Molecular Vision 2002; 8:372-388 <http://www.molvis.org/molvis/v8/a45/> Received 3 September 2002 | Accepted 8 October 2002 | Published 11 October 2002

Download Reprint

Localization of lens intrinsic membrane protein MP19 and mutant protein MP19^To3 using fluorescent expression vectors

Tong Chen, XiaLian Li, Yu Yang, Robert L. Church
 
 

Department of Ophthalmology, Emory University School of Medicine, Atlanta, GA

Correspondence to: Robert L. Church, Emory Eye Center, Rm B5601, 1365B Clifton Road, NE, Atlanta, GA, 30322; Phone: (404) 778-4101; FAX: (404) 778-2232; email: rlchurc@emory.edu
 
Dr. Li is now at the Department of Endocrinology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan, 450052, Peoples Republic of China

Abstract

Purpose: MP19 is the second most abundant major intrinsic protein of the lens fiber cell membrane. A specific heritable mutation at amino acid 15 in the MP19 protein, termed MP19^To3, results in total cataract and microphthalmia in the mouse. The goals of this study were to determine the specific localization of MP19 in the cell membrane and to determine whether the mutant MP19^To3 protein migrates to the cell membrane in a similar fashion to normal MP19.

Methods: MP19 and MP19^To3 cDNAs were cloned into two different sets of expression vectors. The first set was composed of two vectors, pEGFP-N1 and pDsRed2-N1. The first vector expressed green fluorescent protein and the second expressed a red fluorescent protein when transfected into mammalian cells. The two lens membrane protein cDNAs were separately cloned into the vectors so that the cDNA was at the 5'-end of the fluorescent protein coding DNA. These vectors expressed each of the lens proteins fused to the fluorescent protein upon transfection into mammalian cell cultures. The second vector set was a single vector, pcDNA4/TO which must be induced in the transfected cells by tetracycline in order to express the cloned cDNAs. Each of the membrane cDNAs coupled to the fluorescent protein coding region was cut out of the first vector set and cloned into pcDNA4/TO and stable clones were isolated. Each of the prepared plasmids was transfected into human and chick embryo lens epithelial cells and human T-Rex^TM-293 cells. The fluorescent cells were viewed using confocal and episcopic-fluorescence microscopy.

Results: Each of the transfected plasmids expressed fluorescent protein in all three cell lines. MP19 was observed to transport to the cell membrane. When compared to the distribution of another, separate fusion protein consisting of a signal peptide that targets to cell membranes fused to EGFP, MP19 did not distribute uniformly on the membrane, but appeared to localize into "spots" or pools of fluorescent material around the cell membrane. In contrast, MP19^To3 protein appeared to not distribute to the cell membrane; it instead appeared to collect in a particular subcellular compartment within the cell.

Conclusions: The distribution of MP19 and MP19^To3 in the cell appeared to be quite distinct. MP19 was observed to distribute to the cell membrane while MP19^To3 did not. The fact that the MP19^To3 did not traffic to the membrane, instead appearing to be trapped within a subcellular compartment within the cell sheds further light on the cause of the cataract and microphthalmia observed in the MP19^To3 mutation, and further sheds information on the pathway of MP19 transport to the cell membrane.

Introduction

MP19 is the second most abundant lens intrinsic membrane protein [1]. The N-terminal amino acid sequence of MP19 is completely conserved among a large number of mammals [1-4], and appears to be lens fiber cell specific. MP19 can be phosphorylated by protein kinase, is probably a receptor for calmodulin [4-6], and appears to associate with galectin-3 at the cell membrane [7,8]. Immunolocalization studies have identified MP19 in both junctional and non-junctional lens regions of fiber cell plasma membranes [9,10]. However, at present, the specific role of MP19 in the lens remains unknown.

In mice, a specific heritable mutation of MP19, termed MP19^To3, leads to total cataract and microphthalmia [11,12]. A single G-> T transition mutation within the first coding exon (exon 2) of the MP19 gene (termed Lim 2) results in the substitution of the normally encoded glycine with valine at amino acid 15 within the polypeptide sequence. This results in a total disruption of the development of the lens, leading to total opacity and at least 50% reduction in overall size of the lens and eye [12]. The MP19^To3 cataractous mouse mutation demonstrated that MP19 has a critical function in maintaining lens homeostasis during normal lens development. Recently [13], it was reported that a human, relatively mild, autosomal recessive presenile cataract was caused by a single amino acid substitution in the third transmembrane domain of the MP19 molecule (F105V).

The goals of this study were to determine the specific localization of MP19 in the cell membrane and determine whether the mutant MP19^To3 protein migrates to the cell membrane in a similar fashion to normal MP19.

Methods

Generation of cDNA mammalian expression vectors

MP19 and MP19^To3 cDNAs were cloned into two different classes of mammalian expression vectors. The first class was composed of two similar vectors, pEGFP-N1 and pDsRed2-N1. These vectors are in the family of Living Colors Fluorescent Protein Reporter Systems (CLONTECH, Palo Alto, CA). The pEGFP-N1 vector expressed green fluorescent protein and the pDsRed2-N1 vector expresses red fluorescent protein following transfection into mammalian cells. Both MP19 and MP19^To3 cDNA were restriction-digested by Hind III and BamH I, then inserted into the multiple cloning site (MCS) in vectors pEGFP-N1 and pDsRed2-N1. The MCS is located between the immediate early promoter of CMV and the EGFP or DsRed2 coding sequence. MP19 and MP19^To3 thus were expected to be expressed as fusions to the N-terminus of EGFP or DsRed2 if they are in the same reading frame as EGFP or DsRed2 and there are no intervening stop codons. These vectors thus were expected to express MP19 or MP19^To3 fusion protein upon transfection into mammalian cells, due to the strong CMV promoter.

The second vector class consisted of a single vector, pcDNA4/TO (Invitrogen, Carsbad, CA). MP19 and MP19^To3 cDNA, previously cloned into the pEGFP-N1 and pDsRed2-N1 vectors (above) were cut out of these vectors, with each fluorescent (green or red) protein coding region included, using Hind III and Not I, and then were cloned into pcDNA4/TO. The pcDNA4/TO vector contains two tetracycline operator 2 (TetO2) sites within the human cytomegalovirus immediate-early (CMV) promoter for tetracycline-regulated expression of the gene of interest. The Tet repressor is expressed from the pcDNA6/TR plasmid. In the absence of tetracycline, expression of MP19 or MP19^To3 is repressed by the binding of Tet repressor homodimers to the TetO2 sequences. Addition of tetracycline to the cells derepresses the hybrid CMV/TetO2 promoter in pcDNA4/TO and allows expression of MP19 and MP19^To3. Stable clones were isolated following selection with 200 μg/ml of Zeocin.

Cell culture, plasmid transfection and generation of stable cell lines

Three different cell types were used in these studies. Initially, chick embryo lens epithelial cell (CLE) cultures [14] were used. These cultures were isolated and cultured as described by Borras et al. [14]. The cultures were grown for about 1 week before transfection in order to allow the cells to reach confluence and produce lentoid bodies. These cells were used for transfections of the pEGFP-N1 and pDsRed2-N1 vectors.

The human lens epithelial (HLE) cell line SRA 01/04 [15] was also used for transfections of the pEGFP-N1 and pDsRed2-N1 vectors. DMEM media with 10% fetal bovine serum (FBS) and 2X Antibiotic-Antimycotic (PSF, Invitrogen) were used to feed the human lens epithelial cells.

Cell line T-Rex^TM-293 (Invitrogen), a human transformed primary embryonal kidney cell line, was used for transfections of all of our lens cDNA vectors. This cell line stably expresses the tetracycline (Tet) repressor. T-Rex^TM-293 cells were grown in DMEM media with 10% FBS and 2X PSF, plus an extra 5 μg/ml blasticidin was added to the cultures in order to maintain the Tet repressor in the cells.

The newly generated cDNA expression vectors were amplified and further purified using Wizard®Plus Maxipreps kits (Promega Corp., Madison, WI). The plasmids were transfected into CLE cells, T-Rex^TM-293 cells, and HLE cells using FuGene 6 (Roche Diagnostics Corporation, Indianapolis, IN).

Transfected CLE and HLE cells were observed for the presence of fluorescent product between 24 h and two weeks following transfection, using either confocal or episcopic-fluorescence microscopy (Epi-fl microscopy).

Stable clones of MP19-EGFP-pcDNA4/TO, MP19^To3-EGFP-pcDNA4/TO, MP19-DsRed-pcDNA4/TO and MP19^To3-DsRed-pcDNA4/TO were selected by the addition of 200 μg/ml zeocin to the DMEM medium. Single cell clones were picked and further grown in medium containing zeocin. Each of the expression cell lines was seeded onto glass coverslips or into tissue culture dishes. Upon reaching confluence, or at the correct cell number per plate, 5 μl of 1 mg/ml of tetracycline was added to induce the expression of recombinant protein. The fluorescent cells were viewed using confocal and Epi-fl microscopy.

Generation of Mem-pEGFP, trans Golgi-pEGFP, Mem-pDsRed2, and trans Golgi-pDsRed2 expression vectors

Two different subcellular localization expression vectors encoding fusions of EGFP or DsRed and localization signals for either the membrane or trans Golgi were constructed. The first, for localization to the cell membrane, consisted of a fusion protein composed of the N-terminal 20 amino acids of neuromodulin (GAP-43 [16]) and the green fluorescent protein (EGFP) or red fluorescent protein (DsRed2). This neuromodulin fragment contains a signal for posttranslational palmitoylation of cysteines 3 and 4 that targets to cellular membranes. PCR was used to amplify the neuromodulin peptide DNA with Hind III and BamH I restriction sites included at the 5'- and 3'- ends, respectively. This DNA was restriction-digested with the above enzymes and inserted into the multiple cloning site (MCS) in vectors pEGFP-N1 and pDsRed2-N1. The membrane localization signal thus was expected to be expressed as fusions to the N-terminus of EGFP or DsRed2.

The second localization vector consisted of a fusion protein composed of the N-terminal 81 amino acids of human β1,4-galactosyltransferase (β1,4-GT [17]), and the green fluorescent protein (EGFP) or red fluorescent protein (DsRed2). This region of human β1,4-GT contains the membrane-anchoring signal peptide that targets the fusion protein to the trans-medial region of the Golgi apparatus. As with the above membrane localization preparation, PCR was used to amplify the β1,4-galactosyltransferase peptide DNA, with Hind III and BamH I restriction sites included at the 5'- and 3'- ends, respectively. This DNA was restriction-digested with the above enzymes and inserted into the multiple cloning site (MCS) in vectors pEGFP-N1 and pDsRed2-N1. The trans Golgi localization signal thus was expected to be expressed as a fusion to the N-terminus of EGFP or DsRed2.

Generation of Double labeled cell lines

A large number of different fluorescent expression vectors were transfected into the MP19-EGFP-pcDNA4/TO and MP19^To3-EGFP-pcDNA4/TO cell lines in order to produce cell lines which expressed both green and red fluorescent proteins simultaneously. MP19-pDsRed or MP19^To3-pDsRed was transfected into MP19-EGFP-pcDNA4/TO and MP19^To3-EGFP-pcDNA4/TO cell lines. G418 was added (600 μg/ml final concentration) three days after transfection to select for DsRed positive cells. After screening for two weeks, the cells were induced with 5 μl of 1 mg/ml of tretracycline and the fluorescent cells were observed. The Mem-EGFP/Mem-DsRed2 and Golgi-EGFP/Golgi-DsRed2 vectors (described above) were also transfected into the above TO cell lines in order to compare the localization of the membrane or Golgi marker with MP19 and MP19^To3 proteins.

Microscopy

Cells were observed for the presence of expressed fluorescent proteins using both confocal microscopy and episcopic-fluorescence microscopy (Epi-fl microscopy).

Epi-fl microscopy was carried out using a Nikon Optiphot-2 upright microscope with the episcopic-fluorescence attachment EFD-3 (Nikon Inc., Melville, NY). Cells were observed using either a 40x/1.3 or 60x/1.4 oil objective. Cells growing on coverslips were either observed live or fixed with 3.7% paraformaldehyde in Phosphate-Buffered Saline (PBS; 140 mM NaCl, 8 mM Na₂HPO₄, 2 mM NaH_2PO4, pH 7.3) for 10 min at room temperature or with 100% methanol at -20 °C for 10 min. Images were photographed using a Nikon Coolpix 995 digital camera.

Confocal microscopy of living cells was carried out using the Bio-Rad MRC-1024 Confocal Microscope with a 15 mW Krypton laser and LaserSharp2000 software (Bio-Rad Life Science Group, Hercules, CA). An inverted microscope (Nikon Diaphot 300) with Hoffman modulation contrast optics (40x/0.5 water objective) was used to visualize the cells. Both single frame and Z-series (iris diameter of 2.8 mm and step size of 0.5 μm) acquisition were collected. Live cells were photographed using a Nikon Coolpix 995 digital camera.

Further analysis of the confocal and Epi-fl microscopy images was carried out using Adobe Photoshop 7.0 software (Adobe Systems Inc., San Jose, CA). The multiple images generated with Z-series acquisition were further analyzed using VoxBlast 3D visualization and Measurement software (VayTek, Inc., Fairfield, IA) and Amira 2.3 Advanced Visualization, Data analysis, and Geometry Resonctruction software (TGS Inc., San diego, CA).

DNA sequence analysis of fusion vectors

Primers were designed based upon the sequence of each of the expression plasmids pEGFP-N1, pDsRed2-N1, and pcDNA4/TO and were used to sequence the complete inserts from both directions in order to determine with complete assurance that the amino acid sequence of the expressed proteins was correct and not altered during the PCR and cloning process.

The DNA was used as a template in cycle sequencing reactions using Cy5-labeled oligonucleotide primers and the Cy5 AutoCycle Sequencing Kit (Pharmacia Biotech, Piscataway, NJ). These reactions were then electrophoresed and analyzed using an ALFexpress automated DNA sequencer (Pharmacia Biotech). DNA sequences were analyzed using BLAST at NCBI, and DNASIS sequence analysis software for Windows (Hitachi Software, San Bruno, CA).

Results & Discussion

Our initial experiment was to determine the profile of expression of EGFP and DsRed in CLE and HLE without the presence of our cDNA inserts. In Figure 1 we show the expression of EGFP and DsRed in a culture of CLE, using confocal microscopy. The cells were cotransfected with empty vectors pEGFP-N1 and pDsRed2-N1. Figure 1A shows EGFP expression in a select group of cells (green fluorescence), while Figure 1C shows DsRed expression (red fluorescence) in a different group of cells within the identical frame as Figure 1A. Figure 1B shows coexpression of both EGFP and DsRed in a small number of cells (yellow). Occasionally, there appeared to be fluorescence localized to the nucleus, however this was not typical and was attributed to there being more fluorescent protein in the central part of the cell, which is thicker in depth than the cell periphery. Using Epi-fl microscopy, this was never observed.

An essentially identical set of results were obtained using HLE as the transfected cell line. Figure 2 demonstrates that EGFP and DsRed expression could be readily observed in HLE cells. Figure 2A shows EGFP expression (green), Figure 2C shows DsRed expression (red), and Figure 2B shows coexpression of the two fluorescent signals (yellow color). In both chick and human lens cells, expression of either EGFP or DsRed alone demonstrated that these proteins remain in the cell cytoplasm, unrelated to any particular subcellular organelles. This very "smooth" appearing expression in the cytoplasm was typical of nonfusion protein expression of the marker fluorescent proteins.

In contrast, when MP19 cDNA was cloned into the vectors and expression of the MP19-EGFP fusion protein observed, the fluorescence was clearly seen at the cell membrane (Figure 3A,B, CLE and Figure 4A-C, HLE), rather than within the cytoplasm, as observed with empty vectors. This fluorescent staining, either green or red, appeared to be rather punctate in nature rather than a smooth, even staining. In both cell CLE and HLE, it appeared that MP19 was expressed and migrated to the cell membrane in a normal fashion, and this expression and membrane association appeared to not be influenced by the green or red fluorescent protein fused to the end of the MP19 protein molecule.

When HLE cells were transfected with both MP19-pEGFP-N1 and MP19^To3-pDsRed2-N1 cDNAs (Figure 5) for about 24 h, MP19 (Figure 5A, green) was again observed to transport into the cell membrane as seen in Figure 4. However, MP19^To3DsRed expressed protein, from the same cell image as in Figure 5A, was not observed to transport to the cell membrane (Figure 5C). It appeared to be sequestered somewhere inside the cell, and did not migrate to the cell membrane. This difference in migration of the two proteins, MP19 and MP19^To3, is clearly observed in the composite micrograph showing both MP19 and MP19^To3 fluorescence (Figure 5B). This result implies that the G->T transition in the To3 mutation changes the normal pathway of MP19 expression and trafficking to the membrane.

Using cells that originated from lens tissue clearly demonstrates that the procedures used to express lens membrane protein MP19 and mutant MP19 protein MP19^To3 in culture were successful. The chick lens epithelial cell culture system has been used routinely to investigate promoter function for a number of lens protein genes, including crystallins [14] and intrinsic membrane proteins [12,18]. The human lens epithelial cell line, HLE, has been shown to synthesize αA- and βB2-crystallin and aldose reductase [15]. These two cell types, therefore, served as good preparations for testing the various expression vectors.

In order to further study how MP19 traffics to the membrane, and how the MP19^To3 mutation alters this pathway, we cloned MP19 and MP19^To3 cDNAs, along with the coding regions of EGFP and DsRed, into the pcDNA4/TO vector, which allowed us to establish clonal cell lines containing the vectors of interest and which could then be induced to express the fusion proteins upon demand with addition of tetracycline. Following transfection of these plasmids into T-Rex^TM-293 cells for 3 days, 200 μg/ml Zeocin-containing medium was added and the cells were grown in this selection medium for two weeks in order to get stable, clonal cell lines. Cell line MP19greenTO, which was a T-Rex^TM-293 (293) cell line transfected with MP19-pEGFP-pcDNA4/TO, was induced with tetracycline and observed with confocal and Epi-fl microscopy about 24 h following induction. Figure 6 shows that the green fluorescence tagged MP19 trafficked to the cell membrane (Figure 6A is from confocal microscopy and Figure 6B was obtained using Epi-fl microscopy). Using both forms of microscopy, the label clearly migrates to the cell membrane, where it tends to arrange in a punctate manner. All of the three cell types used to express MP19/EGFP protein were similar in the quality of membrane expression. All three cell types appeared to express the protein on all surfaces of the cell, both apical and basal, and certainly between apposing cells. The cells varied slightly in the type of membrane expression, with the two lens cell lines demonstrating much more punctate staining than the 293 cell line. In many cases, all three cell types demonstrated variable sized "plaques" of fluorescence. In no case was there observed a smooth, diffuse localization of MP19 on the membrane.

In contrast, when cell line MP19^To3greenTO, which was a 293 cell line transfected with MP19^To3-pEGFP-pcDNA4/TO, was induced with tetracycline and observed about 24 h later, a totally different picture was observed. The MP19^To3greenTO cells expressed a protein that appeared to not traffic to the cell membrane (Figure 7D-F). Compared to the MP19greenTO line (Figure 7A-C), the MP19^To3greenTO cells appeared to trap the fluorescent protein in a subcellular compartment, possibly the Golgi. Using both confocal (Figure 7A,B,D,E) and Epi-fl (Figure 7C,F) microscopy, the fluorescent label clearly migrated to the membrane in the normal MP19-transfected cells and was clearly trapped in a subcellular compartment in the MP19^To3 transfected cells. The fluorescent signal from MP19^To3 transfected cells was usually so intense that the overall shape of the cell was obscured.

In order to demonstrate that the fluorescent signal, especially the MP19^To3 signal, concentrated in the cell, we carried out time point experiments. Both MP19greenTO and MP19^To3greenTO were seeded on coverslips and observed for fluorescent protein synthesis at various time periods following induction with tetracycline. Within 5 h after induction (Figure 8), both cell lines clearly expressed fluorescent protein. Little difference between MP19green and MP19^To3green was initially observed. However, MP19green, even at this early stage of expression, demonstrated fluorescent protein migrating to the membrane (Figure 8A) while MP19^To3green did not migrate to the cell membrane (Figure 8B). At 15 h following induction with tetracycline (Figure 9), both MP19green and MP19^To3green expression had established their "classic" phenotype, with MP19green migrating to the cell membrane (Figure 9A-D) while MP19^To3green collected in a subcellular compartment (Figure 9E-H).

In order to further demonstrate that MP19green expressed protein migrated to the membrane and that MP19^To3green did not, we used the MP19greenTO and MP19^To3greenTO cell lines and transfected each of these with the empty pDsRed2-N1 vector. This vector expressed DsRed as a cytoplasmic, soluble protein in the cells. After 48 h following transfection with pDsRed2-N1, the cells were induced with tetracycline and observed for fluorescence 24 h later. The MP19greenTO cells, which were double-transfected with pDsRed2-N1, demonstrated both red and green fluorescence (Figure 10A). All of the cells should have expressed MP19green since this is a stable cell line. Few of the cells were successfully transfected with pDsRed2-N1 since the transfection efficiency for these cells was between 10 and 20%. Those cells that coexpressed both fluorescent proteins showed up as red cells with a yellow outline, demonstrating that the green fluorescence was indeed on the membrane. The cells that expressed MP19^To3green and DsRed were dramatically different from the MP19greenTO/pDsRed2-N1 cells (Figure 10B). In this view, the cells that expressed the MP19^To3 mutant protein could only be viewed as an "inclusion body" within the cell, not the cell outline. The cells expressing both MP19^To3green and DsRed demonstrated an overall red cell with one or two inclusion yellow fluorescent signals (Figure 10B). The two cell types, MP19greenTO and MP19^To3greenTO, were obviously different in their mode of expression of normal MP19 and the mutant MP19^To3 proteins.

We next transfected MP19greenTO and MP19^To3greenTO cells with Mem-pDsRed2, a marker for the cell membrane (the N-terminal 20 amino acids of neuromodulin, GAP-43 [14], and the red fluorescent protein, DsRed2). MP19 was observed to migrate to the cell membrane (Figure 11A,B, green fluorescence) in essentially every cell. A certain number of cells were cotransfected with Mem-DsRed and were observed to colocalize with MP19 (Figure 11A,B, yellow fluorescence). In contrast, MP19^To3green was again observed to not migrate to the cell membrane, but was sequestered in a subcellular compartment (Figure 11C,D, green fluorescence). No colocalization of Mem-DsRed with MP19^To3green was observed (Figure 11C,D, red fluorescence).

In order to investigate the nature of the subcellular compartment that MP19^To3 appeared to be trapped within, MP19greenTO cells were cotransfected with MP19^To3-pDsRed2-N1 to produce cells that simultaneously expressed MP19green protein and MP19^To3red protein. Figure 12 shows the results obtained from such an experiment. Every cell expressed MP19green, which migrated normally to the cell membrane. Many cells also expressed MP19^To3red, which collected in a subcellular compartment. In many cases, the subcellular compartment was not uniformly round or oblong; it appeared to be in the form of a tubular or twisting pathway (arrow in Figure 12C and magnified view of inclusion at arrow in Figure 12E). This material appeared to continue to concentrate in the subcellular compartment for several days, building up to very large quantities and appearing to distort the cell itself (Figure 13A, arrows). It appeared that the MP19^To3 protein concentrated to the point that it could be discerned physically. In Figure 13A, the cell appeared to have two separate "bulges". From the confocal fluorescent image (Figure 13B), the MP19^To3 protein was seen to concentrate in the same area as the "bulge" in the cell. By overlaying the two images (Figure 13C), it is evident that the cellular "bulge" was due to a high concentration of MP19^To3 protein trapped within the cell.

After about 7 to 10 days following induction of MP19^To3 expression, the cells concentrated the fluorescent signal to such an extent that it was quite difficult to photograph the Epi-fl images (Figure 14A,D, large arrows). It was also observed that with extended time following induction, the fluorescent MP19^To3 protein appeared to migrate and further concentrate in the tips of the cells (Figure 14B,C, small arrows). Figure 14E shows an enlargement of one of these areas of the cell (from Figure 14C, bottom arrow) which appeared to be further concentrating the MP19^To3 protein. It can be readily seen that the protein appeared to be concentrating in a twisting pathway or tubular structure. Another set of enlarged images demonstrating the nature of these inclusions of MP19^To3 protein is shown in Figure 15. These inclusions fluoresce red due to the DsRed tag on the MP19^To3 protein. The inclusion nature of the MP19^To3 protein is further demonstrated in Figure 16. This is an Epi-fl image taken with the visible light lamp also turned on, revealing the overall cell shape as well as the red fluorescence of the MP19^To3 protein. Images A through C were taken with the focus at the top of the cell (Figure 16A) and then manually traversing through the cell in three total steps to finally focus on the MP19^To3 protein (Figure 16B,C). It is extremely difficult to keep the entire MP19^To3 protein inclusion body in focus since it is so large in size. In almost every case, the MP19^To3 protein inclusion is very close to the nucleus, usually at the elongated end of the nucleus.

Our original thought was that this MP19^To3 protein was being trapped in the Golgi of the cell. In order to further investigate the identity of the subcellular compartment into which the MP19^To3 protein was being trapped, we cotransfected MP19^To3redTo cells with the Golgi-pEGFP-N1 plasmid. This plasmid expresses green fluorescent protein fused to the N-terminal 81 amino acids of human β1,4-galactosyltransferase (β1,4-GT [17]). This region of human β1,4-GT contains the membrane-anchoring signal peptide that targets the fusion protein to the trans-medial region of the Golgi apparatus. The data obtained using confocal microscopy was, in general, rather uninformative. It appeared that a certain amount of colocalization of MP19^To3 protein with Golgi was present (Figure 17). The green fluorescence was Golgi marker while the yellow fluorescence indicated overlap between MP19^To3 protein (red) and Golgi marker (green). In many cases, the yellow fluorescence appeared to be on the side or at one end of the Golgi marker. However, another confocal image at higher magnification demonstrated a slightly different result. In Figure 18A, the green Golgi marker fluorescence appears as a rather diffuse product while the MP19^To3 protein (red, Figure 18B), taken from the same area, is a quite compact inclusion. When the two images are superimposed (Figure 18C), it is seen that MP19^To3 protein appears to partially overlap at only one end of the diffuse Golgi fluorescence. If the MP19^To3 protein inclusion was trapped in the Golgi, complete colocalization of the two fluorescent signals should be seen. The β1,4-GT marker signal does target the Golgi, however, it's main target is the trans Golgi. It is possible that MP19^To3 protein is concentrating in a part of the Golgi other than the trans Golgi region. There are several different, and apparently distinct, regions of the Golgi that might not be marked by our particular Golgi marker.

When cells were transfected only with the Golgi marker plasmid and the expressed protein observed using Epi-fl microscopy (Figure 19A,B), rather than confocal microscopy, it can be seen that the Golgi is composed of branching vesicles and cisternae (the small, orange granules observed throughout the cell are autofluorescent particles). When MP19^To3red protein and β1,4-GTGolgigreen are both expressed simultaneously, it can readily be seen that the red fluorescent MP19^To3 protein is compartmentalized very near the Golgi (green fluorescence, Figure 20), but no overlap of signal is observed. In almost every case, one, or possibly two, MP19^To3 protein inclusions were observed very close to the Golgi (Figure 21A,B), usually between the nucleus and the trans Golgi region. With Epi-fl microscopy, which shows higher resolution of the images compared with confocal microscopy, there is clearly no, or very little, colocalization of the two fluorescent signals. Again, it can be readily observed that the MP19^To3 protein signal is trapped in a type of vesicle that is not simply a "ballon" type structure, rather it collects in a structure that has a very tubular or labyrinth type structure (Figure 21C, arrowheads) which appears to fold back on itself, but which is apparently not part of the actual Golgi itself (or the trans Golgi).

MP19^To3 protein clearly does not collect in subcellular structures that are post trans-Golgi, such as peroxisomes or endosomes. From our data, MP19^To3 protein also does not collect in the trans-Golgi. Ladinsky et al., [19] recently carried out a three-dimensional reconstruction of the Golgi complex using dual-axis, high-voltage EM tomography and have determined, at least in their reconstruction, that the Golgi complex contains stacks of seven cisternae (the final three being trans in nature), bounded on both sides by two types of endoplasmic reticulum (ER), cis and trans. The cis ER lies against the ER-Golgi intermediate compartment (ERGIC), which is a set of tubulo-vesicular membranous elements layered adjacent to the cis-most Golgi cisterna. The ERGIC appears to be a major intermediate compartment between the ER and the Golgi [20-23]. The ERGIC would possibly be a suitable structure for MP19^To3 protein to be sequestered in. This subcellular compartment is large enough and its tubulo-vesicular structure would look very similar to the MP19^To3 protein inclusions observed.

At present, we have no knowlege of the form the MP19^To3 protein takes in the observed inclusion bodies. The mutant protein may be trapped in the subcellular compartment because it is misfolded or partially denatured. There are many examples in nature of proteins misfolding or aggregating within or outside of cells, thus leading to disease [24-28]. Helenius [29] demonstrated that misfolded G protein eventually made its way to the ERGIC where it remained until it was recycled back to the ER.

Another possible mechanism for MP19^To3 protein being sequestered within a subcellular compartment would involve the alteration of some type of signal that normally causes MP19 protein to be transported to the cell membrane. Possibly the To3 cataract mutation is due to the alteration of the normal signal, either changing the signal to localize to the subcellular compartment or the mutation completely ruins the normal signal, leading to compartmentalization and possible degradation. We did not observe any obvious degradation in the MP19^To3TO cells, even after two weeks of induction with tetracycline. Further work is needed to identify the nature of the compartmentalized MP19^To3 protein.

Several other mutations in lens fiber cell membrane proteins have led to cataract formation in both mouse and man (reviewed in [30]). Of these, Lop, a missense mutation in the Major Intrinsic Membrane (MIP) protein of the lens appears to affect trafficking of MIP to the cell membrane [31]. MIP^Lop protein appears to collect within the endoplasmic reticulum (ER) of the fiber cells. Another cataractous MIP mutation called Hfi is caused by a 76-bp deletion that resulted in exon 2 skipping in MIP mRNA. Again, the abnormal protein produced appeared to be trapped in the perinuclear space, possibly the ER [32]. Both of these phenotypes are similar to what occurs with MP19^To3 protein, although MP19^To3 protein does not appear to become trapped in the ER.

The nature of the To3 mutation is now relatively clear. It is probably not a loss-of-function type of mutation. It is more likely that the MP19^To3 protein builds up in the lens fiber cells until it kills the cell, a form of "constipatus terminalis", as it were. The To3 mutation would be considered a detrimental gain of function mutation, probably leading to cytotoxicity. This mutation would, therefore, not be of major use in eliciting the function of MP19 in the lens fiber cell membrane. It is interesting that the cultured cells that have been transfected with the MP19^To3 vector do not appear to have a cytotoxic reaction to the buildup of mutant protein, even after two or three weeks following induction. This is probably due to the fact that the cultured cells are continuously dividing (with a doubling time of about 18 h) and spreading the mutant protein over two cells each doubling time. Lens fiber cells do not undergo cell division since they do not have a functional nucleus and therefore the mutant protein continues to accumulate until cell death occurs.

Acknowledgements

This research was supported in part by National Institutes of Health grants R01 EY11516 and R01 EY12301 to RLC, and T32 EY07092, P30 EY06360, C06 EY06307, the Knights Templar Educational Foundation of Georgia, Inc., and a Departmental Grant from Research to Prevent Blindness, Inc. RLC is presently a Research to Prevent Blindness Senior Scientific Investigator. Some of these data were presented in abstract form at the 2002 meeting of the Association for Research in Vision and Ophthalmology.

References

1. Rao GN, Gutekunst KA, Church RL. Bovine lens 23, 21 and 19 kDa intrinsic membrane proteins have an identical amino-terminal amino acid sequence. FEBS Lett 1989; 250:483-6.

2. Church RL, Wang JH. The human lens fiber-cell intrinsic membrane protein MP19 gene: isolation and sequence analysis. Curr Eye Res 1993; 12:1057-65.

3. Louis CF, Hur KC, Galvan AC, TenBroek EM, Jarvis LJ, Eccleston ED, Howard JB. Identification of an 18,000-dalton protein in mammalian lens fiber cell membranes. J Biol Chem 1989; 264:19967-73.

4. Voorter CE, Kistler J, Gruijters WT, Mulders JW, Christie D, de Jong WW. Distribution of MP17 in isolated lens fibre membranes. Curr Eye Res 1989; 8:697-706.

5. Mulders JW, Voorter CE, Lamers C, de Haard-Hoekman WA, Montecucco C, van de Ven WJ, Bloemendal H, de Jong WW. MP17, a fiber-specific intrinsic membrane protein from mammalian eye lens. Curr Eye Res 1988; 7:207-19.

6. Louis CF, Hogan P, Visco L, Strasburg G. Identity of the calmodulin-binding proteins in bovine lens plasma membranes. Exp Eye Res 1990; 50:495-503.

7. Gonen T, Donaldson P, Kistler J. Galectin-3 is associated with the plasma membrane of lens fiber cells. Invest Ophthalmol Vis Sci 2000; 41:199-203.

8. Gonen T, Grey AC, Jacobs MD, Donaldson PJ, Kistler J. MP20, the second most abundant lens membrane protein and member of the tetraspanin superfamily, joins the list of ligands of galectin-3. BMC Cell Biol 2001; 2:17.

9. Tenbroek E, Arneson M, Jarvis L, Louis C. The distribution of the fiber cell intrinsic membrane proteins MP20 and connexin46 in the bovine lens. J Cell Sci 1992; 103:245-57.

10. TenBroek EM, Johnson R, Louis CF. Cell-to-cell communication in a differentiating ovine lens culture system. Invest Ophthalmol Vis Sci 1994; 35:215-28.

11. Steele EC Jr, Kerscher S, Lyon MF, Glenister PH, Favor J, Wang J, Church RL. Identification of a mutation in the MP19 gene, Lim2, in the cataractous mouse mutant To3. Mol Vis 1997; 3:5.

12. Steele EC Jr, Wang JH, Lo WK, Saperstein DA, Li X, Church RL. Lim2(To3) transgenic mice establish a causative relationship between the mutation identified in the lim2 gene and cataractogenesis in the To3 mouse mutant. Mol Vis 2000; 6:85-94 <http://www.molvis.org/molvis/v6/a12/>.

13. Pras E, Levy-Nissenbaum E, Bakhan T, Lahat H, Assia E, Geffen-Carmi N, Frydman M, Goldman B, Pras E. A missense mutation in the LIM2 gene is associated with autosomal recessive presenile cataract in an inbred Iraqi Jewish family. Am J Hum Genet 2002; 70:1363-7.

14. Borras T, Peterson CA, Piatigorsky J. Evidence for positive and negative regulation in the promoter of the chicken delta 1-crystallin gene. Dev Biol 1988; 127:209-19.

15. Ibaraki N, Chen SC, Lin LR, Okamoto H, Pipas JM, Reddy VN. Human lens epithelial cell line. Exp Eye Res 1998; 67:577-85.

16. Skene JH, Virag I. Posttranslational membrane attachment and dynamic fatty acylation of a neuronal growth cone protein, GAP-43. J Cell Biol 1989; 108:613-24.

17. Watzele G, Berger EG. Near identity of HeLa cell galactosyltransferase with the human placental enzyme. Nucleic Acids Res 1990; 18:7174.

18. Zhou L, Li X, Church RL. The mouse lens fiber-cell intrinsic membrane protein MP19 gene (Lim2) and granule membrane protein GMP-17 gene (Nkg7): Isolation and sequence analysis of two neighboring genes. Mol Vis 2001; 7:79-88 <http://www.molvis.org/molvis/v7/a12/>.

19. Ladinsky MS, Mastronarde DN, McIntosh JR, Howell KE, Staehelin LA. Golgi structure in three dimensions: functional insights from the normal rat kidney cell. J Cell Biol 1999; 144:1135-49.

20. Klumperman J, Schweizer A, Clausen H, Tang BL, Hong W, Oorschot V, Hauri HP. The recycling pathway of protein ERGIC-53 and dynamics of the ER-Golgi intermediate compartment. J Cell Sci 1998; 111:3411-25.

21. Schweizer A, Ericsson M, Bachi T, Griffiths G, Hauri HP. Characterization of a novel 63 kDa membrane protein. Implications for the organization of the ER-to-Golgi pathway. J Cell Sci 1993; 104:671-83.

22. Hauri HP, Kappeler F, Andersson H, Appenzeller C. ERGIC-53 and traffic in the secretory pathway. J Cell Sci 2000; 113:587-96.

23. Gilbert A, Jadot M, Leontieva E, Wattiaux-De Coninck S, Wattiaux R. Delta F508 CFTR localizes in the endoplasmic reticulum-Golgi intermediate compartment in cystic fibrosis cells. Exp Cell Res 1998; 242:144-52.

24. Chiti F, Taddei N, Bucciantini M, White P, Ramponi G, Dobson CM. Mutational analysis of the propensity for amyloid formation by a globular protein. EMBO J 2000; 19:1441-9.

25. Walsh DM, Klyubin I, Fadeeva JV, Cullen WK, Anwyl R, Wolfe MS, Rowan MJ, Selkoe DJ. Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature 2002; 416:535-9.

26. Fandrich M, Fletcher MA, Dobson CM. Amyloid fibrils from muscle myoglobin. Nature 2001; 410:165-6.

27. Ellis RJ, Pinheiro TJ. Medicine: danger--misfolding proteins. Nature 2002; 416:483-4.

28. Bucciantini M, Giannoni E, Chiti F, Baroni F, Formigli L, Zurdo J, Taddei N, Ramponi G, Dobson CM, Stefani M. Inherent toxicity of aggregates implies a common mechanism for protein misfolding diseases. Nature 2002; 416:507-11.

29. Helenius A. Quality control in the secretory assembly line. Philos Trans R Soc Lond B Biol Sci 2001; 356:147-50.

30. Graw J. Cataract mutations and lens development. Prog Retin Eye Res 1999; 18:235-67.

31. Shiels A, Bassnett S. Mutations in the founder of the MIP gene family underlie cataract development in the mouse. Nat Genet 1996; 12:212-5.

32. Sidjanin DJ, Parker-Wilson DM, Neuhauser-Klaus A, Pretsch W, Favor J, Deen PM, Ohtaka-Maruyama C, Lu Y, Bragin A, Skach WR, Chepelinsky AB, Grimes PA, Stambolian DE. A 76-bp deletion in the Mip gene causes autosomal dominant cataract in Hfi mice. Genomics 2001; 74:313-9.