Phosphatidylinositol 3-Kinase in Bovine Lens and Its Stimulation by Insulin and IGF-1

Abstract

purpose. To identify and characterize phosphatidylinositol 3-kinase (PI-3K) in the lens and to study its involvement as a signal mediator in lens epithelial cells exposed to insulin and insulin-like growth factor (IGF)-1, which are known to induce lens epithelial cell proliferation and differentiation into fiber cells.

methods. Concentric fiber cell layers from single bovine lens were prepared by dissolution in buffer. PI-3K activity in capsule-epithelium and fiber cell layers was determined after immunoprecipitation with antibodies against p85, the regulatory subunit of PI-3K. High-performance liquid chromatography on an ion exchange column (Partisil-SAX; Whatman, Maidstone, United Kingdom) was used to identify PI-3K reaction products. Cultured bovine lens epithelial cells were stimulated with insulin or IGF-1, and PI-3K activity was determined after immunoprecipitation with antibody against phosphotyrosine. Association of p85 with other proteins after stimulation was determined in anti-p85 immunoprecipitates by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and Western immunoblot analysis using anti-phosphotyrosine antibody.

results. PI-3K activity was found in both lens epithelial cells and fiber cells. The highest specific activity was found in the capsule-epithelium, but there was considerable activity in other fiber cell layers. Insulin and IGF-1 stimulated the PI-3K activity in epithelial cells in culture by more than 100%, and activation of the enzyme resulted in tyrosine phosphorylation of the p85 subunit. After stimulation, the p85 subunit of PI-3K was associated with 100- and 180-kDa tyrosine phosphorylated proteins.

conclusions. The activation of PI-3K and its association with specific tyrosine-phosphorylated proteins may be important in insulin and IGF-1 signal transduction in lens epithelial cells. The presence of significant PI-3K activity throughout the lens further suggests that this signal transduction enzyme is sustained in fiber cells.

Lens tissue is formed from a monolayer of cuboidal epithelial cells at the anterior surface that proliferate, elongate, and eventually differentiate and become fiber cells. These fiber cells constitute the main body of the lens. Insulin and insulin-like growth factor (IGF)-1 have been shown to induce lens epithelial cell proliferation in chicken, bovine, and rabbit lens and differentiation in rat and chicken lens (reviewed in Reference 1). Other growth factors, such as fibroblast growth factor and platelet-derived growth factor, have also been shown to affect proliferation and differentiation processes in rat and mice lens. ^{1 2} Growth factors bind to their tyrosine kinase receptors and generate signals in the cells that are transduced by several intracellular signal mediators. The actual molecules that participate in the growth factor–mediated signal transduction in lens cells are not known at this point. 

PI-3K is an important component in the signal transduction cascade initiated by various receptor tyrosine kinase specific growth factors. This kinase is a heterodimeric protein with a 110-kDa catalytic subunit (p110) and an 85-kDa regulatory-adaptor subunit (p85) with two Src homology (SH2) domains, ³ that interact with the cytosolic portion of the tyrosine kinase domain of the receptor. PI-3K phosphorylates phosphatidylinositol lipids at the 3′ position of the inositol ring, generating an important group of second messengers, namely, phosphatidylinositol 3-phosphate (PI-3P), phosphatidylinositol 3,4-bis phosphate (PI-3,4P₂), and phosphatidylinositol 3,4,5-triphosphate (PI-3,4,5P₃). ³ Several studies have implicated PI-3K through its products in the regulation of cell proliferation, cellular differentiation, cytoskeletal reorganization, membrane trafficking, prevention of apoptosis, glucose transport, survival and metabolism (reviewed in Reference 4). The present study shows for the first time that insulin and IGF-1 stimulate PI-3K in bovine lens epithelial cells, suggesting its role as a mediator in growth factor–dependent cellular processes such as proliferation and differentiation. We also show that PI-3K is present throughout the lens and that fiber cells retain this enzyme after their differentiation. 

Materials and Methods

Materials

Polyclonal and monoclonal antibodies for the p85 subunit of PI-3K and anti-phosphotyrosine, (anti-PY) and IGF-1 were from Upstate Biotechnology, (Lake Placid, NY). All electrophoresis reagents were from Bio-Rad, (Richmond, CA). [γ³³P]ATP, phosphatidyl [2-³H] inositol-4-mono phosphate (³H-PI-4P), and phosphatidyl[ 2-³H] inositol 4,5 bis phosphate (³H-PI-4,5P₂) were from Amersham (Arlington Heights, IL). Insulin, Wortmannin, Dulbecco’s modified Eagle’s medium (DMEM), phosphatidylinositol (PI), PI-4P and PI-4,5P₂, and antibiotic-antimycotic solutions were from Sigma (St. Louis, MO). 

Epithelial Cell Culture

Lens from calf eyes (Pel-Freeze Biologicals, Rogers, AR), were separated, the capsule with adhering epithelial cells was dissected from the lens, and primary cultures of epithelial cells were cultured, as reported ⁵ in DMEM supplemented with 10% fetal calf serum (Gibco, Grand Island, NY) and 1% antibiotic-antimycotic solution at 37°C in a 5% CO₂-95% O₂. Second-passage cells were used in all experiments. 

Preparation of Cultured Epithelial and Lens Fiber Cell Layer Extracts

Concentric fiber cell layers from a single decapsulated bovine lens (weighing approximately 2 g) were obtained by progressively dissolving the lens ⁶ with gentle and constant agitation in 3 to 5 ml lysis buffer (20 mM HEPES, 2 mM Mg ²⁺, 2 mM EGTA, 2 mM orthovanadate, 2 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 0.1 mM leupeptin, and 0.25 M sucrose, 1% Triton X-100 and 0.5% NP-40). Six layers were prepared from one lens. Layers were numbered consecutively from the periphery to the center of the lens: layer 1 represented the outermost fiber cells of the lens and layer 6 was designated as the nucleus. The capsule with adherent epithelial cells was also collected. All lens cell fractions and stimulated cultured epithelial cells were homogenized in lysis buffer using a glass homogenizer and centrifuged at 100,000g for 30 minutes, and the supernatant was used as the source of PI-3K. All the operations were performed at 4^οC. Protein content of the extracts was determined by the method of Lowry et al. ⁷  

Immunoprecipitation and Western Blot Analysis

Cultured epithelial cell extracts (0.3–0.6 mg protein) or fiber cell extracts (1–10 mg protein) were immunoprecipitated with 3 to 5μ l of anti-p85 or 3 to 5 μg of anti-PY polyclonal antibodies, as described before. ⁷ Proteins present in immunoprecipitates were either assayed for PI-3K activity or subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE; 9% gel), followed by western transfer to nitrocellulose membranes. ⁸ The transferred proteins were probed with anti-p85 subunit or anti-PY antibodies. The protein bands were visualized using an enhanced chemiluminescence kit (Amersham). Negative controls without the primary antibody produced no bands. The chemiluminescence bands on x-ray film were quantified as described earlier. ⁸  

PI-3K Assay

PI-3K activity was determined essentially as described previously, ⁸ using PI as substrate (37.5 μg). In some experiments, PI-4P and PI-4,5P₂ were also used as substrates (37.5 μg each). The PI-3K reaction products were separated on thin-layer chromatography (TLC) plates (Merck, Darmstadt, Germany) and scanned for ³³P-labeled 3′ PI spots using an image analyzer (Packard, Meriden, CT). ⁸ In some cases, radiolabeled spots on plates were scraped and lipids extracted for further identification by high-performance liquid chromatography (HPLC). 

HPLC Analysis of PI-3K Reaction Products

Products of PI-3K were deacylated ⁸ and the glyceroinositol-phosphates separated on an anion exchange HPLC column (Partisil-SAX) ⁹ with a gradient from 0 to 1.0 M (NH₄)₂HPO₄ (pH 3.8) in water for 120 minutes (pump A, water; pump B, (NH₄)₂HPO₄; 0% B for 10 minutes to 25% B with a gradient over 60 minutes, to 50% B over 30 minutes, and 100% B for the last 20 minutes). The compounds were quantified with a radiomatic detector connected on-line. Deacylated ³H-PI-4P and ³H-PI-4,5P₂ were used as authentic standards, because ³H-PI-3P and ³H-PI-3,4P₂ compounds are not commercially available. The identification was based on the fact that the isomers of PI-3P and PI-4P and of PI-3,4P₂ and PI-4,5P₂ have different retention times. ⁸  

Statistical significance of the data was determined using Student’s t-test. 

Results

PI-3K in Bovine Lens Epithelial and Fiber Cells

Initial experiments were conducted to characterize PI-3K activity in lens fiber cells after immunoprecipitating the enzyme with anti-p85 antibody that immunoprecipitates the 85-kDa regulatory subunit (p85) and the associated p110, the catalytic subunit of PI-3K. The enzyme converted the substrate PI to PI-3P (Fig. 1A , lane 3); it also less efficiently phosphorylated PI-4P and PI-4,5P₂ to PI-3,4P₂ and PI-3,4,5P₃, respectively (Fig. 1A , lanes 1and 2). Wortmannin, a selective inhibitor of PI-3K that covalently binds to the catalytic p110 subunit inhibited the enzyme activity (Fig. 1A , line 4). The three PI-3K products separated by TLC were further identified by HPLC analysis (Fig. 1B) and their retention times were compared with deacylated standards of ³H-PI-4P and ³H-PI-4,5P₂. ⁹ The characterized PI-3K activity was distributed throughout the lens (Fig. 1C) . The activity, expressed per milligram lens protein, was higher in the capsule-epithelial layer than in the remainder of the lens; but significant amounts of the activity were present in fiber cell layers. The lower activity in the inner layers was due to dilution of the enzyme protein by massive amounts of crystallins, as can be seen when total activity per layer was calculated. The p85 protein in the immunoprecipitates of the lens layers was also identified after Western blot, with anti-p85 antibody. Similar to enzyme activity, a gradient of p85 protein was observed in lens with the highest concentration in the capsule-epithelial layer (Fig. 1D) . 

Insulin and IGF-1 Stimulate PI-3K Activity in Lens Epithelial Cells

To understand the relevance of PI-3K, we investigated whether, in lens epithelial cells, the growth factors insulin and IGF-1 can modulate its activity. For these studies, bovine lens epithelial cells in culture were incubated with insulin or IGF-1 and the proteins immunoprecipitated with anti-PY antibody to determine whether the PI-3K activity associates with the tyrosine phosphorylated proteins. ¹⁰ Both growth factors caused more than a 100% increase in PI-3K activity immunoprecipitated by anti-PY antibody after 10 and 15 minutes of stimulation and reached near basal levels at 45 minutes (Fig. 2) . To examine the presence of p85 in anti-PY immunoprecipitates, cells were exposed for 10 and 45 minutes to insulin or IGF-1, the cell extracts immunoprecipitated with anti-PY antibody, and the proteins analyzed by Western blot with anti-p85 antibody. A single band with a molecular weight of 85 kDa was found (Fig. 3A ). Stimulation of the cells for 10 minutes caused more than a 200% increase in the intensity of the band (Fig. 3B) demonstrating a correlation between PI-3K activity and an increase in the p85 in the anti-PY immunoprecipitates. 

Association of the p85 Subunit of PI-3K with Other Tyrosine-Phosphorylated Proteins in Stimulated Lens Epithelial Cells

To determine whether insulin and IGF-1 signaling leads to the interaction of p85 with other tyrosine phosphorylated proteins, bovine lens epithelial cells in culture were stimulated with IGF-1, and proteins were immunoprecipitated with anti-p85 antibody and subjected to SDS-PAGE followed by Western blot using anti-PY antibody. In addition to the p85 subunit there were 100- and 180-kDa bands (Fig. 4A ) indicating that these proteins coimmunoprecipitate with the p85. The level of protein phosphorylation after IGF stimulation increased between 5 and 15 minutes. Blots probed with anti-p85 antibody showed the same amount of p85 protein in all samples (Fig. 4B) . Insulin-stimulated lens epithelial cells showed a similar association of p85 with 100- and 180-kDa proteins (data not shown). 

Discussion

Growth factors are important for cellular processes such as proliferation, elongation, and differentiation that occur during the development and growth of ocular lens. The precise pathways that transduce the growth factor signals in lens have not been previously studied. In the current study, we show that in lens epithelial cells, insulin and IGF-1 stimulate PI-3K activity, which can lead to the generation of 3′ PIs, which have been shown to trigger a cascade of signal transduction reactions involved in mitogenic and differentiation processes. ^{3 4} There was more than a 100% increase in PI-3K activity over controls after insulin and IGF-1 stimulation. Similar increases have also been observed in other cell types when stimulated with agonists. Previous studies have shown increased turnover of PI during epithelial cell division and a rapid decline during their differentiation, ¹¹ and it was suggested that phospholipase C–dependent PI turnover pathway may be involved. However, this pathway was not found to be influenced by insulin. ¹² PI and other PIs when phosphorylated by PI-3K are not substrates for phospholipase C but are components of an intracellular signaling system important in the control of cellular functions such as growth and survival. ^{3 4} We also show the tyrosine phosphorylation of the p85 after insulin and IGF-1 stimulation. There was a correlation between phosphorylation of the p85 and activation of PI-3K, suggesting that phosphorylation may be important in the activation of the enzyme in lens epithelial cells. Tyrosine phosphorylation of the p85 subunit has been shown after growth factor activation in other cell types, ^{13 14} although it is not clear whether it is necessary for enzyme activation. 

After insulin and IGF-1 stimulation, p85 is associated with other tyrosine-phosphorylated proteins. We speculate that the 100-kDa protein could be the insulin-IGF receptor β subunit (95–105 kDa) with intrinsic tyrosine kinase activity and the 180-kDa protein, the IRS-1 (165–185 kDa), because these proteins have been shown to be phosphorylated and to form a signaling complex with PI-3K, ^{10 15} facilitating its translocation to the plasma membrane (preliminary experiments to identify IRS-1 with two commercially available anti-IRS-1 antibodies were unsuccessful, probably because of differences in species specificity of the antibodies). Several tyrosine residues of IRS-1 are phosphorylated after growth factor stimulation. ¹⁶ In our experiments we have observed a broad, not well-separated 180-kDa (Fig. 4B) band that could be a result of differential phosphorylation of tyrosine residues on IRS-1 molecules. 

We also show that fiber cells retained PI-3K activity after differentiation. The metabolic activity of epithelial cells is high, because these cells proliferate, elongate, and, finally, differentiate into fiber cells to keep the lens growing. In addition, fiber cells depend on this epithelial cell layer for the maintenance of transparency. This could be the reason for high activity of PI-3K in epithelial cells, and this enzyme may play a role in these processes. However, it is interesting to note that PI-3K activity is not confined only to the epithelium and elongating and differentiating superficial fiber cells and that a considerable activity is also present in inner fiber cell layers (it should be noted that during lens dissolution the amount of fiber cell protein recovered in the inner layers 3 and 4 was higher than in other layers). Taking into account the enzyme activity and the amount of p85 found in each fraction (Fig. 1) , we calculated the theoretical specific activity of the enzyme in epithelial cells as 3750 cpm/unit p85. This activity decreased to approximately 1500 cpm/unit p85 in superficial fiber cell layer (layer 1) and was maintained at similar values up to the nucleus, indicating that this protein did not undergo significant modifications in the fiber cells that could result in loss of catalytic activity. The relevance of PI-3K presence in fully matured and even in aged fiber cells is not clear at present. It can be argued that the enzyme is present as a result of the absence of active turnover of proteins in lens. However, several enzymes have also been shown to be active and participate in many reactions that have relevance to lens function. In chicken lens, insulin and IGF-1 receptors have been shown to decrease but not completely disappear. ^{17 18} Therefore, it is possible that PI-3K–mediated signal transduction is sustained in fiber cells and that this enzyme may serve as a survival factor from its role in a wide array of functions. 

In conclusion, we have shown that in lens epithelial cells, insulin and IGF-1 induce tyrosine phosphorylation of the p85 subunit, as well as activation of PI-3K. Insulin and IGF-1 have been shown to augment fibroblast growth factor–induced fiber cell differentiation. ^{19 20} The activation of PI-3K and its binding to specific tyrosine-phosphorylated proteins may be an important signaling mechanism in the regulation of insulin and IGF-1 actions in lens, such as proliferation of epithelial cells and their differentiation into fiber cells. The presence of PI-3K throughout the lens implicates sustained PI-3K signal transduction in fully differentiated fiber cells. The role of this enzyme during differentiation induced by growth factors and its function in maintaining the normal metabolism of fiber cells is currently under investigation. 

 Supported by National Institutes of Health Grant EY-06635 and Core Grant EY-02377.

 Submitted for publication July 15, 1999; revised October 1, 1999; accepted October 26, 1999.

 Commercial relationships policy: N.

 Presented in part at the annual meetings of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, 1998 and 1999.

 

Figure 1.

 

Characterization of PI-3K in bovine lens. (A) Synthesis of different 3′ PIs by PI-3K. The PI-3K activity in anti-p85 immunoprecipitates from fiber cells extracts was assayed using PI-4,5P₂ (lane 1), PI-4P (lane 2), and PI (lanes 3 and 4) as substrates and the products separated by TLC. Twenty nM Wortmannin (Wt, lane 4), a specific inhibitor of PI-3K, inhibited the formation of PI-3P. (B) Separation and identification of PI-3K products from crude lens fiber cells extracts by HPLC. The bands were extracted from a TLC plate, and the deacylated products, gPI-3P (tracing 2), gPI-3,4P₂ (tracing 4), and gPI-3,4,5 PI₃ (tracing 5) derived from PI-3P, PI-3,4P₂, and PI-3,4,5P₃ were run and compared with deacylated standards. A mixture of gPI-3P and authentic standard ³H-gPI-4P is shown in tracing 1. Tracing 3 represents the elution profile of a mixture of gPI-3,4P₂ and authentic standard ³H gPI-4,5P₂. (C) Protein fractions from different lens regions were prepared as described in the Methods section. PI-3K activity was assayed using PI as substrate. PI-3P bands on TLC plates were quantified by image analysis and expressed as counts per million per milligram protein. The values ± SD correspond to two different experiments. The average total amount of activity in each layer was indicated above the corresponding bar. (D) Identification of p85, the regulatory subunit of PI-3K in different lens layers. Cell extracts (1 mg protein) of each layer were immunoprecipitated with a polyclonal anti-p85 antibody, subjected to SDS-PAGE, and identified by Western blot analysis with monoclonal anti-p85 antibody (the amount of protein used for immunoprecipitation in layer 5 [∗] and nucleus [∗∗] was four and eight times higher than in the other fractions). The experiment was repeated once with similar results. The density units obtained for the p85 band of capsule-epithelium per milligram protein, were normalized to 1 and the total amount of p85 in each layer was calculated, relative to the amount present in capsule-epithelium. The total units thus calculated for p85 in each layer are: capsule-epithelium (2.0 units), layer 1 (19.4), layer 2 (19.2), layer 3 (34.6), layer 4 (33.0), layer 5 (11.4), and nucleus (4.2).

Figure 2.

 

PI-3K activation after insulin and IGF-1 stimulation in bovine lens epithelial cells. Epithelial cells in culture were incubated overnight in serum-free medium and then treated with insulin (100 nM) or IGF-1 (50 nM). Cell lysates containing the same amount of protein were subjected to immunoprecipitation with a polyclonal anti-PY antibody. Immunoprecipitates were then assayed for PI-3K activity, the enzyme product PI-3P was separated by TLC, and the radiolabeled PI-3P band quantified with an image analyzer. Because different experiments were performed with insulin and IGF-1, control (without growth factors) bars for each condition are shown separately. Values are expressed as counts per million per milligram protein and are the average of three experiments ± SD. *, Significant difference (P < 0.05) compared with unstimulated cells.

Figure 3.

 

Presence of the p85 subunit of PI-3K in anti-PY immunoprecipitates after insulin and IGF-1 stimulation. (A) Cultured epithelial cells were incubated overnight without serum and then incubated with insulin (100 nM) or IGF-1 (50 nM). Proteins were immunoprecipitated with a polyclonal anti-PY antibody, separated by SDS-PAGE, transferred to nitrocellulose membrane, and probed with a monoclonal anti-p85 antibody. (B) Quantification of the p85 protein band by densitometric scanning. The experiment was repeated once with similar results.

Figure 4.

 

Coimmunoprecipitation of tyrosine-phosphorylated proteins with p85 after IGF-1 stimulation. Lens epithelial cells in culture were incubated with IGF-1 (50 nM) for up to 45 minutes. Cell lysates were immunoprecipitated with a polyclonal anti-p85 antibody, resolved by SDS-PAGE, and analyzed by Western immunoblot analysis with monoclonal anti-PY (A) or anti-p85 antibodies (B). This experiment was repeated once with similar results. Lane C, controls (without IGF-1).