| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1From the Laboratory of Molecular and Developmental Biology, National Eye Institute, National Institutes of Health, Bethesda, Maryland.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
METHODS. Serial analysis of gene expression (SAGE) was performed on postnatal day (PN)9 and adult mouse (6 week) total corneas. The expression of selected genes was analyzed by in situ hybridization.
RESULTS. A total of 64,272 PN9 and 62,206 adult tags were sequenced. Mouse corneal transcriptomes are composed of at least 19,544 and 18,509 unique mRNAs, respectively. One third of the unique tags were expressed at both stages, whereas a third was identified exclusively in PN9 or adult corneas. Three hundred thirty-four PN9 and 339 adult tags were enriched more than fivefold over other published nonocular libraries. Abundant transcripts were associated with metabolic functions, redox activities, and barrier integrity. Three members of the Ly-6/uPAR family whose functions are unknown in the cornea constitute more than 1% of the total mRNA. Aquaporin 5, epithelial membrane protein and glutathione-S-transferase (GST) 
-1, and GST 
-4 mRNAs were preferentially expressed in distinct corneal epithelial layers, providing new markers for stratification. More than 200 tags were differentially expressed, of which 25 mediate transcription.
CONCLUSIONS. In addition to providing a detailed profile of expressed genes in the PN9 and mature mouse cornea, the present SAGE data demonstrate dynamic changes in gene expression after eye opening and provide new probes for exploring corneal epithelial cell stratification, development, and function and for exploring the intricate relationship between programmed and environmentally induced gene expression in the cornea.
Formation of the mature cornea depends on an intrinsic developmental program as well as later extrinsic cues from the environment that initiate after eye opening 2 weeks after birth. The anterior surface of an adult cornea is covered with stratified, squamous epithelium that resides on top of a collagen-rich stroma containing a small number of specialized fibroblastic cells known as keratocytes. A one-cell-thick endothelial layer resides at the posterior surface of the cornea, adjacent to the anterior chamber of the eye.2 Like other sites in the body covered with stratified squamous epithelium, such as the epidermis, gut, and esophagus,3 the corneal epithelium undergoes continuous self-renewal, turning over, on average, every 7 to 10 days under normal homeostasis,4 5 and increasing after injury.6 7 8 9 10 A delicate balance exists between proliferating basal cells, early differentiated suprabasal cells, and terminally differentiated superficial squamous cells, ultimately proceeding through a complex differentiation program while migrating centripetally from a peripheral reservoir of stem cells and upward to the anterior surface to be sloughed.11 12 13 14 15 The mature corneal epithelium, therefore, is composed at all times of cells in various stages of differentiation. Approximately 2 weeks after birth, stratification of mouse corneal epithelium commences, with the expansion from 1–2 to 8–10 cell layers by 6 weeks of age.
The onset of epithelial stratification coincides with two other major postnatal developments in the mouse cornea. First, the eyelid opens exposing the eye to new and changing levels of oxygen, light, moisture, and pathogens. Coincident with eye opening and stratification is the upregulation of several abundant soluble corneal proteins including aldehyde dehydrogenase III (ALDH3) and transketolase (TKT), which constitute 44% and 10% of the total soluble protein of the adult mouse cornea, respectively.16 17 ALDH3 mRNA is not detected at birth, but increases robustly by postnatal day (PN)14 (Kays and Piatigorsky, unpublished results, 1996). A time course of ALDH3 protein showed very small amounts of ALDH3 protein, beginning at PN9, with a sharp increase at PN13, coincident with the time of eye opening (Davis and Piatigorsky, unpublished results, 2002). A sizable increase in the levels of mRNA and protein for TKT has also been observed in the mouse cornea at PN14.16 Although environmental and developmental stimuli are thought to contribute to the increase in levels of these proteins,18 19 the identity of the molecules that are important in their regulation is unknown.
Indeed, the gene expression profile of the cornea has not been examined in detail yet. Neither the molecular determinants that subserve the various protective and optical functions of the transparent cornea nor the molecular components that distinguish the cornea before and after eye opening are known. Of particular interest is that both the developmental regulation and the function of the transparent cornea involve a tight interaction between intrinsically programmed and environmentally controlled events. This interaction is especially clear by the profound structural (i.e., epithelial stratification) and biochemical (i.e., major increases in ALDH3 and TKT gene expression) changes that take place precisely at eye opening, the time at which the corneal surface comes in direct contact with the oxidative and other stresses of the environment. Consequently, in the present study, we used the powerful method of serial analysis of gene expression (SAGE)20 to identify the repertoire of genes expressed in the mouse cornea before eye opening at PN9 and in the mature cornea at 6 weeks after birth. SAGE provides several advantages over the more commonly used microarray analysis. It detects without bias all the genes that are expressed, not just those that are represented on the microarray chip, allows quantitative analysis of mRNA contents for each species, and provides probes for further analysis, such as gene isolation, in situ hybridization, and comparison with expressed sequence tags (ESTs) from other sources.
Our results provide many new findings of fundamental importance that contribute to the molecular description of the mouse cornea. These data give the first gene expression signature for the mature mouse cornea and classify the expressed genes into broad provisional functional classes. In addition, we found that the large repertoire of genes expressed before eye opening was surprisingly different quantitatively from that expressed in the mature cornea. Indeed, two thirds of the expressed genes differed at the two stages, with only one third overlapping the two stages. Of the numerous (n > 200) transcripts in the cornea that change expression levels after eye opening, a small cluster (n = 25) of transcription factors were identified that may mediate corneal development, providing specific directions for future experiments. Finally, in situ hybridization provided new molecular markers for distinct layers of the stratified corneal epithelium. One of these, glutathione S-transferase -1, is the first molecular marker to be reported for middle wing cells of the corneal epithelium. In summary, in addition to giving a comprehensive profile of expressed genes in the mature mouse cornea, the present SAGE libraries demonstrate global and unexpectedly dynamic changes in corneal gene expression after eye opening and provide new probes for exploring corneal epithelial cell stratification, development and function, and for exploring the intricate relationship between programmed and environmentally induced gene expression in the cornea.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
SAGE Library Construction, Sequencing, and Analysis
Total RNA was isolated from 24 adult and 98 PN9 whole corneas, using a total RNA isolation kit (SNAP; Invitrogen, Carlsbad, CA) according to the manufacturer’s directions, except that the corneas were first homogenized in lysis buffer in small glass homogenizers. Each library was constructed from 20 µg of total RNA using a SAGE kit (I-SAGE; Invitrogen) according to the manufacturer’s instructions, except that the amplification of the ditags was performed in two stages.21 The first round of PCR included 27 cycles of amplification, and the 100-bp product was purified and used as template for the second round of PCR, which included 9 cycles of amplification.
Bacterial colonies were transferred with sterile toothpicks into 100 µL low-salt antibiotic medium (Zeocin; Invitrogen) and incubated at 37°C for 2 hours. The bacterial culture (5 µL) was amplified with the -47 primer (CEQ2000 Quick Start; Beckman-Coulter, Fullerton, CA) and reverse primer (5'CAGGAAACAGCTATGACC3') for 30 cycles. Amplified PCR products were purified with a PCR cleanup kit (Qiagen, Valencia, CA) and eluted with 10 mM Tris (pH 8.0). Templates were sequenced with the -47 primer using a sequencing kit (CEQ2000 Quick Start; Beckman-Coulter) on a DNA analysis system (CEQ2000; Beckman-Coulter).
Sequencing results were analyzed using sequence-analysis software (SAGE2000 ver. 4.0; Invitrogen). Using this software, tag numbers in the two libraries were compared pair-wise with a test based on a Monte Carlo simulation of tag counts.22 Tag counts were normalized to 62,206 unless otherwise indicated in the table footnotes. The sequences of novel tags (tags that have no database match) are indicated in the "Description" columns.
Tag-to-gene assignments were made using the reliable mouse tag database of July 2, 2002 (Mus musculus UniGene Build 108; http://www.ncbi.nlm.nih.gov/UniGene; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). In the event of multiple gene assignments, several strategies were used to determine the most likely assignment for a tag. The "probable extra base," determined by the sequence-analysis software (SAGE2000; Invitrogen), was used to eliminate some of these. Also, if a tag matched only one deposited sequence in a UniGene cluster, that gene assignment was eliminated. If a single most likely gene assignment could not be determined, the tag was eliminated from further consideration. If a gene was represented by more than one tag, the counts of all tags were added to obtain a count for the gene.
The number of duplicate ditags were 3276 (10.2%) and 1659 (5.3%) in the PN9 and adult libraries, respectively. The GC content of a randomly selected group of tag sequences was 46.7%.
In Situ Hybridization
Fresh, frozen, 10 µm eye sections from 6-week-old mice were fixed, treated with proteinase K (0.2 µg/mL PBS) for 8 to 10 minutes, and processed for in situ hybridization. Riboprobes were synthesized using a digoxygenin (DIG) RNA labeling kit (Sp6/T7; Roche Molecular Biochemicals, Indianapolis, IN) with linearized, proteinase-K–treated, plasmid cDNA templates. Random sequencing of a total corneal C57Bl/6 cDNA library provided us with cDNAs encoding epithelial membrane protein 1 (EMP; X98403; bases 782-1215), aquaporin 5 (AF087654; bases 960-1533), SLURP1 (NM_020519; bases 24-525), calcyclin (X66449; bases 106-526), calizzarin (NM_009112; bases 16-422), glutathione S-transferase (GST) -4 (NM_010357; bases 22-437), GST -1 (U80819; bases 62-548), annexin A2 (BC003327; bases 31-517), and PSCA (AF319173; bases 14-504). Hybridizations were performed as described.23 The reaction was allowed to proceed until purple color was visible, approximately 30 to 60 minutes, at which time reactions for both the sense and antisense riboprobes generated from a given cDNA were terminated.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Table 1 shows a summary of the sequenced SAGE libraries. The 64,272 PN9 and 62,206 adult total tags corresponded to 19,544 PN9 and 18,509 adult unique tags. Unique (different) SAGE tags continued to accumulate at the end of our analysis at the rate of 22% in both libraries. This measurement indicated that we had sampled more than 78% of the unique transcripts present in the cornea at each of the two developmental stages. When tags observed only once were eliminated from the analysis to compensate for possible sequencing errors, 6535 PN9 and 5877 adult unique tags with a frequency of more than 1 (n > 1) remained.
|
Abundant Tags Reflect Corneal In Vivo Functions
The most abundant tags encountered in the corneal SAGE libraries from the two developmental stages, minus those corresponding to ribosomal genes, are shown in Tables 2A (PN9) and 2B (adult). A ranking, the absolute number of times a tag was encountered (count), a corresponding UniGene number, and a descriptive name, where possible, was assigned to the top 100 tags in each library. In addition, each transcript was placed in a functional category that best represented its primary known or presumed function in the cornea. The categories were selected to reflect the normal physiological function of the cornea. Genes with protein products that play a role in redox, barrier, or transcriptional activities were separated into three categories. A fourth category comprised genes with housekeeping, metabolic, and signal transduction functions (HMT). The last category, termed unknown, represents tags that correspond to named genes with unknown function, ESTs, and novel tags.
|
-1 and -2), calizzarin, calcyclin, calpactin, SPARC, destrin, annexins (A1, A2, and A8), epithelial membrane protein1 (EMP1), and aquaporin 5.
|
Cornea Enriched Genes: Defining the Cornea at the Molecular Level
To reveal the genes that distinguish the cornea from other tissues, each of the two corneal libraries was compared with eight publicly available mouse SAGE libraries from limb, brain, heart, and embryonic stem cells. A tag was considered enriched in the cornea if it was encountered five times more frequently in the corneal library under examination than in each of the other eight libraries. The corneal libraries contained 334 PN9 and 339 adult tags fitting this description; 172 of these tags were expressed in both stages of development. A complete tabulation of the cornea enriched genes is available in Supplemental Tables 2A and 2B, accessible online at http://www.iovs.org/cgi/content/full/45/2/429/DC1.
The proportion of enriched genes from the adult library comprising each of the major functional divisions is charted in Figure 1 . More than half of the enriched genes were placed in the unknown function category, whereas the remainder represented genes with known functions, many of which have been shown previously to play important roles in the cornea. For example, the cornea-specific roles of keratin 12, BIGH3, mucin 4,29 keratocan, and lumican30 31 are well documented. Of the enriched genes, 17% mediated housekeeping functions, 18% contributed to barrier integrity, and 4% were involved in redox activities. A small, but significant fraction of enriched genes appear to play a role in regulating gene transcription (described later). A similar functional distribution of corneal-enriched tags was observed at PN9.
|
A diverse collection of expression patterns was observed for the nine genes examined, using antisense riboprobes (Fig. 2) . The control sense PSCA riboprobe did not hybridize to the cornea (Fig. 2I) , nor did any of the other sense riboprobes tested (data not shown). SLURP1 and PSCA (Figs. 2A 2B , respectively), two Ly6/uPAR family members, were expressed throughout the cornea. A hybridization signal of uniform intensity was observed in the cells of epithelium, keratocytes, and endothelium. In contrast, three calcium-binding proteins—calcyclin (Fig. 2C) , calpactin (Fig. 2D) , and annexin 2 (data not shown)—were restricted to the epithelial portion of the cornea where expression was homogeneous in all layers of the epithelium. Aquaporin 5, EMP1, GST -1, and GST -4 were also localized exclusively to the epithelial portion of the cornea. Moreover, their expression patterns were restricted to different layers of the epithelium. Aquaporin 5 mRNA was observed in all anterior cell layers including the suprabasal cells; however, it was not detected in the basal cell layer of the epithelium (Fig. 2E) . EMP1, in contrast, was expressed at detectable levels throughout the epithelium, but highly enriched in the most superficial cells (Fig. 2F) . The most distinctive patterns of expression were for the two GSTs: GST -1 (Fig. 2G) was expressed only in the most superficial, epithelial cell layer, while GST -4 (Fig. 2H) mRNA was found in the center, but not the basal or the superficial layers of the corneal epithelium.
|
|
0.01). A total of 109 transcripts were downregulated and 96 transcripts were upregulated in the adult relative to the PN9 cornea. These transcripts are shown in Supplemental Tables 3A and 3B, which are available online at http://www.iovs.org/cgi/content/full/45/2/429/DC1. Differences in expression levels ranged from 4- to 143-fold (change multiples for each tag are listed in the supplemental tables) with the most dramatic changes occurring in the downregulated transcripts. ESTs accounted for 34% and 35% of the up- and downregulated transcripts, respectively. No sequence match was found for 6 (upregulated) and 15 (downregulated) tags, indicating that a significant number of putative novel genes show temporal changes in their expression levels. Any biological significance of these differences (or an apparent lack of a difference) in gene expression must be qualified by considering that the cornea is made up of different cell types whose proportions change over the period examined in this report. More information is needed to determine whether these differences reflect changes in expression levels per cell type. The differentially expressed transcripts were assigned to one of the functional categories described earlier (Fig. 1) . Although the functions of the majority of the differentially expressed transcripts were unknown, the assignment of the remaining transcripts revealed some general trends in cellular processes over this period. Most of the genes that were expressed at greater levels at PN9, relative to adult stages, were in the barrier integrity category. It appears that the extracellular stroma is synthesized early, in that six different procollagens subunits along with lysyl oxidase, glypican3, Ednra, and lumican are downregulated in the adult cornea. In contrast, a greater number of genes devoted to housekeeping/signal transduction/general metabolism were expressed at greater levels in the adult versus the PN9 cornea. The second most abundant class of differentially expressed genes was that involved in regulating gene transcription.
Identification of Transcripts that Regulate Gene Transcription
Two hundred and thirty-four different genes (n > 1) involved in transcriptional regulation were detected in the combined libraries. A complete list, including the UniGene number, gene name, and abundance of each of these transcripts can be found in Supplemental Table 4, which is available online at http://www.iovs.org/cgi/content/full/45/2/429/DC1. Genes that mediate both site-specific and global transcriptional regulation were identified, as were proteins that modulate transcription independent of DNA-binding such as coactivators and repressors and binding proteins. Also included in this category were general transcription components, such as RNA polymerases and elongation factors.
A large number of the transcripts contained an Ets-, helix-loop-helix (HLH)-, zinc finger-, homeobox-, or basic leucine zipper (bZIP)-domain (Table 3) . Also, members of the interferon regulatory factor (IRF), SOX, and nuclear receptor families were well represented (Table 3) . The six most abundant transcriptional regulators at PN9 included myocyte enhancer factor 2 (MEF2; n = 38), kruppel-like factor 4 (KLF4; n = 33), Pax 6 (n = 28), retinoblastoma binding protein 4 (n = 25), cold shock domain protein A (n = 25), and HMG17 (n = 25). The six most abundant transcriptional regulators at the adult stage included IRF1 (n = 51), KLF4 (n = 48), serum response element binding factor-2 (SREBF-2; n = 25), Ets-like factor 3 (ELF3; n = 22), MEF 2 (n = 22), and suppressor of variegation 3-9 homolog 1 (n = 22).
|
|
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Representatives of the major functional classes of the cornea are among the 100 most abundant transcripts. More than half of these transcripts fall into the HMT category with more than 40 encoding distinct ribosomal components, reflecting a major requirement for protein synthesis. Abundant genes that may play prominent roles in cell signaling in the cornea include calmodulin-like protein,35 Rack-1,36 and stratifin.37 38 Components of other signaling pathways including Notch,39 40 Wnt,41 TGF,42 FGF,43 Src kinase,44 and EGF/ERB-B245 46 were present at lower amounts. The importance of some of these pathways in eye development has been described,47 48 49 50 51 52 53 54 but their precise role in the postnatal cornea awaits further investigation.
A significant number of the transcripts encode structural proteins that presumably impart mechanical strength and barrier properties to the cornea. Corneas lacking K12, one of the five different keratins to appear in the top 100 most abundant transcripts, are fragile, underscoring the importance of this corneal-specific keratin in conferring mechanical strength to the cornea.55 A cytoskeletal role for microtubules is attested to by the abundance of transcripts encoding actin, myosin, tubulin, and the actin-binding proteins thymosin beta 456 and destrin.57 Likewise, the abundant procollagens are pivotal to the formation of the tough, extracellular-rich stroma.58 59 The importance of calcium in cellular signaling60 and in epidermal stratification61 62 63 is well documented, and the presence of several abundant calcium-binding proteins including calizzarin, calcyclin, and calpactin and the possible role of MEF2 in mediating calcium-dependent signaling pathways64 portends a significant role for calcium in the cornea. Although EMP1 has been implicated in regulating cell growth and differentiation,65 66 67 the intense localization of EMP1 to the most superficial layer of the corneal epithelium suggests that it may play a role in barrier formation as has been shown for another family member.68 Proper water balance in the cornea maintains transparency and is mediated by water channels located principally in the endothelium.69 Although aquaporins-1, -2, and -3 were identified in the SAGE library, the role of the abundant aquaporin 5 is unclear, as it was shown to localize exclusively in the upper layers of the corneal epithelium.
More than one tenth of the 100 most abundant corneal transcripts are involved in defensive mechanisms that reduce the detrimental effects of environmental insults70 known to be factors in corneal disease.71 Most of these transcripts, including ALDH3, TKT, GST -1 and GST -4, ferritin heavy chain, and hexokinase 1, were increased after eye opening suggesting that exposure to environmental stimuli induced this category of transcripts. ALDH3 and TKT mRNAs represent approximately 1% each of the total mRNA in the adult cornea by SAGE. Corroborating the numbers obtained by SAGE, 2.9% and 0.5% of 1103 randomly selected clones from a traditional cDNA library, constructed from adult mouse total cornea, corresponded to ALDH3 and TKT, respectively (Norman B, Davis J, Piatigorsky J, unpublished data, 2001). K12 was the most abundant sequence in the cDNA screen, representing 3% of the total clones. However, ALDH3 and TKT constitute 20% to 40% of the total water-soluble protein,16 17 suggesting that their expression is under posttranscriptional regulation and/or that protein stabilization is a major regulatory mechanism controlling the protein profile in the cornea. A yeast analysis showed no correlation between levels of mRNA and levels of the corresponding protein,72 supporting the idea that quantitative mRNA data, as generated by SAGE, may not reflect corresponding protein levels.
The nonoverlapping pattern of expression of the two abundant GSTs is, to the best of our knowledge, unprecedented in the cornea. Expression in two separate layers of the anterior corneal epithelium suggests an underlying biochemical difference between these layers of epithelial cells. The promoters of these two genes may be useful in revealing a differentiation program in the cornea similar to that observed in skin.73 Although these two GSTs are members of the phase II detoxification enzyme family important in xenobiotic inactivation,74 75 76 they may serve different functions in the cornea—both bind glutathione, but only GST -4 is enzymatically active.77
We were unable to assign a corneal function to three abundant transcripts belonging to the Ly6/uPAR family.78 79 80 81 PSCA, a membrane protein, is best known as an indicator of prostate cancer.82 No function has been described for the abundant hypothetical CD59 antigen containing protein; however, its namesake CD5983 is an important inhibitor of the complement cascade in the eye.84 SLURP1 is a secreted Ly6/uPAR family member resembling snake venom neurotoxins.85 Mutations in SLURP1 and a newly discovered relative, SLURP2, which is present in lower amounts in the corneal libraries, are associated with the skin diseases Mal de Meleda86 and psoriasis vulgaris,87 respectively. The overlap between genes expressed in the cornea and in the skin, such as SLURP1 and -2, calmodulin-like protein, and stratifin, is noteworthy.88 89 90 Several transcription factors were identified in the corneal SAGE libraries that have known roles in epidermal development, including mOVO,91 CUX,92 TSC-22,93 grainyhead,94 95 and p63.96 97
More than 200 genes that regulate transcription were identified by SAGE and one third of them are enriched in the cornea. The abundant Ets, KLF, and IRF families of transcription factors stand out. Ets factors are important in the transcriptional regulation of genes involved in diverse biological functions.98 99 Among the six ets family members found in this study, four are enriched in the cornea. ELK3 and Ets domain protein are expressed exclusively at PN9, whereas two others, ELF3 and ets homologous factor, are two of the most abundant transcription factors in the cornea at both developmental stages. ELF3, an epithelium-specific ets transcription factor,100 regulates expression of genes associated with differentiation101 and may play a similar role in corneal epithelial differentiation.102 ELF3 is also essential for normal gut and uterine epithelial cell maturation.103 Other transcription factors pivotal to intestinal epithelial cell development104 are also expressed in the cornea, including Nkx2.3, Tcf-4, Pax6, and KLF4,105 perhaps reflecting a common course of self-renewal throughout life.
The abundant KLF4 is one of the six members of the kruppel-like factor (KLF) zinc-finger–containing family members expressed in the mouse cornea as well as in the epithelia of skin, lung, and the gastrointestinal tract.106 107 Goblet cells in the colon fail to differentiate in KLF4 null mice,108 and the mice die shortly after birth because of a perturbation in the cornified envelope, a late-stage structure in epidermal differentiation necessary to establish the barrier function of the skin.109 Other studies suggest that KLF4 may also have an important function in regulating growth-specific genes in epithelia.110 111 The role of KLF4 in the cornea has not been investigated.
The importance of an immune surveillance system in the cornea is suggested by the wide variety and abundance of IRFs,112 113 expressed in the SAGE libraries. IRF1,114 the most abundant IRF in the SAGE libraries, is an important regulator of IFN-mediated antibacterial response.115 IRFs may also regulate the cell cycle, cell transformation, and apoptosis.116
The ability of the cornea to respond to continuously changing environmental conditions is suggested by the expression of redox-, hypoxia-, and light-sensitive transcriptional regulators whose binding sites reside in several corneal gene promoters.19 117 118 119 Numerous transcriptional components of two major universal response pathways mediating oxidative stress,120 121 NF-
B (tags for p50, p65, nuclear factor of activated T-cells 5, I-B, Bcl, and IKKß) and AP-1 (tags for JunB; Mafs K, B, F, and G; JDP2; Fra2; and Nrf1), were identified in the SAGE corneal libraries. AP-1 and NF-B play other major roles in inflammation,122 123 cell differentiation and proliferation, and cell death.124 The importance of these factors in eye development is illustrated by the phenotypes that develop when their expression is altered.125 126
The PAS-domain–containing genes, hypoxia-inducible factor1a (HIF1a) and period 1 and 2 (Per1 and 2), considered to be environmental sensors of oxygen and light,127 128 129 respectively, were detected in the corneal libraries. Of particular interest is the observation that ALDH3 expression is downregulated under hypoxic conditions.19 118
Finally, we have identified 25 transcriptional regulators exhibiting differential expression during the developmental period examined in this study. Further studies are needed to examine their role(s) in epithelial stratification and in the regulation of the abundant corneal proteins, ALDH3 and TKT.
In conclusion, the present SAGE libraries provide a foundation for the complexity and a catalog of genes expressed in the mouse cornea, including many genes implicated in corneal disease. The analysis shows a surprising dynamism of gene expression patterns in the postnatal cornea, while delineating a genomic signature of the mouse cornea. Finally, and most important, it has opened a plethora of avenues to pursue for understanding the development, maintenance, and disease of the mouse cornea.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the National Eye Institute.
Submitted for publication May 8, 2003; revised July 7 and August 19, 2003; accepted September 15, 2003.
Disclosure: B. Norman, None; J. Davis, None; J. Piatigorsky, 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.
Corresponding author: Joram Piatigorsky, National Eye Institute, National Institutes of Health, 7 Memorial Drive, Building 7, Room 100A, Bethesda, MD 20892; joramp{at}nei.nih.gov.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
