Studies on the isolation and composition of human ocular mucin

Eq

Eyr

Res. (1988)

47, 185196

Studies CHIEN-CHYOU

on the Isolation and Composition Human Ocular Mucin W. CHAO*,SUSAN

M. BUTALA

AND

of

ANTHONY

HERP~

Department of Ophthalmology, University of Pittsburgh School of Medicine. Pittsburgh, PA 15213, U.S.A. and t Department of Biochemistry, New York Medical College, Valhalla. NY 10595, U.S.A. (Received

15 January

1987 and accepted in revised form

5 January

1988)

A method for the isolation and purification of human ocular mucin from the brief saline extract of human ocular mucus is reported. Initial purification of ocular mucin was achieved by sequential chromatography of the saline-soluble mucus extract from an individual donor’s mucus pool on columns of Sephadex G-50 and Sepharose CL-4B. A portion of such mucin isolate was subjected to quantitative analysis of the 0-seryl (threonyl))N-acetylgalactosaminyl linkage, characteristic of mucins, by alkaline p-elimination and tritiated borohydride reduction. Following Rio-Gel P-2 filtration. the mucin isolate whose cleaved oligosaccharides contained tritiated galactosaminitol > 0.5 ,&i mg-i, a value that represents at least 64% of that observed for bovine and ovine submaxillary reference mucins, was considered to be mucin-rich. These isolates were subjected to further purification on Sephacryl S-500 and DEAE-Trisacryl M column chromatographies. The purified mucin had a minimum molecular weight of 120 kDa. It consisted of 25-30 % protein and 54-55 o/u carbohydrate. Its amino acid and carbohydrate compositions are characteristic of a mucin structure. The purity of the mucin was verified by SDS-gradient PAGE. Upon isoelectric focusing, polydispersity/microheterogeneity were exhibited in the pZ range WM.6. Key ovords: human ocular mucin; gel and ion exchange chromatographies: 0-glycosidic linkage: amino acid composition : carbohydrate composition.

1. Introduction Mucus glycoproteins (mucins) are generally considered to be the major constituents of the human ocular mucus, and are essential in maintaining the stability of the tear film (Holly and Lemp, 1977). The ocular mucins, which are mainly secreted by the conjunctival goblet cells, provide lubrication, act, as receptors for hormones and viruses, remove dust particles, and protect against mechanical injury (Wright and Mackie, 1977; Fatt, 1978). Ocular mucins have been identified histologically (Matsumoto and Mimura, 1974; Srinivasan, Worgul, Iwamoto and Merriam, 1977) by periodic acid-Schiff (PAS) staining following gel electrophoresis (Moore and Tiffany, 1979), and more recently, by chemical studies (Moore and Tiffany, 1981; Chao, Vergnes and Brown, 1983a, 1983b; Chao and Vergnes, 1983). Isolation procedures of ocular mucins have been hampered by the scarcity of ocular mucus specimen and by the abundance of contaminating components, notably tear and serum proteins, lipids, and glycosaminoglycans (Moore and Tiffany, 1981; Chao, Vergnes and Brown, 1983a; Chao and Brown, 1986). Our continuing efforts to isolate a pure ocular mucin have included investigation of the parameters relating to time, type, and sequence of solvent extraction, the quality of an individual donor’s mucus based on mucin content, and column chromatographic procedures for purification. A detailed methodology of isolation and purification, coupled with preliminary characterization of human ocular mucin, is reported in this paper. * To whom correspondence 203 Lothrdp St.. Pittsburgh, 0014~4835/88/080185+

should be addressed PA 15213, U.S.A.

12 $03.00/O

at Rm. 911 The Eye and Ear Institute

of Pittsburgh,

0 1988 Academic

Press Limited

(‘.A’. IV. (‘HA0

1X6

2. Materials

and

F:T i\l, Methods

Bephadex G-50-80, Sepharose (‘T,-4B-200. and SrJ)hacryI S-T,OU were l)ur(*haserl frown Pharmacia Fine Chemicals, Piscataway. N.J. Bio-Gel P-2 was from Bio-Rad. Richmond. (‘A, and JIEAE-Trisacryl M from LKB. Gaithersburg. MT>. Bovine and ovine submaxillarymucins (BSM and OSM) were isolated in the glycoprotein research lab at ?ie\v York Mrtlical College. Valhalla. NY. using the method of Tettamanti and Pigman (1968). Sodium /“HI borohydride (347.8 mCi mmol-‘) was purchased from New England Nuclear (‘orl).. l
of human

ocular

muws

Collection of the ocular mucus and the procedure for the saline extraction of the mucus were as previously reported (Chao, Vergnes and Brown, 1983a). Briefly, mucus was collected from the outer canthus of the eyes of adults of both sexes 20-45 yr of age. individual]? pooled, and stored at -20 “C. All volunteers were free from eye disease and had no medication. No attempts were made to subdivide the donors according t,o any additional parameters. The dry weight of the crude mucus pools varied from 4- to 15 “A. The dry mucus contained approx. 49 % proteins, 20 % lipids, and 22 %I carbohydrates (Chao, Vergnes and Brown, 1983a). Saline-soluble mucus extra&o,, The individually pooled mucus specimen, upon accumulation of approx. 50 mg wet wt. was gently stirred with five volumes of ice-cold 0154 M NaCl containing @02 % NaN,. 10 nlM Na,EDTA, and 2 mM phenylmethylsulfonyl fluoride for a period of 30 min at 4 “C. The insoluble mucus residue was recovered by centrifugation at 12000 g for 10 min, and was reextracted three more times as described above. The supernatants were combined and stored at -2O’C. Preliminary

mucin

isolation

by gel jiltrations

The saline-soluble mucus extract from an individual donor was dialyzed against water in the cold, lyophilized, dissolved in a small volume (25 ml) of @05 M Tris-HCl-002 % NaN,, pH 7.4, and applied to a column of pre-equilibrated Sephadex G-50 (1.6 x 85 cm). The column was eluted with the same buffer at 4 “C at a flow rate of 7 ml hr-‘. Fractions (2.5 ml) were monitored for protein absorbance at 280 nm. The void volume peak was collected, dialyzed against water, lyophilized, dissolved in 0.154 M NaCla.O2% NaN,, applied t,o a column of Sepharose CL-4B (1.6 x 85 cm) equilibrated with the same buffer. and eluted as before. The void volume fraction was pooled, dialyzed against water, lyophilized, and a portion of the sample was subjected to analysis of seryl(threonyl)-N-acetylgalactosaminyl type 0-glycosidic linkages. Alkaline

borohydride

treatment

The void volume fraction from the Sepharose CL-4B mucin isolate of an individual donor. about 0.5 mg, was subjected to reductive alkaline p-elimination in the presence of 3H-labeled sodium borohydride. Samples of BSM and OSM used as reference mucins were treated similarly. Treatment was selectively chosen from three methods (Takasaki and Kobata, 1974; Downs, Peterson, Murty and Pigman. 1977; Ogata and Lloyd, 1982), with minor modifications. Each sample was incubated with @2 ml of @6 M NaB3H, (7 mCi mmol-I) 4.1 N NaOH at 45 “C! for 20 hr. Afterwards, the reaction was stopped with one drop of octanol and neutralized with glacial acetic acid. Following removal of exchangeable tritium as 3H,0 and excess boric acid by repeated evaporation in vacua, the sample was applied to a column (1 x 10 cm) of Dowex 5OW-X8, (H’ form) for removal of salt and cleaved peptides using water as the eluant. The reduced [3H]-oligosaccharides were fractionated at room temperature on a Bio-Gel P-2 column (1.6 x 85 cm) equilibrated and eluted with @5 M ammonium acetate, pH 68. Fractions (2.5 ml) were monitored for 3H radioactivity by a Beckman model LS 1800 liquid scintillation counter using Beckman Ready-Solv scintillation fluid. The major fractions were analyzed after acid hydrolysis and re-N-acetylation

HUMAN

OCVLAR

>1rC I S

(Takasaki and Kobata, 1974) for cleaved and reduced sugars by paper chromatography a solvent system of butanol-ethanol-water (lo:3 : 5, v/v) at room temperature. quantitation of labeled sugars involved cutting the chromatograms into 05-cm-wide for radioactivity measurement. Gel and

ion

exchange

187 using Microstrips

chromatographies

Mucin-rich isolates from individual donors were pooled and fractionated at 4 ‘(1 on a column of Sephacryl S-500 (1.6 x 85 cm) equilibrated with 0154 M NaCl+OZ % NaN, and eluted under the gel filtration conditions described above. The fractions were monit’ored at 280 nm for protein. After dialysis and lyophilization. t’he major fraction in the included volume was re-dissolved in 0.05 M Tris-HCl. pH 7.0. and subjected to ion-exchange chromatography on a column (1.6 x 185 cm) of DEAE-Trisacryl M previously equilibrated with the same buffer. The column was eluted with a gradient of increasing Na(‘l concentration. from 09 to 95 M. at a flow rate of 20 ml hr’. Immunological

method

Double immunodiffusion was performed according to the method of Ouchterlony and Nilsson (1978). Specific antisera to serum proteins were purchased from Accurate Chemical Scientific Co., Westbury, NY. Ocular mucoisolate was a high molecular weight (> lo5 Da) frartion isolated from the saline-extractable human ocular mucus by Sepharose CT&B chromatography as described previously (Chao, Vergnes and Brown, 1983a). Antisera to human tear and ocular mucoisolate were prepared as previously dasrribed (Chao. Butala. Zaidman and Brown. 1987). Analytical

,naethods

Protein was assayed by the method of Lowry. Rosebrough, Farr and Randall (1951). For amino acid analysis, samples were hydrolyzed in 6 N HCl at 107 “C under N,, for 24 hr (Chao. Vergnes and Brown, 1983a). After acid removal, amino arids were analyzed on a Beckman System 890M fiequencer. Factors used to correct for the destruction of serine and threoninr under these hydrolytic conditions were 1.06 and 1.03, respectively, based on our control experimental data. Neutral hexose was assayed by the phenol-H,SO, method of Dubois. Gilles, Hamilton. Rebers and Smith (1956). Sialic acids were measured by the thiobarbiturie acid method of Aminoff (1961) with N-acetylneuraminic acid as the standard. Neutral and amino sugars plus sialic acid were analyzed by GLC. using the method of Clamp, Bhatti and Chambers (1971), as previously described (Chao. Vergnes and Brown, 1983a). Briefly, samples were methanolyzed in 1 N methanolic HCl for 24 hr at 100 “C; followed by acetylation and trimethylsilylation. The trimethylsilylated carbohydrate derivatives were apphed to a Packard Model 419 gas chromatograph with a glass-lined injection port. The derivatized sugars were separated on a silanized glass column (2 mm i.d. x 6 ft), packed with 106200 mesh Chromosorb WHP coated with 3% SE-30 (Supelco. Bellefonte, PA) with temperature programming from 120- to 200 “C at 05 ‘C mini. Internal standards (mannit,ol and perseibol) were added to each sample prior to methanolysis. Model glycoproteins usrd were fetuin (22 “/u carbohydrate) and bovine submaxillary mucin (56 % carbohydrat,e). Sulfate was determined by the method of Terho and Hart’iala (1971). and uranic acid by the procedure of Bitter and Muir (1962). DNA was assayed by the diphenylamine method of Richards (1974). Lipid content was assayed by the method of Marsh and Weinstein (1966). SDS-PAGE was performed according to the Laemmli system (Laemmli, 1970) on a 51.5 % gradient running gel with a st’acking gel containing 3 % acrylamide and 62 % agarose with minor modifications as described previously (Chao. Vergnes and Brown, 1983a). The molecular weight of the isolated mucin was estimated by SDS--PAGE using a range of acrylamide concentrations of 3.5 ; 5 ; 65 ; 8 ; and 95 %. similar to the procedure of Segrest and Jackson (1972). Plots were made of the log molecular weight vs. relative mobility of standard proteins (Bio-Rad), which included ovalbumin, 45 kDa: bovine serum albumin. 66.2 kDa: phosphorylase B, 92.5 kDa; P-galactosidase, 116.2 kDa; myosin. 200 kDa. Agarose horizontal isoelectric focusing, in the p1 range 3.8-8.6, was performed as described previous13 (Chao and Butala, 1986).

3. Results

and

Discussion

Het,ween the two solvent systems previously used for the extraction of human ocular mucus, more muck was found in the saline than in the urea extracts ((rhao and Brown, 1986). In this st#udy, saline was therefore chosen to extract’ the IIIU(~LIS. In examining the time rourse of saline extraction, an individual donor’s mucus was divided into two portions, one subjected to brief extraction over a 2-hr period, the other to exhaust,ive extraction for 1 day. We concluded that brief extraction of the ocular mucus was most suitable in exkacting readily soluble proteins and glycoproteins while minimizing lipid (glycolipid) levels in the saline extract. Specifically, in comparing the composit,ion of the two methods of solubilizat’ions by saline from seven individual donors. based on dry weight. the briefly ext’racat’ed soluble mucus contained on average similar amounts of proteins (5’i *4’%, vs. 55f5%). hexose (12-&4”/, vs. 13&3%). hexosamine (9+3%, vs. 10f3%), and sialic acid (11 k 3 % vs. 9 &- 3 %). but significantly less lipid (4 + 2 % vs. 14 f 4 ‘!A) than when extract’ion was performed for 24 hr. The decrease in the percentage of lipid should simplify the isolation of ocular mucin.

“t 4

“0 1.2-

4

ELUTION

VOLUME

(ml

1

FIG. 1, Sephadex O-50 column (l%rfr85 cm) chromatography of the saline 4 mg dry wt. eluted with 0.05 M Tris-HC1-002 % NaN,, pH 7.4. at 4 “C. Flow monitored for protein at 280 nm.

Initial

isolation

extract of ocular mucus. rate :7 ml hr-‘. Fractions

of muck

Figure 1 shows a typical elution profile of the saline-soluble mucus extract following chromatography on Sephadex G-50. Four fractions were resolved, of which the major one was eluted in the void volume. Yields resulting from application of 4 mg dry mucus extract to the column were: Fl, 2.4 mg; F2, @5 mg; F3, O-9 mg; F4, @2 mg. By double immunodiffusion against various specific antisera to human tear. serum

HLTivIAS

O(‘I-LAR

Ml-(‘IN

1x9

proteins and ocular muroisolate (see Materials and Met#hods), the void volume fraction was found to contain predominant’ly ocular mucosubstance plus serum albumin. and fractions 2-4 to contain 18 tear and serum proteins which have been previously ident,ified (Chao. Butala. Zaidman and Brown. 1987). Lysozyme was found in fract,ions 3 and 4, amounting to about 0.24 mg. Through our preliminary evaluations of the chromatographic methods for mucin isolation and based on earlier studies by other researchers (Jenssen and Smidsrod. 1982: Snyder, Nadziejko and Herp, 1982), lysozyme, presumably due to its basic nature, was found to bind with mucoisolate. The elution process of Sephadex Ci-50 gel filtration at mildly alkaline pH and 1 Y’ was chosen to dissociat’e lysozyme from ocular mucosubstance. It is an effective step in mucin purification.

ELUTION

VOLUME

(ml 1

Fro. 2. Sepharose CL-4B column (1.6 x 85 cm) chromatography of the Sephadex G-50 fraction. 2.4 mg. &ted with 0.154 IV! N~CMKC~% NaN, at 4 “C. Flow rate: 7 ml hr-‘.

void

volume

The void volume fraction was subsequently subjected to chromatography on Sepharose CL4B (Fig. 2). a procedure previously developed for the separation of ocular mucosubstance from lower molecular weight tear and serum proteins (Chao, Vergnes and Brown, 1983a). Four fractions were resolved. Again, fractions 2-4 contained both tear and serum proteins, of which serum albumin was predominant in fract.ions 2 and 3. The mucin was eluted in the void volume and amounted to approx. 1.4 mg, representing approx. 35% of the dry mucus extract. Relative compositional 12 % lipids, 1% sulfate, and 54 % analysis revealed, on average, 33% proteins, carbohydrates, which includes 3 % uranic acid. Selection

of hiyh-quality

muci~

The ocular mucus sample a daily mucus secretion of (Ehlers, Kessing and Norn, among individual donors, it

isolates

from an individual donor represents an accumulation of about 2.2 ,ul collected over an extended period of days 1972). Since there are large variations in mucin content’ is necessary in the early stages of mucin purification to

190

(‘. (‘. \V. (‘HA0

20

40 ELUTION

60

80

loo

VOLUME

I20

140 (ml

160 )

ET

AI,

204060608000l2Ot4Ol6O ELUTION VOLUME

(ml

J

FIQ. 3. Bio-Gel P2 elution patterns of labeled oligosaccharides derived from alkaline tritiated borohydride treated mucin samples, 0~5 mg each. (a) BSM; (b) OSM; (c) muck isolate of low total “H count, @21 PCi mg-‘. and (d) of high total 3H count. @66 ,&i mg~‘. Eluant : WO5 M ammonium acetate. pH 6-8. Column size: 1.6~85 cm.

distinguish between those mucus specimens displaying chemical profiles characteristics of mucin and those which are poor in mucin and therefore of lower quality. For this purpose, a small portion of the Sepharose CL-4B mucin fraction isolated from an individual donor was subjected to quantitative analysis of the 0-Ser(Thr)-GalSAc linkage characteristic of mucin by alkaline borohydride treatment (see Materials and Methods). The cleaved tritiated oligosaccharide solution was fractionated on Bio-Gel P-2. Typical column profiles of the ocular mucin samples, showing both low and high extreme tritiated counts, as well as profiles of the reference mucins BSM and OSM, are shown in Fig. 3. The solutions of BSM and OSM, after alkaline p-elimination, each produced two major peaks at identical elution positions, whereas the ocular mucin sample solut,ion showed two minor peaks and one major peak. The major peak appeared at the elution position corresponding to the second-eluting peak of the model mucins. The average total count in ,uCi mg-’ for BSM and OSM was about @78 and 05’6, respectively. The count for the ocular mucin samples varied from @21 to 0.66. However. about 75 % of the tested samples were above 05 ,uCi mg-‘. Following hydrolysis and re-N-acetylation (see Materials and Xethods). the products of the oligosaccharide alditols were analyzed by descending paper chromatography (Fig. 4). The major fraction of each of the reference submandibular mucins appears to contain only N-acetylgalactosaminitol, indicating thats each of these fractions resulted exclusively from the cleavage of 0-glycosidic linkage. Likewise, the major fraction of the ocular mucin samples appears to contain primarily N-acetylgalactosaminitol. In addition, a small amount of N-acetylglucosaminitol was detected as indicated by the shoulder occurring on the left side of the peak. This may have possibly resulted from the cleavage of N-glycosidic linkages of N-acetylglucosaminyl asparagine of the mucin isolate (Rasilo and Renkonen, 1981; Ogata and Lloyd, 1982; Hounsell, Pickering, Stoll, Lawson and Feizi, 1984). Based on these analyses, further ocular mucin isolation was conducted using mucin

ILGaINA+,,

600 400 2 z 5 F 2:: 0

200

i

600

0

(a )

GalNAc,,

(b)

c) 200

* 0

k

400

L 0

IO DISTANCE

20 30 FROM ORIGIN (cm)

Fm. 4. Graphs of radio paper chromatograms of [3H]alditols, N-acetylgalactosaminitol, and Nacetylglucosaminitol in the hydrolysate of the oligosaccharides fraction produced by alkaline borohydride treatment. The sugars were separated by chromatography on Whatman No. 1 paper in butanol-ethanol-water (10 : 3 : 5, v/v) for 40 hr. Deoxyribitol was used as an internal marker. Spots were visualized with periodatebenzidine reagent. (a) BSM or GSM; (b) ocular mucin isolate. *Peak of Sacetylglucosaminitol.

,-

O

20

40

60

60 ELUTION

100

120

VOLUME

(ml)

140

160

I60

FIG. 5. Sephacryl S-500 column (1.6 x 85 cm) chromatography of pooled, high-quality 5 mg, eluted with 0.154 M NaC1-002 % NaN,, at 4 “C. Fractions were monitored for

isolates

from

individual

donors

whose

cleaved

oligosaccharides

mucin

isolates,

protein at 280 nm.

contained

tritiated

galactosaminitol with a value greater than @5,&i mg-‘. This value corresponds to about two-thirds of the tritiated oligosaccharidescount for the reference mucins BSM and OSM. purijcation of mucin by gel and ion exchangechromutographies Sepharose CL-4B mucin-rich isolates from individual donors were pooled and further fractionated on a column of Sephacryl S-500 (Fig. 5). This procedure Further

(‘4’.

192

\V. (‘HA0

I”‘I’ 1

;\I 1

separated the muck from any I)NA that was present in the void volume frat*tion. based on analysis by the diphenylamine method (Richards, 1974). and also from proteins in t,he lower molecular wright’ fractions c~luted in the inc*luded volumr~. Typica,lly. upon gel filtration of 5 mg of a pooled Sepharose CL-4B mucin isolate. we recovered an average of 24 mg of mucinrontaining fraction, OY! mg of the fraction containing DNA, and 0.2-0.4 mg of low molecular weight protein fractions. The Sephacryl S-500 mucin fraction was subjected to a final purification by DEAETrisacryl M ion exchange chromatography. The mucin fraction was eluted in the gradient at approx. 0.2 M NaCl (Fig. 6). The yield of the major frac.tion represented 50-60X of the mucin isolated upon Sephacryl S-500 chromatography.

0.5

0.4

0.3

0.2

f : =

0. I

ELUTION

VOLUME

(ml

)

FIG. 6. Further purification of mu& (5.4 mg) isolated from Sephacryl S-500 by DEAE-Trisacryl ion exchange chromatography. Column size: 1.6 x 185 cm. Elution was carried out at 4 ‘C with Tris-HC1, pH 7.0, using a gradient of N&l from CO-05 M. Flow rate : 20 ml hr-‘.

Preliminary

characterization

of ocular

M @05

M

muck

SDS-(515%) gradient-PAGE of the isolated ocular mucin under reduced conditions displayed a single intense PAS-positive band in the molecular weight region > 2 x lo5 Da, which stained only faintly with Coomassie Brilliant Blue R-250. These staining patterns were virtually identical to those observed in our earlier studies with ocular mucin isolates (Chao. Vergnes and Brown, 1983a). However, in contrast to the previous staining patterns, neither protein-positive nor PAS-positive bands were visualized in the lower molecular weight region of < lo5 Da. This indicated that no other protein contamination was evident based on this procedure. Upon isoelectric focusing, polydispersity and microheterogeneity were revealed in the pI range 5w6. Such a pattern was also noticed for the reference mucins BSM and OSM, and has been assumed to be the result of variability in oligosaccharides chain length (Herp, Wu and Moschera, 1979). The minimum molecular weight of the mucin isolate, corrected for the presence of 56 % carbohydrate (see Table I), was estimated to be 120 K.

HUMAS

OCTLAR

MI’(‘IS

TABLE Chemical

composition

of human

Component Interest

of

Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Half cystine Valine Methionine Isoleurine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine Fucoae Mannosc Galactose Glucose N-Ac-galactosamine N-Ac-glucosamine Sialic acid

ocular

J

much,

Human Pool 1

64 158 137

90 99 80

bovine and ovine submaxillary ocular

Pool 2

68 48

c.5 26 .56

20 30 21

mol amino 40 50 202 106 74 179 97 < *5

BSM

16

140 196

101 201 139

< 5 6<5 <.i

< 5 61 c.5

14 41 11

13

22 21 '3

16

2” ‘3 40 10 9

58

< 5 30

42

35

‘7 23 IX

61

41

-14

21 195

"5

20 1X9

33

45 224

276

97 41 211 142 144

105

Relative 31 54 15

0-Ser

45

29

Linkage

19 144 183 56

59 I17 180 125

32

53 164 156

OSM

acid

21 20 29

Protein Carbohydrate Lipid (Thr)-GalNAc

Pool 3

63 156 153 103 87

16

l!W 132

81 262

37

16 15 2.’24 35 6

12 15 239 19 264

Percentage 32 56

31

33

53

58

53

18

12

I1

14

vs. hydroxyamino 49

48

mucins

mucin

Mol/iOOO

99 71 if, 51 <5 26 47

< .i

193

acid (%) 68

62

The chemical composition of ocular mucins isolated from three different pools along with the compositions of BSM and OSM used as reference mucins are given in Table I. Among the amino acids analyzed, a mucin character was reflected by the high levels of the hydroxyamino acids serine and threonine (295,309-, and 282 mol per 1000 mol amino acids for pools 1, 2, and 3, respectively, compared with 336 and 327 for BSM and OSM, respectively), and by the low amounts of the aromatic amino acids phenylaianine and tyrosine (50, 37, and 44, for pools 1, 2, and 3 respectively, and 30 and 31 for BSM and OSM, respectively). It should be noted that while the total amounts of hydroxyamino acids are similar in the three pools, the relative amounts of threonine and serine in pool 3 (80 and 202) differed considerably from those in pools 1 (158 and 137) and 2 (156 and 153). In addition, the level of glycine in pool 3 (179) was substantially higher than in pools 1 and 2 (80 and 99, respectively). Trace levels of the sulfur-containing amino acids half-cystine and methionine were found in the ocular mucin from all three pools and in the reference mucins. Protein accounted for about 31-, 29., and 32%, carbohydrate for 54., 53-. and 56%, and lipid for 15-, 1%.

191

(‘.A’. 11’. (‘HA0

ET

Al,

and 12 %, for pools 1. 2. and 3, respectively. Analysis of t,he monosaccharide content for both mucin pools revealed fucose, galactose. S-acetylgalactosamine. N-acetylglucosamine, and sialic acid (61. 195, 164, 156. and 105 for pool 1. 41, 97. 211. 142, and 144 for pool 2. and 44. 189, 224, 190, and 132 for pool 3). Analysis of the 0-glycosidic linkage between seryl and threonyl and N-acetylgalactosaminyl gly(*o peptides for pools I. 2, and 3 indicated t’hat about 45549% of the total hydroxyamino acids are 0-glycosidically linked as compared with 6X- and 62 %)(I. respectively, in BSM and OSM. These data closely reflect a composition characteristic of a mucin structure (Reid and Clamp, 1978). In addition to 0-glycosidic linkages, the ocular mucin appears to contain N-glycosidic linkages, as indicated by the presence of the small amount of mannose and the detection of the cleaved N-acetylated glucosaminitol by paper radio-chromatography resulting from alkaline borohydride reduction (Fig. 4b). Such a structural feature has been found in the mucin isolated from mouse submandibular gland (Denny and Denny. 1982). Unlike the reference mucins BSM and OSM. t,he ocular mucin isolated from all sample pools contained glucose (53., 41., and 45 mol per 1600 mol amino acids for pools 1,2, and 3, respectively). Glucose-containing glycoproteins are rare (Butler and Cunningham, 1966; Spiro, 1967). I n most cases, the presence of glucose may signify contamination by glycolipids (Butler and Cunningham, 1966; Spiro, 1967). It is well known that mucus secretions contain lipids (Slomiany et al., 1982), some of which may be covalently linked to the glycoproteins (Slomiany, Takagi, Liau, Jozwiak and Slomiany, 1984). Further studies are needed to elucidate the potential structural involvement of glucose and/or lipids with the ocular mucin. The data presented should be regarded as preliminary in nature until other parameters affecting mucin quality have been explored. It is believed that the composition of amino acids in ocular mucin, particularly with respect to hydroxyamino, aromatic, and sulfur-containing amino acids may vary with the conditions used for mucus extraction and with the procedures for isolation and purification (Horowitz, 1977). In addition, a donor parameter based on gene expression may also affect amino acid composition (Bretscher, 1971). 0 ur compositional data also revealed pronounced variations in the type and proportions of individual sugars of t,he oligosaccharides in the ocular mucin. It has been established that among non-ocular mucins, variations are common in the size, composition, and distribution of oligosaccharide side-chains attached to the protein core (Horowitz, 1977). These microheterogeneities, further demonstrated by the exhibition of multiple isoelectric points by the ocular mucin, are believed to be metabolically closely regulated processes, allowing for distinct biological and physical functions of extracellular mucins (Gallagher and Corfield, 1978). A number of factors may contribute to these heterogeneities, among which blood group specificity is considered to be particularly significant (Horowitz, 1977). Knowledge of the carbohydrate structure is essential in understanding the biophysical properties and physiological functions of mucins (Herp, Wu and Moschera, 1979). In future studies involving isolation of ocular mucin, we shall attempt to rorrelat,e carbohydrate structure with such parameters as donor blood t’ype% sex. age, and season of collection. However, ocular mucus of human origin is scarce and difficult to obtain. In addition, it is an aggregate of mucus strands associated with lipids, tear components, glycosaminoglycans, and cellular debris (Chao, Vergnes and Brown, 1983a; Chao and Brown, 1986); which further complicates the isolation and purification of mucin. This paper describes a method which we feel is particularly

HUMAN

OCVLAR

bft:crx

suitable for the isolation of small quantities of human ocular 3 mg. Other methods for the purification and characterization the process of being evaluated.

195

mucin, between l- and of ocular mucin are in

ACKKOWLEDGMENTS This investigation was supported by grants (EY 04934), and Research to Prevent Blindness.

from the National

Institutes

of Health

REFERENCES Aminoff, D. (1961). Methods for the quantitative estimation of N-acetylneuraminic acid and their application to hydrolysates of sialomucoids. Biochem. J. 81, 384-92. Bitter, T. and Muir, H. M. (1962). A modified uranic acid carbazole reaction. Anal. Biochem. 4, 3304. Rretncher, M. 8. (1971). Initiation and termination of the polypeptide chain. In Protein Bynthesis, Vol. 1. (Ed. McConkey, E. H. ). Pp. 89-129. Marcel Dekker: New York. Butler, W. T. and Cunningham, L. W. (1966). Evidence for the linkage of a dissaccharide to hydroxylysine in tropocollagen. J. Biol. Chem. 241, 3882-S. Chao. C.-C. W. and Brown, S. I. (1986). Macromolecular components of human ocular mucus. In The Preocular Tear Film in Health, Disease, and Contact Lens Wear. (Ed. Holly, F.). Pp. 33140. Dry Eye Institute, Lubbock, TX. Chao, C.-C. W. and Butala, S. M. (1986). Isolation and preliminary characterization of tear prealbumin from human ocular mucus. Curr. Eye Res. 5, 895-901. Chao, C-C. W., Butala, S. M.. Zaidman, G. and Brown. S. I. (1987). Immunological study of proteins and mucosubstance in saline soluble human ocular mucus. Invest. Ophthalmol. Vis. 9ci. 28, 54G54. Chao. C.-C. W. and Vergnes, J.-P. (1983). Correction of tyrosine content in the amino acid analysis of human ocular mucus and its fractions. Ezp. Eye Res. 37, 531-2. Chao, C.-C. W., Vergnes, J.-P. and Brown, S. I. (1983a). Fractionation and partial characterization of macromolecular components from human ocular mucus. Exp. Eye Res. 36, 139-50. Chao, C-C. W., Vergnes, J.-P. and Brown, S. I. (1983b). 0-Glycosidic linkage in glycoprotein isolates from human ocular mucus. Exp. Eye Res. 37, 53341. Clamp. J. R., Bhatti, T. and Chambers, R. E. (1971). Determination of carbohydrate in biological materials by gas-liquid chromatography. In Methods of Biochemical Analysis, Vol. 19. (Ed. Glick, D.). Pp. 229-344. Interscience: New York. Denny, P. A. and Denny, P. C. (1982). A mouse submandibular sialomucin containing both N- and 0-glycosylic linkages. Carbohydr. Res. 110, 305-14. Downs, F., Peterson, C., Murty, V. L. N. and Pigman, W. (1977). Quantitation of the pelimination reaction as used on glycoproteins. Znt. J. Pept. Protein Res. 10, 315-22. Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A. and Smith, F. (1956). Colorimetrie method for determination of sugars and related substances. Anal. Chem. 28, 35&6. Ehlers. N.. Kessing, S. V. and Norn, M. S. (1972). Q uantitative amounts of conjunctival mucous secretion and tears. Acta Ophthulmol. 50, 21&14. Fatt. T. (1978). The tears and the lids. In Physiology of the Eye. Pp. 205-10. Butterswort,h: Boston. Gallagher, J. ‘I’. and Corfield. A. P. (1978). M ucin-type glycoproteins-new perspective on their structure and synthesis. Trends Bioch.em. Sci. 3, 3841. Herp. A.. Wu, A. M. and Moschera. J. (1979). Current concepts of the structure and nature of mammalian salivary mucous glycoproteins. Mol. Cell Biochem. 23, 2744. Holly. F. J. and Lemp, M. A. (1977). Tear physiology and dry eyes. &‘urp>. Ophthalmol. 22. 69-87. Horowitz, M. I. (1977). Purification of glycoproteins and criteria of purity. In The Glycoconjugates. Vol. I. (Eds Horowitz, M. I. and Pigman. W.). Pp. 15-34. Academic Press: New York.

196

(‘2’.

IV. (‘HA10

ET

241,

The effect Hounsell, E. F.. Pickering, il‘. .I ., Stall. M. S.. Lawson. i 1. 11. and Feizi. 1‘. (1984). of mild alkali and alkaline borohydridr on thcx carbohydratr ;tnd ljrptidr moirties 01’ fetuin. Biochem. SOC. Tratls. 12, 607-10. Jenssen. A. 0. and Smidsrod. 0. (1982). Preparat,ion of t~nzymically a(ti\-(h Iysozymr from sputum and its distriblltion between the sol ant1 gel phases. ficcr. .I. KPS~ I)i,u. 63. 5X4-90. Laemmli. I’. K. (1970). (‘leavage of structural prot,eins during the assrml)l~of the head ot bacteriophage T4. N&we (Londo,L) 227. 680-5. Lowry. 0. H.. Rosebrough, N’. .I.. Farr. A. L. and Randall, R.J. (1951). Protein measurement with the Foiin phenol reagent. J. Riol. (‘Hun/. 193. 265-75. Marsh. J. B. and Weinstein. 1). B. (1966). Simple charring method for determination of lipids. J. Lipid Res. 7. 754-6. Matsumoto, K. and Mimura, Y. (1974). Histochemical studies on t,he conjunctival goblet cells. Connect. Tis.wrs 6, 85590. Moore, J. c’. and Tiffany. J. M. (1979). Human ocular mucus. Origins and preliminaq characterization. E.rp. Eye Res. 29, 291-301. Moore, J. C. and Tiffany. J. h1. (1981). Human ocular mucus. Chemical studies. Exp. Eye Res. 33. 203-12. Ogata. S. and Lloyd, K. 0. (1982). Mild alkaline borohydride treatment of glycoproteins-a method for liberating both NJ- and O-linked carbohydrate chains. Anal. Biochertt. 119. 351-9. Ouchterlony, 0. and Nilsson. A. (1978). Immunodiffusion and immunoelectrophoresis. In Handbook of Experimental Zmmunoloyy. (Ed. Wier. D. M.). Pp. l-44. Blackwell Scientific Publications : Oxford. Rasilo. M.-L. and Renkonen, 0. (1981). Mild alkaline borohydride treatment liberates Nacetylglucosamine-linked oligosaccharide chains of glycoproteins. FEBS Lett. 135, 38-42. Reid. L. and Clamp. *J. R. (1978). The biochemical and histochemical nomenclature of mucus. Br. Med. Bull. 34(l), 5-8. Richards, G. M. (1974). Modification of the diphenylamine reaction giving increased sensitivity and simplicity in the estimation of DNA. Anal. Biochem. 57. 369-76. Segrest, J. P. and Jackson, R. L. (1972). Molecular weight determination of glycoproteins by polyacrylamide gel electrophoresis in sodium dodecyl sulfate. Tn Methods itt Enzymology, Vol. 28. (Ed. Ginsburg, V.). Pp. 54-63. Academic Press : New York. Slomiany. A., Murty, V. L. N.. Aono. M., Snyder, C!. E., Herp, A. and Slomiany. B. L. (1982). Lipid composition of tracheobronchial secretions from normal individuals and patients with cystic fibrosis. Biochim. Biophys. Acta 710, 10611. Slomiany, B. L.. Takagi. 8.. Liau. Y. H., Jozwiak. Z. and Slomiany. A. (1984). In vitro acylation of rat gastric mucus glgcoprotein with [3H]palmitic acid. J. Biol. Chem. 259. 11997-12000. Snyder, C. E., Nadziejko, C!. E. and Herp, A. (1982). Binding of basic proteins to glycoproteins in human bronchial secretions. Id J. Biochem. 14, 895-8. Spiro, R. G. (1967). The structure of the disaccharide unit of t,he renal glomrrular basement membrane. J. Biol. Chem. 242, 4813-23. Srinivasan, B. D.. Worgul. B. V.. Iwamoto, I. and Merriam. G. R,.. .Jr. (1977). The conjunctival epithelium. Ophthalmic Res. 9. 65-79. Takasaki, S. and Kobata, A. (1974). Microdetermination of individual neutral and amino sugars and N-acetylneuraminic acid in complex saccharides. J. Biochetn. (Tokyo) 76, 783-9. Terho, T. T. and Hartiala, K. (1971). Method for determination of the sulfate content of glyoosaminoglycans. il nul. Biochetn. 41, 47 l-6. Tettamanti. G. and Pigman. W. (1968). Purification and characterization of bovine and ovine submaxillary mucins. Arch. Biochem. Biophys. 124. 41-50. Wright, P. and Mackie. I. A. (1977). Mucus in the healthy and diseased eye. Trans. Ophthalmol. Sot. I:.K. 97. 1 7.