|
|
||||||||
1 From the Departments of Ophthalmology and 2 Biochemistry, Jichi Medical School, Tochigi, Japan; the 3 Institute for Molecular and Cellular Biology, Osaka University, Osaka, Japan; and 4 Schepens Eye Research Institute, Harvard Medical School, Boston, Massachusetts.
| Abstract |
|---|
|
|
|---|
METHODS. Sixteen human donor corneas that had no history of corneal disease, infection, or intraocular surgery were used within 7 days of death. Total RNA was extracted from corneal endothelial cells with attached Descemet membranes. A 3'-directed cDNA library was constructed from mRNA by using a pUC19-based primer. These sequences were compared with each other to determine their frequency and were searched against GenBank for identification. To identify novel specific and abundant transcript genes in corneal endothelial cells, the novel genes were compared with an expressed sequence tag database, the expected sequence extended, and 5' rapid amplification of cDNA endspolymerase chain reaction cloning performed.
RESULTS. The human corneal endothelial cDNA library showed that the most abundant transcript was prostaglandin D2 synthase. The remaining transcript genes that were present in abundance consisted of lactate dehydrogenase-A, gene signature (GS) 3582, which is a novel gene without a known function, and matrix Gla protein. The full-length sequence of GS3582 showed similarity to genes obtained in ovary and TESTIS.
CONCLUSIONS. A human corneal endothelial cDNA library was constructed. An expression profile of corneal endothelium provides probes to monitor physiologic and pathologic conditions of this tissue in terms of gene expression.
| Introduction |
|---|
|
|
|---|
In the human eye, gene expression profiles have been investigated for the corneal epithelium,3 conjunctival epithelium,4 trabecular meshwork,5 and retina.6 Much of the corneal stroma, the CE, and iridocorneal angle (trabecular meshwork) are derived from neural crest, whereas the corneal and conjunctival epithelium and sensory retina are derived from surface ectoderm and neural ectoderm, respectively, suggesting that the genes expressed in the CE may be distinct from those in the epithelium of the cornea or conjunctiva. A cDNA library from cultured rabbit CE7 has been characterized. However, human CE (HCE) does not regenerate in vivo, reflecting an essential difference from the regenerative capacity of rabbit CE.8 Therefore, gene expression profiles of noncultured HCE could provide important details on the in vivo function of HCE.
Unique gene expression analyses based on 3'-directed cDNA contain a short sequence of mRNA polyA+ tails through the nearest MboI site. The information is called a gene signature (GS) and consistently represents the mRNA population in a given tissue.9 10 The MboI site is used, because GATC is the most uniformly distributed recognized sequence for commercially available four-base cutters in human gene sequences. By comparing the expression profile for a particular tissue with profiles from other cells, genes can be categorized into those that are active in many other source tissues.
In this study, we describe the gene expression profile of the HCE by using the GS system and we identify the full-length sequence of a novel gene. These results may provide important candidate genes for the study and understanding of the physiological and metabolic functions of the HCE.
| Materials and Methods |
|---|
|
|
|---|
Construction of a 3'-Directed cDNA Library
Total RNA was extracted from the homogenized solution. PolyA+ RNA was prepared by using an mRNA purification kit (QuickPrep; Amersham Pharmacia Biotech, Tokyo, Japan) according to the manufacturers instructions. A 3'-directed cDNA library was constructed using polyA+ RNA, as described previously.11
Briefly, pUC19-based vector primer was used for cDNA synthesis, and the 5' side of double-stranded was trimmed off with MboI (GATC) to decrease the transforming efficiency bias caused by size difference.12
Cohesive termini were created at the other end of the vector molecules by cleaving with BamH1 and were self-ligated. By transforming them into Escherichia coli DH5 (Toyobo, Osaka, Japan) and using one tenth of the mixture, we obtained approximately 2.5 x 105 independent transformants. With this shortening of the cDNAs using MboI, we could diminish the transforming efficiency bias caused by size differences.10
Sequence and Data Analysis
The transformed colonies were selected randomly and lysed. The cDNA moiety was then amplified by polymerase chain reaction (PCR), using a primer pair that flanked the insert (5'-GTTTTCCCAGTCACGACGTTG-3'; 5'-ACCATGATTACGCCAGCTTG-3'), as described elsewhere.12
The PCR products were subjected to a cycle sequencing reaction, using dye primer M13, and analyzed by autosequencer (model 373A; PE Biosystems, Foster City, CA). Inserted sequences shorter than 20 bp were eliminated from analysis because of inadequate length to discriminate among approximately 105 transcripts at their 3' termini. Inserts that had no unique sequences or had more than 5% ambiguous nucleotides were also eliminated. Sequences were compared to each other using the FastA program,13
and identical sequences were grouped together as a single gene species. Sequences with more than 95% identical bases throughout the query sequences were regarded as identical. For recurring sequences that appeared more than twice, homologies were searched in a daily updated expressed sequence tag database (dbEST) using the Basic Local Alignment Tool (BLAST) of the National Center for Biotechnology Information (NCBI, Bethesda, MD; available in the public domain at http://www.ncbi.nlm.nih.gov/BLAST).
Full-Length Sequencing
The novel and abundant transcript gene, GS3582, in our established HCE cDNA library, corresponded to the daily updated dbEST and extended the expected sequence. 5'-Rapid amplification of cDNA ends (RACE)-PCR cloning was performed using a kit (Sure-RACE; Origene Technologies, Inc., Rockville, MD) that allows isolation of the 5' sequence of the target transcript. The RACE panels in the kit consist of double-stranded cDNA from 24 individual human tissues arrayed in a multiwell plate and provides two contiguous adapter-specific primers at the 5' end. Outer primer for first-round PCR, gene-specific primer (GSP)-1: 5'-TTTCCGCAACATTCTCCTTTT-3') and inner primer for second-round (nested) PCR, GSP2: 5'-TTCTGTGTTTGGCTTGGT-3') were designed from the expected sequence.
The RACE cDNAs were amplified with DNA polymerase (KOD-Plus; Toyobo). After incubation at 94°C for 3 minutes, the first round of PCR was performed with 5 cycles of 94°C for 3 seconds, 63°C for 30 seconds, and 72°C for 3 minutes and then 15 cycles of 94°C for 30 seconds, 60°C for 30 seconds, and 72°C for 6 minutes, followed by 72°C for 6 minutes in a PCR thermal cycler (model MP; Takara, Kyoto, Japan). After incubation at 94°C for 3 minutes, the second round of PCR was performed with 35 cycles of 94°C for 30 seconds, 62°C for 30 seconds, and 72°C for 3 minutes, followed by 72°C for 6 minutes. Among human cDNAs in different tissues, PCR products 300 to 400 bp in length were circularized and subcloned into a vector (pBluescript; Stratagene, La Jolla, CA). Randomly selected plasmids were amplified and purified into DNA with a kit (Plasmid Midi; Qiagen, Hilden, Germany). The products were amplified with T7 promoter primer (T7 HT Primer; Toyobo) with 25 cycles of 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 4 minutes after incubation at 96°C for 30 seconds. Amplified products were sequenced with the autosequencer (ABI Prism 310 Genetic Analyzer; PE Biosystems). The resultant sequence was extended to the expected sequence, and the analysis of the total GS3582 sequence was concluded.
| Results |
|---|
|
|
|---|
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
The CE is metabolically active and therefore requires nutrients for its function. Glycogen and glucose from the aqueous humor are the main energy sources for endothelial cells. Glucose is catabolized through aerobic pathways, including the tricarboxylic acid cycle and hexose monophosphate shunt, and anaerobic pathways. However, whether the CE has the capacity to store glycogen is not known.1 The active genes such as lactate dehydrogenase A (n = 28), NADH dehydrogenase (n = 10), and adenosine triphosphatase (ATPase, n = 8) express abundantly, suggest dynamic glucose metabolism in the corneal endothelium. The control of corneal stromal hydration sustained by pump and barrier function in the corneal endothelium is essential for transparency. The abundant adenosine triphosphatase (ATPase) and carbonic anhydrase gene (n = 15) may explain the pump function in relation to Na-K, ATPase.
Physiologic extracellular matrix calcification is restricted to bones, teeth, and the hypertrophic zone of growth plate cartilage. Matrix Gla (n = 17) is the protein associated with protection against calcification in soft tissues.16 Nonsense mutations in the human matrix Gla protein gene cause Keutel syndrome, a rare autosomal recessive disorder characterized by diffuse cartilage calcification.17 In mice, deletion of matrix Gla protein causes cartilage calcification and arterial calcification.16 In the CE, calcium ions in the aqueous humor are necessary for the maintenance of the barrier function of the CE. The existence of matrix Gla protein gene in the HCE suggests that the protein may metabolize calcium ions that maintain barrier function and control stromal hydration while protecting calcium deposit in the cornea. SPARC/osteonectin (n = 6) in secretory and extracellular proteins also could be associated with the barrier function of the CE.18
PGD is a neuromodulator that participates in the sleepwake cycle, thermoregulation, and odor responses.19 There are two enzymes that catalyze the conversion of PGH2 into PGD2: hematopoietic PGD synthase and lipocalin-type PGD synthase.20 21 Not only brain, but also the epididymis and testis express a high amount of lipocalin-type PGD synthase protein.22 23 24 In the eye, lipocalin-type PGD synthase is synthesized within the epithelial cells of the iris-ciliary body and pigment epithelium and is secreted into the aqueous and vitreous humors, respectively.25 26 27 The gene transcript of PGD2 synthase (n = 68) is the most abundant transcript gene in the present study, suggesting that CE covering the anterior chamber is also the major source of PGD2 synthase secretion and that PGD2 may be involved in the regulation of intraocular pressure.28
It is of interest that ß-amyloid protein precursor gene (APP, n = 10) is detected in the CE. Disruption of the normal function of the APP is known as a leading cause of Alzheimer disease.29 After a wound to the cornea, APP is upregulated in the basal cells of the corneal epithelium that is actively migrating,30 31 but not in the normal human corneal epithelium,3 implying a role for mediating epithelial migration during re-epithelialization. Because in vivo the HCE does not show any regenerative ability, it heals only by migration, suggesting a critical role for APP in CE migration. Further study is needed to reveal the exact mechanism that mediates migration in the HCE.
GSs found only in corneal endothelium included damage-specific DNA-binding protein 2 (DDB2; n = 5). Ultraviolet (UV) radiation-damaged DNA-binding activity requires the expression of two subunits, p127, DDB1 (n = 8), and p48, DDB2,32 which are detected in the HCE library, but not in corneal epithelium.3 Clinically, xeroderma pigmentosum group E cells, induced by a nonsense mutation of DDB2,33 are deficient in global genomic repair.34 Moreover, decreased cell density and increased coefficient of variation in cell area have been reported in patients with xeroderma pigmentosum.35 Nonregenerative HCE, different from regenerative corneal epithelium, may be necessary to prepare an efficient genomic system for repair of DNA damaged by UV radiation, with the aid of DDB2 protein.
As for the previously established cDNA library of anterior segment tissues of the eye, our library of noncultured HCE is not entirely compatible with that of cultured rabbit CE.7 This may be due to not only species differences, but also to the samples originating from cultured versus noncultured cells. There are few common gene expressions among the cDNA libraries derived from HCE and human corneal epithelium.3 Perhaps this is not surprising, given the differences in embryologic origins between these two cell types, the former derived from neural crest and the latter from surface ectoderm. In contrast, abundant genes, including lactate dehydrogenase and matrix Gla protein, in the CE coincide well with those in the human trabecular meshwork5 cDNA library, which also has a neural crest origin. Moreover, cDNA libraries of human iris-ciliary body and CE have similar gene expression profiles in our BodyMap (data not shown). These results suggest that ocular tissues originating from the neural crest surrounding the anterior chamber can provide common proteins for the maintenance of aqueous humor physiology.
We chose GS3582 (n = 21), a novel gene with a high rate of expression in the HCE, and isolated a full-length cDNA. As shown in Figure 2 , GS3582 does not possess a signal peptide at N termini, indicating that GS3582 is not a secretory protein and could be classified in the intracellular organelle or cytoplasm. By using the Web server at NCBI, significant homology (E value of 3e-98) was observed between GS3582 and the ovary-specific acidic protein (GenBank accession number, AF329088).14 The ovary-specific acidic protein was so named because of the restricted expression to the ovary by Northern blot analysis of the other systemic tissues not including human corneas. Moreover, a similar full-length sequence was also obtained in the mouse testis (accession number: AK006339 in NCBI). These results suggest apparent polymorphisms for GS3582. GS3582 is found in the human corneal epithelium, but not retina or iris in our BodyMap, and in the human cDNA of pituitary, prostate, testis, and ovary. Further studies should be conducted to reveal possible biological functions shared among these tissues and cornea. It is of interest that the ovary, testis, and anterior chamber all represent immune-privileged tissues. As such, the sharing of specific gene products among these tissues may provide some insight into similarities they share in terms of immunity.36
In summary, we have constructed a cDNA library of the HCE based on analysis of 2663 clones and have reported a full-length sequence of a novel gene showing the similarity to genes obtained in the ovary and the testis. Our cDNA library of HCE could provide new insights into the physiology of HCE.
|
| Footnotes |
|---|
Supported by grants from Scientific Research from the Ministry of Education, Japan, (RS, SY), Comprehensive Research on Aging and Health, Ministry of Health, Labour and Welfare, Japan (TK), and National Eye Institute Grant EY-12963 (MRD).
Submitted for publication August 23, 2001; revised January 10, 2002; accepted January 25, 2002.
Commercial relationships policy: N.
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: Satoru Yamagami, Department of Ophthalmology, Jichi Medical School, Yakushiji 3311-1, Minamikawachi-machi, Kawachi-gun, Tochigi 329-0498, Japan; yamagami{at}jichi.ac.jp.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
N. Ebihara, S. Yamagami, L. Chen, T. Tokura, M. Iwatsu, H. Ushio, and A. Murakami Expression and Function of Toll-like Receptor-3 and -9 in Human Corneal Myofibroblasts Invest. Ophthalmol. Vis. Sci., July 1, 2007; 48(7): 3069 - 3076. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Ebihara, S. Yamagami, S. Yokoo, S. Amano, and A. Murakami Involvement of C-C Chemokine Ligand 2-CCR2 Interaction in Monocyte-Lineage Cell Recruitment of Normal Human Corneal Stroma J. Immunol., March 1, 2007; 178(5): 3288 - 3292. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Kinouchi, T. Kinouchi, T. Hamamoto, T. Saito, A. Tavares, T. Tsuru, and S. Yamagami Distribution of CESP-1 Protein in the Corneal Endothelium and Other Tissues. Invest. Ophthalmol. Vis. Sci., April 1, 2006; 47(4): 1397 - 1403. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yamagami, N. Ebihara, T. Usui, S. Yokoo, and S. Amano Bone Marrow-Derived Cells in Normal Human Corneal Stroma Arch Ophthalmol, January 1, 2006; 124(1): 62 - 69. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Mimura, S. Yamagami, S. Yokoo, Y. Yanagi, T. Usui, K. Ono, M. Araie, and S. Amano Sphere Therapy for Corneal Endothelium Deficiency in a Rabbit Model Invest. Ophthalmol. Vis. Sci., September 1, 2005; 46(9): 3128 - 3135. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Uchida, S. Yokoo, Y. Yanagi, T. Usui, C. Yokota, T. Mimura, M. Araie, S. Yamagami, and S. Amano Sphere Formation and Expression of Neural Proteins by Human Corneal Stromal Cells In Vitro Invest. Ophthalmol. Vis. Sci., May 1, 2005; 46(5): 1620 - 1625. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yokoo, S. Yamagami, Y. Yanagi, S. Uchida, T. Mimura, T. Usui, and S. Amano Human Corneal Endothelial Cell Precursors Isolated by Sphere-Forming Assay Invest. Ophthalmol. Vis. Sci., May 1, 2005; 46(5): 1626 - 1631. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. S. Rabinowitz, L. Dong, and G. Wistow Gene Expression Profile Studies of Human Keratoconus Cornea for NEIBank: A Novel Cornea-Expressed Gene and the Absence of Transcripts for Aquaporin 5 Invest. Ophthalmol. Vis. Sci., April 1, 2005; 46(4): 1239 - 1246. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Gonzalez, M. Caballero, P. B. Liton, W. D. Stamer, and D. L. Epstein Expression Analysis of the Matrix GLA Protein and VE-Cadherin Gene Promoters in the Outflow Pathway Invest. Ophthalmol. Vis. Sci., May 1, 2004; 45(5): 1389 - 1395. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yamagami, S. Yokoo, T. Mimura, and S. Amano Effects of TGF-{beta}2 on Immune Response-Related Gene Expression Profiles in the Human Corneal Endothelium Invest. Ophthalmol. Vis. Sci., February 1, 2004; 45(2): 515 - 521. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. I. Tomarev, G. Wistow, V. Raymond, S. Dubois, and I. Malyukova Gene Expression Profile of the Human Trabecular Meshwork: NEIBank Sequence Tag Analysis Invest. Ophthalmol. Vis. Sci., June 1, 2003; 44(6): 2588 - 2596. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Yamagami, S. Yamagami, T. Inoki, S. Amano, and K. Miyata The Effects of Proinflammatory Cytokines on Cytokine-Chemokine Gene Expression Profiles in the Human Corneal Endothelium Invest. Ophthalmol. Vis. Sci., February 1, 2003; 44(2): 514 - 520. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |