Characterization of Gene Expression in Human Trabecular Meshwork Using Single-Pass Sequencing of 1060 Clones

Abstract

purpose. To study the gene expression profile of the human trabecular meshwork (HTM).

methods. A polymerase chain reaction (PCR)–amplified cDNA library was constructed using RNA from the TM of a 67-year-old normal, perfused human eye. A total of 1060 clones were randomly selected for sequencing of one end. These sequences were searched against nonredundant GenBank and dbEST databases for similarity comparison by using a FASTA file and the BLASTcl3 program. Relative expression patterns of those clones that matched other expressed sequence tags (ESTs) were determined using the National Center for Biotechnology Information (NCBI) Unique Human Gene Sequence Collection (UniGene) database.

results. Of the 1060 clones analyzed, 519 (48.9%) had sequences identical with known genes, 125 (11.8%) matched ESTs, and 189 (17.8%) did not match any database sequences. Of the remaining clones, 31 (3%) corresponded to mitochondrial transcripts and 196 (18.5%) to repetitive and noninformative sequences. It is notable that some of the genes highly represented in this library are not ubiquitously expressed in other tissues, which suggests a potentially important role in the HTM. As evidence for the presence of true novel genes in the library, one of the clones was fully sequenced. This clone comprised a complete open reading frame of 966 nucleotides, and its deduced amino acid sequence corresponded to a protein 33% similar to the MAS-related G-protein–coupled receptor.

conclusions. The identification of the more highly expressed genes in HTM and the discovery of novel genes expressed in this tissue provides basic information for further research on the physiology of the TM and for the identification of glaucoma candidate genes.

The trabecular meshwork (TM) is known to play a critical role in the maintenance of the intraocular pressure (IOP) of the eye. In humans, approximately 85% to 90% of the aqueous humor generated by the ciliary body exits the eye through Schlemm’s canal (SC) and the collector channels after being filtered by the TM. The distal TM generates most of the normal resistance to aqueous humor outflow, and it is also the site of the abnormal increased resistance that leads to elevated IOP and glaucoma. ^{1 2 3}  

Despite its small size, the TM has a complex architecture, including several layers and cell types that are morphologically and functionally different. The more anterior portion, the nonfiltering meshwork, consists of three to five elongated beams covered by cells. The posterior portion, or filtering meshwork, consists of three layers: the uveal meshwork, formed by irregular strands; the corneoscleral meshwork, formed by 8 to 15 layers of interconnected flat beams; and the outermost layer, the juxtacanicular region, formed by two to five layers of loosely arranged cells just underneath SC. 

Human TM (HTM) samples include three main cell types: (1) trabecular cells, which cover the beams and strands of the nonfiltering, uveal, and corneoscleral meshwork regions, originate from the neural crest, and may be predominately involved in phagocytosis and tissue repair; (2) juxtacanalicular cells, which are randomly distributed within the extracellular matrix of the juxtacanalicular meshwork, are not attached to basal membranes, and exhibit long, slender processes that can reach the SC cells; they are believed to originate from the perivascular cells of Rouget; (3) and SC inner wall endothelial cells, which constitute the last barrier for the aqueous humor before it reaches the lumen of SC. These cells appear to be of vascular origin and have been implicated in modulating the resistance to aqueous humor outflow. ^{1 4 5 6 7} In addition to these three main cell types, electron microscopy of HTM sections often reveals the presence of other cells, such as macrophages. ⁸ These cells are found within the intertrabecular spaces, are usually loaded with pigment granules, and probably leave the meshwork through paracellular routes of the inner wall. ⁹  

The TM is implicated in functions such as phagocytosis, extracellular matrix dynamics, secretion, cytoskeletal reorganization, and detoxification of aqueous humor. ^{10 11} It is also believed to be involved in the maintenance of the immune privilege in the eye. ^{12 13 14} However, most of the specific mechanisms by which the TM accomplishes these functions are not well understood. 

The biologic features of TM, like those of any tissue, are determined largely at the level of gene expression. The identification of the genes expressed in TM constitutes a necessary step toward the understanding of its physiology. A powerful approach to the analysis of expression in a particular tissue or cell type involves single-pass partial sequencing of clones from a cDNA library together with the comparative analysis of the resultant expressed sequence tags (ESTs) with entries in public databases. A number of sequencing projects have provided a large collection of human ESTs from a variety of tissues. ^{15 16 17 18} However, the gene expression profile in the HTM has not been studied—in part, because of difficulties in obtaining enough RNA from such a small tissue that in turn preclude the construction of cDNA libraries by conventional methods. Polymerase chain reaction (PCR)–based methods for construction of cDNA libraries have been developed to circumvent these problems. ^{19 20 21}  

However, libraries generated using standard PCR methods tend to include a number of artifacts because of the overcycling of the PCR reaction. In addition, these libraries result in an enrichment of small cDNA clones that do not possess the entire open reading frame (ORF) of the genes. Instead, a recently developed, exponential PCR-based method generates high-quality cDNA libraries, enriched for full-length clones. The new method has been used to make good cDNA libraries from mouse blastocyst ²² and human CD34⁺ hematopoietic stem–progenitor cells. ²³ It also has been successfully used in our laboratory to generate HTM cDNA probes for the hybridization of gene arrays. ^{24 25} Using this new approach, we constructed a cDNA library from the TM of a 67-year-old human donor with no history of glaucoma, and analyzed 1060 clones by single-pass sequencing. 

For this first analysis of the HTM gene expression profile, we chose a TM sample that had been perfused under organ culture conditions. Perfusion of the anterior segment of the eye is one of the best and most commonly used models for studying outflow physiology. ^{26 27 28} Although perfused HTMs may not exactly reproduce the pattern of gene expression of the in vivo tissue, they are one of the closest. The model helps recover TM cell activity and maintains the expression of genes responsive to flow and IOP. In comparison, gene expression of a nonperfused HTM sample from a postmortem donor diverges from the in vivo expression pattern because of ongoing cell death and absence of IOP and aqueous humor flow. 

Our results provide the first profile of gene expression in a human TM. This information will undoubtedly help in understanding the normal physiology of the TM and will provide new candidate genes potentially involved in the pathophysiology and the susceptibility to glaucoma. 

Materials and Methods

Perfused Human Anterior Segment Organ Culture

One pair of normal human eyes was obtained from The National Disease Research Interchange (NDRI), a nonprofit organization engaged in the procurement and distribution of human tissues for biomedical research in the United States. It was obtained with the signed consent of the patient according to the Tenets of the Declaration of Helsinki. The age of the donor was 67 years, the eyes were dissected within 30 to 40 hours of death, and no glaucoma was diagnosed. To revive the tissue and recover cell activity, the anterior segments were perfused as described earlier. ^{26 27 28} Briefly, eyes were bisected at the equator and the lens, iris, and vitreous removed. The anterior segment was then clamped to a modified petri dish and perfused at 3μ l/min of constant flow using a microinfusion pump. The culture medium was Dulbecco’s modified Eagle’s medium containing 100 U/ml penicillin, 0.1 mg/ml streptomycin, 170 μg/ml gentamicin, and 250μ g/ml amphotericin B (DMEM+). Anterior segments were maintained at 37°C in 5% CO₂. IOPs were continuously monitored with a pressure transducer connected to the dish’s second cannula and recorded with an automated computerized system. For this experiment, perfusion was performed for a total of 27 hours. At this time, the outflow facility baseline was stable and had a value of 0.101μ l/min per mmHg. At the end of the experiment, anterior segments were frozen in liquid nitrogen within 2 minutes of turning off the pumps and stored at −70°C for later dissection of the TM and RNA extraction. 

Light and Electron Microscopy

A second pair of eyes from a 32-year-old normal donor was perfused in similar conditions and used for morphology analysis. After 48 hours in culture, whole anterior segments were fixed by overnight perfusion with 2% glutaraldehyde at a pressure of 15 mm Hg. The anterior segments were then cut meridionally into quadrants and immersed in the same fixative until processing. From each of two quadrants, wedge-shaped specimens containing the angle region with the TM were cut and embedded in glycol methacrylate using an embedding kit (JB-4; Polysciences, Warrenton, PA). Blocks were counterstained with eosin and hematoxylin and sectioned to 2 μm. At least two sections from each of the two quadrants of the processed eyes were analyzed by light microscopy to examine the morphology of the tissue. 

For electron microscopy, wedges of meridional sections of the TM were postfixed in 1% osmium tetroxide, block stained in 2% uranyl acetate, dehydrated, and embedded (Spurr embedding medium; Polysciences). Ultrathin sections were stained with potassium manganate and lead and examined with a transmission electron microscope (1200Ex; JEOL, Tokyo, Japan). 

Construction of an HTM cDNA Library

RNA extraction and exponential amplification of the reverse transcribed cDNA were performed as previously described. ^{24 25} Briefly, the TM from a single eye was dissected and its RNA extracted with an RNeasy kit (Qiagen, Chatsworth, CA). One forth of the total RNA was lyophilized and used as template for the generation of a reversed transcribed first cDNA using a DNA synthesis kit (SMART-PCR; Clontech, Palo Alto, CA) and Superscript II, RNase H⁻ reverse transcriptase (Gibco, Gaithersburg, MD) to a final volume of 50 μl. The number of cycles needed for exponential phase amplification of this cDNA was determined by running a series of 16 μl analytical PCR amplifications at 17, 20, and 23 cycles using the same kit. To prevent overrepresentation of clones with short inserts (which would ligate to the vector more efficiently than larger ones), the resultant cDNAs were size fractionated using agarose gel electrophoresis. For that, the PCR-amplified sample was purified and concentrated in an elution column (QIAquick; Qiagen) to 30 μl and loaded onto a single well of a 2% low-melting agarose gel. Four separate slices corresponding approximately to molecular weights of 0.6 to 1 kb (fraction I), 1 to 2 kb (fraction II), 2 to 3 kb (fraction III), and more than 3 kb (fraction IV) were cut from the gel and melted at 65°C for 10 minutes. The cDNA from each slice was purified and concentrated to a total of 30 μl with an elution column (QIAquick; Qiagen). One tenth of each eluted cDNA was used for ligation into a cloning vector (pCR2.1-TOPO; Invitrogen, Carlsbad, CA) followed by transformation of Escherichia coli competent cells (TOP10F′ One Shot; Invitrogen) and plating onto ampicillin-agar plates with 5-bromo-4-chloro-3-indoyl-β-galactosidase (X-gal). A similar number of white cell colonies from each fraction were randomly selected for DNA sequencing. Glycerol stocks from each of these clones have been kept under identification numbers HTM1-0001 to HTM1-1060. 

Template Preparation and Sequencing

Plasmid purification and automatic DNA sequencing were performed by Midland Certified Reagent (Midland TX). Briefly, plasmid DNAs were purified by the alkaline lysis method and 0.5 to 1 μg used as template for cycle sequencing reactions (Thermosequenase kit; Amersham, Arlington Heights, IL) with IRD800-labeled M13 as the forward primer. Electrophoresis was then conducted (model LS4200; Li-cor, Lincoln, NE). For clones in which the M13 primer produced a poor-quality sequence, a second reaction was prepared using the M13 reverse primer. Because the SMART PCR library cloned by the TA method is not directional, the ESTs were generated from either the 5′ or the 3′ end of the cDNA clones. In the case of full-length cDNA sequencing, custom-made oligonucleotides were made inside the clone insert, and the full sequence was obtained by overlapping readings of both strands of the cDNA. 

Sequence Analysis

Sequences corresponding to the vector and to the PRC primer used to construct the library were deleted manually from each clone. All sequences were arranged in a unique FASTA file and searched against databases using the network BLAST client program (BLASTcl3) ²⁹ of the National Center for Biotechnology Information (NCBI; URL: http://www.ncbi.nlm.nih.gov/blast/blast.cgi). The searches were performed between September and November 1999. 

The first sequence search was conducted using BlastN against all entries in the nonredundant GenBank database. Sequences with an E -value expected (expected number of matches to be found merely by chance) ³⁰ equaling zero were considered to identify known genes or to have partial similarity to known genes. Sequences with E-values close to zero, were manually analyzed for overall similarity or identical stretches that could indicate a different gene from the same family, small sequencing errors, alternative splicing, chimeric clones or presence of vector sequences. Sequences including fragments with similarity with repetitive sequences (e.g., ALU, LINE) were considered perfect matches only when BLAST searches performed after eliminating the nonrepetitive part of the sequence also provided an E-value equal to zero. 

DNA sequences for which no match was found in the nonredundant GenBank database were subjected to a second search with the BlastN against the dbEST database. From those, DNA sequences that matched dbESTs were searched against the Unique Human Gene Sequence Collection (UniGene) database (URL: http://www.ncbi.nlm.nih.gov/UniGene/) to determine whether they were integrated in a gene cluster. Expression patterns and chromosomal localization of these DNA sequences found to belong to a cluster were thus obtained. 

Last, DNA sequences found to contain no similarities to any database were translated in all reading frames and further searched against the nonredundant protein database using the BLASTX program. The same DNA sequences were also searched for the presence of open reading frames using the ORF Finder program (http://www.ncbi.nlm.nih.gov/gorf/gorf.html). 

Hybrid Panel Analysis

To determine the chromosomal location of the gene represented by HTM1-0025 cDNA, we used the hybrid panel G3 from Stanford University (Stanford, CA). We designed primers for the specific PCR amplification of a fragment of the gene. Positive amplification was seen only with the human genomic DNA control but not with the hamster DNA. After analyzing PCR reactions from the 83 hybrids in the G3 panel, we submitted the results to the SHGC server ([email protected]). 

Results

Morphology of the Perfused HTM

The perfused human anterior segment organ culture keeps the majority of the cells of the TM tissue alive. ²⁸ In our experiments, after a period of 30 to 40 hours after death, the cells of the TM were revived by the flow of serum-free medium pumped through the outflow pathway at the constant rate of 3 μl/min. Figure 1 shows the histologic examination of a representative pair of human anterior segments perfused for 48 hours. The light microscopy photograph (Fig. 1A) reveals the well-preserved architecture of the TM tissue and the presence of a good number of cells in all three layers. Although some loss of the uveal cells was evident, a large percentage of cells from the corneoscleral and juxtacanalicular regions were conserved. The electron microscopic analysis (Figs. 1B 1C 1D) confirmed the healthy appearance of the cells. In the corneoscleral layer, cells properly lined the trabecular beams (Fig. 1B) and the juxtacanalicular layer was intact. SC was well formed, and the endothelial cells of the inner wall appeared to form tight intercellular contacts. 

Characteristics of the HTM Library and General Data of the ESTs

The electrophoretic analysis of the amplified HTM cDNA yielded the expected smear pattern with visible, discrete bands corresponding to the more abundant transcripts of the tissue (Fig. 2) . The smear extended from approximately 0.2 to more than 5 kb and showed clear, visible bands that indicate high abundance of full-length cDNAs. To determine whether significant genomic DNA contamination was present in the library, we amplified library genes using primers spanning short introns as well as other genes not found in HTM. We found that the corresponding PCR products from those amplifications never included intron sequences (even in cases in which introns were only a few hundred base pairs long). In addition, PCR reactions with primers for genes not expressed in the HTM did not generate any product (data not shown). 

The analysis of 1060 ESTs revealed the five groups of cDNA sequences: Group I (519 ESTs, 48.9%) showed identity with the sequences of the nonredundant GenBank database (using BlastN and E values of zero). They were labeled known genes. Group II (125 ESTs, 11.8%) matched well to EST sequences from unknown genes reported in other tissues in the public domain database (dbEST). They were labeled known ESTs–unknown genes. Group III (189 ESTs, 17.8%) exhibited no significant similarity to known genes or known ESTs in the public databases and were thus defined as novel ESTs. Group IV (31 ESTs, 2.9%) were of mitochondrial DNA. Group V (196 ESTs, 18.5%) included either repetitive elements (ALU, L1, MER), short tandem repeats, or clones with none or short inserts with no similarity in any database (192 ESTs). Four ESTs were found to be chimeras (β-mannosidase, Bir L37, ferritin L, proline-rich calmodulin) with no introns, possibly the result of rearrangements during the cDNA synthesis or PCR amplification. These were classified as uninformative sequences. Finally, in this library there were no sequences corresponding to rRNA (Table 1) . 

Some of the genes were represented in the library by more than one clone. The level of redundancy of the different cDNAs is a reflection of the relative abundance of expression of each of the library genes. The total number of unique cDNA species in each of the five groups is represented in the right column of Table 1 . All clones from the same gene (identical or nonoverlapping) corresponded to one cDNA species. The 519 ESTs of group I corresponded to 338 unique cDNA sequences and the 125 ESTs of group II to 120; the 189 novel ESTs of group III were all found to be single-copy cDNAs. 

Other potentially relevant genes were present in this library but did not show in the 1060 clones analyzed by single-pass sequencing. For instance, the presence of TIGR/MYOC as well as that of other olfactomedins was confirmed by PCR, but none of the 1060 clones corresponded to these genes. Despite the importance of TIGR/MYOC in glaucoma, this result is not surprising. In our laboratory, Northern blot analysis consistently shows that the expression of this gene in the HTM is lower than that ofα B-crystallin (represented by only one clone in the library). 

Gene Expression Profile in HTM

Group I: Clones Similar to Known Genes.

The 519 group I sequences matched to known genes with E-values of zero represented 338 different genes. The description and accession number of each of these genes, along with their distribution into nine different functional categories are summarized in Table 2 . The number of clones and cDNA species from each category are represented in Figure 3 . 

The genes more highly represented in the HTM library (by four or more clones) are summarized in Table 3 . The right column of this table shows the number of ESTs reported in the UniGene database for each of these genes. This information helps to differentiate between genes with ubiquitous high expression in most tissues and those in which high expression in TM may be associated with the specific functions of the tissue. 

Group II: Clones Similar to Reported ESTs from Unknown Genes.

Of the 1060 clones sequenced, 125 (11.8%) produced sequences matching ESTs from unknown genes previously reported in other tissues. Ninety-three clones can be classified as part of clusters in the UniGene database, whereas 32 where identical with ESTs from the GenBank with no UniGene cluster. Table 4 shows the complete number of HTM ESTs identical with UniGene clusters from unknown genes. The table also includes their chromosomal localization (when available), number of ESTs in the cluster, and information regarding their similarity to other known genes. In addition, those ESTs in the cluster showing a pattern of expression restricted to few tissues are also indicated. 

Group III: Novel ESTs. Clones with No Similarity to Sequences in the GenBank.

A relatively high number of clones, 189 (17.8%), showed no similarity with any entry of the public domain databases and were classified as previously unreported or novel HTM ESTs. Recently, a human hematopoietic cell line exhibited a 14.3% of novel ESTs, ²³ whereas a mouse blastocyst had only 2%. ²² However, a similar percentage of novel sequences have recently been reported in another eye tissue, a rabbit corneal endothelial library. ¹⁷ To demonstrate that some of the unreported ESTs truly corresponded to novel genes previously unreported in other tissues, we selected one of them, clone HTM1-0025, for full characterization. The complete insert of this clone was sequenced in both orientations using the forward and reverse M13 flanking primers. The sequence revealed an ORF of 966 nucleotides (nt) flanked by a 5′ untranslated region (UTR) leader of 58 nt and a 3′ UTR of 343 nt (Fig. 4A ). The deduced amino acid sequence of the ORF corresponded to a protein of approximately 33 kDa (322 amino acids). The 5′ leader cDNA contains a stop codon 31 nt upstream of the first methionine of the ORF which identifies this ATG as the first translated codon. Moreover, the nucleotides surrounding this ATG match the consensus sequence for the translation initiation region of vertebrate mRNAs. ³¹ On the 3′ end, the cDNA contains a putative polyadenylation signal located 22 nt upstream of a polyA tail. 

The protein encoded by this cDNA had no match in any database and was therefore classified as a novel gene expressed in the human TM. The closest similarity found for the new protein was that of a 33% with the MAS-related G-protein–coupled receptor (343 amino acids, GenBank accession number P35410; Fig. 4B ). ³² The similarity extended along most of the ORF and had a 14–amino acid stretch of identical residues close to the C-terminal end. The hybrid panel analysis of this clone provided a lod score of 7.28 with the marker SHGC-6030, and located the gene to chromosome 11. 

Discussion

With the Human Genome Project nearing completion, a major goal in biology is to provide a link between the genes identified in structural genomic studies and the physiology and pathophysiology of all tissues and cell types. A major step in providing such a link is the identification of ESTs from each particular tissue. Although a number of sequencing projects have already provided a large collection of human ESTs, none of them has focused on the tissues of the outflow pathway. In this study, we present a profile of the gene expression in the HTM based on the analysis of 1060 clones. 

The percentage of ESTs obtained in group I (known genes), group II (dbESTs), and group III (novel sequences) in the HTM library was not substantially different from those in other tissues. ^{16 17 18 22 23} The classification of the known genes by function (Table 2 , Fig. 3 ) also revealed a general pattern similar to that previously described for an average of 37 human tissues. ³³ The functional group showing a relatively higher representation in the TM library was the one for cell signaling and cell communication. This finding may be an indication of the importance of sensing changes in IOP and sending signals to other cells to respond to these changes. The group with a higher level of gene redundancy was the one involved in extracellular matrix synthesis and degradation, which could be a reflection of the importance of extracellular matrix dynamics in the TM metabolism. Although some genes coding for proteins involved in vesicle dynamics were present in the library, genes involved directly in phagocytosis were not highly represented. This could be the result of the perfusion model used. Because the perfusion medium did not contain aqueous humor proteins, DNA, and/or debris, the mechanisms to eliminate them may have been underregulated. 

In group I (known genes), some of the observed most abundant transcripts are usually not present at such high frequency in other tissues and may therefore provide useful clues about the biologic features of TM. The high expression of other genes, such as the glycolytic enzymes LDHA, GAPDH, and TPI is expected, given that TM obtains most of its energy though anaerobic metabolism. ³⁴ More surprising is the high abundance of transcripts coding for proteins such as matrix GLA (MGP), chitinase 3-like 1 (CHI3L1), apolipoprotein D, small inducible cytokine (SCYA20), regulator of G protein signal, stromelysin 1, and from the uncharacterized genes KIAA0258 and DKFZp586O0118. 

Matrix GLA is the first protein clearly associated with protection against calcification of soft tissues. ^{35 36} High levels of matrix GLA expression are observed in response calcification of human vascular cells in vitro ³⁷ and in atherosclerotic lesions. ^{38 39 40 41} The presence of matrix GLA as one of the more abundant transcripts in our library may indicate that protection against calcification is important in TM. Calcification is associated with the age-related loss of elasticity and contractility in some tissues and could have a similar effect on TM, impairing the ability of TM to regulate outflow. Of note, dexamethasone treatment that can result in reversible glaucoma, is known to induce calcification in blood vessels. ⁴²  

Chitinase 3-like or human cartilage protein gp-39 is another extracellular protein expressed in cartilage and strongly induced in macrophages from atherosclerotic lesions. ⁴³ This protein is also one of the antigens implicated in the autoimmune disease, rheumatoid arthritis. ^{44 45} Although its function is not yet clear, its involvement in tissue remodeling and in the pathologic course of blood vessel disease makes this gene a good candidate to play a potential role in TM conductivity and the pathology of glaucoma. 

It is more difficult to speculate about the possible role of such proteins as ApoD, SCYA20, and RGS5, because of our incomplete knowledge about their function in other tissues. However, data suggesting the involvement of ApoD in neuronal degeneration and regeneration, ⁴⁶ SCYA20 in inflammation and vascular permeability, ^{47 48 49} and RGS5 in the modulation of rectifying potassium (GIRK) channels ^{50 51} suggest possible links to TM physiology. 

Stromelysin 1 already has been identified as an important protein in the regulation of outflow resistance. ^{52 53 54} Stromelysin 1 can be induced by a number of factors. ^{55 56 57 58 59 60} However, it is not usually found as such an abundant transcript in noninduced tissues, with only 21 ESTs reported in the UniGene cluster. Its relative abundance in the TM library (approximately 0.4%) is consistent with the concept that the TM’s ECM is under constant remodeling. 

The presence of four clones representing each of the uncharacterized genes KIAA0258 and cDNA DKFZp586O0118 is particularly interesting, because none of them appears to be highly expressed in other tissues, and therefore they may be important for some specific functions of TM. This finding further emphasizes the need for high-throughput sequencing of HTM cDNAs. Some of the key genes essential for the physiology of this tissue may not yet have been fully characterized. 

The identification of ESTs from group II (previously reported ESTs from uncharacterized genes), and group III (novel ESTs not previously reported in other tissues) is particularly important for two reasons: First, given the incomplete penetrance, the influence of nongenetic factors, and the late onset of the disease, an efficient search for glaucoma susceptibility genes necessitates not only positional cloning but also a candidate gene approach. The identification of novel genes expressed in HTM together with their chromosome location may greatly contribute to the identification of glaucoma genes. Second, because glaucoma does not appear to be associated with problems in other tissues, proteins involved in this disease may have a TM-preferred tissue expression pattern. 

ESTs from uncharacterized genes (group II) can be analyzed for chromosome location and pattern of tissue expression using the information available in databases such as UniGene (Table 4) . Some of the ESTs found in our TM library are located in chromosomes with regions linked to glaucoma. ^{61 62 63} As new linkage analysis provides more precise information about all glaucoma loci and more extensive analysis of TM libraries provides more information about ESTs from this tissue, it will be possible to select candidate genes from this group of clones. Furthermore, other ESTs from this group belong to UniGene clusters that appear to have a pattern of expression restricted to very specific tissues. This is the case of clusters such as Hs.97431 which includes only four ESTs, all from a testis library; Hs.132591, which includes five ESTs, all from brain; and Hs.199612, which has only one EST from kidney. This observation suggests that these genes may have some specific functions associated with the particular physiology of these tissues. 

The identification of novel ESTs (group III) is particularly important in highly specialized and poorly studied tissues such as TM. However, analysis of novel, previously unreported ESTs has to be undertaken with some caution. cDNA libraries are never totally free of artifacts, which can include chimeric clones and/or fragments of genomic DNA. The presence of four chimeric clones in our library indicates that some of the novel clones could also be artifacts. However, four chimeras in 519 clones in the known gene pool represents a small percentage, probably not higher than that normally found in conventional non-PCR–amplified libraries. More important, the characterization of clone HTM1-0025 shows that true novel genes can be discovered within the subset of novel ESTs. The features of this clone (complete ORF, similarity with a known family, and presence of the polyadenylation signal) confirm its identity as a novel gene and not as an artifact. 

Finally, when evaluating single-pass sequencing analysis of cDNA libraries such as the one presented in this study, it is important to remember that any single library has some bias associated with the specific source of RNA. For instance, HTM samples acquired from postmortem donors may not accurately reflect the in vivo expression profile because of selective cell death (uveal layer), absence of IOP, and no aqueous flow. Using perfused TM solves some of these problems by recovering cell activity and maintaining IOP and aqueous flow. However, a limitation of this system is that the TM cells are exposed to tissue culture medium, not to physiological aqueous humor. They therefore do not have the exposure to some components, such as growth factors and reactive oxygen species, that is likely to affect their gene expression. In addition, it is important to consider the high degree of individual variability associated with the analysis of human samples. The particular history of the donor, medication, age, sex, and race also affect gene expression. Further analysis of more individuals is needed for a definitive characterization of the HTM gene expression profile. 

In summary, our study describes the first expression profile of known and novel genes in a human TM. We identified 338 known genes, 120 ESTs previously reported in other tissues, and 189 new ESTs. The relative higher abundance of some of the gene species suggests new potential functions of the TM tissue, such as the prevention of cell calcification. The identification of novel ESTs should provide new tools for understanding TM physiology and for identifying genes involved in the genetic susceptibility to glaucoma. 

 Sequence data reported in this article have been deposited in the GenBank database under accession numbers: BE439390-BE440238.

 Supported by National Eye Institute Grants EY11906 and EY01894, and Research to Prevent Blindness. TB is a Jules and Doris Stein Research to Prevent Blindness Professor Awardee.

 Submitted for publication March 22, 2000; revised June 23, 2000; accepted July 5, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Meridional sections of a 48-hour perfused TM from a human donor eye acquired after death. (A) Light micrograph of the filtering TM showing all three layers plus SC. Electron micrographs of the same TM: (B) trabecular cells covering the beams in the corneoscleral layer; (C) cells of the juxtacanalicular layer and inner wall of SC; and (D) corneoscleral layer showing a wandering cell, possibly a macrophage, within the intertrabecular space. Magnifications: (A) ×400; (B, D) ×3000; (C) ×4000.

Figure 2.

 

Size fractionation of HTM PCR-amplified cDNA. Two microliters of the diluted SMART-PCR cDNA were amplified for a 19-cycle template in 100μ l PCR reaction. The optimum number of cycles was previously estimated in analytic reactions stopped at 17, 20, and 23 cycles. The entire product of the preparative reaction was concentrated and loaded in a 2% low-melting-point agarose gel for size fractionation. Four cDNA fractions of the size shown were eluted from the gel and cloned separately into a vector (TOPO PCR2.1; Invitrogen, Carlsbad, CA). A similar number of clones from each sublibrary were randomly selected for DNA sequencing. M, molecular weight marker.

Table 1.

 

Summary of the ESTs Found in the HTM Library

cDNA Category	Clones			cDNA Species
		n	%	n	%
I Known genes	519	48.9	338	52.2
II Known ESTs/unknown genes	125	11.8	120	18.5
III No match in databases	189	17.8	189	29.2
IV Mitochondrial transcripts	31	2.9
V Uninformative sequences	196	18.5
Total	1060		647

Table 2.

 

ESTs Representing Known Genes in the HTM cDNA Library Classified by Function

Gene Name	Accession Number	Frequency
Cell division
GRO1 oncogene	X54489	1
Histone H3F3B	NM_005324	2
Histone H3F3A	X05857	1
EXT-like protein 2 (EXTL2)*	AF000416	2
Prothymosin-α (PTMA)	M14483	1
Nucleolar protein (B23), nucleophosmin (NPM1)	M23613	1
Homo sapiens DEAD/H (DDX5)	NM_004396	1
Id-2H Helix-loop-helix protein (ID2)	D13891	1
SET translocation (myeloid leukemia-associated) (SET)	NM_003011	1
Smoothened oncogene (SMO)	AF114820	1
Enhancer of rudimentary (Drosophila) homologue (ERH)	NM_004450	1
Transforming, acidic coiled–coil containing protein 1 (TACC1)	AF049910	1
TAX1BP1	U33821	1
Nuclear mitotic apparatus protein 1 (NUMA1)	Z14229	1
Apoptosis
Caspase 4 (CASP4)	NM_001225	1
BCL2/adenovirus E1B 19 kDa-interacting protein 3-like (BNIP3L)	AB004788	1
Apo1-human (MER5)	D49396	1
Apoptosis-related RNA binding protein (NAPOR3)	AF090693	1
Cell signal and cell communication
Small inducible cytokine (SCYA20)	NM_004591	4
Interleukin-6 (IL6)	M54894	3
Interleukin-8 (IL8)	Z11686	3
Interleukin-1-β (IL1B)	M15330	1
Angiopoietin-like (CTD6)	Y16132	2
Transforming growth factor-β (TGFB)	M60315	1
Tumor necrosis factor receptor 2 (TNFR2)	U52165	1
Natural killer cell enhancing factor A (NKEFA)	L19184	2
Fibroblast growth factor-2 (FGF2)	NM_002006	2
Fibroblast growth factor-7 (FGF7)	NM_002009	3
Pre-B cell–enhancing factor (PBEF)	U02020	2
Macrophage-specific colony-stimulating factor (CSF1)	M37435	1
Migration inhibitory factor (MIF)	M25639	1
Monocyte chemoattractant protein-1 (JEPR)	X60001	1
Human cellular growth-regulating protein (CELGROR)	L10844	1
Insulin-like growth factor-binding protein-3 (IGFBP3)	M35878	1
Insulin-like growth factor binding protein-5 (IGFBP5)	L27556	1
Intracellular adhesion molecule 1 (ICAM1)	M24283	1
Intracellular adhesion molecule 2 (ICAM2)	M32334	1
Annexin I (ANX1)	NM_000700	1
Annexin II (ANX2)	NM_004039	3
Integrin, β1 (ITGB1)	NM_002211	1
Integrin, β4 binding protein (ITGB4BP)	Y11435	1
Integrin, αE (ITGAE)	L25851	1
OB-cadherin-1, cadherin 11 (CDH11)	D21254	2
Transmembrane 4 superfamily member 1	M90657	1
Wilms’ tumor-related protein, laminin receptor (QM)	M64241	2
Calmodulin-I (CALM1)	U16850	1
Calcineurin Aα (CALAA)	L14778	1
S100 calcium-binding protein (S100E)	Z18950	2
Stanniocalcin (STC1)	NM_003155	2
cAMP responsive element modulator (CREM)	D14826	1
cAMP phosphodiesterase (CAMPHOS)	L12052	1
A kinase anchor protein (AKAP149)	NM_003488	1
Orphan G-protein–coupled receptor (RDC)1*	U67784	1
Regulator of G-protein signal (RGC5)	NM_003617	4
RAB2	AF070629	1
RAL A small GTP-binding (RALA)	X15014	1
GAB2	AB018413	1
RAP1B	X08004	1
P2X4 Purino receptor (P2RX4)	U83993	1
Angiotensin II Receptor (AIR)	L48211	1
T-cell receptor β chain dopamine-β-hydroxylase-like	U66059	1
Guanine nucleotide–binding protein 11 (GNG11)	NM_004126	2
ADP-ribosylation factor 4 (ARF4)	NM_001660	1
Lectin, galactoside-binding, Soluble, 1 (LGALS1)	NM_002305	1
Tumor protein D52-like 2 (TPD52L2)	NM_003288	1
Sphingolipid activator (SPHINO)	M81355	1
Cell surface glycoprotein (CD44)*	L05424	1
Protein tyrosine phosphatase receptor type M (PTPRM)	NM_002845	1
Protein phosphatase 1b (PPM1B)	NM_002706	1
Protein phosphatase 2a 65–kDa regulatory subunit	M65254	1
Protein phosphatase 1, catalytic Subunit, β Isoform (PPP1CB)	NM_002709	1
Inositol 1,4,5-triphosphate 3-kinase B (ITPKB)	NM_002221	1
Lithium-sensitive myo-inositol monophosphatase A1 (IMPA1)	AF042729	1
Inositol 1,4,5-triphosphate receptor, type 2 (TPR2)	NM_002221	1
Secretogranin I, Chromogranin B (CHGB)	NM_001819	1
Cell structure/motility
α-Tubulin	K00558	4
β-Tubulin	AF070600	4
α-Actin (ACTA2)	M33216	2
β-Actin (ACTB)	X63432	2
Actin like 6	AB015907	1
Thymosin β-4 (THYB4)	M17733	2
Vimentin (VIMENT)	X56134	2
Syntenin (SYCL)	AF000652	2
Myosin regulatory light chain	D50372	1
Myosin light chain 3 Nonmuscle (MLC3NM)	M31212	1
Profilin 2 (PFN2)	NM_002628	1
Arg/Abl-Interacting Protein	AF049885	1
Dynein light intermediate chain 2 (LIC2)	AF035812	1
Arp2/3 protein complex p21-Arc	AF006086	1
Extracellular matrix synthesis/degradation
Chitinase 3-Like 1/cartilage gp-39 (CHI3L1)	NM_001276	9
Matrix GLA (MGP)	M58549	9
Collagenase inhibitor, EPA glycoprotein	M59906	7
Stromelysin 1 (MMP3)	J03209	4
Macrophage elastase (MMP12)	NM_002426	2
Decorin (DCN)	NM_001920	2
Collagenase (MMP1)	X05231	1
Osteopontin (OSTP)	D14813	1
Pro-α-1 (V) Collagen (PA1V)	M76729	1
Fibulin 1 (FBLN1)	U01244	1
Cell/organism defense
β-2-microglobulin (B2M)	NM_004048	8
Human MHC protein homologous to chicken B complex protein	M24194	2
MHC class I HLA-Bw62	M28203	1
MHC class II DPw3-α-1 chain	M27487	1
Trophoblast STAT utron	AF080092	1
Protectin, MEM43 complement-inhibitory protein (CD59)	M34671	1
Complement decay-accelerating factor (DAF)	M15799	2
X2 box repressor A	U22680	1
λ-Immunogloblin light chain	D87018	2
Complement component 2 (C2)	L09706	1
Complement component 3 (C3)	J04763	1
H factor 1 (complement) (HF1)	M17517	1
Plasma protease (C1) inhibitor (C1NH)	M13656	2
Serum amyloid A (SAA)	M26152	1
Natural resistance-associated macrophage protein 2 (NRAMP2)	AF064484	1
Interferon-induced protein 17 (IFI17)	J04164	2
Interferon-inducible 1-8U Gene	X57352	1
Interferon-induced tetratricopeptide protein (IFI60)	AF083470	1
Homo sapiens hepatitis B virus X interacting protein (XIP)	AF029890	1
Ribonuclease L (RNASEL)	NM_002937	1
Heat-shock protein 90 (HSP90)	D87666	5
Heat-shock 70-kDa protein 10 (HSC71)	Y00371	3
Heat-shock 10-kDa protein 1 (HSP10)	X75821	1
Tumor rejection antigen (gp96) 1, heat-shock protein gp96 (TRA1)*	NM_003299	1
αB Crystallin (CRYAB)	NM_001885	1
Chaperonin containing T-complex polypeptide 1, β-subunit	AF026293	1
Tubulin-specific chaperone (TBCA)	NM_004607	1
Mangano-superoxide dismutase (MnSOD/SOD2)	X14322	3
Cu/Zn superoxide dismutase (SOD1)	X02317	1
Catalase (CAT)	X04096	1
Thioreductase-dependent peroxide reductase SP-22	AA987769	1
Anionic glutathione-S-transferase (GSTpi1)	X15480	2
Microsomal glutathione-S-transferase 3 (MGST3)	NM_004528	1
Glutathione-S-transferase M2 (GSTM2)	NM_000848	1
Thioredoxin (TXN)	X77584	2
Putative thioredoxin-like protein	AJ010841	1
Glutaredoxin, thioltransferase (GRX)	X76648	1
Ferritin L (FTL)	M11147	5
Ferritin H (FTH)	M11146	1
Metallothionein II (TISO2)	V00594	2
Esterase D	AF112219	1
Dihydrodiol dehydrogenase 1 (AKR1C1)	NM_001353	1
Caeruloplasmin (CP2)	X69706	1
Lipoprotein-associated coagulation inhibitor (CILA)*	J03225	1
Endothelial plasminogen activator inhibitor (PAI1)	X04429	1
DNA-dependent protein kinase DNA repair protein complex	U47077	1
Gene/Protein expression
Translation elongation factor 1-α (EEF1A1)	NM_001402	19
Transcription elongation factor-β (TCEB1L)	NM_003197	1
Translation elongation factor β (EEF1B2)	NM_001959	1
Transcription factor AREB6	D15050	1
Transcription factor TFIID	U57693	1
Translation initiation factor 3, Subunit 7 (EIF3S7)	NM_003753	1
Translation initiation factor 4a, Isoform 1 (ARF4)	NM_001660	1
Translation initiation factor 4b (INTFA4B)	X55733	2
Translation initiation factor 4 Gamma (EIF4G1)	NM_004953	2
Translation elongation factor-1-γ (EEF1G)	NM_001404	1
Translation initiation factor 2, subunit 3 (EIF2S3)	L19161	1
Ribosomal protein S2 (RPS2)	NM_002952	1
Ribosomal protein S3a (RPS3A)	NM_001006	1
Ribosomal protein S7 (RPS7)	AF077042	1
Ribosomal protein S15a (RPS15A)	NM_001019	1
Ribosomal protein S16 (RPS16)	NM_001020	1
Ribosomal protein S19 (RPS16)	NM_001022	1
Ribosomal protein S20 (RPS20)	NM_001023	1
Ribosomal protein S27 (RPS27)	NM_001030	2
Ribosomal protein, Large, P0 (RPLP0)	NM_001002	3
Ribosomal protein L5 (RPL5)	NM_000969	2
Ribosomal protein L6 (RPL6)	NM_000970	5
Ribosomal protein L19 (RPL19)	NM_000981	1
Ribosomal protein L23a (RPL23A)	U43701	1
Ribosomal protein L24 (RPL24)	NM_000986	4
Ribosomal protein L30 (RPL30)	L05095	2
Ribosomal protein L37 (RPL37)	L11567	1
Ribosomal protein L39 (RPL39)	NM_000997	2
Ubiquitin (UBC)	M26880	3
SMT3B protein ubiquitin-like (SMT3B)	X99585	2
Ubiquitin-homology domain protein PIC1, sentrin (SMT3C)	U83117	1
MMS2, ubiquitin-conjugating enzyme	AF049140	1
Ubiquitin-activating enzyme E1C (homologous to yeast UBA3) (UBE1C)	AF046024	1
Ubiquitin carboxyl-terminal Esterase L1 (UCHL1)	NM_004181	1
DEAD/H (ASP-GLU-ALA-ASP/HIS), RNA helicase, 68KD (DDX5)	NM_004396	2
Small nuclear ribonucleoprotein Snmp Sm D2 (SNPRD2)	NM_004597	1
Small nuclear ribonucleoprotein D3 Polypeptide (18-kDa) (SNRPD3)	NM_004175	1
hnRNP-E1 (HSRNPE1)	X78137	1
hnRNP core protein A1 (HSRNPA1)	X06747	1
RNA polymerase I 16–kDa Subunit	AF077044	1
Polymerase (RNA) II (POLR2B)	NM_000938	1
Ribonuclease P	AF001175	1
RCH1, karyopherin-α2 (KPNA2)	U09559	1
Karyopherin α4/Importin α3 (KPNA4)	NM_002268	1
26S Proteasome ATPase Subunit (MIP224)	AF038965	2
ζ proteasome chain (PSMA5)	NM_002790	1
Glutamine-tRNA synthetase (QARS)	NM_005051	1
Tryptophanyl tRNA synthetase (IFNWRS)	M77804	1
Metaxin 2 (MTX2)	AF053551	1
Signal recognition particle 19-kDa (SRP19)	NM_003135	1
Splicing factor, ARG/SER-rich 10 (SFRS10)	NM_004593	1
9G8 splicing factor (9G8SF)	L22253	1
α NAC transcriptional coactivator	X80909	1
Cullin 1 (CUL1)*	NM_003592	1
Zinc finger protein 9 (ZNF9)	NM_003418	1
CRSP9 cofactor required for Sp1 transcriptional activation, subunit 9 (33-kDa)	AF104251	1
Metabolism
Glyceraldehyde-3-phosphate-dehydrogenase (GAPDH)	M33197	11
Lactate dehydrogenase A (LDHA) (LDHH)	X02152	7
Triosephosphate isomerase (TPI)	M10036	5
Phosphoglycerate mutase 1 (PGAM1)	NM_002629	2
Phosphoglycerate kinase (PGK)	L00160	2
Phosphoglucomutase 1 (PGM1)	NM_002633	1
Enolase 1 (ENO1)	NM_001428	1
Neuron-specific γ-2 Enolase (ENOG)	M22349	1
Aldehyde dehydrogenase 1 (ALDH1)	K03000	1
Aldehyde reductase 1, aldose reductase (ALDR1)	NM_001628	1
Mannosidase αB, lysosomal (MAN2B1)	U05572	1
Glutamate dehydrogenase 1 (GLUD1)	NM_005271	1
Phosphorylase kinase, γ2 (PHKG2)	NM_000294	1
Spermidine N1 acetyltransferase (SAT)	NM_002970	3
Protein kinase C Inhibitor-I (HINT)	U27143	3
UDP-glucose dehydrogenase (UDGH)	NM_003359	2
NADH Ubiquinone oxidoreductase Subunit B13 (B13)	U53468	2
NADH dehydrogenase (ubiquinone) 1β4 (NDUFB4)	NM_004547	1
Hydroxyacyl-coA dehydrogenase (HADHB)	NM_000183	1
Fatty acid-coenzyme A ligase (FACL3)	NM_004457	1
11 β-Hydroxysteroid dehydrogenase	M76664	2
Farnesyl pyrophosphate synthetase (FDPS)	NM_002004	1
3-Hydroxy-3-methylglutaryl-coA reductase (HMGAB)	M62633	1
Carnitine palmitoyltransferase II (CPT2)	U09646	2
High-density lipoprotein binding protein	AI141861	1
Adipocyte lipid-binding protein (ALBP)	J02874	1
Apolipoprotein D (APOD)	M16696	6
Apolipoprotein J (APOJ)	J02908	1
Dihydrodiol dehydrogenase 1 (DDH1)	NM_001353	3
Phospholipase A2 cytosolic (PLA2)	M68874	2
Phospholipase A2, mitochondrial (PLAG2A)	NM_000300	1
Cyclooxygenase-2 (hCox-2)	U04636	1
ATPase, H+ transporting, lysosomal (ATP6H)	NM_003945	3
CGI-11 homologue to Sus scrofa 54-kDa vacuolar H(+)-ATPase subunit	AF132945	2
Cytochrome C oxidase subunit VIIC (COX7C)	NM_001867	1
Cytochrome C oxidase subunit VIIA Polypeptide 2 (COX7A2)	NM_001865	1
Cytochrome C oxidase subunit VIB (COXVIB)	X54473	1
Cytochrome C oxidase subunit IV	U90915	1
Somatic cytochrome C	M22877	1
ATP synthase subunit C (ATPSCP2)	D13119	1
Mitochondrial ATPase-coupling factor 6 (MATPSY)	M37104	1
S-Adenosyl homocysteine hydrolase	U82761	1
S-Adenosylmethionine syntase	X68836	1
Proline synthetase	AB018566	1
Pyrroline-5-carboxylate synthase	U68758	1
Cysteine dioxygenase	D85777	1
Nuclear-encoded mitochondrial serine hydroxymethyltransferase	L11932	1
Deiodinase, Iodothyronine, type II (DIO2)	NM_000793	1
Omithine aminotransferase (OAT)	NM_000274	1
Deoxycytidylate deaminase (DODDA)	L39874	1
Uridine phosphorylase (UP)	NM_003364	1
(Guanine-7-) methyltransferase (RNMT)	NM_003799	1
Carbonic anhydrase III (CAIII)	M29458	1
α-2,3-Sialytransferase, sialyltransferase 4a (SIAT4A)	AF059321	1
Pterin carbinolamine dehydrogenase	AF082858	1
Solute carrier family 16, member 4 (SLC16A4)	NM_004696	1
ADP/ATP carrier protein (ANT2)	L78810	1
Calpain (Ca2+)-activated cysteine protease (CAPN2)	NM_001748	1
Site-1 protease subtilisin-1 (SP1)	NM_003791	1
Coatomer protein complex, subunit α (COPA)	NM_004371	1
Reticulocalbin 2 (RCN2)	NM_002902	1
Ribophorin II (RPN2)	NM_002951	1
Caveolin-1 (CAVEOL)	AF095593	1
Homo sapiens mouse SKD1 homologue (SKD1)	NM_004869	1
COPII component (SEC23A)	X97064H	1
Unclassified
Ariadne homologue (ARI)	AJ130976	1
Erythrocyte membrane protein 72, stomatin (EPB72)	M81635	2
Integral transmembrane protein 1 (ITM1)	L38961	1
Novel gene similar to C. elegans hypoth 55.2-kDa protein F16A11.2	AL050255	2
P97 Homologue to bovine BCNT	D85939	1
Zona pellucida B Protein	U05781	1
GS1 (protein of unknown function)	M86934	1
Expressed in fibroblasts of periodontal ligament PL108	AB019409	2
LR8 Expressed by a subpopulation of human lung fibroblasts	AF115384	1
Ring zinc finger (RZF)	AF037204	1
AF1q (expressed in leukemic and immature hematopoietic cells)	U16954	1
Okadaic acid-inducible phosphoprotein (OA4818)*	AF069250	1
CGI-21 *	AF132955	1
HS1 binding protein HAX-1 (HAX1)	U68566	1
23-kDa highly basic protein (23KDHBP)	X56932	1
NEL-related protein 2	D83018	1
Nuclear protein (NP220)	D83032	1
Homo sapiens TDE homologue (DIFF33)	U49188	1
OTK27 mRNA, highly homologous to yeast nuclear protein NHP2	AF091076	1
RNA-binding protein regulatory subunit (DJ1)*	AF021819	1
N33 candidate tumor suppressor gene	U42349	1
RRM RNA-binding protein GRY-RBP (GRYRBP)	AF037448	1
Type II membrane protein similar to CD69 (CLECSF2)	AB015628	1
Divalent cation tolerant protein (CUTA)	AF106943	1
Homo sapiens homologue of yeast (S. cerevisiae) ufd2 (UFD2)	NM_004788	1
4F5rel candidate modifying gene for spinal muscular atrophy	AF073298	1
gp25L2 protein (GP25L2)	X90872	1
HSPC003	AF070659	1
HSPC013/cDNA DKFZp566G1246	AF077037	1
HepG2 3′ region Mbol cDNA	AA862447	1
JWA protein	AF070523	1
Josephin MJD1, Machado–Joseph disease (SCA3)*	U64821	1
Similar to human 7S L/1 EST (7S L pseudogene)	M20910	1
WSB-1	AF069313	1
SH3 domain–binding glutamic acid-rich protein-like (SH3BGRL)	NM_003022	1
SH3 domain–containing proline-rich protein (P85SPR)	NM_003899	1
Human lupus p70 (Ku) autoantigen	J04611	1
CaaX box 1 (CXX1)	AF038168	1
mRNA, chromosome 1 specific	AB007956	1
Testis-enhanced gene transcript (TEGT)	NM_003217	1
Clone 25071 and 25177 mRNA	AF131742	1
Clone 25022 mRNA	AF131802	1
DJ0614C 10/Hs. 56729 lymphocyte-specific protein 1	AC006453	1
Hypothetical protein	AJ012409	1
Major nuclear matrix protein (MNMP)	M63483	2
Placental transmembrane protein (Diff 33)	U49188	1
KIAA0005	D13630	1
KIAA0020	D13645	1
KIAA0054	D29677	1
KIAA0062	D31887	2
KIAA0164	D79986	1
KIAA0205	D86960	1
KIAA0258 *	D87447	4
KIAA0325	AB002323	1
KIAA0394	AB007854	1
KIAA0592	AB011164	1
KIAA0770	AB018313	1
KIAA0804	AB018347	1
KIAA0853	AB020660	1
KIAA0941	AB023158	1
KIAA0959	AB023176	1
cDNA DKFZp564P142	AL049976	1
cDNA DKFZp564D116	AL050018	2
cDNA DKFZp564L0822	AL049949	1
cDNA DKFZp564O1716	AL050265	1
cDNA DKFZp564P0462	AL080097	1
cDNA DKFZp566B183 *	AL050272	1
cDNA DKFZp566E144	AL110239	1
cDNA DKFZp566E2346	AL050073	1
cDNA DKFZp566H073	AL050060	1
cDNA DKFZp566J0124	AL050038	1
cDNA DKFZp586O0118	AL049389	4

*  EST in HTM library shows a possible alternative splicing with respect to the GenBank entry.

Figure 3.

 

Number of clones and cDNA species from each functional group described in Table 2 .

Table 3.

 

Summary of the Genes More Highly Represented in the HTM cDNA Library

Gene Name	Accession Number	ESTs in the HTM Library (n)	ESTs in UniGene Cluster (n)
Translation elongation factor 1-α (EEF1A1)	NM_001402	19	7810
Glyceraldehyde-3-phosphate-dehydrogenase (GAPDH)	M33197	11	1072
Chitinase 3-like 1/cartilage gp-39 (CHI3L1)	NM_001276	9	70
Matrix GLA (MGP)	M58549	9	132
β-2-Microglobulin (B2M)	NM_004048	8	984
Collagenase inhibitor, EPA glycoprotein (TIMP1)	M59906	7	440
Lactate dehydrogenase A (LDHA) (LDHH)	X02152	7	339
Apolipoprotein D (APOD)	M16696	6	204
Heat-shock protein 90 (HSP90)	D87666	5	2078
Ferritin L (FTL)	M11147	5	2421
Ribosomal protein L6 (RPL6)	NM_000970	5	1001
Triosephosphate isomerase (TPI)	M10036	5	679
Small inducible cytokine (SCYA20)	NM_004591	4	16
Regulator of G-protein signal (RGS5)	NM_003617	4	78
α-Tubulin	K00558	4	2699
β-Tubulin	AF070600	4	1008
Stromelysin 1 (MMP3)	J03209	4	21
Ribosomal protein L24 (RPL24)	NM_000986	4	576
KIAA0258	D87447	4	42
cDNA DKFZp586O0118	AL049389	4	60

Table 4.

 

UniGene Clusters for ESTs from Unknown Genes Represented in the HTM Library

UniGene ID	Accession Number	Frequency in the HTM Library	Chromosome Localization			ESTs in the UniGene Cluster (n)	Comments
Hs.3353	AA947562	1	Chr.5	D5S470-D5S487	110	Weakly similar to MCJ
			Chr.11	D11S1309-qTEL
Hs.3576	AI362255	1	Chr.20	D20S192-D20S177	235
Hs.3585	AI268262	1	Chr.14	D14S70-D14S281	57
Hs.3726	AI343831	1	Chr.18	D18S482-D18S71	52
Hs.3903	AI680796	1	Chr.17	D17S840-D17S1352	142
Hs.4206	AI089890	1	Chr.X	DXS990-DXS1059	97	Weakly similar to pp21
Hs.4210	AA740440	1	Chr.1	D1S2790-D1S2640	47
Hs.6226	W69563	1	Chr.16	D16S512-D16S515	72
Hs.6671	AA397735	2	Chr.4	D4S400-D4S1534	101	Highly similar to COP9 complex subunit 4
Hs.6945	AA127733	1	Chr.21	D21S260-D21S261	89
Hs.7099	AI469900	1	Chr.4	pTEL-D4S412	38
Hs.7298	AI589870	1	Chr.3	D3S3606-D3S3554	77
Hs.7535	AI144241	1			236	Highly similar to COBW-like placental protein
Hs.7887	AI262856	1	Chr.19	D19S225-D19S874	47
Hs.8249	AA481610	1			173	Weakly similar to SH3BGR
Hs.9973	AA740442	1			96
Hs.11156	W44483	1			120	Weakly similar to M. musculus PRAJA1
Hs.11463	AI025736	1	Chr.1	D1S2843-D1S417	126	Weakly similar to adenylate kinase 5
Hs.11637	AI672170	1	Chr.5	D5S471-D5S393	30
Hs.12101	AA424619	1	Chr.17	D17S933-D17S800	68
Hs.13895	H78107	1	Chr.2	D2S121-D2S110	23
Hs.15984	AI085974	1	Chr.X	DXS990-DXS1059	113	Weakly similar to gene pp21
Hs.17917	AI342481	1	Chr.11	D11S4194-D11S926	68
Hs.18778	AI380877	1	Chr.11	pTEL-D11S1318	94
Hs.22744	AI281368	1	Chr.19	pTEL-D19S413	34	Weakly similar to zinc finger protein 177
Hs.23655	AI472209	1	Chr.6	D6S453-D6S311	21
Hs.24183	AA740757	1			23
Hs.25023	AA191708	1	Chr.16	D16S3031-D16S3139	17
Hs.30558	AI307352	1			8
Hs.32148	AA629979	1	Chr.15	D15S157-qTEL	41
Hs.32352	AA662205	1	Chr.8	D8S560-D8S1820	80
Hs.34641	AA195063	1	Chr.4	D4S421-D4S1564	94
Hs.41271	AA236316	1	Chr.3	D3S1603-D3S1271	42
Hs.42824	AI130817	1	Chr.3	D3S1294-D3S1314	33
Hs.43847	AA495953	1	Chr.12	D12S333-D12S325	82	Weakly similar to splicing factor, ARG/ SER-RICH 7
Hs.44787	AA436003	1	Chr.12	D12S1596-D12S333	22
Hs.53656	AA935145	1	Chr.7	D7S2545-D7S2420	24
Hs.54943	Z78396	1	Chr.11	D11S1318-D11S909	102	Moderately similar to rat small zinc finger-like protein (TIM9b)
Hs.55189	AI279508	1			46
Hs.55940	AA421950	1			52
Hs.56178	H29087	2			3	Brain and heart
Hs.61364	AA258280	2			95
Hs.61635	AA316181	1	Chr.7		19	PAC clone DJ1121E10 from 7q21.1-q2
Hs.70333	AI186850	1	Chr.10	D10S197-D10S588	284
Hs.74313	AA976737	1	Chr.20	D20S182-D20S106	187
Hs.75847	AI264166	1	Chr.15	D15S146-D15S117	230
Hs.79530	AI631745	1	Chr.3	D3S1303-D3S3576	185	Weakly similar to hypothetical 30-kDa protein (D. melanogaster)
Hs.80857	AI341828	1	Chr.2	D2S176-D2S1897	75	Moderately similar to thymidine diphosphoglucose 4,6-dehydratase (C. elegans)
Hs.85222	AA926776	1	Chr.12	D12S99-D12S358	110	Weakly similar to R27090_2
Hs.85944	AI168347	1	Chr.10	D10S1786-D10S541	3	Lung and pool of melanocyte+ heart+ uterus
Hs.89271	AA284079	1			2	Normalized pooled tonsil germinal B-cells
Hs.94949	AA256769	1	Chr.2	D2S292-D2S145	42
Hs.97372	AA398546	1			4	Heart, lymph node, testis
Hs.97431	AA398575	1			4	Testis
Hs.102773	AA599292	1	Chr.9	D9S176-D9S279	46
Hs.103262	AA464846	1			11	Brain, germ cell, heart, prostate, uterus
Hs.106747	AA576835	1			35
Hs.107381	AA418277	1	Chr.4	D4S408-qTEL	80	Weakly similar to F38A5.1 (C. elegans)
Hs.109253	AI189007	2	Chr.20	D20S182-D20S106	118	Highly similar to N-terminal acetyltransferase complex ard1 subunit
Hs.110445	AI160162	1	Chr.7	D7S645-D7S2415	59
Hs.111515	AI192085	1	Chr.20	D20S197-D20S109
			Chr.18	D18S58-D18S70	362
Hs.111650	AA399376	1	Chr.3	D3S1555-D3S1299	105
Hs.113660	AI338842	1	Chr.3	D3S3606-D3S3554	18	Weakly similar to X-linked retinopathy protein
Hs.120051	AA837018	1			7	Germ cell, kidney, lung, pineal gland, uterus
Hs.120644	AA742659	1			5	Ear, lymph node, pancreas, uterus
Hs.121849	W67941	1	Chr.16	D16S422-qTEL	316	Weakly similar to GEF-2 protein
Hs.123218	AI655749	1			12	Brain, germ cell, lung, lymph node, ovary, pancreas, placenta
Hs.125035	AA722809	1	Chr.17	D17S922-D17S798	80	Contains ALU repeat
Hs.126403	AI359890	1			6	Brain, pancreas, pool lung+ testis+ b-cel
Hs.127345	AI368409	1			9	Other, brain, heart, kidney, pancreas
Hs.130522	AI498849	1			4	Brain, lung
Hs.132591	AA776274	1			5	Brain
Hs.132967	AI084948	1			14
Hs.137158	AA630955	1			8	Other, adipose, bone, muscle, placenta, prostate
Hs.151903	AA252446	1	Chr.4	D4S412-D4S1601	85
Hs.154554	AA093825	1			16	Weakly similar to ANKYRIN
Hs.165363	AI377892	1			9	Other, adrenal cortex, kidney
Hs.168541	AI554001	2			98
Hs.169014	AA917597	1			6	Brain, kidney, testis, vascular
Hs.169341	AA381200	1	Chr.8	D8S505-D8S519	62	Weakly similar to phosphatidic acid phosphohydrolase type-2c
Hs.169544	W35301	1			14
Hs.170247	AA872710	1	Chr.1	D1S474-D1S439	55
Hs.171889	AI138560	1	Chr.12	D12S346-D12S78	65
			Chr.14	D14S1038-D14S290
Hs.175941	AI332994	1	Chr.7	D7S2545-D7S2420	23	PAC clone DJ0593H12 from 7p31
Hs.178662	AI679407	1	Chr.16	D16S3114-D16S405	136	Chromosome 16 BAC clone CIT987SK-A-589H1
Hs.199612	AI672479	1			1	Kidney
Hs.200402	AI370915	1	Chr.19	pTEL-D19S413	32
Hs.206174	T72362	1	Chr.6	D6S1712-D6S407	7	Weakly similar to L82C (D. melanogaster)

Figure 4.

 

(A) Complete sequences of the insert from clone HTM1-0025 and deduced amino acid sequence of the ORF. The stop codon upstream from the putative initiation codon, the stop codon at the end of the ORF, and the putative polyadenylation signal are underlined. (B) Alignment of the deduced amino acid sequence from clone HTM1-0025 and the MAS-related G-protein (GenBank accession number P35410). +, conservative replacements.

The authors thank Laura-Leigh Rowlette, Ewa Worniallo, Rahul Garg for excellent technical assistance, especially with the perfused organ culture system, electron microscopy, and computer assistance.