Potential pro-inflammatory impact of scleral lens midday fogging on human corneal epithelial cells: An in vitro study

Abstract

Purpose:

Midday fogging (MDF) occurs when particulate material accumulates in the fluid reservoir (FR) beneath scleral lenses (SL), and its impact on epithelial cells is unknown. This study examines the in vitro pro-inflammatory effect of the FR on human corneal epithelial cells in varying degrees of MDF.

Methods:

Normal SL neophytes were recruited to wear SL 8 h daily for 4 days. Following 8 h on days 1 and 4, optical coherence tomography (OCT) images were acquired for MDF quantification using ImageJ, and the FR was collected. FR samples from the same eye were later pooled, diluted 2-fold and applied on human telomerase-immortalized corneal epithelial (hTCEpi) cells cultured on Terasaki microwell plates. Tumor necrosis factor (TNF)-α and culture media were used as positive and negative controls, respectively. After a 30-minute treatment, the nuclear factor-kappa B (NF-κB) pathway was measured by NF-κB-p65 immunofluorescence and images were analyzed with ImageJ. Pearson’s correlation was conducted to determine the association between median nuclear fluorescence and MDF.

Results:

Fourteen FR samples with a mean volume of 22 ± 16 μl were tested. Mean MDF severity following 8 h of SL wear was 25 ± 17 units (range 7 – 64). The median nuclear fluorescence (NF-κB-p65 translocation) in cultured hTCEpi cells ranged from 31.43 to 45.16 while the negative and positive controls were 44.71 ± 1.72 and 108.77 ± 68.38, respectively. Although a potential positive trend between MDF and median nuclear fluorescence was observed, Pearson’s correlation analysis revealed no significant association (r = +0.48, P = 0.09).

Conclusions:

The results suggest that the FR can trigger NF-κB-p65 translocation in hTCEpi cells, which may be associated with MDF severity. This study introduces the use of Terasaki microwell plates for immunofluorescence studies of the FR. The technique is simple, minimizes sample usage, and does not require expensive instrumentation.

1. Introduction

Scleral lenses (SL) are widely prescribed for the management of corneal irregularity, complicated refractive error, and ocular surface disease [1,2]. Conversely to corneal gas permeable contact lenses, SL are large diameter rigid contact lenses that vault over the cornea and maintain a fluid reservoir (FR) between the posterior lens surface and the cornea. During SL wear, the FR keeps the cornea hydrated and protected from the shearing forces of the lids and environment [3–6]. However, the FR is vastly different than the natural tear film in thickness, composition, and fluid properties [3,7–11]. For example, the thickness of the FR ranges from 100–1000 μm in depth over the cornea in successful wearers [12,13], compared to the 10 μm thickness of the natural tear film. Initially composed of the SL filling solution (usually sterile saline), this mixes with the ocular surface tears upon application to form the FR. The natural tear film includes proteins, lipids, and metabolites, which presumably become trapped in the FR with the lens applied. Moreover, due to the suction and settling of SL, there is minimal to no lens movement associated with SL wear. This is believed to cause a reduction or even a lack of tear exchange [14–16]. Stagnation of the tear fluid might favor the accumulation of cellular debris and metabolic byproducts in the FR, creating visual disturbances and potentially favoring inflammatory complications [7,15,17].

The existent knowledge on FR composition remains limited, but its importance is underscored by its prolongued proximity to the ocular surface and the potential for it to impact corneal health. To determine the composition of the FR, recent studies have used techniques such as ELISA/Luminex, mass spectrometry, and flow cytometry to reveal the presence of a variety of lipids [18], proteins [7], and cells [19] in the FR. One study measured the polymorphonuclear cell (PMN) content in the FR, finding several thousands of neutrophils in FR samples which are not typically detected in daytime tears [19]. These studies seem to indicate there is a “closed eye” environment present during SL wear, as these mediators and cells naturally increase in number on the ocular surface overnight [20,21].

One primary complication of SL wear is midday fogging (MDF), in which cellular debris and particulate material accumulate in the FR during lens wear. MDF is reported in an estimated 26–46 % of SL wearers [17,19,22,23], resulting in reduced vision and unknown physiologic effects [15,17,19,23]. Additionally, MDF has been associated with a disruptive impact on daily life with wearers complaining of blurry vision, lens fogging and discomfort which requires the lenses to be removed and cleaned multiple times per day. While it is unclear how MDF can impact the ocular surface, specifically the cornea, it is reasonable to hypothesize that the presence of MDF debris (which includes an increase in leukocytes [19]) could trigger an inflammatory response in corneal epithelial cells. Given these associated complications with MDF, there is a need to understand the impact of MDF on the ocular surface for proper risk management and to develop mitigation strategies for this frequent SL complication.

The study of any possible direct association between MDF and inflammation is challenging. Individual pro-inflammatory cytokines and tear cells in the FR can be identified [7], yet trying to define the pro-inflammatory properties of FR based on its composition is still difficult considering the potential interactions of its multiple components. In this regard, the use of the FR in a functional assay rather than attempting to characterize its composition may be more practical and effective. Nuclear factor-kappa B (NF-κB) is a rapid response inducible transcription factor present ubiquitously in cells involved in inflammatory reactions. It exerts its effect by inducing the expression of cytokines, chemokines, cell adhesion molecules, growth factors, and immunoreceptors [24,25]. Inactive NF-κB-p65 protein is present in the cytoplasm complexed with an inhibitory protein, IκB. Incoming signals from the cell surface cause IκB degradation and release of NF-κB-p65 which then can translocate to the nucleus and promote the expression of proinflammatory mediators with some of them (e.g., IL-1α, IL-1β, TNF-α) being able to activate the pathway again.

The purpose of this study was to examine the potential pro-inflammatory effect of the FR on human corneal epithelial cells in vitro, and to determine if there is a correlation with MDF. The hypothesis is that greater levels of MDF will trigger an increasing inflammatory response (i.e., more translocation of NF-κB-p65) in corneal epithelial cells treated with FR samples. This is the first study that assesses the capability of FR to trigger the translocation of protein NF-κB-p65 from the cytoplasm to the nucleus in corneal epithelial cells. In addition, a novel technique is proposed for in vitro studies of FR by using cell culture in Terasaki microwell plates followed by immunofluorescence.

2. Materials and methods

2.1. Study participants and SL wear

This study was compliant with the tenets of the Declaration of Helsinki and was approved by the University of Houston’s Institutional Review Board. The experiments were done using tear samples from 9 normal subjects that were recruited to participate in a study to investigate lipid composition in MDF [18]. All enrolled subjects signed an informed consent prior to participation and all study visits and experiments were performed at the University of Houston, College of Optometry (UHCO). Inclusion criterion was a normal ocular surface including the cornea and conjunctiva. Soft contact lens wear was permitted but subjects discontinued lens wear for at least 3 days prior to beginning the study. All subjects were SL neophytes and were excluded if they were unable to wear SL for 8-hours per day or had a history of any ocular disease or surgery including refractive surgery within the past 2 years.

Subjects were fitted with custom SLs based on scleral topography (sMap, Precision Ocular Metrology) and refraction calculations. Lenses were manufactured in a high Dk (125 barrers) gas permeable material (Optimum Extreme, Contamac, UK), and toric or spherical landing curvatures were automatically designed by the sMap software and adjusted as needed in the SL fitting and training sessions (pre-experimental) to provide the most optimal fitting SL for each subject. For FR sample collection, subjects wore lenses for 4 days, applying them using sterile saline (Purilens, Lifestyle, Inc.) in the morning and coming in for the visit after 8 h of lens wear on days 1 and 4. At each study visit, anterior segment-optical coherence tomography (AS-OCT) images were acquired to grade MDF, and the FR was collected. The participants were asked to keep a log of their SL wear time.

2.2. AS-OCT acquisition and processing

AS-OCT images (Cirrus OCT^™, Carl Zeiss, Germany) were acquired to measure the FR depth (corneal apical clearance) and to quantify MDF severity. FR depth was calculated at the corneal apex using software calipers, measuring the distance between the posterior SL surface and the anterior apical corneal surface. To quantify MDF severity, the AS-OCT image from each eye was processed using a custom ImageJ [26] protocol in which the net gray value in the FR was calculated as calibrated optical units as we have previously described [18]. Briefly, in each image, the region of interest was selected as the FR area, gray value calculated, and then another area on the outer boundary of the image was measured and subtracted as background; MDF scores were reported as calibrated units of optical density (Fig. 1). Each image was scored by two masked investigators and averaged for the final score. Repeatability and reproducibility were evaluated by comparing the intra- and inter-observer measurements. The scores for each eye at visits on day 1 and 4 were averaged for the final MDF score (since FR samples from days 1 and 4 were pooled for the in vitro analysis).

Fig. 1. MDF quantification.

Images from the AS-OCT were exported to ImageJ, where the grayscale was inverted and the region of interest (ROI, the FR area, excluding the central reflex) and a background region (BG, area of image outside of the lens/cornea boundaries) were selected (A). Each ROI was measured for mean gray pixel density (theoretical range 0–100), with lower scores indicating less severe MDF and higher scores more severe MDF (B-D).

2.3. Fluid reservoir collection

The FR was collected from each eye using an established method [7,18]. Briefly, participants were asked to tip their chin so that their face was parallel to the ground, and the investigator carefully removed each SL using a small plunger removal tool. The fluid captured in the basin of the SL was then collected using a micropipette and immediately stored in a low-protein binding microcentrifuge tube at −80 °C until use in the experiments.

2.4. Cell culture and treatment

Human telomerase-immortalized corneal epithelial cells (hTCEpi) [27] were seeded at 2×10³ cells/well and cultured in Terasaki 60-micro-well plates (Greiner, Bio-One, Monroe, NC) using serum-free keratinocyte basal medium supplemented with keratinocyte medium bullet kit (KGM-2 Lonza, Walkersville, MD) for 24–36 h at 37 °C in 5 % CO₂ until approximately 95 % confluence. Each Terasaki microwell is approximately 1 mm in diameter and holds a maximum volume of 20 μl, ideal for treatments of cells with small volumes. After the hTCEpi cells reached approximately 90–95 % confluence, FR samples were thawed, pooled (days 1 and 4 from the same eye) and diluted 2-fold in KGM-2 culture media and applied on the cell monolayer (10 μl per well) in technical duplicates. Tumor necrosis factor-α (TNF-α, 10 ng/ml) and culture medium alone were used as positive and negative controls, respectively. All cells were treated for 30 min at 37°C in 5% CO₂ [28].

2.5. Immunostaining

A procedure described previously [29] was adapted to be used on the Terasaki microwell plate. Briefly, after treatment, the FR was carefully removed using a micropipette and cells were fixed with 4 % paraformaldehyde. After 15 min, fixative was removed, and cells were permeabilized with 20 μl of cold methanol and treated with 20 μl blocking solution (15 % goat serum in PBS) at room temperature. After 2 h, cells were incubated overnight at 4°C with rabbit anti-human p65 subunit of NF-κB primary antibody (Abcam, 5 μg/ml) or rabbit IgG isotype control (Invitrogen, 5 μg/ml). Next, cells were incubated with Alexa fluor 488-labeled goat anti-rabbit secondary antibody (Invitrogen, 5 μg/ml) at room temperature. After 1 h, nuclear-specific dye 4, 6-diamidino-2-phenylindole (DAPI, 0.1 μg/ml) was applied for cell nucleus identification. All washes between steps were done using 20 μl of phosphate buffered saline per well. DAPI (blue channel) and Alexa fluor 488 (green channel) images were acquired using the EVOS 5000 fluorescent microscope (Thermofisher, Waltham, MA) under the same magnification and time of exposure conditions.

2.6. Image analysis and NF-κB-p65 translocation

Image analysis was done using ImageJ [26], DAPI staining was used to locate the nuclei boundaries and Alexa 488 intensity was measured within the nuclear area [30]. Briefly, for each sample, an 8-bit DAPI-image (blue) was used to define the nuclei as regions of interest (ROIs) and was overlayed on the top of the corresponding 8-bit Alexa 488 image (green). Next, the Alexa 488 mean intensity of each DAPI-defined ROI and the median intensity per image were obtained (Fig. 2). The median Alexa 488 nuclear intensity per sample resulted from averaging the median values of the technical duplicates. The median rather than the mean intensity was determined as a more appropriate measure given that the median is not affected by exceptionally high or low values. In fact, the translocation of NF-κB-p65 protein from the cytoplasm to the nucleus is a dynamic process that does not occur in a synchronously manner, therefore the range of nuclear intensity values within a single image is quite broad.

Fig. 2. NF-κB translocation measurement.

DAPI staining was used to define cellular nuclei as regions of interest (ROI) (1); next, nuclear ROIs were overlayed on Alexa 488 labeled-NF-κB-p65 cells (2); finally, NF-κB-p65 fluorescence intensity data was obtained from each DAPI-defined nuclear ROI (3). Scale bar: 50 μm.

2.7. Data and statistical analysis

Statistical analyses were performed using GraphPad Prism v10.0.2. The normality of outcomes were evaluated using the Shapiro-Wilk test, and significance was set at P ≤ 0.05 for all outcomes. MDF scores and FR depth were averaged between day 1 and day 4 for each eye and Pearson’s correlation was used to determine the strength (r) and significance of the association between all the variables analyzed.

3. Results

3.1. Study participants and SL parameters

A total of 14 FR samples from 9 subjects had sufficient volume for in vitro testing and were included in this analysis. Enough FR volume was obtained from both eyes (5 subjects) or one eye (4 subjects). The study subjects’ demographics and SL parameters are shown in Table 1. The mean SL diameter was 16.1 ± 0.3 mm. All SL were designed to vault the cornea (including limbus) and land on the conjunctival surface without blood vessel impingement or blanching. All lenses contained a central 8.5 mm wide optic zone and a 2.1 mm wide adjacent peripheral curve. The radii of curvature in these zones depended on the SL power and sagittal depth (SAG). Average lens SAG was 4369 ± 288 μm, and 9 lenses (64 %) were designed with toric (rotationally asymmetric) landing zones, 5 lenses (36 %) with spherical (rotationally symmetric) landing zones; mean landing zone toricity for subjects wearing toric lenses was 165 ± 92 μm.

Table 1.

Subjects Demographics and SL Parameters.

Subjects	Outcome
Age (range)	21–26
Gender, number female (%)	5 (55)
Hours SL wear, Day 1 Mean ± SD	8.07 ± 2.22
Hours SL wear, Day 4 Mean ± SD	8.09 ± 2.16
SL Parameter
Brand	Europa
Manufacturer	Visionary Optics
Material (Company)	Optimum Extreme (Contamac)
Dk, barrer	125
Power range, diopter (SE)	+1.00 to −5.50
Diameter range, mm	15.5 – 16.5
SAG range, μm	3961 – 4832
Toric landing zone, # lenses (%)	9 (64)

3.2. AS-OCT analysis: Fluid reservoir depth and MDF scores

MDF scores were strongly correlated between eyes (r = +0.94; P < 0.001) and between days (r = +0.94; P < 0.001). Intra-observer repeatability of the measurements was 0.97 and 0.93, respectively, for the two examiners, and inter-observer repeatability was 0.91 (data not shown). Mean FR depth was 336 ± 81 μm and lenses settled an average of 95 ± 57 μm after 8 h of wear. Mean MDF scores were 25 ± 17 units (Fig. 3). MDF scores were compared to initial FR depth, amount of SL settling, hours of SL wear to determine correlation, with no significant correlations found between MDF and any variable (Table 2).

Fig. 3. Fluid reservoir (FR) volume, FR Depth, and Midday fogging (MDF).

Descriptive data (A) and scatter plots (B) of 14 FR samples collected from 9 healthy subjects who wore SL for 4 days. Horizontal line on graphs represent median values.

Table 2.

Correlation of MDF to the other measured outcomes.

Compared outcome	Correlation to MDF	P-value
Initial FR depth	+0.53	0.06
Amount of SL settling	−0.02	0.94
Hours of SL wear	−0.07	0.78
NF-κB-p65 translocation (median nuclear fluorescence)	+0.46	0.09

3.3. Translocation of NF-κB-p65 in human corneal epithelial cells

The average FR volume collected was 22 ± 16 μl (range 4 – 50 μl). In the hTCEpi cells cultured with FR from SL wearers, the median nuclear fluorescence as indicator of NF-κB-p65 translocation was 38.68 ± 4.71 (range 31.43 – 45.16) with a mean number of ROIs per well of 1124 ± 98 (range 995 to 1330). The positive control (TNF-α) showed a median nuclear fluorescence of 108.77 ± 68.38 while the culture media (negative control) showed 44.71 ± 1.72 (P ≤ 0.001) (Fig. 4). The mean number of ROIs per image for the positive and negative controls was 1059 ± 331 (range 825 to 1293) and 1083 ± 22 (range 1067 to 1098), respectively. NF-κB-p65 immunostaining of the controls and samples with MDF scores 7 and 64 are also shown in Fig. 4. The correlation between increased NF-κB-p65 translocation and MDF severity showed a positive trend but this was not statistically significant (r = +0.48, P = 0.09) (Table 2, Fig. 5). Likewise, FR depth as indicator of corneal apical clearance did not correlate with MDF (r = +0.53, P = 0.06) (Table 2, Fig. 5).

Fig. 4. Exposure of HCEC to Scleral Lens (SL) Fluid Reservoir (FR) and NF-κβ translocation.

hTCEpi were cultured in the presence of FR (10 μl) for 30 min and immuno-stained for NF-κB-p65 expression. Median nuclear fluorescence as indicator of NF-κB translocation was 44.71 ± 1.72, (negative control, untreated, UNT), 108.77 ± 68.38 (positive control, TNF-α, 10 ng/mL) (A). Corneal epithelial cells treated with FR samples with MDF scores of 64 (a) and 7 (b) illustrate the potential association between MDF and NF- κB translocation (B). Median nuclear fluorescence is compared between the FR and the negative (media) and positive (TNF-α) controls (C). Scale bar: 50 μm.

Fig. 5. Correlation analyses.

The potential associations between MDF and FR Depth (A), MDF and NF-κB Median Nuclear Fluorescence (B) and FR Depth and NF-κB Median Nuclear Fluorescence (C) were analyzed by Pearson’s correlation. The r and P values are shown on each plot.

4. Discussion

These data suggest a possible positive association between MDF and NF-κB-p65 translocation in human corneal epithelial cells cultured in the presence of FR (r = +0.48, P = 0.09), however the observed trend was not statistically significant. If real, this association could result from a greater abundance of pro-inflammatory molecules in the FR with increasing MDF. We recently showed an increase in nonpolar lipids in MDF [18], and other studies have shown increased polymorphonuclear leukocytes (likely neutrophils) in the FR during MDF [19], which we have also confirmed in experiments in our lab (unpublished data). Furthermore, we have previously reported an increase in inflammatory proteases (e.g., MMP-9, MMP-10) in the FR [7], compared to basal tears, which could provide a stimulus to the corneal epithelial cells to activate the NF-κB pathway. The FR might also contain damage associated molecular pattern (DAMPS) derived from cellular damage caused by local inflammation on the ocular surface under the SL which could favor in vitro NF-κB-p65 translocation in epithelial cells in culture. Importantly, the presence of DAMPs in FR could fuel inflammation itself contributing to establish a vicious cycle commonly seen in ocular surface inflammatory disorders. Future studies in this field might provide relevant mechanistic information and are promising.

When looking at the other variables, a positive trend was observed between MDF and the FR depth (apical corneal clearance) created by the vaulted lens over the cornea, but again it was not found statistically significant. Similarly, a study that compared subjects with MDF (n = 11) and without MDF (n = 12) found no difference in mean post-lens thickness by OCT [15]. Another study conducted in habitual scleral lens wearers (n = 19) reported that those with fogging had an average central clearance that was about 50 μm greater than the average clearance for those without fogging (P = 0.047) [19] and postulated that hypoxia caused by increased central clearance could cause the inflammation that in turn causes MDF. Although our data could not confirm or reject the association between MDF and inflammation, there was an interesting positive trend between FR depth and NF-κB-p65 translocation that could indicate that having a deeper FR is associated with more pro-inflammatory mediators during SL wear. Future studies will need to be done to determine the levels of inflammatory mediators and/or cells in the FR of different depths.

Studies on FR composition point toward it containing a mixture of ocular surface cellular debris (which accumulates when a SL prevents its mechanical removal by the upper lid during a blink), tears, and the filling solution added to the lens reservoir before the lens is applied to the ocular surface. A study reported visual acuity decrease during SL (filled with preserved saline solution) due to the increasing FR turbidity [31]. Further, benzalkonium chloride, a common preservative in contact lens solutions, has been reported to cause an inflammatory response [32]. In addition, solutions with electrolyte composition that differ from normal tears including saline solution have been reported to increase corneal epithelial cell sloughing [33]. This would favor the accumulation of cellular debris in the FR and potentially trigger inflammation as the tear fluid in patients with MDF has been shown to contain increased levels of of MMPs [7] and leukocytes [19].

In this study, FR samples were collected from healthy SL wearers and their potential in vitro pro-inflammatory activity was determined by measuring NF-κB-p65 translocation after they were added to HCEC cultures. The NF-κB pathway is one of the most ubiquitous transcription pathways that initiates the cellular production of inflammatory cytokines. When activated, this transcription factor moves from the cytoplasm into the nucleus and starts the synthesis of inflammatory cytokines such as IL-6, IL-8, TNF-α which play a key role in ocular surface inflammation.

Studies using tears in cell culture techniques are very scanty and mostly conducted on artificial tears [34,35]. In this regard, our study’s contribution to the field is reflected by the introduction of a novel, small-scale method to study the impact of FR samples on ocular surface cells in vitro. Due to the finite volume of FR samples (e.g., 5–30 μl), it is essential to have an efficient technique that requires small volumes of FR. The use of Terasaki microwell plates for examination of small volumes of FR allows analysis of individual samples avoiding the need of pooling from several subjects. The technique might also be applicable for multiple immunostainings and for the use of primary cell cultures which being derived directly from tissue are challenging to grow at high yields.

While the study is unique by utilizing Terasaki microwells to study the impact of FR on NF-κB activation, the small sample size is a limitation. Larger studies including established and diseased SL wearers (e. g., keratoconus patients) and longer follow-up periods are required to confirm the results of the present study. Although valuable as exploratory, given the observed data, more subjects would have rendered more robust results. Also, to keep the FR intact, samples were immediately frozen and did not undergo any pre-processing step (e.g., centrifugation) prior storage to remove its particulate content (including cells and cell debris) from its soluble components. Although the freeze–thaw step might have favored cell lysis and amplified the basal levels of cellular debris and pro-inflammatory molecules in the samples, this increase would have been proportional to the number of cells in the samples which still represents a potential indicator of MDF [19]. More recent experiments in our lab have shown that there is a distinct cellular composition in the FR, so future studies in this field of research should consider evaluating the cellular and non-cellular content of FR samples separately.

In conclusion, MDF is very common among SL wearers as particulate material and cellular debris collect between the posterior surface of the lens and the front of the ocular surface, leading to symptoms of blurred vision, discomfort or even redness and irritation [23]. MDF-associated inflammation and its potential clinical implications in SL wear remains elusive, and while this study does not show a definitive correlation between MDF and FR-led inflammation in corneal epithelial cells, it does provide evidence in favor of a potential benefitial impact of limiting the amount of MDF that the corneal epithelium is exposed to. Minimization of particulate accumulation and inflammation will be critical to benefit SL wearers by extending uninterrupted wear time, avoiding discomfort, and increasing the quality of life of patients who commonly present with complex ocular conditions. Clinical guidelines to reduce MDF as much as possible are supported by this study.