Spatial Characteristics of Receptor-Induced Calcium Signaling in Human Lens Capsular Bags

Abstract

purpose. Despite recent improvements in intraocular lens (IOL) design, posterior capsule opacification (PCO) arising from lens cell growth remains a major problem. Calcium signaling has been shown to play a major role in driving human lens cell growth, and therefore it is necessary to understand the underlying mechanisms.

methods. Calcium signaling was studied in capsular bags (ex vivo) removed from donors who had undergone earlier cataract surgery. Fresh capsular bags were also produced from intact donor lenses and cultured in serum-free EMEM for up to 8 weeks. Both preparations were loaded with Fura-2, and ratiometric imaging of cytoplasmic calcium was performed using epifluorescence techniques. Changes were monitored in response to 10 μM ATP (adenosine triphosphate), 10 μM acetylcholine, and 10 ng/mL epidermal growth factor (EGF), and data were collected from equatorial, posterior, and anterior regions. Calcium transients were also recorded from anterior epithelial specimens in response to pilocarpine.

results. All equatorial cells of ex vivo bags responded to ATP and EGF, but not to acetylcholine, and this pattern was maintained in the cultured bags. Posterior capsule cells of both preparations also had similar properties, in which a large proportion of the cells responded to ATP and EGF, but not to acetylcholine. Conversely, most anterior cells of the in vivo bags responded to pilocarpine, whereas no cells in the cultured bags responded. All cells in the fresh anterior epithelium responded to pilocarpine.

conclusions. Ex vivo capsular bags retain the region-specific calcium-signaling characteristics of the native lens. Apart from losing M1 muscarinic expression properties, the in vitro capsular bags also reflect region-specific signaling properties and therefore provide a good model for the investigation of the contribution of calcium-signaling to PCO.

Modern extracapsular cataract surgery involves removing a circular disc of anterior capsule and expelling the opaque fiber mass through the opening. The remaining anterior capsule and entire posterior capsule then forms a capsular bag, into which a plastic intraocular lens (IOL) is inserted to restore focus. However, a proportion of viable lens epithelial cells (LECs) remain attached to the anterior capsule after the operation, despite scrupulous surgical techniques and the use of specially shaped IOLs that physically retard the movement of cells within the bag. ^{1 2 3} The surviving LECs can grow across the available surfaces of the capsular bag and IOL, including the previously cell-free posterior capsule. ^{4 5} LECs can induce light scatter by forming regions of multilayered cell aggregates and causing contraction of the posterior capsule in a process called posterior capsule opacification (PCO). All these features can be seen in the dark-field image in Figure 1 of a capsular bag removed from a donor who had undergone earlier cataract surgery. There are three morphologically distinct regions of the capsular bag. The equator corresponds to the equatorial regions of the intact native lens and holds over part of its circumference the arms (haptics) of the IOL. The posterior region comprises the center of the posterior capsule, and, if there are heavy infiltrations of cells forming capsular wrinkles in this area, there is sufficient loss of visual quality for Nd-YAG laser treatment to be necessary to remove the light-scattering regions. Although all capsular bags contain viable cells, only a proportion has sufficient numbers of posterior capsular changes to necessitate laser treatment. The interface region, which is off the visual axis has dense light-scattering regions in all capsular bags examined, ^{1 2 3} and these are formed from cellular and extracellular debris sandwiched between the anterior and posterior capsules. Despite improvements in surgical techniques, IOL design and laser therapy, PCO remains a significant burden on health budgets of developed countries and prevents modern surgery from being widely available in underdeveloped countries. It is important therefore to produce a range of therapies to reduce the incidence of PCO. Possibly the most promising drug interventions are designed to induce apoptosis of lens cells or to inhibit lens cell growth on the posterior capsule. ^{6 7}  

Intracellular calcium concentration ([Ca²⁺]_i) is known to play a key role in both lens cell growth ^{8 9} and cell death, ^{10 11} and these findings are synonymous in many other cell types (for a review see Ref. ¹² ). Furthermore, application of the endoplasmic reticulum Ca-ATPase inhibitor thapsigargin to block the [Ca²⁺]_i-signaling pathway totally inhibits lens cell growth in human capsular bags cultured in protein-free medium and in the presence of serum. ⁷ This indicates that [Ca²⁺]_i signaling has a major role to play both in autocrine and paracrine growth control. More recently, experiments with this in vitro model have shown that a number of molecules influence the progression of PCO including transforming growth factor (TGFβ), ¹³ fibroblast growth factor (FGF), ¹⁴ and hepatocyte growth factor (HGF). ¹⁵ Cultured human LECs proliferate extremely well in both protein-free medium and medium containing serum, and replicate features of clinically observed PCO, suggesting that lens cells respond to a variety of autocrine and paracrine stimuli. ¹⁶ Several growth-promoting substances have been detected in the aqueous humor of humans, including epidermal growth factor (EGF), basic fibroblast growth factor (bFGF), and TGFβ. ¹⁷ Moreover, a breakdown of the blood–aqueous barrier after surgery increases the amounts of growth factors, such as those in the aqueous humor. ¹⁷ In vivo, external signaling molecules in the ocular fluids are likely to have a key role in the regulation of lens cell survival and proliferation and the progression of PCO. Analysis of human donor capsular bags with implanted IOLs has shown that viable lens cells exist for many years after cataract surgery, ⁵ and, furthermore, a recent study has revealed that all donor bags analyzed showed a very high level of cell proliferation and matrix deposition. ¹⁸ Although cell distribution, cytoskeletal ultrastructure, and cell growth have been well described, investigating receptor-induced [Ca²⁺]_i cell signaling in different regions of the capsular bag is necessary to identify the extracellular signals involved in modulating [Ca²⁺]_i. For example, in the intact human lens, it has been shown that activation of tyrosine-kinase receptors mobilizes [Ca²⁺]_i only in equatorial cells, whereas muscarinic receptors are active only in the central anterior region. ¹⁹ We chose therefore to examine the response characteristics of these two well-described, regionally specific receptor systems and also included, as a positive control, the G-protein agonist ATP, as it has been found to signal in both anterior and equatorial regions of the intact human lens. 

The data obtained in the present investigation of ex vivo lens capsular bags obtained from patients who had undergone earlier cataract surgery identified G-protein- and tyrosine-kinase–coupled receptor agonists that induced changes in [Ca²⁺]_i concentration. Ex vivo capsular bag preparations provide the best model with which to compare how well receptor activity is preserved within selected regions of cultured capsular bags. These data provide the first evidence that postoperative lens cells in capsular bags retain much of the functional receptor-mediated [Ca²⁺]_i-signaling activity of epithelial cells within the intact lens. It is intriguing that as far as the anterior and equatorial epithelial cells are concerned, the signaling receptors also appeared to be confined largely to the same regions as they are in the native human lens. In vitro capsular bags cultured in serum-free medium for 8 weeks lost their ability to respond to acetylcholine (ACh), but retained functional ATP- and EGF-induced [Ca²⁺]_i signaling throughout this period. 

Methods

Freshly Isolated Ex Vivo Capsular Bags

Capsular bags with IOLs were obtained from human eyes donated for corneal transplantation from the East Anglian Eye Bank (Norwich, UK) after donors had undergone cataract surgery. All capsules used in this study were obtained from donors aged more than 65 years, and the time elapsed from cataract surgery to removal of the capsules was between 6 months and 7 years. As has been observed in a previous study ¹⁸ all capsular bags contained evidence of viable cells (Fig. 1) . The capsular bag was dissected from the globe, placed posterior-side down in a small plastic chamber used for calcium imaging, and bathed in 35°C artificial aqueous humor (AAH), with the following composition (in mM): 130 NaCl, 5 KCl, 5 NaHCO₃, 1 CaCl₂, 0.5 MgCl₂, 5 glucose, and 20 HEPES, adjusted to pH 7.25 with NaOH. Eight entomological pins (Watkins and Doncaster Ltd., Kent, UK) were inserted through the edge of the capsule into the base of the chamber to secure the preparation in place. The IOL was then dissected from the capsular bag. This arrangement allowed calcium imaging of cells residing on the posterior capsule, equator, and the combined anterior-posterior interface region (see Fig. 1 ). Ten capsular bags from donor eyes were used in this study and are described as ex vivo capsular bags when appropriate. 

Cultured In Vitro Capsular Bag Model

The model described by Liu et al. ²⁰ was used. After removal of corneoscleral discs for transplantation purposes, human donor eyes obtained from the East Anglian Eye Bank were used to perform a sham cataract operation. The resultant capsular bag was then dissected from the zonules and secured on a sterile 35-mm plastic Petri dish. Eight entomological pins were inserted through the edge of the capsule to retain its circular shape. The capsular bag was bathed in sterile serum-free Eagle’s minimum essential medium (EMEM) and incubated at 35°C in a 5% CO₂ atmosphere. Capsular bags were cultured for 2 to 8 weeks in serum-free EMEM and used in calcium-imaging experiments to determine receptor activity after certain time periods in culture. The capsular bags were then removed from the culture dish, placed in the calcium-imaging chamber, bathed in AAH (35°C), and imaged in exactly the same way as the ex vivo capsular bags. Twelve in vitro capsular bag preparations were used in this study and are described as in vitro capsular bags to differentiate them from the capsular bags removed from donors who had undergone earlier cataract surgery. 

Freshly Isolated Anterior Lens Epithelium

Human donor material was obtained as described, and the lens dissected from the zonules and placed anterior-side down into the chamber used for imaging. The center of the posterior capsule was punctured and an incision made across the diameter of the posterior capsule. Pins were inserted at the edge of the capsule to secure it at either end of the incision. Small cuts were then made in the capsule near the pins, so that most of the posterior capsule could be removed. The remaining capsule was then further secured with six additional pins and the fiber mass removed with forceps. Any remaining lens fiber fragments were then removed from the surrounding capsule by irrigation with AAH at 35°C. This preparation was used to develop a protocol for identifying the M1 receptor subtype that we could apply to the capsular bag preparations. Six different concentrations of ACh and pilocarpine were applied in total, and six donor epithelia were needed to map out the whole range. 

Measurement of [Ca²⁺]_i

Both the capsular bag and isolated lens epithelium preparations were loaded with the acetoxymethylester (AM) form of 3 μM Fura-2/AM for 40 minutes at 35°C. The lens cells were then washed in AAH for 20 minutes to allow complete de-esterification of the dye. Ratiometric imaging of [Ca²⁺]_i took place on the stage of an epifluorescence microscope (model TE-200; Nikon, Tokyo, Japan) fitted with a ×20 objective. In all lens preparations, data were collected from regions of interest (ROIs) consisting of approximately 10 to 12 cells and acquired as a running ratio average. Using this epifluorescence approach, it was not possible to distinguish between anterior and posterior cell contribution in the interface region (see Fig. 1 ). All preparations were continuously perifused with AAH at 35°C. Solutions were administered by a two-way tap, and every effort was made to ensure that solution turnover time in the chamber was kept at a constant rate (approximately 10 seconds). ACh and ATP were applied for 30 seconds and EGF for 2 minutes, as the latter response was significantly slower in time course. The three agonists were applied in random order. Cells were excited alternatively with light of 340- and 380-nm wavelengths. Resultant fluorescent emissions at both wavelengths were collected by a charge-coupled device (CCD) camera at 510 nm, and sampled every 2 seconds. The resultant fluorescence emissions from each lens preparation were left as ratios, as it was not possible to obtain fully calibrated data from all preparations (see Ref. ¹⁹ for a full discussion). A response was considered to be significant if the change in ratio elicited was greater than 0.05. This research followed the tenets of the Declaration of Helsinki regarding the use of human tissue in scientific research. 

Results

All ex vivo capsular bags examined showed evidence of recolonization of the anterior capsule and cell growth on the posterior capsule. It was therefore possible to study [Ca²⁺]_i-signaling characteristics of three distinct regions (Fig. 1) . The ultrastructural characteristics have been described in detail elsewhere. ¹⁸ The equatorial region of the capsule (Fig. 1) corresponds to the equatorial, differentiating cells of the intact lens, and the G-protein receptor agonist ATP elicited a robust, transient response from cells within this region (Fig. 2A) . The tyrosine-kinase receptor agonist EGF, in contrast, produced a much slower response, both in the rising and recovery phases of the response. The differences in dynamics of the two responses have been discussed in detail in a previous publication. ¹⁹ There was very little response to ACh from cells in this region. 

The posterior capsule (Fig. 1) was devoid of epithelial cells before the operation, and hence all the cells studied had reached this region through a combination of cell migration and division. The cells were large and irregular in shape and resembled tissue-cultured human lens cells. ^{18 21} Again, there were pronounced responses to both ATP and EGF (Fig. 2B) and the dynamics of the two responses were similar to those observed in the equatorial region. Again, there was little or no response to ACh from cells in this region. 

Cells in the anterior-posterior capsule interface region (Fig. 1) were the most heterogeneous of cells in all three regions. The anterior face of the capsule was covered with cells with a regular, hexagonal morphology, very similar in size and appearance to the mature epithelial cells originally residing on the anterior capsule. Cells on the underlying posterior capsule were much larger and flatter and resembled more the cells in the anterior-posterior capsule interface region. Interposed between the two regions were large strands of extracellular debris. ¹⁸ With the present epifluorescence optics, it was not possible to resolve the [Ca²⁺]_i dynamics of cells in the two layers separately. Large, reproducible responses were obtained from all three agonists tested (Fig. 2C) , and again the time courses of the ATP and EGF responses were similar to those observed in the equatorial and the anterior-posterior interface regions. 

Because the cells in all regions were not homogeneous in appearance, the responses to the three agonists were quantified in terms of the number of ROIs responding within the field of view. The total area in the field of view at any time was approximately 10⁴ μm² and 10 to 12 ROIs were sampled individually within this area. In the equatorial region (Fig. 1) , all ROIs responded to ATP and EGF, whereas none responded to ACh (Fig. 3) . In the posterior region (Fig. 1) , most ROIs responded to EGF and ATP, whereas only a few responded to ACh. The pattern was quite different in the anterior-posterior interface region (Fig. 1) where most of the ROIs responded to all three agonists, including ACh. 

Because it is important to know which of the muscarinic subtypes are present in the ex vivo and in vitro bags, we developed a pharmacological protocol to distinguish between the two most likely candidates M1 and M3. ²² Pilocarpine preferentially activates M1 receptors, ²³ rather than M3, and the relative sensitivity of the M1 receptors can be seen from the data shown in Figures 4A and 4B . The concentration–response characteristics were mapped out in cells from the freshly dissected anterior epithelium that are known to contain only the M1 receptor subtype (Fig 4B) . In the freshly isolated anterior epithelium 100% of cells respond to ACh and pilocarpine, and it should be noted (Fig. 4A) that 10 μM ACh and 100 μM pilocarpine elicited identical responses. This was also true in the anterior-posterior interface region of the ex vivo capsular bag, indicating that the response largely arises from the activation of M1 receptors (Fig. 4C) . 

Capsular bags created from intact donor lenses (in vitro bags) were cultured for 2 to 8 weeks in serum-free EMEM and the [Ca²⁺]_i-signaling characteristics of the same three regions examined. All the ROIs in the equatorial region responded to both ATP and EGF throughout the entire culture period and concomitantly, there was no response to ACh (Fig 5A) . Cells began to arrive in the field of view on the posterior capsule within 2 weeks, ¹⁶ and at that time more than 20% of the ROIs responded to ACh (Fig. 5B) . With increasing time in culture (and a greater coverage of the posterior capsule) a greater proportion of the cells respond to EGF, whereas a declining proportion responded to ACh. In the early stages of culture, cells in the anterior-posterior region show predominantly the characteristics of mature anterior cells, as they respond well to ACh and ATP, but poorly to EGF (Fig. 5C) . By the end of 8 weeks of culture, the cells in this region were indistinguishable from cells growing on the posterior capsule and, indeed, the pattern of responses resembled those of equatorial cells. 

Discussion

Despite recent improvements in intraocular lens design, PCO still remains a significant problem in the developed world ² and largely prevents modern cataract surgery’s being made widely available in other parts of the world. It is essential, therefore, that model systems be available, first to understand the cell biological processes involved in the events that follow cataract surgery and second to provide a base where technologies to combat PCO can be developed. Many systems are available from human tissue cultured cells ²⁴ to in vivo animal models. ²⁵ The former, although providing valuable information concerning specific molecular pathways, do not retain the native geometry of the capsular bag. The in vivo models, in contrast, recapitulate the spatial arrangement of the capsular bag, but a detailed interruption of molecular pathways is difficult, if not in some cases, impossible. Moreover, there are species differences in the array of membrane receptors that could be important in modifying the capsular bag after injury. ²⁶ The human capsular bag developed in this laboratory ²⁰ maintains the spatial arrangement of cells within the bag and, importantly, can be maintained for long periods (>1 year) in serum-free medium, thus permitting a detailed investigation of the mechanisms driving PCO. ¹⁶ In this study we compared [Ca²⁺]_i cell-signaling characteristics of three defined regions within capsular bags removed from donor eyes (where earlier cataract surgery had been performed) with signaling responses obtained from the same regions in cultured capsular bags. In the latter case, the bags were derived from intact lenses, and the sham cataract operation performed under laboratory conditions. ²⁰ We decided to investigate [Ca²⁺]_i-signaling characteristics, because previous experiments with the capsular bag system ⁷ have shown that when [Ca²⁺]_i signaling is inactivated, all growth within the bag is totally inhibited. Furthermore, we have recently reported the distribution of the same signaling mechanisms in the intact lens, ¹⁹ and so it was possible to compare directly the native lens and the ex vivo capsular bag. 

The data from the ex vivo capsular bags are summarized in Figure 3 and, remarkably, many of the signaling characteristics of the native lens were maintained within this severely disrupted system. The pattern at the equator was identical with that of the intact lens, in that there was no response to ACh, but all cells responded to the G-protein receptor agonist ATP and the tyrosine-kinase receptor agonist EGF. These data imply that there are very strong conservative forces maintaining the native identity of cells within the equatorial region and this identity is quite different from native anterior cells that respond to ACh, but not to EGF. ¹⁹ The ex vivo capsular bag maintains a strong muscarinic receptor response in the anterior-posterior interface region, and this presumably originates in the anterior cells, in that those cells, which were clearly visible on the posterior capsule, do not respond to ACh (Fig. 3) . The interfacial response is derived from the activation of M1 muscarinic receptors (Fig. 4C) , precisely as is the case in native anterior cells (Fig. 4A) . It is interesting that anterior cells in the ex vivo bag are smaller than those on the posterior and retain the regular hexagonal pattern of native anterior cells. ¹⁸ The limit of resolution of the epifluorescence technique does not permit us to conclude whether a [Ca²⁺]_i response in the anterior cells spreads to the posterior cells. However, it is unlikely that it does, because there was no transfer of signal from anterior to equatorial cells, for example (see also Ref. ¹⁹ ). Furthermore, [Ca²⁺]_i signals can be confined to specific regions within a single cell (see Ref. ²⁷ for a review of this point). 

Cells within the posterior region of the capsular bag responded well to ATP and EGF, but showed little response to ACh. They had the characteristics therefore of dividing equatorial cells rather than quiescent anterior epithelial cells. The ex vivo bag represents the outcome of the in vivo culture of lens cells. It is interesting to compare the characteristics that remain from an in vitro culture of the capsular bag. The same equatorial characteristics found in the ex vivo preparation were maintained in vitro, in that all cells responded to ATP and EGF, but there was no response to ACh (Fig. 5A) . It is unlikely that the cues maintaining the assembly of receptors are derived externally from the aqueous humor; more likely, they are intrinsic within the substance of the lens itself. Signals obtained within the posterior region of the in vitro bags (Fig. 5B) at the end of the culture period were again very similar to ex vivo signals and so are characteristic of active, dividing cells. In the initial stages of culture more than 20% of cells on the posterior capsule responded to ACh (and pilocarpine) and so these cells must have originated from the anterior rather than posterior population. This is the first occasion when the origin of any of the cells on the posterior capsule has been positively identified. 

Although the cells on the anterior capsule of cultured capsular bags retain their normal hexagonal appearance, ^{3 20} they lose the muscarinic response entirely after 6 weeks of culture and, indeed, PCR studies of capsular bags maintained in serum-free or serum-containing medium for 8 weeks or more show that muscarinic receptor expression has been lost (Duncan et al., unpublished data, 2001). In contrast, human lens cells cultured on plastic lose M1 expression but show increased expression of M3. ²² M1 receptor expression in a range of cell types is very transitory, ²⁸ and it will be a challenge to devise culture conditions that retain M1 expression both in the capsular bag and on plastic. It would also be interesting to know why the expression of M3 is suppressed in cells within ex vivo bags and also in cells within in vitro bags. 

These studies have shown that ex vivo capsular bags retain many of the regional [Ca²⁺]_i signaling characteristics of native lens tissue. This includes the maintenance of anterior cell responses to ACh and equatorial responses to EGF. Dividing and migrating cells on the posterior capsule of both ex vivo and in vitro capsular bags retain a strong EGF signal, but do not respond to ACh. These data suggest that EGF has a prime role to play in driving lens cell growth, and it will be interesting to see whether both paracrine and autocrine control mechanisms are involved. 

 Supported by Fight for Sight, Biotechnology and Biological Sciences Research Council (BBSRC), and Humane Research Trust.

 Submitted for publication July 7, 2003; revised August 22, 2003; accepted September 6, 2003.

 Disclosure: D.J. Collison, None; L. Wang, None; I.M. Wormstone, None; G. Duncan, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

(A) Dark-field image of an ex vivo donor human lens capsular bag with implanted IOL. (B) Diagrammatic representation of the structure of a capsular bag in sectional view. Both (A) and (B) show the equatorial region, the central region of the posterior capsule, and the anterior-posterior interface region of the capsular bag, where the anterior capsule and posterior capsule are in close apposition. Viable lens cells exist in all three regions.

Figure 2.

 

[Ca²⁺]_i responses elicited by receptor agonists in spatially distinct regions of the ex vivo human capsular bag. (A) The lens cells in the equatorial region responded well to ATP and EGF but responded only very slightly to ACh. (B) Lens cells residing on the posterior capsule showed a response pattern similar to those at the equator, responding to ATP and EGF but not to ACh. (C) In contrast, lens cells residing between the anterior and posterior capsules showed a robust response to all three receptor agonists. In (A–C), ATP and ACh were administered to the capsular bag in the external bathing medium in 30-second pulses, whereas EGF was applied for 2 minutes. (A) and (C) are representative traces from one lens capsule but have been repeated, with six capsules used from three independent donors. Trace (B) was repeated on three lens capsules. Note that the three traces correspond to a single ROI comprising approximately 10 to 12 cells.

Figure 3.

 

Proportion of lens ROIs responding to receptor agonists in different regions of the ex vivo lens capsular bag. In the equatorial region ATP and EGF produced a response in 100% of cells, whereas ACh failed to initiate a significant response. Most of the cells on the posterior capsule responded to ATP and EGF, but very few responded to ACh. Cells in the anterior-posterior interface capsular region responded well to ACh and ATP, but only approximately one third of cells also responded to EGF. Note that data from 10 to 12 ROIs were acquired from each of the three regions per donor capsular bag to compare the percentage responding.

Figure 4.

 

Pharmacology of the ACh-induced [Ca²⁺]_i response in the freshly isolated human lens epithelium (A, B) and in the anterior-posterior interface region of the ex vivo human lens capsular bag (C). (A) The M1-selective muscarinic receptor agonist pilocarpine induced a response in the isolated anterior epithelial preparation. (B) Concentration–response characteristics of pilocarpine and ACh in the freshly isolated human lens epithelium. The EC₅₀ of pilocarpine is 9 μM and of ACh is 0.6 μM. Data are normalized to 1 mM ACh and are expressed as the mean ± SEM (n ≥ 3 for each data point). (C) Pilocarpine produced a response in the anterior-posterior interface region of the ex vivo capsular bag as in the freshly isolated human lens epithelium. (C) was repeated on at least three independent ex vivo capsular bag preparations. Note that the traces in (A) and (C) correspond to a single ROI comprising approximately 10 to 12 cells.

Figure 5.

 

Receptor-induced [Ca²⁺]_i responses in spatially distinct regions of in vitro human lens capsular bags in culture. (A) Equatorial region showing robust responses throughout to EGF and ATP. (B) Response in the posterior capsule to ACh declined with time, whereas that to EGF increased. (C) Region between anterior and posterior capsules showed a marked decline in response to ACh and an increase in response to EGF. Data in (A–C) were computed as in Figure 3 and are expressed as the mean ± SEM (n ≥ 3); 12 capsules were used.

The authors thank the Norfolk and Norwich University Hospital Eye Bank for providing donor globes and Pam Keeley for invaluable assistance.