Molecular Vision 2003; 9:360-396 <>
Received 2 May 2003 | Accepted 21 August 2003 | Published 22 August 2003

Expression profiling and gene discovery in the mouse lens

Michael A. Wride,1,2 Fiona C. Mansergh,2,3 Steffan Adams,3 Rebecca Everitt,4 Stephanie E. Minnema,4 Derrick E. Rancourt,4 Martin J. Evans2,3

1Cardiff University, Cell and Molecular Research Group, School of Optometry and Vision Sciences, Cardiff, Wales, UK; 2Cardiff Institute of Tissue Engineering and Repair, Cardiff Medicentre, Heath Park, Cardiff, Wales, UK; 3Cardiff University, Mammalian Genetics Research Group, Cardiff School of Biosciences, Cardiff, Wales, UK; 4University of Calgary, Department of Biochemistry & Molecular Biology, Calgary, Alberta, Canada

Correspondence to: Michael A. Wride, Cardiff University, Cell and Molecular Research Group, Department of Optometry and Vision Sciences, Cardiff, Wales, CF10 3NB, UK; Phone: +44 (0)29 2087 0054; FAX: +44 (0)29 2087 4859; email:


Purpose: Defects in the development and physiology of the lens can result in cataracts (opacification of the lens), which are currently treatable only by surgical removal. The lens is also an excellent system for understanding fundamental biological processes such as cellular differentiation and ageing. Here, microarrays have been used to gain insights into global patterns of gene expression in the mouse lens. Lens gene expression compared to non-lens tissues has been investigated in order to identify genes preferentially expressed in the lens and lenses of different ages have been compared to identify differentially regulated genes.

Methods: Genes expressed in the lens were identified using mouse GeneFilters® microarrays (GF400; ResGenTM). Each array comprises 5,184 mouse cDNAs representing sequence-verified known genes and uncharacterized ESTs spotted onto a nylon membrane. Target RNA (33P labeled) from lens and non-lens samples was hybridized to the arrays. The proportion of genes involved in various biological processes was investigated using Onto-ExpressTM to search for GeneOntologyTM terms associated with them. Differential gene expression was investigated using K-means clustering analysis. Expression of known and uncharacterized genes selected from the arrays was investigated further using semi-quantitative RT-PCR.

Results: 1,668 genes were expressed in one or more of newborn, 7 day old, and adult mouse lenses at levels significantly above background. Raw data and bioinformatics data relating to these genes have been published herein. There were 543 (33%) known genes, 124 (7%) had some similarity to known genes, 400 (24%) were functionally uncharacterized, and the remaining 601 (36%) genes were novel (matching only existing ESTs). Onto-ExpressTM identified genes involved in various biological processes including several categories containing greater numbers of genes than would be expected by chance, such as transcription regulation and G-protein coupled receptor signaling genes. Semi-quantitative RT-PCR confirmed preferential expression of several genes in the lens compared to non-lens tissues and genes exhibiting significantly higher expression in the 7 day lens compared to either adult or newborn lenses. Expression in the lens of 10 genes involved in apoptosis was also confirmed and, intriguingly, expression of hemoglobin isoforms (Hba-a1, Hba-X, Hbb-b1, Hbb-b2, and Hbb-Y) was confirmed using isotype specific primers. Finally, we confirmed the expression in the lens of all additional novel, uncharacterized and known genes tested.

Conclusions: The present work has provided insights into global patterns of gene expression in the lens and the expression of a significant number of genes has been confirmed using semi-quantitative RT-PCR. Genes preferentially expressed in the lens compared to non-lens tissues have been identified as well as genes differentially expressed between lenses at different ages. Gene expression profiling and gene discovery in the lens are essential prerequisites for future functional studies aimed at gaining insights into the potential roles of these genes in lens development, maturation, physiology, and pathogenesis (using targeted mutagenesis in mice, for instance).


Animal models are available that provide important knowledge regarding lens physiology and development [1]. However, new insights into normal vertebrate lens development and physiology and the mechanisms underlying cataract formation will be facilitated by systematic searches for, and investigations into the functions of, genes that are expressed in the maturing and adult lens or which, when mutated, cause cataract. For example, a mutagenesis screen has been very informative in this regard [2] revealing that mutations to various crystallins (including cryga, crygb, and cryge), pax-6, pax-2, mitf, lim2, and cx50 all cause cataracts.

The development of microarray technology provides an unprecedented and uniquely powerful approach for examining global patterns of gene expression, for intimating new roles for known genes, and suggesting roles for novel or uncharacterized genes in particular processes. Indeed, microarrays have an immense amount of potential for ophthalmic research [3]. The present study was instigated to systematically examine gene expression in the normal mouse lens during various stages of lens maturation. Recent studies have used microarrays to investigate gene expression changes in the lenses of pax-6 knockout and pax-6 over-expressing transgenic mice [4-6] and the lens epithelial cells of selenite injected rats [7]. We are also currently investigating differential gene expression during cataract progression in the Sparc knockout mouse (Unpublished data), the major phenotype in which is age-onset posterior cortical cataract [8,9].

Here, we have used a radioisotopic approach, using the mouse GF400 GeneFilters® array from ResGenTM, to compare relative gene expression levels between lenses of different ages and compared to various non-lens tissues. We aimed to identify genes potentially preferentially expressed in the lens and potentially differentially expressed at different stages of lens maturation. We also aimed to gain insights into global patterns of gene expression in the lens and the biological processes in which these genes are involved.

The microarrays used here are relatively inexpensive, there is a high proportion of novel and uncharacterized genes represented on them, and the use of radioactive detection methods means that they are highly sensitive, avoiding the need for RNA amplification. We have analyzed the data using Onto-ExpressTM [10] to determine in which biological processes the genes expressed in the lens are involved. We have used J-Express [11] to carry out K-Means clustering analysis [12] in order to identify groups of genes exhibiting particular expression profiles.

Using microarray analysis, we have identified 1,668 genes that are potentially expressed in the lens and, of these genes, only 33% are known. Numerous novel and uncharacterized genes were identified as well as genes not previously known to be expressed in the lens, including a number involved in apoptosis and, intriguingly, various isoforms of hemoglobin. We have also sought to identify genes preferentially expressed in the lens compared to non-lens samples and genes differentially expressed between lens samples.

This study provides new insights into the global pattern of gene expression underpinning the transparency associated with the lens, thereby providing a rich resource to the lens research community as a whole for the further investigation of genes expressed in the lens and elucidation of their functions. Identification of such genes is an essential prerequisite for future functional studies aimed at gaining insights into the roles of these genes in lens development, maturation, physiology, and pathogenesis by various methods including targeted mutagenesis in mice. Furthermore, such insights may facilitate the development of pharmaceutical and/or gene therapy based approaches for treatment of cataracts in humans, which will be valuable given the extent of cataract surgery waiting lists [13]. Furthermore, it has been pointed out in the recent Vision Problems in the US report that "ongoing research into the normal healthy functioning of the eye's lens may help us better understand the causes of cataract and how they might be prevented". Such knowledge will contribute towards the aim of Vision 2020 to eliminate avoidable blindness by the year 2020.


RNA extraction from tissues

The maintenance and treatment of animals were in full compliance with animal care guidelines comparable to those published by the Institute for Laboratory Animal Research (Guide for the Care and Use of Laboratory Animals). These were standard laboratory animal care protocols approved by the University of Calgary's Animal Care Committee. RNA was obtained from the following mouse tissues (strain CD1); newborn lenses, 7 day old lenses, adult lenses, skeletal muscle, liver, brain, 14.5 day embryo, and embryonic stem (ES) cells (strain D3). Tissues were dissected under RNAse-free conditions.

Lenses were placed in TRIzol® reagent (Invitrogen Corp., Paisley, UK) and immediately homogenized with a Dounce tissue grinder (Wheaton Science Products, Millville, NJ). Lenses from each of the various stages were pooled to obtain sufficient RNA for the array experiment. Several different pools of lens RNA were collected and aliquots of these were used for array analysis and subsequently for semi-quantitative RT-PCR. Other tissues were placed in TRIzol® and immediately homogenized using a mechanical homogenize (Polytron PT1200; Kinematica Inc. Cincinnati, OH). ES cells were grown in tissue culture in the presence of leukemia inhibitory factor (LIF) using standard procedures [14,15]. RNA was then prepared according to the manufacturer's instructions (Invitrogen). The amount of RNA was determined using a spectrophotometer and running samples out on 1% denaturing agarose gels checked RNA quality. Only high quality RNA preparations, as judged by the clarity of the ribosomal bands, were used in the labeling reaction.


GeneFilters® (GF400) microarrays (8) were obtained from ResGenTM. Each array comprises 5,184 mouse cDNAs representing sequence-verified genes and UniGene clusters spotted onto a nylon membrane. The cDNAs represent both known and novel genes. A catalogue of all the genes spotted on the GF400 GeneFilters® array is available at the ResGenTM web site. The title for each clone spotted is that of the UniGene cluster and the manufacturer also provides the accession number for each cDNA spotted.

Microarray hybridization

dCTP (33P-labeled) was obtained from NEN® Life Science Products (Boston, MA). Prehybridization of GF400 filters, preparation of 33P labeled cDNA, hybridization, and washing of the filters was carried out according to the manufacturer's instructions (ResGenTM). Hybridizations were repeated four times per sample and filters were stripped as per the manufacturer's instructions and sequentially hybridized with different samples.


After washing, filters were wrapped in cling wrap. Wrapped filters were then placed in a StormTM phosphoimager for several days and images of the hybridized filters were obtained using ImageQuantTM software (Amersham Biosciences, Quebec, Canada). Images were saved for import into PathwaysTM 3 software.

PathwaysTM 3 software

Image analysis was carried out using PathwaysTM 3 software (ResGenTM). All the microarray images were imported into a project folder in the software. Microarray data were normalized using datapoint normalization in which the mean intensity value for all the spots on all the microarrays uploaded was calculated by the software and used as a normalization factor. The software divides the raw intensity value of each spot by the mean intensity value of all spots on all the microarrays. In this way, the mean intensity value across all the microarrays receives a normalized value of 1.

Variation of the data

Hybridizations were carried out 4 times for each condition (3 times for ES cells) using a hybridization-stripping-hybridization procedure. The consistency of hybridization for the same condition was checked manually during the acquisition of microarray images using ImageQuantTM. Only images from filters with reproducible spot patterns between repetitions, with clear spots and low background were imported into PathwaysTM 3 software for further analysis.

Data from each repetition of each experimental condition were exported from PathwaysTM 3 into Microsoft® Excel and mean and standard deviation (SD) values were obtained for each data point. Since the SD alone does not measure the relative variation of a set of repetitions, we calculated the coefficient of variation (CV). The CV expresses the standard deviation as a percentage of the mean (SD/mean x 100 [16]) calculated from each repetition of an experiment. The CV is good for assessing data variability because it takes account of the fact that a small SD may actually be a large percentage of the mean of a small number, whereas it may be only a small percentage of the mean for a larger number. In the microarray studies described here, we considered that the smaller the CV value for each spot the greater the reliability of the mean intensity value for that particular spot.

Filtering of data for export to Onto-ExpressTM and J-Express

In Excel, the data were filtered by removing spots that fell at or below the mean normalized background intensity plus two SDs of the background. The normalized mean background intensity value for each array was calculated by the PathwaysTM 3 software. Export of these data to Microsoft® Excel allowed the mean normalized background intensity value across all the arrays to be calculated (0.189). The SD of the background intensity calculated from all the arrays was 0.158. Thus, the mean normalized background intensity plus two SDs was determined to be 0.505 (0.189 + 2 x 0.158). All spots with normalized intensity values less than or equal to 0.505 were eliminated from the data set. Therefore, the data set was reduced from 5,184 to 1,668 ESTs.

Bioinformatic management of the data

All 1,668 genes expressed at mean background plus 2 SDs of background in lens samples have been placed in a searchable web-accessible database based on one that we previously developed in the course of a different project [17]. We have called our publicly accessible database the Cardiff Array Database (CARD), available through the Wride laboratory web page. This raw data is included in Appendix 1. In the CARD, bioinformatic information for each EST is available via links from the accession numbers, UniGene clusters, LocusLink information, mouse genome Informatics (MGI) for known genes, and GeneOntologyTM terms. The data are listed in descending order of the maximum lens intensity value for each EST identified using the array. The microarray spot intensity profiles across all the samples tested are available by clicking on the microarray icon in the "Data" column. The development of the database will be ongoing.


Accession numbers for each of the 1,668 ESTs were imported into Onto-ExpressTM [10], which was used to search for GeneOntologyTM terms relating to the known biological process of these genes, thus correlating identified genes with their functional characteristics. This software also calculates a significance value (p) for the number of times genes in a particular category are identified, based on the total number of genes in that category on the array. This value indicates whether there are higher numbers of genes identified in a particular category than one would expect by chance.


The data were exported from PathwaysTM 3, were logged (base 10), and were then imported into J-Express for further analysis [11]. J-Express facilitates clustering analysis of genes identified using microarrays. Clustering analysis involves the use of computer algorithms designed to identify genes with similar expression profiles within the data generated by a microarray experiment. These genes are then grouped into different clusters containing groups of genes with similar expression profiles. The entire set of clusters generated using the algorithm represents all possible clusters identified within the data. We analyzed our data by K-means clustering analysis using the Euclidean distance matrix. In K-means clustering, the user defines the number of clusters that the software should find within a data set. Generally, the number of clusters to find will depend on the number of samples for which microarray data has been obtained and the extent of gene expression profile variation within the data set. In the case of K-means, the user may run the algorithm a number of different times with different numbers of user-defined clusters until he/she is satisfied that an appropriate number of clusters are identified (for a detailed review of various types of clustering analysis see [12]). When all the samples were analyzed, to identify genes potentially preferentially expressed in the lens samples compared to the non-lens samples, 36 user-defined clusters (K) were identified. For the lens samples, 24 user-defined clusters (K) were identified. In both cases, the maximum number of iterations was set to 200 and a random seed was generated before running the algorithm. In each case, the profiles of the different clusters identified were different from each other, but each individual cluster contained genes with similar expression profiles. Clusters were selected for further analysis if the mean spot intensity of the three lens samples indicated at least two-fold higher expression than the mean value for the non-lens samples. In the lens only comparisons, clusters of genes were selected if the intensity value in the 7 day old lens exhibited at least two-fold higher expression than either the newborn or adult sample (whichever was highest). J-Express was also used to carry out hierarchical clustering and similar results were obtained as with K-means (in terms of gene composition of the clusters identified, data not shown); therefore, K-means was the primary means of clustering analysis in this study.

Gene selection from clusters

Each gene in each cluster selected was examined in more detail in order to decide if that gene should be selected for follow-up using semi-quantitative RT-PCR. For genes potentially preferentially expressed in the lens, in order to be selected for follow-up, the minimum value of expression of a gene in all the lens samples was required to exhibit at least two-fold higher expression than the maximum value for that gene in the non-lens samples. Genes were then ordered in priority for follow up, based on these ratio values. For example, in cluster 17, the ratio of the γA-crystallin intensity value in the 7 day old lens (140.3), compared to the highest value in the non-lens samples (14.5 day embryo, 0.3) was 467.7.

In the case of the clusters exhibiting highest expression at 7 days, genes were selected for follow up based on the ratio of the intensity value of a given gene in the 7 day lens sample divided by the next largest intensity value in either the newborn or adult sample. For example, the EST AI428498 exhibited a 19.9 fold higher expression in the 7 day lens sample (2.8) compared to either the adult and newborn lens samples (0.1). The CV for each spot was also considered in selecting genes from clusters for follow up using semi-quantitative RT-PCR. We aimed to select genes for follow up, which had lower (less than 75%) mean CV values across all repetitions for all samples tested. Finally, we considered both the ratio values and the CV values together. Genes showing high mean intensity ratios between the relevant samples and low mean CV values were prioritized when considering which genes to follow up using semi-quantitative RT-PCR.

Semi-quantitative RT-PCR

Semi-quantitative RT-PCR was used to further investigate the expression of genes selected for follow up from the microarray. All PCRs for a given gene were performed at least three times with the same set of primers in order to confirm the reproducibility of the data. PCRs were performed from cDNAs formed in a reverse transcription reaction from the same, pooled RNA sample used on the microarrays.

RNA samples were initially digested with RNAse free DNase I (Ambion, Huntingdon, UK) to remove contaminating genomic DNA. The RT reaction was then carried out using the SuperScriptTM First-Strand Synthesis System for RT-PCR (Invitrogen) according to the manufacturer's instructions. Briefly, 2 μg of total RNA from each RNA sample was used in the RT-reaction and no-RT controls (no RT enzyme added) were run in parallel.

PCR primers for each gene picked from the array were designed as follows. Image clones representing ESTs selected from the array were ordered from ResGenTM or the Human Genome Mapping Project Resource Centre (HGMP; Hinxton Cambridge, UK). These clones were sequenced from both ends (T3 and T7, or SP6) using the BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA) and the University of Calgary and Cardiff University sequencing facilities. The sequence of ESTs was confirmed using BLAST 2 sequences at NCBI by blasting the sequence obtained against the sequence of the EST chosen from the array. It was essential to check clone fidelity, since it has been reported that there is a degree of error associated with IMAGE clones [18]. Clones where the sequence did not match that identified on the array were discarded. In fact, only two clones ordered from ResGenTM were incorrect. One clone did not contain a gene insert in the vector and the other had an incorrect insert. All clones ordered from HGMP were correct when sequenced. Thus, it can be assumed that only a relatively small amount of error exists in the data set of 1,668 genes identified here. The IMAGE clone-specific 5' and 3' sequences in the database, and, where they existed, full-length reference sequences in the database, along with the sequences we generated in the lab, were used to obtain a consensus sequence using the cap 3 sequence assembly program. PCR primers were designed using the consensus sequences generated as described above for each EST picked for follow up using the Primer3 program from the Canadian Bioinformatics Resource. The primers and conditions used for each EST picked for follow up are listed in Table 1. The appropriate IMAGE clone and genomic DNA prepared from ES cells using standard protocols [19] were used as positive controls (not shown) every time a PCR was run. Negative (no RT) and water blank controls were also used each time in the PCR reactions. An even concentration of cDNA samples was achieved using actin and GAPDH primers in PCRs to generate even bands across the sample range using varied numbers of cycles. Samples were normalized with respect to each other based on the lowest number of cycles that gave even bands when using the concentration controls.


MIAME compliance

Every effort has been made here to present the results of our microarray analysis according to the Minimum Information about a Microarray Experiment (MIAME) standards [20]. The raw normalized data is included in Appendix 1 and also available from the CARD. We are unable to provide pre-normalized data from the PathwaysTM 3 software due to the technical limitations of the program. However, original gel files of microarray images are available upon request from the authors.

Variation of the data

A representative assembled image of the same section of the four hybridization repetitions of the newborn lens microarray is presented in Figure 1A. The hybridization patterns for this section of filter are consistent with each other. We also present in Figure 1B a representative scatter plot (a screen shot from the PathwaysTM 3 software) comparing the data from filters 1 and 2 (both newborn lens sample repetitions) excluding below background spots. It is apparent that the data, for the most part, show a low degree of variability. However, the scatter plot is tighter at high intensities, but spreads at lower intensities, indicating increased variability in the repetitions with low intensity spots. The means, SDs, and CV for each repetition of each of the 1,668 spots on the array are presented in the CARD.

Bioinformatic analysis of the data

Using the database that we set up, we searched UniGene using the accession numbers of the relevant ESTs on the array. Of the genes we identified at above background levels using the microarray, 543 (33%) genes were known, 124 (7%) had some similarity to known genes, 400 (24%) were functionally uncharacterized, and the remaining genes 601 (36%) were novel (only matching existing ESTs).

We also carried out a lens expression survey for all of the known genes amongst the set of 1,668. However, the only UniGene database that contains sufficient numbers of lens ESTs to be representative is H sapiens. We therefore searched for readily identifiable human homologues of our known mouse genes using HomoloGene via links from the mouse UniGene cluster in question. The human homologues of 250 (46%) of our known mouse genes contained lens derived ESTs within the human UniGene clusters. The remaining 293 (54%) of the known genes either did not have lens ESTs within the human cluster (the majority) or were not homologous to any human UniGene cluster (the minority). We also carried out searches of OMIM. Using this approach, only 9 known genes were noted as expressed in the lens.

Biological properties of the genes identified in the lens

1,668 genes were expressed in at least one of the lens samples at intensity levels above the mean normalized background plus 2 SDs of the background. Onto-ExpressTM version 2 [10] was used to search for GeneOntologyTM terms relating to the known biological process of these genes.

Considering the p values calculated by Onto-ExpressTM for the biological processes, the following results were obtained for groups containing at least 4 genes at p<0.05. The most significant group of genes contained those involved in the G-protein coupled receptor protein signaling pathway (p=0.014, 8 genes), followed by heat shock response genes (p=0.035, 4 genes), sensory organ development genes response (p=0.035, 4 genes), and genes involved in transcription regulation (p=0.047, 31 genes). For details of genes contained in these groups see Table 2.

Most highly expressed genes

The 50 most highly expressed genes identified in the lens samples, in order of descending maximum lens intensity value, are presented in Table 3. As would be expected, crystallins represent the most highly expressed group of genes and they also have the greatest lens-specific expression (considering their intensity values compared to the non-lens samples). Numerous other genes are also expressed at high levels, although the intensity of expression in lens versus non lens samples varies depending on the particular gene in question.

Clustering analysis

K-Means clustering analysis was used to identify genes on the array potentially preferentially expressed in the lens compared to non-lens samples and potentially differentially regulated between the lens samples at different ages. K-means was set to identify 36 separate clusters in the data set (Figure 2). Of these, four clusters (7, 14, 17, and 31) were selected as potentially containing genes preferentially expressed in the lens samples relative to non-lens samples (Figure 3). The genes identified in each of these clusters are presented in Table 4.

K-means was set to identify 24 individual clusters on lens samples alone and two clusters (L4 and L11) were selected for follow up using semi-quantitative RT-PCR (Figure 4). These clusters contained genes exhibiting potentially highest expression in 7 day old lenses compared to both newborn and adult lens samples (Figure 5). All genes were followed up from cluster L4, whereas only one gene (AI452115) was followed up from cluster L11. The genes identified in each of these clusters are presented in Table 5. No clusters were identified, which showed significant up-regulation in newborn or adult lenses compared to the other stages examined.

Semi-quantitative RT-PCR

Semi-quantitative RT-PCR was used to further investigate the potential preferential expression in the lens of genes indicated by the K-means clustering analysis (Figure 6) as well as genes with potentially significantly higher expression at 7 days, compared to newborn and adult lens samples (Figure 7). The genes tested and those confirmed as preferentially expressed in the lens compared to the non-lens samples are presented in Table 4 (marked with an asterisk). The genes tested and those confirmed as up regulated in the 7 day lens samples compared to adult and newborn samples are presented in Table 5 (marked with an asterisk). Those tested but not confirmed are marked with a sharp symbol ("#"). Genes not tested using semi-quantitative RT-PCR during the course of these studies are not marked.

We confirmed preferential expression in the lens of 2 out of 3 genes tested in cluster 17 (67%), 3 out of 8 genes tested in each of clusters 7 and 14 (37.5%), and 1 out of 4 genes tested in cluster 31 (25%). Considering the lens samples alone, we confirmed higher expression in the 7 day old lens samples compared to either adult or newborn samples in 9 out of 11 genes tested (82%). Only two genes from cluster L4 were not confirmed as differentially expressed.

We identified 10 apoptotic genes (as determined using Onto-ExpressTM) using the microarray (Table 2). Using semi-quantitative RT-PCR, we confirmed the expression of all 10 of these apoptosis genes in the lens samples (Figure 8). These genes are presented with their normalized intensity values in Table 6.

Using Onto-ExpressTM, we identified a biological process category defined as "oxygen transport". Upon closer examination, the EST in this category (AI323916) matched the hemoglobin beta adult major chain (Hbb-b1). Bearing in mind the possibility of sequence similarity between hemoglobin isoforms, we therefore hand designed primers to different mouse hemoglobin isoforms (Hba-a1, Hba-X, Hbb-b1, Hbb-b2, and Hbb-Y). The primers were designed using 3' sequence and in each case this sequence was incompatible with the 3' ends of all other isoforms, thus ensuring their specificity. Expression in the lens of each isoform was then confirmed using semi-quantitative RT-PCR (Figure 9).

We also randomly selected genes from the list of 1,668 that were expressed at above background levels in order to test their expression in the lens. The results of these additional experiments are presented in Figure 10 and Figure 11. Although, we were not always able to confirm preferential expression in the lens or differential expression between lens samples using semi-quantitative RT-PCR, we were always able to confirm expression in at least one of the lens samples for all the genes we tested. Expression of an additional 19 novel and uncharacterized genes (Figure 10) and 11 known genes (Figure 11) identified using the arrays was confirmed in the lens using semi-quantitative RT-PCR. These genes are presented with their normalized intensity values in Table 7 and Table 8.


The systematic analysis of gene expression profiles, gene discovery and gene characterization in the lens are priorities for gaining insights into the roles of these genes in lens development, maturation, physiology, and pathogenesis. Here, we have used nylon-filter based microarrays to gain insights into global patterns of gene expression in the lens. We have used GeneOntologyTM analysis to determine the biological processes in which these genes are involved and we have also compared relative gene expression levels between lenses of different ages and compared to various non-lens tissues using clustering analysis. We have successfully identified genes preferentially expressed in the lens, compared to non-lens samples and differentially expressed in the lens at different stages of maturation. We have also identified expression in the lens of significant numbers of genes involved in apoptosis and also various hemoglobin isotypes, the expression of which in the lens has not previously been reported. Finally, using semi-quantitative RT-PCR we confirmed expression in the lens of all of the additional novel, uncharacterized and known genes we elected to test.

"I-gene" microarrays containing eye genes as probes are now becoming available [21]. However, these resources were not available when we began this study. Therefore, in this study, we chose to use ResGenTM nylon-based filter arrays because of their relative inexpensiveness, the sensitive nature of the radioactive labeling technique employed, and the high number of novel and uncharacterized ESTs that they contain. This system has been highly effective in identifying a significant number of genes (1,668) potentially expressed in the lens. Many of these genes are uncharacterized and many of them are known genes whose expression has not previously been recognized in the lens.

Expression in the lens of a significant number of these genes has been confirmed in the present study using semi-quantitative RT-PCR. The data generated represent an excellent resource for the lens community as a whole. In order to facilitate access to these data, we have established a web accessible database, the CARD which contains bioinformatics information relating to all 1,668 genes expressed at above background levels in at least one of the lens samples on our arrays as well as raw microarray data for all 5,184 spots on the arrays, including mean, SD, and CV values. Indicative of the usefulness of this study is the fact that we were able to confirm expression in the lens of all of the genes we selected from the set of 1,668 using semi-quantitative RT-PCR. Furthermore, of the 1,668 genes from the microarray, a known biological process could only be assigned for 543 genes, thus the remaining genes (approximately 67%) remain novel or functionally uncharacterized. Thus, this study has made a significant contribution towards gene discovery in the lens.

GeneOntologyTM analysis using Onto-ExpressTM

Using Onto-ExpressTM, we were able to gain insights into the numbers of genes that were uncharacterized and known and, amongst the known genes, to which biological processes they could be assigned (Table 2). The most significant group of genes belonged to the G-protein coupled receptor protein signaling pathway (p=0.014, 8 genes). GPCRs constitute the largest single family of cell surface molecules involved in signal transduction and they are involved in numerous physiological functions and pathological conditions when their function is altered [22].

Heat shock response genes and genes involved in sensory organ development were both significant categories of genes according to Onto-ExpressTM (p=0.035, 4 genes in both cases). Hsp40 is a small heat shock protein that has recently been shown to be down-regulated using microarrays in Pax-6 heterozygous mouse lenses [5]. HSP47 is thought to possess a collagen specific molecular chaperone function [23] and thus may be important in maintaining collagen molecular integrity in the lens, for example, during synthesis of the lens capsule. The small heat shock protein chaperonin-10 was also expressed in the lens and its function deserves further attention. All the genes identified as being involved in sensory organ development were crystallins.

Significant numbers of transcription factors were also identified in numbers greater than one would expect by chance (p=0.047, 31 genes). Amongst the known transcription factors identified here using the microarray, published data confirms the presence of and/or a role for several of these in lens development or physiology. For example, YY1 is associated with the nuclear matrix of lens epithelial cells and may have a role in the genetic regulation and development of the lens [24]. Retinoblastoma 1 (Rb1) and retinoblastoma-like 2 were identified here. Knockout mutants of Rb1 result in a block in lens differentiation (reviewed in [25]). Stat1 is also expressed and factor(s) responsible for lens cell proliferation in vivo activate the Jak-STAT-signaling pathway [26]. Finally, the Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain, 2 was identified here from the microarray and Cbp/p300 is a known activator of crystallin transcription [27]. Thus, this transactivator may be involved in regulating transcriptional activation of crystallin during lens fiber cell terminal differentiation. It is likely that many of the other transcription factors identified will also have functional roles in the lens.

Amongst the 50 most highly expressed genes identified (Table 3) were, as would be expected, several crystallin isoforms (γA-, γC-, γE-, and βA1-crystallin). The expression in the lens of another highly expressed gene the RIKEN cDNA 2810455F06 gene with weak similarity to ankyrin has been confirmed using semi-quantitative RT-PCR (Figure 10). Ankyrins appear to be important in the lens in maintaining contact between lens fiber cells, and targeted disruption of ankyrin-B leads to disorganisation of lens fibers in mice [28] suggesting that the ankyrin like gene identified here may also be important in this process. We have also identified high expression of ferritin light chain 1 in the lens and the NEIBank study [29] has also identified this gene amongst its most highly expressed transcripts in construction of a human lens cDNA library.

Some genes identified here in our microarray experiments, in the biological process categories just described, are known to be expressed in the lens. These genes have known functional roles in lens physiology and/or development (crystallins, lens epithelial protein, major intrinsic protein of eye lens fiber, ankyrin, collagen IV, and ferritin). Our lens gene expression survey demonstrated that human UniGene clusters representing 250 of the known genes we identified contained lens derived ESTs, thus demonstrating their expression in the lens. The remaining 293 known genes show no evidence of lens expression. The fact that we have identified a significant number of genes with prior evidence of lens expression supports the methodological approach used for lens gene identification. On the other hand, considerable numbers of genes have been identified which were previously not noted as expressed in the lens. This supports the utility of the approach we used for lens gene discovery.

Clustering analysis

Onto-Express was very informative in terms of gaining insights into the biological processes in which genes, whose expression was identified in the lens using the microarray, were involved. However, these data only indicated whether a gene was expressed in at least one of the lens samples (newborn, 7 day old, or adult lens) at levels significantly above background, but gave no indication as to the profiles of gene expression across all the samples examined. We therefore wished to analyze the microarray data further in order to identify genes that were preferentially expressed in the lens compared to non-lens samples and/or differentially expressed between the lens samples. We therefore elected to use the program J-Express in order to carry out clustering analysis on our data. Clustering analysis searches for groups of genes that have similar expression profiles when comparing the various microarrays performed on different samples. Genes exhibiting similar profiles in terms of relative intensities are then grouped together in clusters. Here, we used K-means clustering analysis [12] to identify potential groups of genes with the appropriate profiles that we were interested in identifying.

Using semi-quantitative RT-PCR, we confirmed the preferential expression in the lens of 9 of the genes we identified using clustering analysis on the microarray data. Overall, the rate of confirmation of genes preferentially expressed in the lens was approximately 50%, although the best cluster containing genes expressed in the lens at high levels was cluster 7 at 67% (although, given that this cluster contains crystallins, it is expected that this percent would be higher if we had tested more of these genes), whereas cluster 14 was the least reliable at 25%. The lens only clusters were very reliable in confirming up-regulation in the 7 day lens compared to either newborn or adult lenses of 9 genes out of 11 tested (82%). The reliability for a cluster to yield genes preferentially expressed in the lens or genes up-regulated in the 7 day old samples depended on several factors. Calculation of the CV proved to be an effective method for selecting genes for follow up. In general, the lower the mean CV for a given set of repetitions of a set of spots, the more reliable the mean normalized spot intensity for that set of spots and the greater the chance of confirming preferential/differential gene expression. The mean spot intensity value (indicating the level of expression of a given gene), was also an important factor to consider in selecting genes for follow up. For example, we found that genes preferentially expressed in the lens at higher levels such as βA1-crystallin, major intrinsic protein of eye lens fiber, and the lens epithelial cell protein were confirmed using semi-quantitative RT-PCR, whereas genes expressed at lower levels were not as reliable. Finally, the ratio value obtained was also important in whether a gene could be confirmed as preferentially/differentially expressed using semi-quantitative RT-PCR. For example, when comparing the lens samples with the non-lens samples, cluster 7 has crystallin genes expressed at very high levels resulting in high expression ratios due to low expression in non lens samples. Preferential/differential expression of genes with higher ratio values was more readily confirmed. We tested all the potentially differentially regulated genes by semi-quantitative RT-PCR. However, in order to determine the maximal number of genes preferentially expressed in the lens compared to non-lens samples, testing of all genes in all of the clusters would have to be carried out and this should yield significantly more genes preferentially expressed in the lens.

Genes preferentially expressed in the lens

Several genes known to be preferentially expressed in the lens were identified using clustering on the microarray data, including Mip, Lenep, Col4a1 and Cryba1 (Figure 6). The identification of these genes using clustering and the confirmation of their preferential expression in the lens using semi-quantitative RT-PCR also help to confirm the utility of the methodological approaches we have utilized here. Several additional known and uncharacterized genes were also confirmed as preferentially expressed in the lens using these approaches. Ceramide kinase (Cerk) belongs to a new class of lipid kinases that are distinct from sphingosine and diacylglycerol kinases [30]. Thus, the lens may provide a particularly amenable system in which to determine the physiological roles of Cerk in intracellular signaling. Lysyl-oxidase-like (Loxl) requires activation by bone morphogenetic protein-1 (BMP-1) before it becomes functionally active in cross-linking of collagen and elastin substrates [31]. BMPs have a role in lens fiber cell differentiation [32,33], suggesting that Loxl may also have a role in this process. CD24a is expressed during the maturation of several hematopoietic lineages [34] and its expression is high in immature precursor cells and low or absent in terminally differentiated cells, suggesting a potential role for this factor in lens fiber cell differentiation. It will be interesting to ascertain the expression patterns of this gene in the lens using in situ hybridization. It may be expected, based on the above, that expression of CD24a would be highest in the lens epithelial and/or equatorial stem cell population rather than the terminally differentiating lens fiber cells. The dead ringer homologue 2 (Dri2) encodes a protein with homology to the retinoblastoma-binding proteins Rbp1 and Rbp2, to the DNA-binding protein Bright (a B-cell-specific trans-activator) and the Drosophila dead ringer gene product [35]. Since knockouts of the retinoblastoma gene (Rb1) result in a block in the lens fiber cell differentiation pathway [25], this implies that Dri2 may contribute to the transcriptional regulation of genes involved in differentiation and tissue-specific expression in the lens. Finally, Pin4 (parvulin) has peptidyl-prolyl cis/trans isomerase has been implicated in cell cycle and transcriptional control and is involved in protein folding [36,37].

Genes up-regulated in the 7 day lens

Using K-means clustering on the lens samples alone, genes were also identified which had potentially higher expression in the 7 day old lens samples compared to either of the newborn or adult lens samples. Differential expression of 9 out of 11 of these genes was confirmed using semi-quantitative RT-PCR. It was noteworthy that no clusters were identified that exhibited significantly higher expression in either of the newborn or adult lenses compared to each other. Since we investigated gene expression using whole lenses, this is perhaps not surprising because a cross section through a lens represents a spectrum of stages of maturation of cells. In the adult lens, only the epithelial cells and the outermost secondary lens fiber cells would be expected to be transcribing RNA, since lens fiber cells lose their nuclei during differentiation. The expression profile of adult lenses compared to newborn lenses is not significantly different according to our clustering analysis, perhaps because a newborn lens will also be transcribing its RNA in epithelial cells and newly formed lens fiber cells. However, there is a significant amount of increase in the size of the mouse lens from newborn to 7 days (Unpublished data). This increase in growth will be brought about by an increase in lens epithelial cell proliferation and, therefore, an increase in the differentiation of a larger number of lens epithelial cells into secondary lens fiber cells. Thus, it is a reasonable suggestion that genes expressed at high levels in the 7 day lens, compared to the adult and newborn lenses, will potentially have important roles in lens growth and differentiation. However, conclusive evidence of this would require more detailed microarray analyses involving comparisons of labeled RNA extracted from compartments of individual lenses of the same age (central epithelial cells, proliferating epithelial cells, equatorial stem cells, outer lens fiber cells, and lens fiber cells losing nuclei and organelles). Such microarray studies would have to be correlated with investigations into relative rates of lens epithelial proliferation and differentiation into lens fiber cells during the early post-natal stages of mouse lens maturation.

Of the genes differentially regulated within the lens samples alone, Werner syndrome (OMIM entry 277700) homologue is a DNA helicase and Werner syndrome (a premature ageing disorder, [38]) is accompanied by cataract. The Werner protein is thought to act by resolving structural alterations to DNA that occur during recombination, replication and repair or during DNA damage. The up-regulation of expression of the Werner syndrome homologue here in the 7 day old mouse lens suggests that these functions may be particularly important during the early growth phase and of the lens. The EST (AI448302) with 74% amino acid identity to a human hypothetical protein (KIAA0543) containing two zinc fingers and a krueppel-associated box could represent a novel transcription factor. If this gene is indeed a zinc finger transcription factor, it may be an up-stream regulator of many other genes expressed in the lens. Moreover, zinc finger proteins expressed in the mouse lens have been identified as candidates for cataract [39]. Gtpbp2 is a novel GTP binding protein [40], suggesting roles for this signaling molecule in intracellular signaling cascades involved in lens growth and differentiation. The EST (AI428498) matching DNA segment, Chr 15, ERATO Doi 747, expressed with 80% amino acid identity to a putative human G-Protein coupled receptor (FLJ10060) could also be involved in such signalling pathways. Rab6-interacting protein 2 belongs to a family of Rab6 GTPases regulating intracellular transport at the level of the Golgi complex [41]. Interestingly, in Lowe Oculocerebrorenal Syndrome (OMIM entry 309000), vesicular transport in the Golgi apparatus is disrupted and these defects lead to cataracts in Lowe syndrome patients [42], suggesting an important role for efficient intracellular transport at the level of the Golgi in the maintenance of lens transparency. Thus, further investigation of the role of Rab6-interacting protein 2 in the regulation of intracellular transport in the Golgi in lens cells may lead to new insights into the role of the Golgi in lens maturation and cataractogenesis. Fibromodulin (OMIM entry 600245) is a member of a family of small interstitial proteoglycans and it participates in the assembly of the extracellular matrix through its ability to regulate collagen fibrillogenesis [43]. During these early post-natal growth phases it would be expected that there would be rapid turnover of lens capsule components as the lens capsule is remodeled to expand as the lens grows. The integrity of the lens capsule extracellular matrix is essential for lens transparency, as highlighted by the Sparc null mouse [8,9] in which the lens capsule is weakened to the extent that it eventually bursts [44]. Indeed, alterations in collagen fibrillogenesis were observed in fibromodulin knockout mice, although no lens defects were noted [45]. The gene encoding the nicotinic acetylcholine receptor-associated 46 kDa protein was also up regulated in the 7 day old lens [46]. Since nicotine use is an important risk factor for cataract [47], it is likely that further investigations into the physiological role of nicotinic signaling pathways during lens development and maturation may lead to important insights into the role of these endogenous pathways in nicotine-induced cataract. Finally, roles for the remaining two uncharacterized genes identified (AI452115 ESTs and AI450802 4930429Rik) remain to be determined.

Apoptosis gene expression in the lens

A transparent lens requires the programmed removal of lens fiber nuclei and organelles emanating from the center to the periphery of the lens throughout life. Our previous work and that of others has demonstrated that this phenomenon has many molecular and morphological features in common with apoptosis, providing a uniquely synchronous system for the study of apoptosis in a developmental context [48-50]. We and others have also demonstrated that pro- and anti-apoptotic genes (tumor necrosis factor (TNF), bcl-2 family members, and caspases) are expressed in the lens and are implicated in lens fiber cell denucleation [51-54]. Recent work implicates a mitochondrial permeability transition involving caspase-9, apaf-1, and cytochrome c in this process [55] and the NEIBank project has revealed that a number of genes involved in apoptosis are expressed in the adult human lens [29].

Apoptosis contributes towards cataract [56]. It is involved in anterior polar cataracts [57] and in TGF-β induced cataract [58]. Since lens fiber cell denucleation continues throughout life, we suggest that age-onset cataracts could also result from disruption in the apoptotic-like mechanisms in lens fiber cells, resulting in aberrant nuclear degradation and lens opacity. Up-regulation of apoptosis gene expression in lens epithelial cells is also implicated in cataract progression [7] and recently it has been demonstrated that there are differences in caspase activity between lens epithelial cells from different types of cataract [59].

We demonstrate here using the arrays and semi-quantitative RT-PCR that a number of additional apoptosis-related genes are expressed in the lens. A greater understanding of the role of these in lens development, maturation and ageing will be essential for gaining important insights into their potential roles in cataract. The protein encoded by the Cradd gene has the ability to bind caspase-2 through its NH-2 caspase homology domain and TNF receptor interacting protein (RIP) through its COOH-terminal death domain [60]. Cradd binds to the pro-domain of caspase-2 and recruits it into the apoptotic-signaling complex [61]. We have previously reported that caspases are expressed in the chick embryo lens and that an inhibitor of caspase-2 (z-VDVAD-FMK) inhibited nuclear degeneration by approximately 60% in lentoids formed by the differentiation of chick lens epithelial cells in culture [52]. This observation suggests an important role for this pathway in the denucleation of lens fiber cells. Casp6 has also been identified here in the lens. In our previous work [52], we demonstrated that a peptide inhibitor of Casp6 (Z-VEID-FMK) reduced the number of TUNEL-labeled nuclei per unit area of differentiated lens epithelial cell culture. The Casp6 inhibitor reduced labeling by the same extent (70%) as two universal caspase inhibitors (Boc-D-FMK and Z-VAD-FMK). Thus, Casp6 appears to be a pivotal executioner caspase during lens fiber cell denucleation at least in chick and, given that its expression in the lens has been confirmed here, its role in mouse lens development merits further attention.

We also suggested previously that members of the tumor necrosis factor (TNF) family of ligands and receptors may be involved in the denucleation process in secondary lens fibers [51]. Here, Ltb (a TNF ligand family member) was identified in the lens using the arrays and its expression was confirmed using semi-quantitative RT-PCR. The Ltb receptor is capable of promoting apoptosis [62]. Moreover, the protein product of another gene identified here, Tiaf1, is able to inhibit TNF mediated apoptosis [63] and human TIAF1 was also identified in the NEIBank lens project [29]. Pea 15 has recently been shown to be involved in apoptosis induced by TNF-related apoptosis inducing ligand [64] and could perhaps serve a similar role in modulating TNF family member apoptotic pathways in the lens. These data support the suggestion that TNF signaling pathways exist in the lens and may be important in lens fiber cell differentiation and/or denucleation [51].

Gilz encodes a transcription factor involved in the regulation of apoptosis [65]. Furthermore, Gilz is able to bind to and inhibit the action of the transcription factor AP-1 [66]. AP-1 is one of the major transcription factors involved in regulation of crystallin expression in the lens [67], suggesting that, in addition to a potential role in apoptosis, Gilz may have a role in regulating crystallin expression during lens fiber cell differentiation. Pmaip1 (Noxa) is a bcl-2 family member containing only bcl-2 homology-3 (BH-3) domains and is a mediator of p53 dependent apoptosis [68], an interesting observation given the role of p53 in apoptosis during lens development [69].

Fsp27 was initially isolated as a factor up-regulated during adipocyte differentiation [70] and its human homologue is CIDE-3 [71]. It contains CAD domains present in proteins implicated in post-mortem DNA fragmentation and also contains a CIDE-N domain (found in caspase-activated (CAD) nuclease). This domain induces DNA fragmentation and chromatin condensation during apoptosis, and it is present in the cell death activator proteins CIDE-A and CIDE-B, which are inhibitors of CAD nuclease. Finally, the expression in the lens of two anti-apoptotic factors has been confirmed here. Api5 is a gene encoding a protein that possesses a leucine zipper domain and which prevents apoptosis upon growth factor withdrawal [72], while Stambp is a essential for protection from apoptosis in neuronal cells [73] and may have a similar role in the lens.

Hemoglobin expression in the lens

A fascinating and unexpected result of these studies was that isotypes of hemoglobin (Hba-a1, Hba-X, Hbb-b1, Hbb-b2, and Hbb-Y) were shown to be expressed in the lens. We have recently confirmed expression of hemoglobin in mouse lenses during additional microarray experiments ongoing in our laboratory (Unpublished data). Furthermore, we have also demonstrated expression in the mouse lens of a chaperone (erythroid associate factor; eraf), which is important for the correct folding of alpha-hemoglobin [74] (Unpublished data).

A major question raised by these results is what role(s) could hemoglobins have in lens physiology and maturation if they are also expressed at the protein level? We are presently considering several possibilities. The lens has recruited stress proteins and metabolic enzymes as crystallins through evolution; this is called gene-sharing [75]. Therefore, hemoglobin may have been recruited by the lens as a crystallin. Another possibility is a role for hemoglobin in lens iron homeostasis. A chelatable pool of potentially redox-active iron is present at increased concentrations in human cataractous lenses [76]. Transition metal-mediated hydroxyl radical production may play a role in the etiology of age-related nuclear cataract by damaging lens proteins [77]. The major regulator of free iron concentration in the lens is ferritin and ferritin light chain 1 was one of the 50 most-highly expressed genes identified here using the ResGenTM array (Table 3). Under physiological conditions, ferritin synthesis is finely regulated at the translational level by iron availability. Dysregulation of this process can result in hyperferritinaemia leading to cataract [78]. A reasonable hypothesis would be that hemoglobin may also be involved in iron chelation in the lens. Indeed, heme oxygenase 1 (an essential enzyme involved in heme catabolism; OMIM entry 141250) is induced in cultured lens epithelial cells exposed to hyperbaric oxygen and this leads to an increase in chelatable iron in the lens [79].

Hemoglobins could also have roles in the lens as oxygen transporters. Age-related cataract may result from a reduction in the transport efficiency of metabolites within the lens [80]. Increased oxidation of nuclear proteins may result from reduced efficiency of oxygen transport mechanisms in removing oxygen from the center of the lens during ageing and cataract progression. This would lead to increases in half-lives of reactive oxygen species and to increased post-translational modification of proteins in the core of the lens. Reduced efficiency of transport mechanisms may also prevent an adequate influx of antioxidants to the lens interior, thereby compounding this problem.

Another possibility is a role for hemoglobin as an oxygen "sink". The lens is an avascular, anoxic structure and is supremely adapted to low oxygen concentrations. Indeed, hyperbaric oxygen therapy causes cataract [81] and recent work suggests that the vitreous body normally protects the lens from opacification as a result of excessive oxygen reaching it from the retina [82]. Any reduction in hemoglobin in the lens during ageing or due to mutations in, or post-translational modifications to, up-stream genes that act to maintain hemoglobin expression may lead to a reduced ability to buffer oxygen, leading to production of reactive oxygen species and further post-translational protein modification in the lens leading to age-onset cataract.

Most interestingly, recent work reveals that hemoglobin is pro-apoptotic in cancer cell lines, acting through suppression of bcl-2 and activation of caspases [83]. It is therefore possible that hemoglobin has a pro-apoptotic role in the lens. Indeed, one proposal for the initiation of lens fiber cell denucleation is that, given a decreasing oxygen gradient from the periphery to the center of the lens, the sensing by lens fiber cells of a crucial oxygen concentration may be the trigger for denucleation [84]. Sequestration of oxygen by hemoglobin could influence oxygen concentrations in lens fiber cells, thereby influencing denucleation. It may be significant that during erythrocyte maturation denucleation occurs in most species (but not in birds) as erythroblasts differentiate into mature erythrocytes. This process is analogous to that occurring in lens fiber cell differentiation [49] and could implicate hemoglobin in both processes in which denucleation is occurring. Furthermore, cataracts have been associated with sickle cell anemia [85] and thalassaemia [86], both of which are caused by mutations in hemoglobin, although there is no evidence as to whether the cataracts observed are a direct effect of hemoglobin mutations or indirect effects of the disease the mutations cause.

Expression of additional uncharacterized genes in the lens

Expression in lenses at various ages of a selection of uncharacterized genes from the microarray, in addition to those previously described, was confirmed using semi-quantitative RT-PCR (Figure 10). We found no similarity to anything in the database using blastx with the contig of the UniGene cluster Mm.24263 containing the EST AI661325. AI465543 in UniGene cluster Mm.182328 representing the RIKEN cDNA 4921516M08 gene has no similarity to known genes. AI426191 in UniGene cluster Mm.27900 (Mus musculus, clone IMAGE:1379146, mRNA, partial cds) had no similarity to anything in the database using blastx. Similarly, blastx of the contig from UniGene cluster Mm.27981 containing the EST AI426997 showed no similarity to anything in the database and neither did blastx of the contig of Mm.23689 (Expressed sequence AW555464) containing AI425997.

AI323902 (in UniGene cluster Mm.46397) represents the RIKEN cDNA 3100002B05 and when blastx was carried out using this nucleotide sequence it was revealed to have 20% identity (40% positive) to Vps53p (required for vacuolar protein sorting; Saccharomyces cerevisiae). This gene encodes a protein involved in protein sorting in the yeast late Golgi compartment [87]. It is of interest that the aquaporins (a family of water channel proteins) are delivered to distinct sites in epithelia [88] and defects in the trafficking and sorting of aquaporins have been shown to result in dominantly inherited cataracts [89]. Thus, the structure and function in the lens of this uncharacterized gene with a potential role in protein sorting merits further attention.

AI414484 in UniGene cluster Mm.11535 representing RIKEN cDNA 0710008A13 has 22% identity to Bmp2-induced gene (Mus musculus) and 22% identity to apoptotic protease activating factor 1 (apaf-1; Rattus norvegicus). BMP-2 has recently been shown to have a role in the initiation of lens fiber cell differentiation [32], thus the structure and function of this potential downstream BMP-2 induced gene in the lens is of interest. A potential involvement for this gene (considering its similarity to rat apaf-1) in BMP-2 induced apoptosis in the lens would be worth investigating.

AI451475 is in UniGene cluster Mm.24729 representing ESTs, Moderately similar (88%) to A55163 nucleolar protein Nopp140 (R. norvegicus) and 93% identical to human Dyskerin (OMIM entry 300126). Nopp140 shuttles between the nucleolus and the coiled bodies and thereby chaperones the transport of other molecules [90]. Thus, we have identified another gene potentially involved in important molecular transport processes in the lens.

Blastx of the EST AI464327 revealed 51% identity to mouse tumorous imaginal discs homolog precursor (htid-1). It is very interesting that tumorous imaginal discs 1 gene products are mitochondrial proteins, which can modulate apoptotic signal transduction within the mitochondrial matrix [91] suggesting that this EST may represent a novel gene involved in the regulation of the apoptotic pathway in the lens.

Expression of additional known genes in the lens

Expression of a selection of known genes was also confirmed using semi-quantitative RT-PCR (Figure 11). The ATPase, class V, type 10A gene encodes a potential phospholipid-transporting ATPase [92]. These authors point out that normal cells contain an unequal distribution of the amino-phospholipid phosphatidylserine (PS) across their plasma membrane lipid bilayers with the inner leaflet containing more PS than the outer. During apoptosis, PS becomes localized to the outer leaflet of the plasma membrane [93] where it acts as a recognition signal for phagocytes that digest apoptotic cells. As discussed above, secondary lens fiber cells undergo a process very similar to apoptosis when they lose their nuclei and other organelles. However, we have demonstrated previously using fluorescein conjugated annexin-V (which binds to PS) [51] that the loss of nuclei in lentoids in culture is not accompanied by the flipping of PS from the inner to the outer leaflet of the membrane. This suggests that there are important differences between classical apoptosis and the molecular events of lens fiber cell differentiation with regards to cell surface changes. The ATPase, class V, type 10A could therefore be important in the lens in ensuring that PS remains sequestered in the inner leaflet of the lens fiber cell plasma membrane during differentiation despite the apoptotic-like events that are occurring in these cells at this time.

CD34 (OMIM entry 142230) is a sialomucin-like adhesion molecule that is expressed on a few percent of primitive bone marrow cells. CD34 plays an important role in the formation of progenitor cells during both embryonic and adult hematopoiesis as revealed by CD34 gene knockout mice [94]. It would be interesting to see if this gene is associated with the lens epithelial stem cell population in the equatorial region of the lens in which lens epithelial cells continuously develop into lens fiber cells throughout life.

Serine/threonine kinase 23 belongs to a family of kinases including SRPK1, which is responsible for phosphorylation of splicing factors during the cell cycle in vivo [95]. Tissue-specific alternative splicing of transcripts is very important in expanding the repertoire of gene functions [96]. Therefore, investigation of the potential involvement of this factor in alternative splicing in the lens would be interesting, particularly given reports that alternative splicing of transcripts is significant in the lens [29].

Coatomer protein complex, subunit zeta 2 is a component of coated vesicles that are involved in vesicular trafficking in the early secretory pathway [97]. Thus, as discussed previously, since trafficking and transport within the lens is very important the function of these factors in the lens merits further attention.

Finally, expression of two factors involved in cell proliferation has been confirmed using semi-quantitative RT-PCR. GTP binding protein 3 may be involved in regulating guanosine triphosphate-dependent nuclear events that are associated with cell proliferation and which may be important in growth and development [98]. PCTAIRE-motif protein kinase 3, along with the related molecule PCTAIRE-1 is a member of a novel subfamily of cdc2/CDC28-related protein kinases [99]. The protein encoded by this gene may be important, therefore, in controlling rates of cell proliferation and cell growth in the lens.


We have provided an overview of gene expression in the mouse lens at various ages using microarrays. Using semi-quantitative RT-PCR, we have confirmed expression in the lens of all of the known and uncharacterized genes that we elected to follow up from the microarray experiments. Thus, these data are an excellent source of knowledge about the expression of known and novel genes in the lens and are a rich resource for researchers with interests in the roles in the lens of the particular classes of genes identified here.

Such a study as this can only provide a "snap-shot" of the complete pattern of gene expression in the lens during development and in the adult, but this work has generated a great deal of data which can be used to generate hypotheses regarding the roles of particular genes in the lens. Indeed, some new hypotheses have been developed here. This is particularly relevant to our interest in the commonalities between lens maturation and apoptosis. The present study supports the suggestion that the lens may represent a particularly amenable system in which to investigate the apoptotic signaling cascade.

It is anticipated that these data will be a useful resource to the lens research community as a whole and that this knowledge will be complementary to studies investigating gene expression in human lenses and other model species such as chick. We anticipate that improved knowledge of the profile of gene expression in the mouse lens will lead to insights into the roles of these genes in lens development, maturation, ageing and cataractogenesis.

Our future studies will include more detailed investigations into the patterns of expression of genes in the lens identified here (with particular emphasis on apoptosis genes and the hemoglobins) as well as functional studies, including the generation of targeted and/or transgenic mice. In the long term, such studies may lead to the development of pharmaceutical and/or gene therapy based approaches for treatment of cataract in humans.


MAW was funded by a post-doctoral fellowship from the Alberta Heritage Foundation for Medical Research (AHFMR) while in the Rancourt laboratory and by the Wellcome Trust while in the Evans laboratory. FCM was funded by post-doctoral fellowships from the AHFMR, Alberta Cancer Board (ACB) and the Canadian Institutes of Health Research (CIHR) while in the Rancourt laboratory and by the Wellcome Trust while in the Evans laboratory. DER is an AHFMR scholar. This work was supported by grants to DER from CIHR, AHFMR, and NIH, to MJE from the Wellcome Trust and to MAW from The Royal Society and National Eye Research Centre. Some of this work was previously presented at the European Vision and Eye Research meeting in October 2002 in Alicante, Spain (Wride MA, Mansergh FC, Everitt R, Minnema SE, Rancourt DE, Evans MJ. Expression profiling and gene discovery in the mouse lens using microarrays. Ophthalmic Research 2002; 34(S1):29).


1. Francis PJ, Berry V, Moore AT, Bhattacharya S. Lens biology: development and human cataractogenesis. Trends Genet 1999; 15:191-6.

2. Favor J, Neuhauser-Klaus A. Saturation mutagenesis for dominant eye morphological defects in the mouse Mus musculus. Mamm Genome 2000; 11:520-5.

3. Wilson AS, Hobbs BG, Speed TP, Rakoczy PE. The microarray: potential applications for ophthalmic research. Mol Vis 2002; 8:259-70 <>.

4. Chauhan BK, Reed NA, Zhang W, Duncan MK, Kilimann MW, Cvekl A. Identification of genes downstream of Pax6 in the mouse lens using cDNA microarrays. J Biol Chem 2002; 277:11539-48.

5. Chauhan BK, Zhang W, Cveklova K, Kantorow M, Cvekl A. Identification of differentially expressed genes in mouse Pax6 heterozygous lenses. Invest Ophthalmol Vis Sci 2002; 43:1884-90.

6. Chauhan BK, Reed NA, Yang Y, Cermak L, Reneker L, Duncan MK, Cvekl A. A comparative cDNA microarray analysis reveals a spectrum of genes regulated by Pax6 in mouse lens. Genes Cells 2002; 7:1267-83.

7. Nakajima T, Nakajima E, Fukiage C, Azuma M, Shearer TR. Differential gene expression in the lens epithelial cells from selenite injected rats. Exp Eye Res 2002; 74:231-6.

8. Gilmour DT, Lyon GJ, Carlton MB, Sanes JR, Cunningham JM, Anderson JR, Hogan BL, Evans MJ, Colledge WH. Mice deficient for the secreted glycoprotein SPARC/osteonectin/BM40 develop normally but show severe age-onset cataract formation and disruption of the lens. EMBO J 1998; 17:1860-70.

9. Norose K, Clark JI, Syed NA, Basu A, Heber-Katz E, Sage EH, Howe CC. SPARC deficiency leads to early-onset cataractogenesis. Invest Ophthalmol Vis Sci 1998; 39:2674-80.

10. Khatri P, Draghici S, Ostermeier GC, Krawetz SA. Profiling gene expression using onto-express. Genomics 2002; 79:266-70.

11. Dysvik B, Jonassen I. J-Express: exploring gene expression data using Java. Bioinformatics 2001; 17:369-70.

12. Quackenbush J. Computational analysis of microarray data. Nat Rev Genet 2001; 2:418-27.

13. Minassian DC, Reidy A, Desai P, Farrow S, Vafidis G, Minassian A. The deficit in cataract surgery in England and Wales and the escalating problem of visual impairment: epidemiological modelling of the population dynamics of cataract. Br J Ophthalmol 2000; 84:4-8.

14. Bain G, Ray WJ, Yao M, Gottlieb DI. Retinoic acid promotes neural and represses mesodermal gene expression in mouse embryonic stem cells in culture. Biochem Biophys Res Commun 1996; 223:691-4.

15. Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, Wagner EF, Metcalf D, Nicola NA, Gough NM. Myeloid leukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature 1988; 336:684-7.

16. Devore J, Peck R. Statistics: the exploration and analysis of data. St Paul: West Publishing Company; 1986. p 86.

17. Everitt R, Minnema SE, Wride MA, Koster CS, Hance JE, Mansergh FC, Rancourt DE. RED: the analysis, management and dissemination of expressed sequence tags. Bioinformatics 2002; 18:1692-3.

18. Halgren RG, Fielden MR, Fong CJ, Zacharewski TR. Assessment of clone identity and sequence fidelity for 1189 IMAGE cDNA clones. Nucleic Acids Res 2001; 29:582-8.

19. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor (NY): Cold Spring Harbor Press; 1989.

20. Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach J, Ansorge W, Ball CA, Causton HC, Gaasterland T, Glenisson P, Holstege FC, Kim IF, Markowitz V, Matese JC, Parkinson H, Robinson A, Sarkans U, Schulze-Kremer S, Stewart J, Taylor R, Vilo J, Vingron M. Minimum information about a microarray experiment (MIAME)-toward standards for microarray data. Nat Genet 2001; 29:365-71.

21. Farjo R, Yu J, Othman MI, Yoshida S, Sheth S, Glaser T, Baehr W, Swaroop A. Mouse eye gene microarrays for investigating ocular development and disease. Vision Res 2002; 42:463-70.

22. Marinissen MJ, Gutkind JS. G-protein-coupled receptors and signaling networks: emerging paradigms. Trends Pharmacol Sci 2001; 22:368-76.

23. Natsume T, Koide T, Yokota S, Hirayoshi K, Nagata K. Interactions between collagen-binding stress protein HSP47 and collagen. Analysis of kinetic parameters by surface plasmon resonance biosensor. J Biol Chem 1994; 269:31224-8.

24. Bagchi M, Ansari SA, Lindenmuth DM, van Wijnen AJ, Lian J, Stein JL, Stein GS. Nuclear matrix associated DNA-binding proteins of ocular lens epithelial cells. Mol Biol Rep 1998; 25:13-9.

25. Graw J. Genetic aspects of embryonic eye development in vertebrates. Dev Genet 1996; 18:181-97.

26. Potts JD, Kornacker S, Beebe DC. Activation of the Jak-STAT-signaling pathway in embryonic lens cells. Dev Biol 1998; 204:277-92.

27. Chen Q, Dowhan DH, Liang D, Moore DD, Overbeek PA. CREB-binding protein/p300 co-activation of crystallin gene expression. J Biol Chem 2002; 277:24081-9.

28. More MI, Kirsch FP, Rathjen FG. Targeted ablation of NrCAM or ankyrin-B results in disorganized lens fibers leading to cataract formation. J Cell Biol 2001; 154:187-96.

29. Wistow G, Bernstein SL, Wyatt MK, Behal A, Touchman JW, Bouffard G, Smith D, Peterson K. Expressed sequence tag analysis of adult human lens for the NEIBank Project: over 2000 non-redundant transcripts, novel genes and splice variants. Mol Vis 2002; 8:171-84 <>.

30. Sugiura M, Kono K, Liu H, Shimizugawa T, Minekura H, Spiegel S, Kohama T. Ceramide kinase, a novel lipid kinase. Molecular cloning and functional characterization. J Biol Chem 2002; 277:23294-300.

31. Borel A, Eichenberger D, Farjanel J, Kessler E, Gleyzal C, Hulmes DJ, Sommer P, Font B. Lysyl oxidase-like protein from bovine aorta. Isolation and maturation to an active form by bone morphogenetic protein-1. J Biol Chem 2001; 276:48944-9.

32. Belecky-Adams TL, Adler R, Beebe DC. Bone morphogenetic protein signaling and the initiation of lens fiber cell differentiation. Development 2002; 129:3795-802.

33. Faber SC, Robinson ML, Makarenkova HP, Lang RA. Bmp signaling is required for development of primary lens fiber cells. Development 2002; 129:3727-37.

34. Wenger RH, Rochelle JM, Seldin MF, Kohler G, Nielsen PJ. The heat stable antigen (mouse CD24) gene is differentially regulated but has a housekeeping promoter. J Biol Chem 1993; 268:23345-52.

35. Numata S, Claudio PP, Dean C, Giordano A, Croce CM. Bdp, a new member of a family of DNA-binding proteins, associates with the retinoblastoma gene product. Cancer Res 1999; 59:3741-7.

36. Shaw PE. Peptidyl-prolyl isomerases: a new twist to transcription. EMBO Rep 2002; 3:521-6.

37. Scholz C, Rahfeld J, Fischer G, Schmid FX. Catalysis of protein folding by parvulin. J Mol Biol 1997; 273:752-62.

38. Shen J, Loeb LA. Unwinding the molecular basis of the Werner syndrome. Mech Ageing Dev 2001; 122:921-44.

39. Danciger M, Brady J, Adamson MC, Farber DB, Piatigorsky J, Kozak CA. Chromosomal localization of the genes for five zinc finger proteins expressed in mouse lens. Genomics 1995; 28:39-43.

40. Watanabe M, Yoshida K, Hida M, Kato H, Uchida K, Yamaguchi R, Tateyama S, Sugano S. Cloning, expression analysis, and chromosomal mapping of GTPBP2, a novel member of the G protein family. Gene 2000; 256:51-8.

41. Monier S, Jollivet F, Janoueix-Lerosey I, Johannes L, Goud B. Characterization of novel Rab6-interacting proteins involved in endosome-to-TGN transport. Traffic 2002; 3:289-97.

42. Suchy SF, Olivos-Glander IM, Nussabaum RL. Lowe syndrome, a deficiency of phosphatidylinositol 4,5-bisphosphate 5-phosphatase in the Golgi apparatus. Hum Mol Genet 1995; 4:2245-50.

43. Ezura Y, Chakravarti S, Oldberg A, Chervoneva I, Birk DE. Differential expression of lumican and fibromodulin regulate collagen fibrillogenesis in developing mouse tendons. J Cell Biol 2000; 151:779-88.

44. Yan Q, Clark JI, Wight TN, Sage EH. Alterations in the lens capsule contribute to cataractogenesis in SPARC-null mice. J Cell Sci 2002; 115:2747-56.

45. Svensson L, Aszodi A, Reinholt FP, Fassler R, Heinegard D, Oldberg A. Fibromodulin-null mice have abnormal collagen fibrils, tissue organization, and altered lumican deposition in tendon. J Biol Chem 1999; 274:9636-47.

46. Frail DE, McLaughlin LL, Mudd J, Merlie JP. Identification of the mouse muscle 43,000-dalton acetylcholine receptor-associated protein (RAPsyn) by cDNA cloning. J Biol Chem 1988; 263:15602-7.

47. Taylor HR. Epidemiology of age-related cataract. Eye 1999; 13(Pt 3b):445-8.

48. Dahm R. Lens fibre cell differentiation - A link with apoptosis?. Ophthalmic Res 1999; 31:163-83.

49. Wride MA. Minireview: apoptosis as seen through a lens. Apoptosis 2000; 5:203-9.

50. Bassnett S. Lens organelle degradation. Exp Eye Res 2002; 74:1-6.

51. Wride MA, Sanders EJ. Nuclear degeneration in the developing lens and its regulation by TNFalpha. Exp Eye Res 1998; 66:371-83.

52. Wride MA, Parker E, Sanders EJ. Members of the bcl-2 and caspase families regulate nuclear degeneration during chick lens fibre differentiation. Dev Biol 1999; 213:142-56.

53. Ishizaki Y, Jacobson MD, Raff MC. A role for caspases in lens fiber differentiation. J Cell Biol 1998; 140:153-8.

54. Fromm L, Overbeek PA. Inhibition of cell death by lens-specific overexpression of bcl-2 in transgenic mice. Dev Genet 1997; 20:276-87.

55. Sanders EJ, Parker E. The role of mitochondria, cytochrome c and caspase-9 in embryonic lens fibre cell denucleation. J Anat 2002; 201:121-35.

56. Nickells RW, Zack DJ. Apoptosis in ocular disease: a molecular overview. Ophthalmic Genet 1996; 17:145-65.

57. Lee EH, Wan XH, Song J, Kang JJ, Cho JW, Seo KY, Lee JH. Lens epithelial cell death and reduction of anti-apoptotic protein Bcl-2 in human anterior polar cataracts. Mol Vis 2002; 8:235-40 <>.

58. Maruno KA, Lovicu FJ, Chamberlain CG, McAvoy JW. Apoptosis is a feature of TGF beta-induced cataract. Clin Exp Optom 2002; 85:76-82.

59. Andersson M, Honarvar A, Sjostrand J, Peterson A, Karlsson JO. Decreased caspase-3 activity in human lens epithelium from posterior subcapsular cataracts. Exp Eye Res 2003; 76:175-82.

60. Ahmad M, Srinivasula SM, Wang L, Talanian RV, Litwack G, Fernandes-Alnemri T, Alnemri ES. CRADD, a novel human apoptotic adaptor molecule for caspase-2, and FasL/tumor necrosis factor receptor-interacting protein RIP. Cancer Res 1997; 57:615-9.

61. Cohen GM. Caspases: the executioners of apoptosis. Biochem J 1997; 326 (Pt 1):1-16.

62. VanArsdale TL, VanArsdale SL, Force WR, Walter BN, Mosialos G, Kieff E, Reed JC, Ware CF. Lymphotoxin-beta receptor signaling complex: role of tumor necrosis factor receptor-associated factor 3 recruitment in cell death and activation of nuclear factor kappaB. Proc Natl Acad Sci U S A 1997; 94:2460-5.

63. Chang NS, Mattison J, Cao H, Pratt N, Zhao Y, Lee C. Cloning and characterization of a novel transforming growth factor-beta1-induced TIAF1 protein that inhibits tumor necrosis factor cytotoxicity. Biochem Biophys Res Commun 1998; 253:743-9.

64. Xiao C, Yang BF, Asadi N, Beguinot F, Hao C. Tumor necrosis factor-related apoptosis-inducing ligand-induced death-inducing signaling complex and its modulation by c-FLIP and PED/PEA-15 in glioma cells. J Biol Chem 2002; 277:25020-5.

65. D'Adamio F, Zollo O, Moraca R, Ayroldi E, Bruscoli S, Bartoli A, Cannarile L, Migliorati G, Riccardi C. A new dexamethasone-induced gene of the leucine zipper family protects T lymphocytes from TCR/CD3-activated cell death. Immunity 1997; 7:803-12.

66. Mittelstadt PR, Ashwell JD. Inhibition of AP-1 by the glucocorticoid-inducible protein GILZ. J Biol Chem 2001; 276:29603-10.

67. Ilagan JG, Cvekl A, Kantorow M, Piatigorsky J, Sax CM. Regulation of alphaA-crystallin gene expression. Lens specificity achieved through the differential placement of similar transcriptional control elements in mouse and chicken. J Biol Chem 1999; 274:19973-8.

68. Oda E, Ohki R, Murasawa H, Nemoto J, Shibue T, Yamashita T, Tokino T, Taniguchi T, Tanaka N. Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosis. Science 2000; 288:1053-8.

69. Pan H, Griep AE. Temporally distinct patterns of p53-dependent and p53-independent apoptosis during mouse lens development. Genes Dev 1995; 9:2157-69.

70. Danesch U, Hoeck W, Ringold GM. Cloning and transcriptional regulation of a novel adipocyte-specific gene, FSP27. CAAT-enhancer-binding protein (C/EBP) and C/EBP-like proteins interact with sequences required for differentiation-dependent expression. J Biol Chem 1992; 267:7185-93.

71. Liang L, Zhao M, Xu Z, Yokoyama KK, Li T. Molecular cloning and characterization of CIDE-3, a novel member of the cell-death-inducing DNA-fragmentation-factor (DFF45)-like effector family. Biochem J 2003; 370:195-203.

72. Tewari M, Yu M, Ross B, Dean C, Giordano A, Rubin R. AAC-11, a novel cDNA that inhibits apoptosis after growth factor withdrawal. Cancer Res 1997; 57:4063-9.

73. Ishii N, Owada Y, Yamada M, Miura S, Murata K, Asao H, Kondo H, Sugamura K. Loss of neurons in the hippocampus and cerebral cortex of AMSH-deficient mice. Mol Cell Biol 2001; 21:8626-37.

74. Kihm AJ, Kong Y, Hong W, Russell JE, Rouda S, Adachi K, Simon MC, Blobel GA, Weiss MJ. An abundant erythroid protein that stabilizes free alpha-haemoglobin. Nature 2002; 417:758-63.

75. Piatigorsky J, Kantorow M, Gopal-Srivastava R, Tomarev SI. Recruitment of enzymes and stress proteins as lens crystallins. EXS 1994; 71:241-50.

76. Garner B, Roberg K, Qian M, Eaton JW, Truscott RJ. Distribution of ferritin and redox-active transition metals in normal and cataractous human lenses. Exp Eye Res 2000; 71:599-607.

77. Garner B, Davies MJ, Truscott RJ. Formation of hydroxyl radicals in the human lens is related to the severity of nuclear cataract. Exp Eye Res 2000; 70:81-8.

78. Roetto A, Bosio S, Gramaglia E, Barilaro MR, Zecchina G, Camaschella C. Pathogenesis of hyperferritinemia cataract syndrome. Blood Cells Mol Dis 2002 Nov-Dec; 29:532-5.

79. Padgaonkar VA, Giblin FJ, Fowler K, Leverenz VR, Reddan JR, Dziedzic DC. Heme oxygenase synthesis is induced in cultured lens epithelium by hyperbaric oxygen or puromycin. Exp Eye Res 1997; 65:435-43.

80. Truscott RJ. Age-related nuclear cataract: a lens transport problem. Ophthalmic Res 2000 Sep-Oct; 32:185-94.

81. Palmquist BM, Philipson B, Barr PO. Nuclear cataract and myopia during hyperbaric oxygen therapy. Br J Ophthalmol 1984; 68:113-7.

82. Beebe D, Shui Y, Harocopos G, McKinnon M, Holekamp N, Gordon M. The vitreous body protects the lens from nuclear cataract formation. Ophthalmic Research 2002; 34(S1):89.

83. Brachat A, Pierrat B, Xynos A, Brecht K, Simonen M, Brungger A, Heim J. A microarray-based, integrated approach to identify novel regulators of cancer drug response and apoptosis. Oncogene 2002; 21:8361-71.

84. Bassnett S, Mataic D. Chromatin degradation in differentiating fiber cells of the eye lens. J Cell Biol 1997; 137:37-49.

85. Babalola OE, Danboyi P, Abiose AA. Hereditary congenital cataracts associated with sickle cell anaemia in a Nigerian family. Trop Doct 2000; 30:12-4.

86. Bloomfield SE, Markenson AL, Miller DR, Peterson CM. Lens opacities in thalassemia. J Pediatr Ophthalmol Strabismus 1978 May-Jun; 15:154-6.

87. Conibear E, Stevens TH. Vps52p, Vps53p, and Vps54p form a novel multisubunit complex required for protein sorting at the yeast late Golgi. Mol Biol Cell 2000; 11:305-23.

88. Brown D, Katsura T, Kawashima M, Verkman AS, Sabolic I. Cellular distribution of the aquaporins: a family of water channel proteins. Histochem Cell Biol 1995; 104:1-9.

89. Francis PJ, Berry V, Bhattacharya SS, Moore AT. The genetics of childhood cataract. J Med Genet 2000; 37:481-8.

90. Isaac C, Yang Y, Meier UT. Nopp140 functions as a molecular link between the nucleolus and the coiled bodies. J Cell Biol 1998; 142:319-29.

91. Syken J, De-Medina T, Munger K. TID1, a human homolog of the Drosophila tumor suppressor l(2)tid, encodes two mitochondrial modulators of apoptosis with opposing functions. Proc Natl Acad Sci U S A 1999; 96:8499-504.

92. Halleck MS, Lawler JF JR, Blackshaw S, Gao L, Nagarajan P, Hacker C, Pyle S, Newman JT, Nakanishi Y, Ando H, Weinstock D, Williamson P, Schlegel RA. Differential expression of putative transbilayer amphipath transporters. Physiol Genomics 1999; 1:139-50.

93. van den Eijnde SM, Luijsterburg AJ, Boshart L, De Zeeuw CI, van Dierendonck JH, Reutelingsperger CP, Vermeij-Keers C. In situ detection of apoptosis during embryogenesis with annexin V: from whole mount to ultrastructure. Cytometry 1997; 29:313-20.

94. Cheng J, Baumhueter S, Cacalano G, Carver-Moore K, Thibodeaux H, Thomas R, Broxmeyer HE, Cooper S, Hague N, Moore M, Lasky LA. Hematopoietic defects in mice lacking the sialomucin CD34. Blood 1996; 87:479-90.

95. Gui JF, Lane WS, Fu XD. A serine kinase regulates intracellular localization of splicing factors in the cell cycle. Nature 1994; 369:678-82.

96. Modrek B, Lee C. A genomic view of alternative splicing. Nat Genet 2002; 30:13-9.

97. Futatsumori M, Kasai K, Takatsu H, Shin HW, Nakayama K. Identification and characterization of novel isoforms of COP I subunits. J Biochem (Tokyo) 2000; 128:793-801.

98. Laping NJ, Olson BA, Zhu Y. Identification of a novel nuclear guanosine triphosphate-binding protein differentially expressed in renal disease. J Am Soc Nephrol 2001; 12:883-90.

99. Okuda T, Cleveland JL, Downing JR. PCTAIRE-1 and PCTAIRE-3, two members of a novel cdc2/CDC28-related protein kinase gene family. Oncogene 1992; 7:2249-58.


Wride, Mol Vis 2003; 9:360-396 <>
©2003 Molecular Vision <>
ISSN 1090-0535