Patterns of Mucin Adherence to Contact Lenses

Abstract

purpose. Contact lens wear alters the preocular fluid through factors that include tear deposits on the lens. In the current study, lens-adherent material was extracted to assess whether contact lenses sample mucins from the preocular fluid.

methods. Discarded extended-wear contact lenses were collected from patients with no ocular surface disease. Mucins were extracted in guanidine hydrochloride (GuHCl) with protease inhibitors. After the supernatant was removed, the extraction was repeated twice with the addition of 10 mM dithiothreitol, making a total of three extractions. Mucins were isolated by cesium chloride (CsCl) gradient centrifugation and size fractionated on Sepharose CL2B. Charge distribution was analyzed on ion-exchange chromatography with a lithium perchlorate (LiClO₄) gradient.

results. Contact lens–adherent mucins comprised soluble mucins and mucins that required solubilization by (repeated) dithiothreitol treatment. MUC1, MUC4, MUC2, and MUC5AC mucins eluted mainly at low buoyant densities in extractions from lenses worn long term without disinfection and at successively higher buoyant densities from monthly disposable contact lenses. Mucins with little negative charge, which were observed in all extractions, and very highly negatively charged species, present in the second and third extractions from contact lenses, had no equivalents in tissue-extracted mucins.

conclusions. Mucins adhering to contact lenses are altered forms of intracellular mucins. Different degrees of adherence of mucins to contact lenses may occur, either because of mucin characteristics or after mucin complexation with adherent materials. In the context of good contact lens hygiene, their presence may offer some protection from toxicants in the tear film, because mucins could function as acceptors for charged moieties such as free radicals.

Contact lenses are apposed to the ocular surface, sometimes continuously, for considerable periods. There is good evidence that contact lens wear alters the commensal flora of the ocular surface. ¹ It remains to be established whether the increase in number and diversity of colonizers is a response to the lens material, to an alteration in tear pH, ² to the tear deposits on the lens, or to a combination of these factors. ³ The presence of tear deposits alone is unlikely to control bacterial flora, because the ages of the lens and deposits do not determine the bacterial contamination of the lens. However, an accumulation of degraded tear components, may engender conditions locally that encourage bacterial colonization. 

Bacteriostatic or bactericidal enzymes and enzyme systems in the tears ^{4 5 6 7 8} are important factors in keeping most ocular surfaces culture negative. Surface mucus has a modulatory role in this context, ⁹ whereas mucin structure permits both pro- and antiadhesive properties. ^{10 11 12}  

Mucins have been identified and quantified in human and animal tears. ^{13 14 15 16} In this study, we explored the concept that mucus adhering to contact lenses may serve as an archive of changes (e.g., in glycosylation and size distribution) in mucin molecules that occur at the ocular surface during residence. To test this hypothesis, we first had to show that mucins are part of the tear deposits on lenses, that all mucin species present at the ocular surface can be extracted from these deposits, and that there is an overlap with the distribution of mucins obtained from conjunctival tissue, herein called intracellular mucins. 

Methods

Visualization of Mucins on the Surface of the Contact Lenses

Discarded daily disposable contact lenses were obtained from a normal female volunteer on several occasions. Reactivity with antibodies against mucin peptide core (Table 1) was assessed with the same techniques previously used in vacuum blots of agarose gel electrophoresis of tissue-extracted (intracellular) mucins. ^{17 18} Visualization on lenses was achieved with alkaline phosphatase–linked secondary antibodies (Dako, Glostrup, Denmark) and the 5-bromo-4-chloro-3-indoyl phosphate/nitroblue tetrazolium (BCIP/NBT) substrate system (Dako). Visualization on vacuum blots of 1% agarose gels was achieved with horseradish peroxidase (HRP)–linked secondary antibodies (Sigma, Poole, UK) and diaminobenzidine (DAB, Sigma). 

Collection of Contact Lenses

With the patients’ consent, 167 extended-wear contact lenses worn without disinfection were collected from asymptomatic patients attending the Optometry Department at the Bristol Eye Hospital. Lenses were discarded because of tear deposits that spoiled the lens optics. Most patients wore one of the three types of lens: E74 (filcon 4a polyxylon, 74% water content, DK × 10⁻¹¹ = 33 cm²/sec), I78 (copolymer filcon 3a, 79% water content, DK × 10⁻¹¹ = 64 cm²/sec), and PCM (filcon 4a polyxylon, 79% water content, DK × 10⁻¹¹ = 57 cm²/sec). The clinic population is mostly elderly (age range, 40 to 90, mode 80), with women outnumbering men two to one. All patients were free of ocular surface disease. Fourteen monthly disposable lenses (ocufilcon D, 55% water content) were also collected from a normal 23-year-old woman. These were analyzed separately from the continuously worn lenses. 

To prevent mucin degradation during storage, the contact lenses were kept at 4°C in 4 M GuHCl with protease inhibitors ¹⁹ until processed. 

Mucin Extraction and Purification

It has not yet been established which structural characteristics control the association of mucin with contact lens materials or with other tear material coating a used lens. Therefore, we used techniques that ensure that the whole complement of adhering mucins is extracted. Adherent mucins were grouped according to the density of glycosylation, and hydrodynamic volume distributions were analyzed within these groups. Further analysis of subunit charge was performed on pools of mucins with similar glycosylation density and similar volume in solution. 

To analyze the entire complement of mucins adhering to contact lenses, we used a published protocol to characterize intracellular mucins. ^{18 19 20} Briefly, mucins were extracted in the chaotropic solvent, 4 M GuHCl (Sigma), and each extraction was subjected to CsCl (Sigma) gradient density centrifugation. Contact lenses were pooled chronologically in groups of 15 to 25 and extracted in GuHCl for periods of 3 to 4 weeks to allow solubilization of adherent material. After removal of the supernatant, fresh GuHCl containing 10 mM dithiothreitol (DTT) was added to the same group of lenses, and a second extraction was performed at 4°C, for periods of between 7 and 10 days. A third extraction, with DDT-GuHCl, was obtained from continuously worn contact lenses that had been through the previous two extractions (a pool of 116 contact lenses). The three extractions were analyzed identically and separately. Monthly disposable contact lenses were treated in the same way: extracted in GuHCl followed by GuHCl-DTT. 

CsCl gradient density centrifugation was performed as previously described, ¹⁹ on an analytical centrifuge (Beckman Instruments, Palo Alto, CA) at 110,000g for 24 hours, at 10°C. Aliquots of 0.5 mL were removed, starting with the top of the gradient (most buoyant material) and their densities were determined by weighing. Mucin banding in the gradient is determined by interactions with the solvent dictated by the density of glycosylation and molecular packing. Mature mucins are expected to have buoyant densities between 1.35 to 1.5 g/mL. However, proteolytic cleavage may release fragments of lower and respectively higher glycosylation density and charge than the parent polymer. Lower-buoyant-density mucins can occur by loss of carbohydrate chains. Mucins with similar buoyant densities were therefore fractionated according to their volumes in solution. 

Gel permeation chromatography on Sepharose CL2B columns (Pharmacia-Biotech, Uppsala, Sweden) was used to obtain hydrodynamic volume distributions. Mucins of buoyant densities less than 1.3 g/mL, 1.3 to 1.4 g/mL, 1.4 to 1.5 g/mL, and larger than 1.5 g/mL from each extraction were applied to 80 × 1-cm columns in phosphate buffered-saline (PBS; pH 7.4). Forty 1-mL fractions were collected from each run, which encompass the excluded volume (V_o; fractions 7–12) and the terminal volume (V_t; fraction 25 onward). 

Ion exchange chromatography (MonoQ; Amersham Pharmacia Biotech, Little Chalfont, UK) was performed to assess the charge of mucin subunits. Samples were reduced and alkylated and then dialyzed against buffer MonoQ (6 M urea, 10 mM piperazine, 0.1% 3-([3-cholamidopropyl]dimethylammonio-2-hydroxy-1-propanesulfonate [CHAPS], pH 8). Mucin subunits were eluted with a 0- to 0.4-M linear gradient of LiClO₄ (Sigma) followed by the addition of a column volume of 0.8 M LiClO₄. The conductivity of 0.5-mL fractions was measured with electrodes (Millicel ERS; Millipore, Bedford, MA). 

Mucin Analysis

The presence of mucins was determined with dot blots on polyvinylidene difluoride (PVDF) membrane, ^{17 18 19} by reaction with antibodies to tandem repeat (VNTR) and non-VNTR domains of the mucin peptide core (Table 1) . Reactivity with ocular mucins and its specificity were checked by contact lens histology and Western blots of purified mucins. Wheat germ agglutinin (WGA), which reacts with sialic acids and N-acetylglucosamine, ^{21 22} was used as a general glycosylation detection reagent, and Molucella laevis lectin probed the presence of Tn and sialylTn epitopes. ²³ Reactivity was visualized by using peroxidase- or alkaline phosphatase–linked secondary antibodies, as appropriate. 

The intensity of reactivity with each antibody was quantified by image analysis. Because no suitable pure mucins are available to prepare calibration curves, the patterns of these intensities, but not their magnitudes, can be compared between different antibodies and between preparations. To facilitate comparisons and emphasize the relative ratios of the different subpopulations, all intensities have been scaled to equal maxima and presented as percentages of maximum reaction intensity. 

Results

Reactivity with Anti-mucin Reagents

The presence of mucin on discarded contact lenses is presented in Figure 1 . MUC2, MUC4, and MUC5AC were clearly visible at the margin of the lens (Fig. 1A 1B 1C) , whereas nonspecific binding (secondary antibody only) was minimal (Fig. 1D) . We could not visualize MUC1. Reactivity with anti-MUC2, anti-MUC4, and anti-MUC5AC antibodies was verified in Western blots of purified conjunctival mucins. ¹⁷ Figure 1E displays the different, but overlapping, mobilities of control purified intestinal intracellular and ocular surface MUC5AC and MUC2. Because of a limited tear mucin stock, the mobility of MUC4 could not be compared with membrane-extracted MUC4. ¹⁷  

Extraction in guanidine hydrochloride and buoyant-density centrifugation indicated the presence of MUC1, MUC2, MUC4, and MUC5AC adherent to contact lenses (Fig. 2) . A single extraction did not yield the whole mucin population adherent to the contact lenses. Addition of a fresh portion of guanidine with DTT solubilized more mucins, representing all the mucin gene products shown. A third extraction (with DTT), from the same lenses pooled together (116 contact lenses), also isolated all gene products identified in conjunctival intracellular mucins (Fig. 2) . No further mucin could be extracted from the contact lenses. 

In the first extraction, MUC1 displayed a peak between 1.25 and 1.3 g/mL buoyant density (Fig. 2A) , similar to its membrane-extracted form in the tissue. Reactivity with anti-MUC4 antibodies (Fig. 2B) indicated the presence of this mucin at low buoyant densities and also at 1.3 to 1.4 g/mL, the buoyant density range usually associated with mature mucins. Profiles of MUC2 were similar in conjunctiva ¹⁸ and lens extractions in GuHCl (not shown), whereas MUC5AC peaked at both lower and higher buoyant densities than did the intracellular mucins (Figs. 2C 2D) . Addition of the reducing agent DTT to the extraction buffer caused the solubilization of all mucin species, mainly those of low buoyant density. With the exception of MUC1 (Fig. 2A) , reaction intensities decreased rapidly toward the mature mucin buoyant range. Small additional peaks were noted at buoyant densities greater than those in intracellular mucins (Fig. 2A 2B 2C) . It is notable that in the third extraction, MUC5AC reactivity displayed a trough between 1.3 and 1.4 g/mL (Fig. 2D) , as did the reactivity of MUC2 (not shown). 

Monthly disposable lenses also showed adherent mucins. The patterns were different, however, from those on continuously worn devices. The extraction of nonreduced MUC1 peaked in the ranges 1.2 to 1.3 and 1.4 to 1.45 g/mL, whereas the presence of DTT shifted the peaks of reactivity to 1.3 to 1.4 and 1.45 to 1.5 g/mL (Fig. 3A) . A tendency toward higher buoyant densities was also observed for second-extraction MUC2 (Fig. 3B) and for MUC4 and MUC5AC (results not shown), for which reactivity extended beyond 1.5 g/mL buoyant density. 

Another facet of mucin heterogeneity is the distribution of hydrodynamic volumes, assessed in this study for mucins with similar buoyant densities. Size fractionation on Sepharose CL2B columns of mucins adherent to long-term-wear lenses indicated mainly large (excluded volume, V_o, >4–5× 10⁶ Da) or small (terminal volume, V_t, <10⁵ Da) hydrodynamic volumes in all extractions (Fig. 4) . For example, in the 1.3- to 1.4-g/mL pool (i.e., classic mature mucins), the reactivity with the antibody LUM 5-1 peaked at V_t in all three extractions. V_o reactivity was evident in the first and third extractions, but was missing in the second (Fig. 4A) . In contrast, anti-M1 did not react with mucins in the V_t of the first or second extraction, but peaked in V_o (not shown). These differences between the reactivities of two antibodies against non-VNTR epitopes of MUC5AC suggest cleavage of the peptide core. WGA reactivity peaked at V_t in the first extraction, whereas in the second extraction there was a V_o peak additional to the V_t maximum (Fig. 4B) . A comparison between WGA-positive material and patterns of reactivity with anti-mucin antibodies revealed the presence of WGA-negative mucins. These forms have no intracellular equivalents. 

The distribution of subunit charge was followed by ion-exchange chromatography (MonoQ; Amersham Pharmacia Biotech) to reveal any changes (compared with intracellular mucins) that may have occurred during the mucins’ residence on the ocular surface. Similar WGA profiles were obtained for mucins of similar buoyant densities, regardless of extraction, as exemplified in Figures 5A and 5B . However, charge fractionation of the largest (V_o) mucins within a narrow range of buoyant density led to the identification of low negative-charge subunits and of forms with high negative charge (Figs. 5C 5D 5E 5F) . These extreme subunit charges were absent in intracellular mucins. ¹⁸ Reactivity with anti-MUC2 and anti-MUC5AC antibodies revealed WGA-negative mucins (Figs. 5D 5E) , with mucin-species–specific distribution. For example, in the second extraction 1.3 to 1.4 g/mL MUC4 eluted at higher negative charges than MUC2 (Fig. 5F) . However, MUC2 and MUC5AC with similar buoyant densities showed similar profiles in consecutive extractions. 

Discussion

The hypothesis tested in this study is that, if mucins adhere to contact lenses, they can be investigated to reveal mucin processing on the ocular surface. Smaller mucins point to proteolytic cleavage and shifts in buoyant density and subunit charge reveal changes in oligosaccharide chains. 

Adherent mucins were extracted from extended-wear contact lenses worn by a group of mostly elderly, asymptomatic patients, and were analyzed separately from those on monthly disposable lenses collected from one volunteer. Adherence to contact lenses did not appear selective because all mucin gene products identified in conjunctival cells were detected in the adherent material, and because the buoyant densities of adherent mucins encompassed those of intracellular conjunctival populations. It is possible that each contact lens polymer and mode of wear results in a unique combination of adherent mucin populations, but such an analysis was beyond the scope of this study. 

Different forms of mucins adhered to lens materials to different degrees, as suggested by the different solubilization requirements (Fig. 2) . Mucins adherent to contact lenses formed deposits designated soluble if they did not require DTT for solubilization and insoluble if they did. ^{24 25} Some mucins may have been trapped within the insoluble complexes, even though they themselves were unaffected by DTT. Thus, the presence of monomeric mucins (e.g., MUC1 and MUC4) in the insoluble phase should be interpreted as mucin–mucin complexation. A single extraction with DTT was insufficient to free all adherent mucins (Figs. 2 3) , suggesting the presence of bonds with different reduction sensitivities between mucin molecules or between mucins and lens materials or other adherent molecules. All investigated mucins species occurred in both soluble and insoluble phases. The formation of insoluble mucins may be part of tissue physiology or disease. In the colon most of the mature secreted MUC2 is insoluble. ²⁶ In cystic fibrosis mucus, however, although MUC5AC and MUC5B can be found in the gel, soluble, and insoluble phases, MUC2 appears exclusively in the insoluble material. ²⁷ Our unpublished results indicate that normal surface mucus contains insoluble mucins (Berry et al., unpublished data, 2000). The appearance of insoluble mucins is thus likely to reflect mucin processing, possibly enhanced by contact lens wear. Continuous wear over a prolonged period is not necessary for the adsorption of insoluble mucin aggregates. Regularly cleaned monthly disposable lenses also yielded further mucins after extraction with GuHCl-DTT (Fig. 3) . 

Transmembrane mucins MUC1 and MUC4, which require detergent solubilization from membrane preparations of conjunctival tissue, were found in all extractions from contact lenses. Their presence clearly indicates not only scission from apical membranes, but also diffusion through the preocular gel. These results strengthen the evidence that the preocular gel is the source of tear-dissolved mucins. ^{15 16} Diffusion of polymers with a radius of gyration of approximately 57 nm ²⁸ or the persistent lengths of approximately 30 nm determined for ocular mucins ²⁹ suggest a large pore size for the preocular gel, which would support diffusion of mucins (analogous to the good diffusion of mucins through a 1% agarose matrix, but not in a polyacrylamide gel). 

Although polydisperse, hydrodynamic volumes of mucins adhering to lenses were mostly smaller than tissue-extracted polymers, ^{19 18} indicating proteolytic cleavage during ocular surface residence, including while adherent to the contact lens. Differences between first-extraction MUC patterns and those of their intracellular counterparts can also be interpreted as evidence for the postsecretory processing of ocular mucins. Changes in oligosaccharides could be discerned from the banding of contact lens–adhering mucins in density gradients. This is largely determined by the density of glycosylation, sometimes interpreted as density of charge. Insoluble mucins adhering to continuous-wear contact lenses were mostly underglycosylated compared with their soluble counterparts and also less glycosylated than mucins extracted from monthly disposable lenses and from intracellular mucins (Figs. 2 3) . The notion that glycosylation influences the degree of adherence is also supported by the paucity of MUC2 and MUC5AC of buoyant density 1.3 to 1.4 g/mL in the third extraction (Fig. 2) . However, insoluble mucins adhering to monthly disposable lenses were more densely glycosylated than their soluble analogues. An explanation for this discrepancy may be that disinfection reduces glycolytic enzymes by decreasing bacterial load. Alternatively, this difference could be assigned to the period of adherence on the contact lens. The link between adhesion and molecular charge was further explored by ion-exchange chromatography. Mucins obtained in the third extraction (e.g., V_o WGA-positive mucins 1.3–1.4 g/mL) contained highly negatively charged forms not present in the analogous fraction of soluble mucins. The main WGA-positive material, however, eluted at similar charges in all extractions, suggesting that adhesion is not simply related to molecular charge, but may be mediated by specific moieties. A systematic analysis of the role of contact lens material surface chemistry and charge on mucin adherence was outside the scope of the present study. 

In intracellular mucins, we did not detect any WGA-negative forms. Indeed, WGA profiles of purified mucins overlapped with, and were wider than, individual mucin-elution profiles. In mucins extracted from contact lenses, however, WGA-negative mucins eluted at low negative charge and also at high negative charge. Low-negative-charge mucins may have arisen as a result of selective cleavage of charged moieties, rather than nonselective oligosaccharide stripping, because they were present among mucins with mature glycosylation (>1.4 g/mL). Because charge is carried mainly on oligosaccharide chains, cleavage between a poorly and a highly glycosylated region could account for the presence of extreme charge forms. It is not known whether this cleavage is effected by tear or bacterial enzymes or appears as an artifact of reduction and alkylation in mucins that have already undergone some degradation in the tear film. Appearance of a highly charged form may, however, also follow from “mopping up” by mucins of active radicals, as proposed by Allen ³⁰ some years ago. 

All the mucin species detected in conjunctival tissue adhered to contact lenses and their hydrodynamic volumes and charges were different from intracellular conjunctiva-extracted mucins. In addition to cellular processing during secretion, some of these changes may have been caused by tear film components during residence in the preocular fluid or while adherent to the lens. Interactions between the duration of mucin residence on the contact lens, the material composing the contact lens, and cleansing regimen have yet to be fully analyzed. Within the constraints of an optically perfect contact lens and good lens hygiene, adherent mucins may offer a buffering substrate for enzymes and active radicals of the tear film. 

 Supported by the Guide Dogs for the Blind Association, United Kingdom.

 Submitted for publication July 16, 2002; revised October 1, 2002; accepted October 7, 2002.

 Disclosure: M. Berry, None; A. Harris, None; A.P. Corfield, 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.

 

Table 1.

 

Specificities of Antibodies against Mucin Peptide Core

Mucin	Antibody	M/P*	Peptide Location
MUC1	BC2	M	VNTR 31 , †
MUC2	LUM2-3	P	C terminal to VNTR 32
MUC4	M4.171 and M4.275	M	VNTR 33
MUC5AC	anti-M1	M	C terminal 34
MUC5AC	LUM 5-1	P	C terminal to VNTR 32

Figure 1.

 

Mucins adhering to contact lenses. Mucins were identified by reactivity with antibodies against sequences in the mucin peptide core (Table 1) and visualized with alkaline phosphatase–tagged second antibodies and BCIP/NBT substrate. (A) Reactivity with LUM2-3, an antibody to peptide core of MUC2. (B) Reactivity with anti-M1, an antibody to the peptide core of MUC5AC. (C) Reactivity with anti-MUC4 antibodies. (D) Control lens incubated with alkaline phosphatase-linked secondary antibody. Lenses were scanned at original size. (E) Ocular MUC5AC, detected by LUM5-1 (lane 1: intestinal mucin; lane 2: tears) and MUC2, detected with LUM2-3 (lane 3: intestinal mucin; lane 4: ocular surface mucin) display different, but overlapping, mobilities on 1% agarose electrophoresis. The samples consisted of mucins of buoyant density 1.35 to 1.4 g/mL. Lane RB: high-molecular-weight color markers. HRP-linked secondary antibodies were visualized with DAB.

Figure 2.

 

Banding on density gradients. Buoyant density distribution of mucin gene products in extractions from conjunctival tissue (intracellular mucins) compared with three consecutive extractions from continuously worn contact lenses (CL). (A) Reactivity with BC2 anti-MUC1 in membrane extractions of conjunctival tissue (filled circles), and first (gray circles), second (open diamonds), and third (gray line) extractions from contact lenses. (B) MUC4 detected by M4.171 and M4.275 in tissue (filled circles) and the first lens extraction (gray circles). (C) Cross reaction with LUM5-1 against MUC5AC in tissue (filled circles) and the first contact lens extraction (gray circles). (D) Cross reaction with LUM5-1 against MUC5AC in the second (open diamonds) and third (gray line) extractions from extended continuously worn contact lenses. To ease comparisons between different antibodies and preparations, all profiles have been scaled to equal maxima and arbitrary units have been deleted from the abscissa. The proportion showing maximum reactivity is illustrated, to facilitate comparison.

Figure 3.

 

Buoyant density distribution of mucin gene products in extractions from monthly disposable contact lenses worn by one individual. (A) Cross-reaction with BC2 anti-MUC1 in the first (gray circles) and second (open diamonds) extractions. (B) Cross reaction with LUM2-3 anti-MUC2 in the first (gray circles) and second (open diamonds) extractions.

Figure 4.

 

Hydrodynamic distribution of mucins adherent to lenses. Material of similar buoyant density (i.e., within 1 g/mL) was pooled and fractionated by size-exclusion chromatography on a Sepharose column. V_o mucins eluted in fractions 7 to 12, whereas mucins eluting after fraction 25 were too small to be separated by the column (V_t). (A) Reactivity of mucins of buoyant density 1.3–1.4 g/mL with antibody LUM 5-1 in the first (small open circles), second (black circles) and third (large open circles) extractions. (B) WGA reactivity in material of 1.3–1.4 g/mL buoyant density in the three consecutive reactions produced wider peaks than those of MUC5AC.

Figure 5.

 

Distribution of subunit charge in mucins adherent to extended-wear contact lenses. Patterns of reactivity are shown against the conductivity of the eluate. Vertical dotted line: start of the lithium perchlorate gradient. Reactivity with WGA in: (A) 1.3–1.4 g/mL material in the first (filled circles) and second (open diamonds) extractions; (B) first extraction material buoyant density 1.3 g/mL (filled circles) and >1.5 g/mL (open circles); (C) V_o fraction buoyant density 1.3–1.4 g/mL first (filled circles) and third (open squares) extractions. Note the low negative- charge peaks and the highly charged forms present. Mucin gene products had a wider charge distribution than WGA or Molucella laevis lectin; (D) profiles of MUC5AC detected with anti-M1 (filled circles) and WGA (gray line) in material of buoyant density 1.4–1.5 g/mL, first extraction; (E) MUC2 (open circles) and Molucella laevis (filled circles) in second extraction material of buoyant density >1.5 g/mL; (F) MUC2 (open diamonds) and MUC4 (filled diamonds) in second extraction mucins with buoyant density 1.3–1.4 g/mL.

The authors thank the patients and staff at the Division of Optometry, Bristol Eye Hospital for collecting discarded contact lenses; Fiona Lawrence and Caroline Routledge for saving their disposable lenses for our research; and Jacques Bara, Ingemar Carlstedt, Michael McGuckin, and Nathan Sharon for gifts of research antibodies and lectin.