Opacification of the posterior capsule is a frequent postoperative

Transcription

1 Effect of Cataract Surgery on Aqueous TGF-jS and Lens Epithelial Cell Proliferation Niklas Wallentin, 1 Kerstin Wickstrom, 2 and Claes Lundberg 5 PURPOSE. TO study lens epithelial cell (LEC) regulation after experimental cataract surgery in rabbits. The effect of aqueous humor (AqH) on LECs in vitro was investigated and the AqH concentrations of basic fibroblast growth factor (bfgf) and transforming growth factor-/3 (TGF-j3) were determined. METHODS. Aqueous humor was aspirated immediately before, and 1, 5,, 15, and 30 days after cataract surgery (n = 8 each day). The effect of AqH on LEC proliferation was investigated by the incorporation of [ 3 H]thymidine. The following concentrations were determined in AqH: bfgf, active and total TGF-/3, leukocytes, and total protein. RESULTS. Aqueous humor collected after surgery triggered LEC growth on each occasion, except on (postoperative) day 30, in contrast to the effect of normal AqH. Lens epithelial cell proliferation peaked on day 1, displaying a sixfold increase. The bfgf concentration in AqH increased 6- to -fold after surgery and remained high throughout the experimental period. The AqH concentration of active TGF-/3 decreased sevenfold 1 day after surgery. It then returned to normal levels on day 15. Total TGF-j3 in postoperative AqH was twice as high as that in normal AqH on all days. Lens epithelial cell proliferation correlated with protein (r = 0.685), leukocytes (r = 0.565), and active TGF-j3 (r = ). CONCLUSIONS. The stimulating effect of postoperative AqH'on LEC proliferation may be caused by a reduction in the concentration of TGF-/3. (Invest Ophthalmol Vis Sci. 1998;39:l4l0-l4l8) Opacification of the posterior capsule is a frequent postoperative complication after cataract surgery and is a major clinical problem. Approximately 35% of patients who undergo surgery have posterior capsule opacification (PCO) within 2 years of surgery. 1 The opacification is mainly caused by proliferation and migration of postoperative remnants of lens epithelial cells (LECs) in the posterior lens capsule. Visual acuity can be restored by Nd:YAG laser posterior capsulotomy. However, opening of the posterior capsule may lead to complications such as macular edema and retinal detachment. 2 ' 3 Knowledge of the cause of PCO is incomplete. Activation of LECs, resulting in cell proliferation and migration, may be explained by surgical trauma that causes a breakdown of the blood-aqueous barrier' 1 ' 5 and evokes an inflammatory response. Further, the implanted lens itself may cause a foreignbody reaction. 6 The spatial environment is altered after the lens is removed, which may activate LECs in a wound-healing From the 'Department of Pharmaceutical Biosciences, Division of Pharmacology, University of Uppsala; the 2 Department of Pharmacology, Pharmacia & Upjohn, Uppsala; and the 3 Medical Products Agency, Uppsala, Sweden. Supported by an unrestricted training grant from Pharmacia & Upjohn, Uppsala, Sweden. Presented in part at the Joint European Research Meetings in Ophthalmology and Vision in Montpellier, France, 1996, and at the annual meeting of the Association for Research in Vision and Ophthalmology in Fort Lauderdale, Florida, Submitted for publication July 3, 1997; revised December 9, 1997, and February 26, 1998; accepted March 6, Proprietary interest category: N. Reprint requests: Niklas Wallentin, Department of Pharmacology, Pharmacia & Upjohn, S Uppsala, Sweden. 14 process. 7 Activation of LECs is regulated by various factors that may either enter the anterior chamber from the blood by selectively crossing or leaking through the blood-aqueous barrier, or that may be locally synthesized and actively secreted by tissues lining the anterior and posterior chambers. 8 Factors that exist in aqueous humor (AqH) in vivo and regulate LEC activation in vitro are fibroblast growth factor (FGF), 9 " 1 ' transforming growth factor-j3 (TGF-j3), l2 ~ 14 TGF-a, 1516 insulin-like growth factor, 17 " 19 epidermal growth factor, 20 ' 21 interleukin-1 and -6, and prostaglandin E 2. 7 The purpose of the present investigation was to study LEC regulation after experimental cataract surgery in rabbits. The proliferative effect of AqH collected before and after cataract surgery on rabbit LECs in culture was determined. Aqueous humor was analyzed for leukocyte, protein, bfgf, and total and active concentrations of TGF-/3. MATERIALS AND METHODS Animals Male New Zealand White rabbits (HB Lidkopings kaninfarm, Hasslosa, Lidkoping, and Estuna AB, Norrtalje, Sweden) weighing 2 kg to 2.5 kg were used. The animals were kept under standardized conditions in separate cages and given tap water and food ad libitum. The experiments in this study adhere to the ARVO Statement on the Use of Animals in Ophthalmic and Vision Research. Surgical Procedure: Extracapsular Extraction of the Lens The surgery was performed by the same surgeon (NW) to minimize variability in the technique used. Sterile surgical Investigative Ophthalmology & Visual Science, July 1998, Vol. 39, No. 8 Copyright Association for Research in Vision and Ophthalmology

2 IOVS, July 1998, Vol. 39, No. 8 Effect of Surgery on TGF-/3 and Lens Cell Proliferation 1411 procedures were used. The pupils of the rabbits were dilated and 5 minutes before anesthesia, using topical % phenylephrine hydrochloride (Neosynephrine; Sterling-Winthrop, New York, NY) and 1% cyclopentolate hydrochloride (Cyclogyl; Alcon, Fort Worth, TX). The rabbits were anesthetized with 2 ml ketamine hydrochloride (50 mg/ml, Ketalar; Parke- Davis, Gwent, UK) and 1 ml xylazin chloride (20 mg/ml, Rompun; Bayer, Leverkusen, Germany) administered intramuscularly. Topical application of 0.5% tetracaine (Tetrakain; Alcon) was used for additional local anesthesia. A corneal incision of approximately 2 mm was made close to the limbus with a scalpel (Superblade; Pharmacia, Uppsala, Sweden). Heparin (0.1 ml [5000 IU/ml], Heparin; Lovens, Ballerup, Denmark) and 0.1 ml 1% sodium hyaluronate (Healon; Pharmacia) were injected into the anterior chamber. The corneal incision was then enlarged to 3.2 mm with an angled ophthalmic slit knife (Alcon). Capsulotomy of the capsulorrhexis type was performed with a cystotome by bending the tip of a 27-gauge needle connected to a syringe. The corneal incision was enlarged to approximately mm with a pair of corneal scissors. After lens extraction, using the extracapsular cataract extraction technique, an intraocular lens (one-piece polymethylmethacrylate, model 720 A, +22 D, CeeOn; Pharmacia, Groningen, Holland) was implanted in the capsular bag. Sodium hyaluronate was used to open the capsular bag before insertion of the lens. The pupil was constricted by injection of 0.1 ml 1% acetylcholine chloride (Miochol; Cooper Vision) into the anterior chamber. Sodium hyaluronate was removed by washing with balanced salt solution (BSS; Alcon), and the corneal incision was closed by a continuous x-suture using 9-0 nylon. No anti-inflammatory treatment was used in order to study the natural course of PCO without intervention of drugs. Collection of Aqueous Humor Immediately before surgery, 200 fxl to 300 JLLI per eye AqH was aspirated from the anterior chamber with a 30-gauge needle inserted through the cornea (normal aqueous humor, normal AqH). The AqH was kept on ice until centrifuged at 7300g for 3 minutes. The cell-free supernatant was immediately frozen at 70 C and stored in the freezer until it was analyzed. The normal AqH was then replaced by injection of the same volume of BSS into the anterior chamber. On (postoperative) days 1, 5,, 15, and 30, the rabbits were killed by an intravenous injection of 5 ml pentobarbiturate (Avlivningsvatska; Apoteksbolaget, Malmo, Sweden). Immediately thereafter, a sample of 200 ixl to 300 \x\ per eye AqH (postoperative AqH) was aspirated as described. A -/xl sample was taken for a leukocyte count before centrifugation. The remainder of the AqH was treated as described earlier. Preliminary experiments show that it is essential to keep AqH samples on ice and to minimize the time between aspiration and freezing to promote its activity in later analysis. Repeated freezing and thawing decreases the activity of AqH samples. Leukocyte Quantification and Proliferation Assay The number of leukocytes in the samples was counted using a Spencer chamber (American Optical, Buffalo, NY) and a light microscope. Rabbit LECs were isolated as described elsewhere. 15 The cells were thawed and cultured in a culture flask until subconfluence. The culture medium consisted of Dulbecco's modified Eagle's medium containing 1% antibiotic antimycotic solution (AA), 1% nonessential amino acids, % fetal calf serum (FCS), and 20% nutrient mixture, HAM's F-12 with L-glutamine (HAM's F-12). All culture solutions were obtained from Gibco BRL (Paisley, UK). The cells were then plated in 96-well dishes. To optimize the method of determining the effect of AqH on rabbit LEC proliferation in vitro, several methodological variables were studied: the number of cells plated, the concentration of AqH, the concentration and type of FCS, protein coating of the cell dishes, and the time of the cell culture (data not shown). These pilot studies resulted in performance of the proliferation assay using,000 cells/well that were allowed to attach overnight in noncoated 96-well plates. Thereafter the medium was removed and 0 jul Dulbecco's modified Eagle's medium containing 1% AA, 0.1% gentamicin, 1% nonessential amino acids, 1% L-glutamine, 20% HAM's F-12, and % dialyzed and heat-inactivated FCS was added in quadruplicate, together with % of the AqH samples. For method control specimens, AqH was replaced by ng/ml TGF-a as a positive control, and BSS was used as a negative control. The media were changed every second day. After 96 hours, rabbit LEC proliferation was determined. In a preliminary study, to ensure that the AqH protein concentration alone did not affect LEC proliferation, rabbit albumin was added within the range of the protein content of the various AqH samples. The various rabbit albumin concentrations had no effect on rabbit LEC proliferation (data not shown). Quantification of Proliferation To quantify the proliferative activity of rabbit LECs after addition of AqH, the incorporation of [ 3 H]thymidine was measured. A final concentration of 9.1 /xci/ml [ 3 H]thymidine (1.0 mci/ml; Amersham International, Little Chalfont, UK) was added to each well after 80 hours in culture. After an additional 16 hours, the medium was removed, each well was rinsed twice with 200 /xl phosphate-buffered saline, and 50 JLII trypsin was added. After minutes, the cells were harvested (model well sample harvesting device; Pharmacia). The radioactivity was measured in a liquid scintillation counter (model 1205, Betaplate; Pharmacia). Protein Determination The protein content of normal AqH and postoperative AqH was quantified using a commercial assay (Bio-Rad Protein Assay; Bio-Rad, Richmond, CA) based on the method of Bradford. 22 A standard curve of rabbit serum albumin (Sigma, St. Louis, MO) was used. Concentrations of unknown samples of protein were determined from the linear portion (1.2- \x%/ ml) of the standard curve. Basic Fibroblast Growth Factor Immunoassay Quantification of bfgf in the AqH was carried out using an immunoassay system (Quantikine; R&D, Minneapolis, MN). The assay uses the quantitative "sandwich" enzyme immunoassay technique. Samples of AqH were incubated in microtiter plates coated with a monoclonal antibody specific for human bfgf. An alkaline phosphatase-conjugated polyclonal antibody, specific for human bfgf, was added to sandwich the immobilized bfgf, after which a substrate solution and an amplifier enzyme solution were added. After stopping the

3 1412 Wallentin et al. IOVS, July 1998, Vol. 39, No. 8 q 6 o,i- 0,01 hr bfgf log Dilution 0 FIGURE 1. The parallelism and linearity were compared in dilution series of human recombinant basic fibroblast growth factor (bfgf) and of rabbit postoperative aqueous humor on postoperative day 1. AqH, aqueous humor. color development with 2 N sulfuric acid, the optical density was determined at 490 nm. Concentrations of unknown samples of bfgf were determined from the linear portion ( pg/ml) of a human recombinant bfgf standard curve. According to the SWISSPROT database, the amino acid sequence of rabbit bfgf shows 99-3% homology with the human form of bfgf, and it was thus expected to cross-react with the anti-human antibodies in the immunoassay system. To confirm the cross-reactivity and specificity of the assay, the linearity and parallelism of dilution series of human recombinant bfgf (R&D) and of postoperative AqH were compared. The results showed good linearity and parallelism between human recombinant bfgf and rabbit postoperative AqH (Fig. 1). Transforming Growth Factor-/3 Bioassay The CCL-64 mink lung epithelial cell growth inhibition bioassay was optimized from previously described methods, l21 ''' 23 ' 2 '' in the number of plated cells, incubation times, and the concentration of the sample. Briefly, the cells were cultured in Dulbecco's modified Eagle's medium containing 1% AA, 1% nonessential amino acids, 1% FCS, and 20% HAM's F-12 in a culture flask until subconfluence. The cells were then plated in 96-well dishes and allowed to attach overnight. Thereafter, the cells were starved for 2 days using cell culture medium without FCS. Subsequently, the medium was changed to a medium containing % of the various AqH samples. After 20 hours of incubation, 9-1 jnci/ml (final concentration) [ 3 H]thymidine was added, and the incubation was continued for 1 hour. The incorporated 3 H radioactivity was counted as described. Active and the total concentrations of TGF-/3 in AqH were studied. Latent TGF-/3 was activated by transient acidification of the AqH samples. The ph was lowered to 1.5 for 30 minutes by the addition of 150 mm HC1, followed by reneutralization to ph 7.5 with 150 mm NaOH. Transient acidification had no effect on the bioassay (data not shown). Concentrations of unknown samples of TGF-j3 were determined from a porcine TGF-/3 2 (R&D) standard curve. To confirm the cross-reactivity and specificity of the assay, the linearity and parallelism of a dilution series of porcine TGF-j3 2 and of postoperative AqH were compared. Good linearity and parallelism between porcine TGF-)3 2 and rabbit AqH was noted (data not shown). To estimate how much of the growth-inhibitory activity was caused by TGF-/3, pan-specific TGF-j3 neutralizing antibody or 0 jag/ml (final concentration) control rabbit IgG (R&D) was incubated with the test samples 1 hour before application to the assay cells. To see how sensitive the bioassay was to bfgf, the CCL-64 cells were cultured with increasing concentrations of TGF-J3 and with 0.1 ng/ml human recombinant bfgf. The 50% inhibitory concentration (IC 50 ) values were then calculated. For TGF-/3 alone, IC 50 was 0.5 ng/ml and for TGF-j3 with 0.1 ng/ml bfgf, IC 50 was 1.05 ng/ml. The rabbit LECs were treated the same way as in the proliferation study described earlier, except that AqH was replaced by human recombinant bfgf, porcine TGF-j3 2, or both in concentrations ranging from 0 ng/ml to 0 ng/ml. Statistics Values are expressed as mean ± SD. Statistical differences between normal AqH and postoperative AqH at the same time point were evaluated using Student's paired t-test. The statistical significance between the different time points was determined using one-way analysis of variance (Fisher's protected least significant difference) with multiple range testing according to the method of 95% least significant difference intervals. P < 0.05 was considered significant. Correlation between variables was assessed nonparametrically using Spearman's rank correlation. RESULTS Proliferative Activity of Aqueous Humor after Surgery and Normal Aqueous Humor Rabbit LECs were cultured in the presence of % rabbit normal AqH or postoperative AqH collected 1, 5,, 15, and 30 days after lens extraction. Rabbit LEC proliferation, as measured by [ 3 H]thymidine incorporation, was increased at all time points except on day 30, compared with that in normal AqH (Fig. 2). The proliferative effect of postoperative AqH decreased as time elapsed after surgery. One day after surgery, the postoperative AqH-normal AqH ratio was 6.38, compared with 1.69 on day 30. The mean proliferative activities of rabbit LECs cultured with % HBSS (negative control) or ng/ml TGF-a (positive control) were 157 ± 13 cpm and 947 ± 222 cpm, respectively. Protein Concentration The protein content of postoperative AqH was higher at all time points than in normal AqH (Fig. 3). The concentration of proteins in postoperative AqH decreased after surgery. On day 1, the protein concentration of postoperative AqH was mg/ml; on day 30, the concentration was 7.1 ± 6.0 mg/ml. There was a statistically significant correlation between the AqH protein concentration and the rabbit LEC proliferation (r = 0.685; P < 0.05).

4 IOVS, July 1998, Vol. 39, No. 8 Effect of Surgery on TGF-/3 and Lens Cell Proliferation Normal AqH FIGURE 2. [ 3 H]Thymidine incorporation of rabbit lens epithelial cells incubated with normal aqueous humor (AqH) and postoperative AqH 1, 5,, 15, and 30 days after surgery. The lines represent individual rabbits, and the bars represent the mean of each group (n = 8). *A significant difference, compared with normal AqH (P < 0.05). Postoperative Aqueous Humor Leukocyte Concentration The total number of leukocytes in postoperative AqH on day 1 was 774 cells/jlli (Fig. 4). This number decreased rapidly to 143 cells/;al on day 5. No leukocytes were found in normal AqH. The concentration of leukocytes in postoperative AqH correlated well with the proliferative activity (r 0.56; P < 0.05). Basic Fibroblast Growth Factor Concentration contained significantly higher levels of bfgf than did normal AqH at all postsurgery time points, except on day 1 (Fig. 5). The bfgf concentration in postoperative AqH remained at a high level throughout the experimental period. No correlation between the AqH bfgf concentration and rabbit LEC proliferation was observed. Transforming Growth Factor-/3 Concentration The ability of the AqH samples to inhibit CCL-64 growth was blocked 75% with antibodies against TGF-/3 (data not shown). The AqH concentration of active TGF-/3 decreased sevenfold on day 1. It then returned to normal levels on day 15 (Fig. 6). There was a statistically significant negative correlation between the AqH TGF-/3 concentration and rabbit LEC proliferation (r = ; P < 0.05). Total TGF-/3 in postoperative AqH was approximately twice that in normal AqH on all days (Fig. 7). No correlation between the AqH total TGF-j3 concentration and rabbit LEC Normal AqH I! FIGURE 3- The protein concentration of normal aqueous humor (AqH) and postoperative AqH 1, 5,, 15, and 30 days after surgery. The lines represent individual rabbits, and the bars represent the mean of each group (n = 8). *A significant difference, compared with normal AqH (P < 0.05). 30

5 1414 WaUentin et al. IOVS, July 1998, Vol. 39, No FIGURE 4. Concentration of leukocytes in postoperative aqueous humor 1, 5,, 15, and 30 days after surgery. The data represent means ± SD (n = 8). proliferation was observed. Comparison of the activity of the AqH samples in the CCL-64 growth inhibition assay before and after acidification revealed that a mean of 29% of the TGF-/3 was present in the active form in normal AqH, 1% on day 1 and 13% on day 30 in postoperative AqH. Effect of Transforming Growth Factor-/3 on Basic Fibroblast Growth Factor-Induced Rabbit LEC Proliferation TGF-/3 inhibited the proliferate effect of bfgf on rabbit LECs in a dose-dependent manner (Fig. 8). Transforming growth factor-/3 inhibited rabbit LEC proliferation induced by ng/ml bfgf by 69% (1 ng/ml), 97% ( ng/ml), and 0% (0 ng/ml). DISCUSSION In this study, we found that AqH from rabbit eyes subjected to lens extraction and intraocular lens implantation had a proliferative effect on rabbit LECs in culture, compared with AqH removed before surgery. also contained increased levels of protein, leukocytes, bfgf, and total TGF-/3, and decreased levels of active TGF-/3. Rabbit LEC proliferation correlated positively with the protein content and leukocyte count and negatively with levels of active TGF-/3. The experimental model developed and described in this report offers the advantage of providing a standardized and controlled method for studying the effects of postoperative AqH on LEC proliferation. A similar study has been reported by Reddan et al. 25 However, they investigated the proliferative effect of AqH after a variety of traumas that did not include cataract surgery. The method described in the present study should provide an important complement to the existing in vivo models 26 ' 27 for studying the cause of PCO and determining treatment. The effect of postoperative AqH on rabbit LEC proliferation was increased after surgery, compared with that in normal AqH at all time points studied, except on day 30. The activity was greatest on day 1 and decreased thereafter during the experimental period. These results are in good accordance with those of a previous study of rabbit LEC proliferation in vivo. 28 In that study, LEC proliferation peaked between days 1 and 2, measured by incorporation of the thymidine analog bromodeoxyuridine after surgery. Thereafter, the activity decreased rapidly and stayed at a low level throughout the 2-month study. Our study showed that the AqH leukocyte count and protein increased after surgery, with a peak within 24 hours, to 774 leukocytes//xl and 23 mg/ml, respectively. The time of the peak is in accordance with that in earlier in vivo studies. 29 ' 31 However, in those studies, the leukocyte and protein concentrations were somewhat lower, compared with those in the present study, 5 to 225 leukocytes//xl and 6 mg/ml on day 1, Normal AqH FIGURE 5. Basicfibroblastgrowth factor concentration of normal aqueous humor (AqH) and postoperative AqH 1, 5,, 15, and 30 days after surgery. The lines represent individual rabbits. The bars represent the mean of each group (n = 8). *A significant difference, compared with normal AqH (P < 0.05). 30

6 VS, July 1998, Vol. 39, No. 8 Effect of Surgery on TGF-/3 and Lens Cell Proliferation Normal AqH 5 FIGURE 6. Active transforming growth factor-/3 concentration of normal aqueous humor (AqH) and postoperative AqH 1, 5,, 15, and 30 days after surgery. The lines represent individual rabbits. The bars represent the mean of each group (n 8). *A significant difference, compared with normal AqH (P < 0.05). 30 respectively. This discrepancy is probably owing to the treatment of the rabbits with anti-inflammatory drugs after surgery 29 ' 30 or to the fact that the lens was not extracted and that the intraocular lenses were implanted in the anterior chamber. 31 Among the different growth factors suggested as playing a central role in LEC activation, bfgf and TGF-/3 were examined in this study. Our data show increased bfgf levels in postoperative AqH, compared with preoperative AqH. However, it was somewhat surprising that the bfgf concentration in postoperative AqH stayed at an elevated level throughout the experimental period, whereas the proliferative activity of the postoperative AqH decreased with time. Basic fibroblast growth factor is present in increased concentrations in AqH after cataract surgery in rabbits, according to a previous study by Namiki. In that study, the bfgf concentration also stayed at an elevated level throughout the 8-week experimental period, with a slight decrease after 8 weeks. The preoperative concentration of bfgf in AqH was under their detection limit, pg/ml, in accordance with findings in the present study. However, the levels of bfgf in postoperative AqH in Namiki's study were 15 times higher, compared with our results. In another study, 32 the preoperative concentration of bfgf was 26 pg/ml. Postoperative bfgf was not measured. The reason 300- Normal AqH 0 15 FIGURE 7. Total transforming growth factor-/3 concentration of normal aqueous humor (AqH) and postoperative AqH 1, 5,, 15, and 30 days after surgery. The lines represent individual rabbits. The bars represent the mean of each group (n = 8). *A significant difference, compared with normal AqH (P < 0.05).

7 1416 Wallentin et al. IOVS, July 1998, Vol. 39, No " bfgf TGF-p 2 (lng/ml) TGF-p 2 ( ng/ml) TGF-p 2 (0 ng/ml) 0,1 1 log bfgf (ng/ml) FIGURE 8. Effects of transforming growth factor-j3 2 (TGF-/3 2 ) on rabbit lens epithelial cells cultured with basic fibroblast growth factor (bfgf). The data are means ± SD. for the discrepancy between all three studies could be explained by the different methods used (radioimmunoassay, enzyme-linked immunosorbent assay), the surgical techniques used (phacoemulsification, extracapsular cataract extraction technique), or the treatment of the AqH samples after aspiration. In our study, samples were always kept on ice and deep-frozen shortly after aspiration. Basic fibroblast growth factor in rabbit AqH has been measured previously in our laboratory before and after cataract surgery using the same enzyme-linked immunosorbent assay kit as was used in the present study. The concentrations measured were in accordance with ours (Lilian Torngren, personal communication, July, 1997). Basicfibroblastgrowth factor regulates lens epithelial cell proliferation in vitro. 933 Further, FGF plays a central role in lens morphogenesis and growth, and extracellular FGF is capable of inducing fiber cell differentiation in lens epithelial cells in vitro and in vivo Immunohistochemical studies have revealed acidic FGF and bfgf in the eye in locations consistent with a role in fiber cell differentiation. 36 Furthermore, in vitro studies have demonstrated that proliferation, migration, and differentiation of lens epithelial cells are stimulated by sequentially increased concentrations of bfgf. 11 The elevated level of AqH bfgf that remained after surgery in our study indicates that bfgf does not play an essential role in LEC proliferation, or that a delicate and complex balance between stimulatory and inhibitory growth factors exists. In the present study, the total TGF-j3 concentration in AqH increased after surgery and stayed at an elevated level throughout the experimental period. The active TGF-/3 concentration decreased after surgery but returned to normal levels within 2 weeks. Previous studies have shown that normal rabbit AqH contains TGF-/3: total, 2.25 ng/ml 37 ; total, 2.72 ng/ml 23 ; total, 2.25 ng/ml and active, 0.5 ng/ml 12 ; total, 0.49 and active, 0.28 ng/ml 38 ; and total, 4.5 and active, 2.7 ng/ml. 39 The concentrations of total and active TGF-/3 are considerably lower in the earlier studies than in the present one: total, 90 ng/ml and active, 24 ng/ml. These discrepancies could have the same explanation as the discrepancies in bfgf detailed earlier that is, sample treatment or methods used (incubation times, plated cells, culture media, or type of TGF-/3 standard). In a pilot study, we compared three TGF-/3 standards used in earlier studies with the standard used in the present study. We found that our standard inhibited CCL-64 cells less, compared with the other three standards. Thus, the sample concentrations interpreted from our standard curve were as much as 2.5 times higher than the other standard curves studied. These results show the importance of the TGF-/3 standard used and that absolute TGF-/3 levels vary with the origin, isoform, manufacturer, and batch of the standard. In addition, previously reported levels of total TGF-/3 analyzed in plasma vary considerably between studies. 40 ' 41 This shows the difficulties in comparing absolute values among studies. Further, the main point in the present study was not to determine the absolute concentrations of the various factors. It was the relative levels before and after surgery that were of importance in our study. Earlier studies have shown that TGF-/3 inhibits LEC proliferation in vitro and that the addition of a TGF-j3 2 antibody increases the proliferative effect of preoperative AqH. 13 ' 42 Further, it has been shown that TGF-/3 induces cataractous changes in LECs. These changes include the formation of spindle-shaped cells, capsule wrinkling, the accumulation of extracellular matrix, cell death with features of apoptosis, 43 and expression of a-smooth muscle actin. 44 Transforming growth factor-j3 is synthesized and secreted as a latent complex that is composed of a TGF-/3 dimer, a latency-associated peptide, and a disulfide-linked binding protein. 45 Dissociation of the biologically active TGF-/3 dimer from this complex occurs with changes in ph or through the action of thrombospondin or such proteinases as plasmin. On release, the TGF-/3 dimer can be neutralized by specific association with several proteins, including a 2 -macroglobulin and the proteoglycans decorin and biglycan. 46 Thus, the decrease in active TGF-/3 concentration after surgery may be explained by decreased synthesis, increased breakdown, or neutralization, either by association with the latency-associated protein, complex or by binding to a protein. It has been shown in earlier studies that ocular inflammation significantly decreases the AqH concentration of TGF-/3. In accordance with our results, de Boer et al. 38 showed that the concentration of active TGF-/3 in AqH from patients with uveitis was lower, compared with that in samples of AqH from patients with no signs of ocular inflammation. In a study by Allen et al., 47 it was shown that the concentration of total TGF-/3 was lower in AqH from eyes injected with lipopolysaccharide than in AqH from control eyes. However, the concentration of active TGF-/3 remained at the same level. The reason for the difference between our study, which agreed with de Boer et al., 38 and that of Allen et al. 47 is obscure, but may be explained by the different experimental models used. No correlation was observed between the proliferative activity of postoperative AqH and bfgf levels. An inverse correlation was evident, however, between proliferative activity and the concentration of active TGF-j3. Rabbit LECs cultured with bfgf in combination with TGF-jS showed that TGF-j3 inhibits the proliferative effect of bfgf. Taken together, these results indicate a balance between stimulatory and inhibitory signals in LEC proliferation and causes of PCO. Immediately after surgery, the level of active TGF-/3 drops, thus making it possible for FGF to exert its proliferative effect. When the active TGF-/3 returns to normal levels it will serve as a stop

8 JOVS, July 1998, Vol. 39, No. 8 Effect of Surgery on TGF-/3 and Lens Cell Proliferation 1417 signal for proliferation but, at the same time, as a go signal for other characteristic PCO changes, such as synthesis of extracellular matrix and wrinkling of the capsule (see earlier discussion). Several other studies have shown that bfgf and TGF-j3 act as go and stop signals for various cell types for example, rat astrocytes 48 and bovine retinal capillary cells. 49 This may also be the case for rabbit LECs. In this study we have focused on bfgf and TGF-/3. It is important, however, to realize that other growth factors have been identified in LEC regulation. Transforming growth factoralpha, 1516 insulin-like growth factor, 17 " 19 and epithelial growth factor 20 ' 21 stimulate LEC proliferation in vitro and have been observed in AqH. The significance of these growth factors in PCO development is unclear. In conclusion, the present study describes a novel experimental in vitro model for studying the cause of PCO in a standardized and reproducible fashion. The results show that in rabbits, postoperative AqH has a proliferative effect on rabbit LECs; this effect is greatest on day 1 after surgery. Further, analyses of bfgf and TGF-/3 suggest that in the causes of PCO, there may be an imbalance between growth factors with a proliferative effect: bfgf, and conversely, growth factors with an inhibiting effect: TGF-j3. Acknowledgments The authors thank Keiko Funa, Ludwig Institute for Cancer Research, Biomedical Center, Uppsala, Sweden, for fruitful discussions and for providing the mink lung epithelial cell line (CCL-64) and Sara Wahlberg for excellent technical assistance. References 1. Moisseiev J, Bartov E, Schochat A, Blumenthal M. Long-term study of the prevalence of capsular opacification following extracapsular cataract extraction. / Cataract Refract Surg. 1989;15: Kraff MC, Sanders DR, Jampol LM, Lieberman HL. Effect of primary capsulotomy with extracapsular surgery on the incidence of pseudophakic cystoid macular edema. Am J Ophthalmol. 1984;98: Ober RR, Wilkinson CP, Fiore JV Jr, Maggiano JM. Rhegmatogenous retinal detachment after neodymium-yag laser capsulotomy in phakic and pseudophakic eyes. Am J Ophthalmol. 1986; 1: Grabner G, Zehetbauer G, Bettelheim H, Dorda W. Immediate changes in protein-concentration in aqueous humor induced by intracapsular lens extraction in human eyes. Graefes Arch Klin Exp Ophthalmol. 1979;2: Latour E, Driot JY, Coquelet C, Bonne C. The role of platelets in blood-aqueous barrier breakdown induced by anterior chamber paracentesis in the rabbit. Curr Eye Res. 1989;8: Wolter JR. Cytopathology of intraocular lens implantation. Ophthalmology. 1985:92: Nishi O, Nishi K. Involvement of residual lens epithelial cells in postoperative inflammation after intraocular lens implantation: cause of fibrin reaction. Asia PacJ Ophthalmol. 1992;4: Schlotzer-Schrehardt U, Dorfler S. Immunolocalization of growth factors in the human ciliary body epithelium. Curr Eye Res. 1993; 12: Arruti C, Cirillo A, Courtois Y. An eye-derived growth factor regulates epithelial cell proliferation in the cultured lens. Differentiation. 1985;28: Namiki M. Quantification of basicfibroblastgrowth factor (bfgf) and transforming growth factor a (TGF a) in rabbit aqueous humor after intraocular lens implantation. Nippon Ganka Gakkai Zasshi. 1994;98: McAvoy JW, Chamberlain CG. Fibroblast growth factor (FGF) induces different responses in lens epithelial cells depending on its concentration. Development. 1989; 7: Cousins SW, McCabe MM, Danielpour D, Streilein JW. Identification of transforming growth factor-beta as an immunosuppressive factor in aqueous humor. Invest Ophthalmol Vis Sci. 1991;32: Kurosaka D, Nagamoto T. Inhibitory effect of TGF-/32 in human aqueous humor on bovine lens epithelial cell proliferation. Invest Ophthalmol Vis Sci. 1994;35: l Granstein RD, Staszewski R, Knisely TL, et al. Aqueous humor contains transforming growth factor-j3 and a small (<35OO daltons) inhibitor of thymocyte proliferation. J Immunol. 1990; 144: Wickstrom K, Madsen K. The effect of transforming growth factoralpha (TGF a) on rabbit and primate lens epithelial cells in vitro. Curr Eye Res. 1993;12: Rotatori DS, Kerr NC, Raphael B, et al. Elevation of transforming growth factor alpha in cat aqueous humor after corneal endothelial injury. Invest Ophthalmol Vis Sci. 1994;35:l43-l Tripathi RC, Borisuth NSC, Tripathi BJ, Li J. Quantitation of growth modulatory polypeptides in aqueous humor [ARVO Abstracts]. Invest Ophthalmol Vis Sci. 1992;33: Reddan JR, Wilson-Dziedzic D. Insulin growth factor and epidermal growth factor trigger mitosis in lenses cultured in a serum-free medium. Invest Ophthalmol Vis Sci. 1983;24:409-4l Ibaraki N, Lin LR, Reddy VN. Effects of growth factors on proliferation and differentiation in human lens epithelial cells in early subculture. Invest Ophthalmol Vis Sci. 1995;36: Majima K. Human lens epithelial cells proliferate in response to exogenous EGF and have EGF and EGF receptor. Ophthalmic Res. 1995;27: Parelman JJ, Nicolson M, Pepose JS. Epidermal growth factor in human aqueous humor. Am] Ophthalmol. 1990;9: Bradford MM. Arapidand sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem. 1976;72: Duclos A, Rocha R, Baines G, Antecka E, Deschenes J. Quantification of transforming growth factor-/3 in aqueous humor. Ocular Immunol Inflamm. 1994;2: Kato M, Ishizaki A, Hellman U, et al. A human keratinocyte cell line produces two autocrine growth inhibitors, transforming growth factor-/? and insulin-like growth factor binding protein-6, in a calcium- and cell density-dependent manner. / Biol Cbem. 1995; 270: Reddan JR, Weinsieder A, Wilson D. Aqueous humor from traumatized eyes triggers cell division in the epithelia of cultured lenses. Exp Eye Res. 1979;28: Lundgren B, Jonsson E, Rolfsen W. Secondary cataract. An in vivo model for studies on secondary cataract in rabbits. Acta Ophthalmol. 1992;205(Suppl.): Nishi O, Nishi K, Yamada Y, Mizumoto Y. Effect of indomethacincoated posterior chamber intraocular lenses on postoperative inflammation and posterior capsule opacification./ Cataract Refract Surg. 1995;21: Wickstrom K, Andersson K, Johansson B, Lundgren B, Torngren L, von Malmborg A. Rabbit and monkey lens epithelial cell QJEC) proliferation in situ after lens extraction (JERMOV Abstracts). Vision Res. 1995;35:S Lundgren B, Ocklind A, Hoist A, Hiirfstrand A. Inflammatory response in the rabbit eye after intraocular implantation with poly- (methyl methacrylate) and heparin surface modified intraocular lenses. / Cataract Refract Surg. 1992; 18: Torngren L, Rolfsen W, Lundgren B. PGE 2 and leukocyte levels in aqueous humor after lens extraction and intraocular lens implantation. Ocular Immunol Inflamm. 1993;1: Wickstrom K. Polyamines in rabbit aqueous humor after surgical trauma to the eye. Curr Eye Res. 1991;: Tripathi RC, Borisuth NS, Tripathi BJ. Detection, quantification, and significance of basic fibroblast growth factor in the aqueous humor of man, cat, dog and pig. Exp Eye Res. 1992;54: Chamberlain CG, McAvoy JW. Evidence that fibroblast growth factor promotes lens fibre differentiation. Curr Eye Res. 1987;6:

9 1418 Wallentin et al. IOVS, July 1998, Vol. 39, No McAvoyJW, Chamberlain CG, de longh RU, Richardson NA, Lovicu FJ. The role of fibroblast growth factor in eye lens development. Ann N YAcad Sci. 1991;638: Robinson ML, Overbeek PA, Verran DJ, et al. Extracellular FGF-1 acts as a lens differentiation factor in transgenic mice. Development. 1995;121:5O5-5l De longh R, McAvoyJW. Distribution of acidic and basic fibroblast growth factors (FGF) in the foetal rat eye: implications for lens development. Growth Factors. 1992;6: Streilein JW, Cousins SW. Aqueous humor factors and their effect on the immune response in the anterior chamber. Curr Eye Res. 1990;9: de Boer JH, Limpens J, Orengb-Nania S, de Jong PT, La Heij E, Kijlstra A. Low mature TGF-/32 levels in aqueous humor during uveitis. Invest Ophthalmol Vis Sci. 1994;35: Jampel HD, Roche N, Stark WJ, Roberts AB. Transforming growth factor-/3 in human aqueous humor. Curr Eye Res. 1990; 9: Pfeiffer A, Middelberg-Bisping K, Drewes C, Schatz H. Elevated plasma levels of transforming growth factor-/3l in NIDDM. Diabetes Care. 1996;19: Wang XL, Liu SX, Wilcken DE. Circulating transforming growth factor /3l and coronary artery disease. Cardiovasc Res. 1997;34: Nishi O, Nishi K, Fujiwara T, Shirasawa E, Ohmoto Y. Effects of the cytokines on the proliferation of and collagen synthesis by human cataract lens epithelial cells. BrJ Ophthalmol. 1996;80: Liu J, Hales AM, Chamberlain CG, McAvoy JW. Induction of cataract-like changes in rat lens epithelial explants by transforming growth factor /3. Invest Ophthalmol Vis Sci. 1994;35: Kurosaka D, Kato K, Nagamoto T, Negishi K. Growth factors influence contractility and a-smooth muscle actin expression in bovine lens epithelial cells. Invest Ophthalmol Vis Sci. 1995;36: Miyazono K, Hellman U, Wernstedt C, Heldin CH. Latent high molecular weight complex of transforming growth factor /3l. Purification from human platelets and structural characterization. JBiolChem. 1988;263: l Schulz MW, Chamberlain CG, McAvoyJW. Inhibition of transforming growth factor-j3-induced cataractous changes in lens explants by ocular media and a2-macroglobulin. Invest Ophthalmol Vis Sci. 1996;37:15O Allen JB, McGahan MC, Ogawa Y, Sellon DC, Clark BD, Fleisher LN. Intravitreal transforming growth factor-j32 decreases cellular infiltration in endotoxin-induced ocular inflammation in rabbits. Curr Eye Res. 1996;15: Vergelli M, Mazzanti B, Ballerini C, Gran B, Amaducci L, Massacesi L. Transforming growth factor-j3l inhibits the proliferation of rat astrocytes induced by serum and growth factors. / Neurosci Res. 1995;40: Bensaid M, Malecaze F, Bayard F, Tauber JP. Opposing effects of basic fibroblast growth factor and transforming growth factor-j3 on the proliferation of cultured bovine retinal capillary endothelial (BREC) cells. Exp Eye Res. 1989;48: