Quantitative Evaluation of Human Lens Epithelial Cell Viability and Cytolysis by Distilled Water Ex Vivo

Purpose: To investigate the role of distilled water (DW) in isolated human lens epithelial cells (LECs) viability and lysis ex vivo. Methods: After immersion in DW or balanced salt solution (BSS) for 1-, 2-, and 3-minutes, respectively, the cell viability of LECs was quantitatively evaluated. In addition, the capsule samples soaked in DW or BSS for 1-, 2-, and 3-minutes were combined with rinse for 1 minute to analyze the difference of LECs shedding percentage in each subgroup. The histopathological changes of the samples after treating were observed. Results: The percentage of LECs shed in DW immersion combined with rinse was signicantly higher than in DW immersion alone (p all <0.001). In the subgroup soaked in DW for 3 minutes, the death number, mortality, and the percentage of cell shedding of LECs was the most (p all <0.001). The histopathological changes showed that the cell destruction in the DW subgroup for 1-, 2-, and 3-minutes, and the transmission electron microscope results showed that the cells were partially detached from the capsule in the DW 3 minutes subgroup. Conclusions: Soaking in the DW can cause LECs death, and DW immersion combined with rinse was an effective method to remove LECs. The histopathology changes of treated DW suggested cellular necrosis was one type of LECs death mechanism.


Introduction
Lens capsule opaci cation, as a common complication after cataract surgery, will cause decreased postoperative vision and affect the effect of surgery. At present, it is believed that lens capsule opaci cation is the stimulation of surgical trauma, which leads to the proliferation, migration, and epithelial-mesenchymal transition (EMT) of the residual lens epithelial cells (LECs) in the periphery and equatorial part of the lens 1 . Lens capsule opaci cation includes anterior capsule opaci cation (ACO), leading to anterior capsular contraction and IOL deviation 2 , and posterior capsular opaci cation (PCO) blocking the visual axis, resulting in vision decreased 3 . The rate of anterior capsule brosis and phimosis after cataract surgery was about 0.47-3.3% 4 . During the follow-up of 3-5 years after cataract surgery in adults, the incidence of PCO was about 20-30% 5 . The LECs of children display higher proliferation and migration properties than those in adults 6 , and almost 100% children develop lens capsule opaci cation after surgery, which has a severe impact on visual development 7 .
Lens capsule opaci cation is usually treated with neodymium: YAG (Nd:YAG) laser capsulotomy or secondary surgical capsulotomy. The former is relatively simple and convenient, but there is a risk of complications 5,8 . The latter is an invasive method. Hence, there is a strong need to prevent lens capsule opaci cation after cataract surgery. Currently, there are two possible preventive directions against lens capsule opaci cation: inhibiting the proliferation, migration, and EMT of LECs, and clearing the LECs as far as possible. The former includes using drugs to inhibit the proliferation, migration, or EMT of LECs 9 , implantation of hydrophobic sharp-edged IOLs 10 or a thick endocapsular open ring to inhibit the migration of LECs to the posterior capsule 11 . However, LECs remain within the capsular bag and the causative reason can't be eliminated 12 . The latter approach is mainly via the combination of drugs and surgical techniques to clear the residual LECs in the capsule as far as possible 13,14 , which is an active area of research. However, the potential toxic effects of many drugs on intraocular tissues limit their clinical application 15,16 .
Distilled water (DW) induces cell lysis by hypoosmotic pressure 17 , which has no chemical toxicity to intraocular tissues and can be neutralized quickly by balanced salt solution (BSS). However, the results of current human clinical trials of DW are divergent. The effectiveness and optimal duration of action of DW on LECs are not completely clear. For this purpose, we carried out this study, taking the isolated anterior capsule of patients with age-related cataract (ARC), to explore the role of DW in LECs viability and the clearance effect of DW combined with or without BSS on LECs.

Patients
Samples were obtained after approval from Ethics Committee of Nanjing Drum Tower Hospital and in accordance with the guidelines of the Declaration of Helsinki. All patients signed an informed consent prior to surgery. The anterior capsules (5.0mm-5.5mm in diameter) were obtained from uneventful cataract surgery (age: 69.40±11.52, n=156, ARC) in Department of Ophthalmology of Nanjing Drum Tower Hospital.

Methods
The collected anterior capsules were quickly split into two pieces (312 small pieces of anterior capsule samples). 282 pieces of these samples were divided into 6 groups: negative control group (23 small pieces, no drug treatment), positive control group (10 small pieces, xed directly with 4% tissue cell stationary uid), BSS group, DW group, BSS with rinse group, and DW with rinse group. The samples soaked in BSS or DW alone were used to investigate the effect of DW on LECs viability, and the capsule samples soaked in DW or BSS combined with rinse were used to explore the clearance effect of DW on LECs. 30 pieces were used for histopathological examination.
There were 124 small pieces anterior capsule samples soaked in BSS or DW for 1, 2, or 3 minutes, respectively (61 pieces in BSS, 63 pieces in DW). Samples in the two control groups and BSS/DW group were stained with trypan blue-eosin to measure cell viability. Photographs at 40× under the light microscope were used for the next procedure. 10 nonoverlapping images were selected from each group at nuclear level and cell contour level, respectively, by a light microscope at 400×. The LECs density (per mm 2 ), the number of LECs deaths (per mm 2 ), and LECs mortality (%) were calculated by Image J (Fig. 1).
The calculation formulas were as follows: the LECs density (per mm 2 ) = the number of LECs deaths (per mm 2 ) the LECs mortality (%) = 63 pieces were immersed in BSS for 1, 2, or 3 minutes, respectively, and then the group was rinsed with BSS at the height of 70cm bottle for 1 minute. 62 pieces were in DW immersion for 1, 2, or 3 minutes, respectively, and then the group was rinsed the same as mentioned above. After treatment, 125 pieces mentioned above were stained with trypan blue-eosin, and were photographed under a light microscope at 40×, to compare with the negative control, BSS or DW group. About 5 to 6 pictures can be stitched into an intact piece of capsule. Then the shedding percentage of LECs (%) was calculated by Image J.
Detailed calculation method was provided in Figure 2 and the shedding percentage of LECs (%) was calculated by the formula The staining step of trypan blue-eosin was: 1) 0.04% trypan blue solution for 1 minute, 2) gentle irrigation of BSS for twice, 3) tissue cell xation solution for 10 minutes (except the positive control group), 4) rinsing for 3 minutes in phosphate buffered saline (PBS), 5) staining with eosin for 30 sec, 6) gentle washing with tap water and left to dry at room temperature.
Histopathological examination was performed on 30 pieces samples, which were divided equally into 5 groups (the negative control, BSS 3 min group, DW 1 min group, DW 2 min group, and DW 3 min group).
After treatment, 3 pieces from each group (n=15) were stained with hematoxylin and eosin (HE) and then examined by light microscope. The others were examined by transmission electron microscopy (TEM).
For light microscopic observation,the samples after treatment were immersed in xative solution for over 24 hours, dehydrated, dipped in wax, embedded, sliced, regular HE stained, and sealed with neutral gum in turn, and then were observed under light microscope.
For TEM observation, treated samples were xed with 2.5% glutaraldehyde, stored at 4℃ for more than 12 hours post xed with 1% osmiumtetroxide, washed three times with PBS, dehydrated with a graded ethanol series, treated with pure acetone, treated with a mixture of embedding agent and acetone (v/v=3/1) for 3 hours, embedded, sectioned with an ultrathin slicer, and double stained with lead citrate and uranyl acetate in turn. And then, the sections were observed and photographed after drying. s  u  m  o  f  c  e  l  l  s  p  e  r  s  q  u  a  r  e  i  n  1  0  i  m  a  g  e  s   0 .

Statistical analysis
Statistical comparison was estimated by a one-way analysis of variance (ANOVA) followed by a Dunnett-t test for comparing all groups with the control group. Tukeys test or Games-Howell test was performed for pairwise comparison. Comparisons between groups or subgroups were performed by paired t-test. Statistical tests were two-side with a signi cant level of 0.05.

LECs viability
The anterior capsule LECs of the normal ARC patients in the negative control group exhibited a regular, polygonal-like shape with a small number of dead cells. In the positive control group, blue cell nuclei were distributed everywhere within the eld of view and all cells were dead. The LECs in BSS group with distinct cellular borders showed no signi cant differences compared with the negative control group. In the 3 subgroups in DW group, the number of LECs expansion was becoming greater as the duration time increased, and the dead cells had indistinct cell boundaries (Fig. 3).  The superscript represented that the difference between the group and the negative group was signi cant by Dunnett-t test, letter "a" meant the p values were less than 0.05, and letter "b" meant the p values were less than 0.001.
When the immersion time was the same, the LECs death number and the LECs mortality in DW group were greater than those in BSS group (p all <0.001). In DW group, the LECs death number and the LECs mortality increased with longer immersion times. These data suggested that DW had a stronger destructive ability on LECs than BSS and the ability increased over time, that was, DW was more effective than BSS for disruption of LECs (Fig. 4).

Percentage of LECs shedding
Only a small amount of LECs shedding was detected in the negative control group, BSS group, and BSS combined with rinse group. A fraction of LECs shedding was observed in DW group. The LECs shedding was evident in DW combined with rinse group, and the range of LECs shedding was larger as DW immersion time increased (Fig. 5).
The differences showed no signi cant differences in BSS group, BSS combined with rinse group, and the negative control group (p >0.05). There were signi cant differences in the percentage of LECs shedding among DW group, DW combined with rinse group, and the negative control group (Welch F=386.267, p <0.001). And the percentage of shedding LECs in DW group, DW combined with rinse group was higher than that in the negative control group (p all <0.001). These data gave indications that DW could cause cell death and contribute to LECs clearance (Table 2). The superscript "a" represented that the difference between the group and the negative group was signi cant by Dunnett-t test, and the p values were less than 0.001.
When the time of soaking was the same, the average percentage of shedding LECs in DW combined with rinse group was more than that in DW group (p all <0.001). In subgroups of DW combined with rinse group, the percentage of shedding LECs increased with immersion time (comparisons among subgroups: Welch F=51.990, p <0.001, pairwise comparison between subgroups: p all < 0.005) (Fig. 6).

Microstructure of treated LECs
In the negative control group and BSS immersion for 3 mins group, the LECs with regular cell shape and round nuclei were arranged in a single layer and adhered to the capsule. In DW immersion for 1-, 2-, and 3mins subgroups, the cell integrity disruption of LECs, the cytoplasmic out ow of LECs, and some LECs shedding were observed in DW immersion for 2-and 3-mins subgroups (Fig. 7).
In the negative control group and BSS immersion for 3-mins group, the LECs appeared intact cellular morphology, round or round-like nuclei, complete nuclear membranes, clear nucleolus, evenly distributed chromatin, and normal morphology of cell organelles. And the interconnections of LECs showed ngerlike protrusions and LECs arranged in a single layer and adhered to the capsule. In DW immersion for 1-, and 2-mins groups, it could be seen cytolysis, nuclei with regular morphology, cytoplasmic destruction, swollen cellular organelles, and out ow of the cytoplasm. Disruption of intercellular junctions and loose cell-to-capsule junctions were also observed. In DW immersion 3-mins group, we could see the e ux of intercellular substances, decreased cell volume, nuclear deformation, and multiple sharp neurites, cytoplasmic destruction, and swelling of organelles. Disrupted intercellular junctions and partial separation of LECs to capsule were also viewed (Fig. 8).
Discussion DW induced cell rupture in hypoosmotic conditions, which could theoretically reduce the survival ratio of LECs. A previous study showed that the LECs exposure to DW for 1 min were markedly swollen with grossly morphologically intact cell membranes, and approximately 2 mins of exposure to DW, the complete cell lysis could be obtained 17 . Another study suggested that exposure to DW for 1 min, the LECs kept intact, cell membranes markedly swollen, and by 2 mins, no intact cells were visible 18 . Regrettably, no quantitative analysis was performed in the 2 studies of Crowston et al., the results were observed under microscope.
In this study, quantitative analyses and accurate calculations of the LECs density, the LECs death number, and the LECs mortality were conducted to obtain more precise results by trypan blue-eosin staining and photographing under a microscope 19 , which could provide a more comprehensive assessment of the effects of DW on LECs viability and clearance. The density of cell population in cataractous lenses varied greatly. The mean number was around 4000/mm 2 , but there were lenses with even lower cell counts 20 . In the negative control group of our study, the LECs density was 3829.08±519.16 mm 2 , which was similar to the conclusion of the study of Laspias et al. 20 . Therefore, our calculation method could be considered reliable.
Quantitative analysis showed that the LECs mortality was 70.05% in DW immersion for 3 mins, which was 2.35 times higher than that in DW immersion for 1 min. This result was similar to that of Rękas et al. 21 . There was 70.8% cytolytic destruction of LECs after 3-minute exposure to DW, which was 3.5 times higher than 1-minute exposure to DW 21 . We also set a negative control group, as well as BSS group acting for the same treatment time, to compare with the DW group. The result suggested that soaking in DW for 2 or 3 mins could effectively cause LECs death, and the effect of DW immersion for 3 mins was better. Naturally, there were different study ndings. In the study of Duncan et al., there was 50% cell survival after FHL124 cells exposed for 2 minutes to DW 22 . This could be due to the different materials and different counting methods. In the study of Rabsilber et al., it suggested that using DW was unable to reduce PCO development signi cantly 23 , which was different from ours. This might be caused by the insu cient acting time of DW.
During cataract surgery, the whole capsular bag should be rinsed with irrigation/aspiration (I/A) to remove the lens cortex and viscoelastic agent. Therefore, it might be considered to introduce DW into the capsule combined with stimulated I/A, which could not only fully exploit the capacity of DW to destroy cells, but also the neutrality against DW and clear the LECs with loosened junctions to capsule. In this study, BSS at the height of 70cm was used to rinse the capsule for 1 minute to simulate I/A during the cataract surgery. Results indicated that DW immersion combined with rinse contributed to LECs clearance, and the most effective exposure time was 3 mins. Some research suggested that 360-degree anterior capsular polishing alone was unable to reduce the incidence of PCO 24 , meaning that mechanical anterior capsular polishing alone was unable to clear all LECs in the capsular bag. Our results indicated that rinsing combined with exposure to hypotonic DW could effectively clear LECs.
To minimize the side effects on intraocular tissues, additional protective measures might be necessary. The sealed capsule irrigation (SCI) was designed to temporarily seal the capsular bag to allow drug injection and saline ush to protect the intraocular tissues 8 . Rękas et al. infused DW in SCI for 3 mins after the removal of cortical material and rinsed the intracapsular space with 0.9% NaCl 25 . They recognized that DW irrigated for 3 mins reduced PCO in the long-term follow-up, which was more effective than anterior capsule mechanical cleaning 25 . After nucleus removal and cortical aspiration, Zhang et al. lled the anterior chamber with a continuous infusion of sterile air, and then injected DW or BSS into the capsule for 3 mins and followed by irrigating the capsular bag with BSS with the assistance of vitrectomy machines 13 . It was found that the technique could signi cantly prevent capsular brosis and PCO 13 .
However, SCI was inapplicable to patients with deep anterior chamber or a small pupil 23 , and unsuitable to microincisional (1.8 mm or 2.2 mm) cataract surgery 13 . What's more, additional surgical techniques and equipment were necessary in the forementioned methods, which limited their popularized. DW combined with rinse could effectively remove LECs to prevent the occurrence of capsule opaci cation after cataract surgery, but how this technique could be applied in the clinic still needed further exploration.
Cell death exists in many different forms, such as necrosis, apoptosis, necroptosis, pyroptosis, and autophagy. Necrosis, a common form of pathological cell death, can be identi ed via hitstopathologic changes in both nuclei and cytoplasm. Our results suggested that soaking in DW not only induced LECs death, but also contributed to LECs clearance. We further considered how the LECs might change as soaking in DW, and cytological observations were carried out on treated LECs. HE staining showed cytolytic destruction and partial shedding of LECs in DW immersion for 2-and 3-mins subgroups. In the study of Rękas et al., basement membrane and LECs were occasionally seen in the specimens prepared for microscopic examination 21 . The results were the same as our light microscopic results. TEM results showed LECs destruction in DW immersion group, loose junctions between cells and capsule in DW immersion for 1-and 2-mins subgroups, and partial separation of LECs from capsule in DW immersion for 3-mins subgroup. The lysis and rupture of the cell membrane, another symbol of necrosis, could be found in our HE staining and TEM results. Therefore, it can be concluded that DW can cause LECs death via necrosis. Our results suggested that DW could not only cause LECs death, but also loosen the junctions of cells with the capsule, which could shed LECs from capsule partially. Therefore, the shedding percentage of cells increased after soaking in DW. All the above histocytological ndings were in agreement with the differences of LECs shedding percentage shown by trypan blue-eosin staining.
Nevertheless, the present study still had some limitations. First, DW could cause LECs death but not enough to kill all LECs. Considering DW frequently was used as a solvent, we did not know whether the trial for combining with other drugs to destroy LECs more fully could have been attempted 24 . Second, although DW combined rinse was effective on cell shedding but couldn't make all cell shedding. To this point, part of the reason lied in the fact that partial folding led to some LECs failing to be cleared when the capsule was rinsed with BSS. Third, LECs of anterior capsules were studied in this study, while the cellular morphology and function of LECs of the anterior and equatorial lens capsules were not the same 26 . Therefore, further animal experiments were necessary to explore the clearance effect of DW on LECs of the equatorial lens capsule. The histopathological changes of treated LECs with DW indicated cellular necrosis, which suggested that there were other operative mechanisms of DW aside from the hypoosmotic mechanism. All these aspects await further investigation.
It was concluded that DW immersion can cause LECs death and DW immersion combined with rinse was an effective method to remove LECs. The histopathology changes of treated DW suggested cellular necrosis was one type of cell death. All of these provided theoretical grounding for preventing lens capsule opaci cation after cataract surgery via DW.

Declarations
Availability of data and materials The data are available from the corresponding author upon reasonable request.       Microstructure of LECs in the negative control group (A), BSS group for 3 mins (B), DW group for 1 min (C), DW group for 2 mins (D), DW group for 3 mins (E). HE staining of the anterior capsule samples showed that the cell destruction in the DW subgroup for 1-, 2-, and 3-minutes, and LECs was partially exfoliated in the DW subgroup for 2-, and 3-minutes. (hematoxylin and eosin staining, scale bar=50 μm) 1: The cell showed a regular cell shape, a round nucleus, and a homogeneous cytoplasm. 2: This cell was with cytoplasmic destruction. 3: Shown here was cell shedding. AC: anterior capsule