	NM_013556	AGGTCATCACTATTGGCAACGA	ATGGATGCCACAGGATTCCA

 ACTB was the β-actin housekeeping gene.

Figure 1.

 

Comparisons of microarray data from 3- and 8-week-old mouse sclera. The changes (x-fold) between different age groups are shown on the x-axis. The statistically significant values for a local pooled error (LPE) test of differences between different age group are shown on the y-axis. Horizontal threshold line: P = 0.01; vertical lines: two- and threefold changes, respectively (as labeled). The top, outer sextant points are genes with significant changes.

Table 2.

 

Representative Genes and ESTs Upregulated in 8-Week-Old-Mouse Sclera

Affymetrix ID	UniGene ID	RefSeq ID	Gene Names	Chromosomal Location	Expression Ratio* 8 wk/3 wk	Molecular Function
1422825_at	Mm.75498	NM_013732	Cocaine and amphetamine regulated transcript	chr13 D1	15.2 (2.9)	Neuropeptide hormone activity
1434670_at	Mm.30355	NM_008447	Kinesin family member 5A	chr10:70.0 cM	8.8 (3.4)	Motor activity
1441330_at	Mm.95700	NM_133239	Crumbs homolog 1	chr1:73.0 cM	8.5 (1.1)	Molecular function unknown
1433785_at	Mm.40461	NM_008614	Myelin-associated oligodendrocytic basic protein	chr9:69.0 cM	8.0 (4.7)	Structural constituent of myelin sheath
1417160_s_at	Mm.1650	NM_007969	Extracellular proteinase inhibitor	chr11:C	7.9 (5.7)	Endopeptidase inhibitor
1440256_at	Mm.222680	NM_145840	Regulator of G-protein signalling 9 binding protein	chr7:A3	6.8 (2.6)	Molecular function unknown
1435447_at	Mm.31625	NM_008600	Major intrinsic protein of eye lens fiber	chr10:74.0 cM	6.7 (5.9)	Structural constituent of eye lens
1436470_at	Mm.125521	NM_053271	Regulating synaptic membrane exocytosis 2	chr15:16.0 cM	6.6 (2.1)	Neurotransmitter transport
1424422_s_at	Mm.258142	NM_177041	RIKEN cDNA A930017E24 gene	chr3 F1	6.5 (5.6)	Molecular function unknown
1457083_at	Mm.234182	NM_025463	RIKEN cDNA 1810009A15 gene	chr19 A	6.3 (1.0)	Molecular function unknown
1435392_at	Mm.95281	NM_028220	WD repeat domain 17	Chr8:8 3.1	6.2 (1.2)	Molecular function unknown
1421084_at	Mm.41982	NM_011302	Retinoschisis 1 homolog	chrX:70.0 cM	5.5 (1.2)	Cell adhesion
1455098_a_at	Mm.3667	NM_011707	Vitronectin	chr11 45.09 cM	5.3 (0.7)	Cell adhesion
1426412_at	Mm.4636	NM_010894	Neurogenic differentiation 1	chr2 46.0 cM	5.3 (2.0)	DNA binding, eye morphogenesis
1419414_at	Mm.218764	NM_022422	Guanine nucleotide binding protein 13, gamma	chr17 A3.3	5.2 (2.2)	Heterotrimeric G-protein GTPase activity
1420536_at	Mm.1215	NM_007773	Crystallin, beta B2	chr5 60.0 cM	5.2 (7.3)	Structural constituent of eye lens
1451590_at	Mm.192285	NM_145493	Complexin IV	chr18 E1	5.2 (0.9)	Molecular function unknown
1451989_a_at	Mm.132237	NM_153058	Microtubule-associated protein, member 2	chr18 A2	5.2 (3.6)	Protein binding
1420453_at	Mm.6253	NM_009967	Crystallin, gamma S	chr16 16.1 cM	5.2 (5.0)	Structural constituent of eye lens
1437145_s_at	Mm.46431	NM_026415	RIKEN cDNA 2310002J15 gene	chr2 A3	5.1 (2.6)	Molecular function unknown
1419740_at	Mm.1372	NM_008806	cGMP phosphodiesterase 6B, beta polypeptide	chr5 57.0 cM	5.1 (2.2)	Catalytic activity
1422651_at	Mm.3969	NM_009605	Adipocyte complement related protein	chr16	5.1 (7.0)	Fatty acid beta-oxidation
1451763_at	Mm.23793	NM_007723	Cyclic nucleotide gated channel alpha 1	chr5 41.0 cM	5.0 (1.0)	Ion transport
1419392_at	Mm.279903	NM_011995	Piccolo	chr5	4.9 (1.3)	Calcium-dependent phospholipid binding
1421061_at	Mm.16224	NM_008189	Guanylate cyclase activator 1a (retina)	chr17 D	4.8 (1.1)	Calcium ion binding
1436981_a_at	Mm.260643	NM_011740	Tyrosine 3-/tryptophan 5-monooxygenase activation protein	chr14 8.0 cM	4.8 (5.0)	Monooxygenase activity
1420338_at	Mm.4584	NM_009660	Arachidonate 15-lipoxygenase	chr11 40.0 cM	4.8 (3.9)	Carbohydrate metabolism
1458871_at	Mm.130801	NM_172880	Hypothetical protein A030012E10	chr5 E1	4.7 (4.0)	Molecular function unknown
1431393_at	Mm.242892	NM_029444	RIKEN cDNA 4930447C04 gene	chr12 C2	4.7 (2.6)	Molecular function unknown
1460368_at	Mm.101838	NM_145143	Membrane protein, palmitoylated 4	chr1 C2	4.6 (1.2)	Molecular function unknown
1451826_at	Mm.103669	NM_013877	Calcium binding protein 5	chr7 A2	4.6 (2.0)	Calcium ion binding
1456936_at	Mm.70121	NM_144532	Calcium binding protein 4	chr19 A	4.6 (1.2)	Calcium ion binding
1455266_at	Mm.256342	NM_008449	Kinesin family member 5C	chr2 32.5 cM	4.5 (1.0)	Kinesin complex
1434292_at	Mm.294494	NM_175692	RIKEN cDNA A930034L06 gene	chr2 H2	4.5 (1.7)	Molecular function unknown
1425530_a_at	Mm.203928	NM_025606	Mitochondrial ribosomal protein L16	chr19 B	4.4 (2.0)	Mitochondrial large ribosomal subunit
1435330_at	Mm.447	NM_175026	Expressed sequence AI447904	chr1 H3	4.3 (3.0)	Molecular function unknown
1416828_at	Mm.45953	NM_011428	Synaptosomal-associated protein 25	2 78.2 cM	4.3 (1.4)	Neurotransmitter uptake

Table 2A.

 

Representative Genes and ESTs Upregulated in 8-Week-Old-Mouse Sclera

Affymetrix ID	UniGene ID	RefSeq ID	Gene Names	Chromosomal Location	Expression Ratio* 8 wk/3 wk	Molecular Function
1434987_at	Mm.284446	NM_009656	Aldehyde dehydrogenase 2, mitochondrial	5 F-G1	4.2 (7.1)	Aldehyde dehydrogenase (NAD) activity
1448125_at	Mm.209711	NM_173759	RIKEN cDNA A730017C20 gene	18 D3	4.2 (0.9)	Integral to membrane
1437528_x_at	Mm.5163	NM_009065	Ras-like without CAAX 2	18 B1	4.2 (1.3)	GTPase mediated signal transduction

*  Numbers in parentheses are X-fold changes from all six samples, including the one 3-week-old mouse array chip with large variation.

Table 3.

 

Representative Genes and ESTs Downregulated in 8-Week-Old-Mouse Sclera

Affymetrix ID	UniGene ID	RefSeq ID	Gene Names	Chromosomal Location	Expression Ratio* 3 wk/8 wk	Molecular Function
1418464_at	Mm.29428	NM_013592	Matrilin 4	chr2 94.0 cM	33.5 (12.3)	Calcium ion binding
1416741_at	Mm.7281	NM_015734	Procollagen, type V, alpha 1	chr2 18.0 cM	25.6 (13.9)	Extracellular matrix structural constituent
1423578_at	Mm.20230	NM_009926	Procollagen, type XI, alpha 2	chr17 18.51 cM	25.1 (10.5)	Cell adhesion
1423253_at	Mm.9986	NM_008623	Myelin protein zero	chr1 92.4 cM	20.8 (13.6)	Intercellular junction maintenance
1450661_x_at	Mm.5104	NM_008688	Nuclear factor I/C	chr10 43.0 cM	19.8 (6.5)	DNA replication
1422966_a_at	Mm.28683	NM_011638	Transferrin receptor	chr16 21.2 cM	17.2 (9.7)	Proteolysis and peptidolysis
1421290_at	Mm.46181	NM_013868	Heat shock protein family, member 7	chr4 E1	13.9 (7.3)	Response to stress, cytoskeleton
1450828_at	Mm.212349	NM_080451	Synaptopodin 2	chr3 G3	12.9 (8.4)	Actin binding
1425504_at	Mm.247544	NM_139300	Myosin, light polypeptide kinase	chr16 A1	10.5 (5.7)	Cytoskeleton and ATP binding
1452308_a_at	Mm.207432	NM_178405	ATPase, Na+/K+ transporting, alpha 2	chr1 94.2 cM	10.4 (5.4)	Sodium/potassium-exchanging ATPase activity
1459909_at	Mm.4925	NM_008535	EST	chr8 38.5 cM	10.1 (4.2)	DNA binding
1420854_at	Mm.275320	NM_007925	Elastin	chr5 75.0 cM	10.1 (25.8)	Extracellular matrix
1416238_at	Mm.4345	NM_011587	Tyrosine kinase receptor 1	chr4 50.0 cM	10.1 (3.7)	Protein-tyrosine kinase activity
1454791_a_at	Mm.12145	NM_009030	Retinoblastoma binding protein 4	chr4 D2	10.1 (12.7)	DNA replication
1427385_s_at	Mm.253564	NM_011501	Striamin	chr12 C3	9.5 (4.2)	Actin binding
1451285_at	Mm.277680	NM_139149	Fusion, t(12; 16) malignant liposarcoma	chr7 F3	9.4 (7.5)	DNA binding
1423493_a_at	Mm.9394	NM_010906	Nuclear factor I/X	chr8 38.6 cM	9.4 (4.8)	DNA replication
1449206_at	Mm.20942	NM_008596	Mitsugumin 29	chr3 F3-H2	9.4 (5.2)	Synaptic vesicle
1449888_at	Mm.234875	NM_010137	Endothelial PAS domain protein 1	chr17 E4	9.4 (4.2)	DNA binding
1450576_a_at	Mm.262677	NM_013651	Splicing factor 3a, subunit 2	chr10 43.0 cM	9.0 (2.1)	Nuclear mRNA splicing
1420653_at	Mm.9154	NM_011577	Transforming growth factor, beta 1	chr7 6.5 cM	9.0 (3.3)	Transforming growth factor beta receptor binding
1458087_at	Mm.32087	NM_177707	Hypothetical protein 9830125E18	chr10 D3	8.9 (4.0)	Phorbol esters/diacylglycerol binding
1415959_at	Mm.293373	NM_009204	Facilitated glucose transporter, member 4	chr11 40.0 cM	8.8 (4.5)	Glucose transport
1416572_at	Mm.280175	NM_008608	Matrix metalloproteinase 14	chr14 12.5 cM	8.7 (5.0)	Extracellular matrix, proteolysis and peptidolysis
1448228_at	Mm.172	NM_010728	Lysyl oxidase	chr18 29.0 cM	8.4 (5.7)	Lysyl oxidase
1417634_at	Mm.2407	NM_011435	Superoxide dismutase 3	chr5 31.0 cM	8.3 (3.3)	Superoxide metabolism
1448914_a_at	Mm.795	NM_007778	Colony stimulating factor 1 (macrophage)	chr3 51.0 cM	8.0 (2.6)	Cytokine activity
1450747_at	Mm.248266	NM_016679	Kelch-like ECH-associated protein 1	chr9 A3	7.9 (3.7)	Transcription regulation
1427884_at	Mm.249555	NM_009930	Procollagen, type III, alpha 1	chr1 25.0 cM	7.6 (3.3)	Cell adhesion
1429148_at	Mm.263940	NM_026756	RIKEN cDNA 1110019L22 gene	chr10 C1	7.6 (5.0)	Molecular function unknown
1452787_a_at	Mm.27545	NM_019830	Methyltransferase-like 2	chr7 23.1 cM	7.6 (5.0)	RNA processing
1426251_at	Mm.10104	NM_010445	H6 homeo box 1	chr5 18.0 cM	7.4 (5.3)	Metallocarboxypeptidase activity
1448553_at	Mm.290003	NM_010856	Myosin, heavy polypeptide 6, alpha	chr14 20.0 cM	7.4 (3.9)	Strialed muscle contraction
1448378_at	Mm.289707	NM_007984	Fascin homolog 1, actin bundling protein	chr5 86.0 cM	7.4 (2.2)	Actin binding
1431856_a_at	Mm.34776	NM_028331	C 1q& tumor necrosis factor related protein 6	chr15 E1	7.4 (3.8)	Molecular_function unknown
1448469_at	Mm.4691	NM_010917	Nidogen 1	chr13 7.0 cM	7.4 (2.9)	Calcium ion binding: protein binding
1450085_at	Mm.208919	NM_011923	Mus musculus cDNA clone IMAGE:4485306	chr2 B	7.3 (3.0)	Molecular function unknown
1460633_at	Mm.19091	NM_134129	PRP19/PSO4 homolog	chr19 6.0 cM	7.2 (4.0)	Molecular function unknown
1421289_at	Mm.46181	NM_013868	Mus musculus cDNA clone IMAGE:4989198	chr4 E1	7.1 (6.9)	Response to stress
1420941_at	Mm.20954	NM_009063	Regulator of G-protein signaling 5	chr1 86.5 cM	7.1 (4.6)	Regulator of PDGF and EDG receptor signaling

*  Numbers in parentheses are fold changes from all six samples, including the one 3-week-old mouse array chip with large variation.

Figure 2.

 

Hierarchical clustering of gene expression data from 3- and 8-week-old mouse sclera based on expression level. Genes with a similar level of expression were grouped together along the vertical axis and genes with similar pattern of expression were close to each other on the horizontal axis. The biological replicates for each age group were clearly clustered together. A greater degree of similarity was observed within each age group than between different age groups.

Figure 3.

 

Real-time PCR quantification of differentially expressed scleral genes. Sclera from 3- and 8-week-old mice were microdissected. Eight differentially expressed genes by microarray analysis were selected for real-time PCR analysis. PCR reactions were performed in triplicate for each gene. Product quantity was normalized using β-actin as an internal housekeeping gene control. Data are the mean of quantities in 8-week-old mouse sclera calibrated to quantities obtained from 3-week-old mouse sclera. The results are the means ± SE of three replicates.

Figure 4.

 

Downregulation TGF-β1 and collagen genes in adult mouse sclera. The numbers on the y-axis are x-fold changes in microarray analysis.

The authors thank George Hii at the Real-Time PCR Core Facility of the Children’s Hospital of Philadelphia for assistance with the real-time PCR analysis.