Mucins are the major glycoproteins in mucous secretions of epithelial cells in the salivary, gastrointestinal, respiratory, and urogenital tracts (1-4). They form ...
Vol. 262, No. 23, Issue of August 15, pp. 11339-11344,1987 Printed in U.S.A .
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society for Biochemistry and Molecular Biology, Inc.
Structural Properties of Porcine Submaxillary Gland Apomucin* (Received for publication, August 18, 1986)
Allen E. EckhardtS, Candace S. Timpte$j, John L. Abernethyn, Arazdordi Tournadjell, W.Curtis Johnson,Jr. 11, and Robert L. Hill$ From the Departments of $Biochemistry and TPathology, Duke University Medical Center, Durham, North Carolina 27710 and the 11Department of Biochemistry and Biophysics, Oregon State University, Corvallis, Oregon97331
Porcine submaxillary gland mucin was deglycosylated with a mixture of pure glycosidases to give apomucin containing less than 1%carbohydrate. The resulting apomucin freed of glycosidases was found to contain nine amino acids: threonine, serine, glutamic acid, proline, glycine, alanine, valine, isoleucine, and arginine. Serine, threonine,glycine, and alaninecomprise 77% of the composition. The molecular weight of apomucin was 96,500 as determined by gel filtration inguanidine hydrochloride. Its Stokesradiuswas >68.6 A, a far larger value than expected for a globular protein with M , = 96,500. Circular dichroism spectroscopy of apomucin suggests that it contains 42% aperiodic or “other” structure, 40% @-turns,10%antiparallel pleated sheet, and 8% helical structures. The predicted secondary structure of a 50-residuepeptide from ovine submaxillary gland mucin resembles the circular dichroism predictions, being dominated by turnsthat would leadto an extended nonglobular structure. Analysis for the secondary structure of a 36residue tryptic peptide derived from porcine submaxillary gland apomucin predicts a similar structure. It is concluded that apomucin is likely devoid of traditional secondary structure and serves as a scaffold upon which oligosaccharides are added in 0-glycosidic linkage. When sufficient sialic acid is present in the oligosaccharides, native highly viscous mucin containing about two-thirds carbohydrate by weight is obtained.
are highly negatively charged and rich in sialic acid and/or sulfate. Removal of sialic acid results in a decreased viscosity of solutions of mucin. Mucins can be described as self-associating high molecular weight moleculeswith considerable heterogeneity in size (16). Mucin monomers may also be linked covalently via intermolecular disulfide bonds to form larger molecules (17). This propensity to form aggregates has presented formidable problems in the study of mucin structure. The best characterized mucin is from ovine submaxillary gland (16, 18). Ovine apomucin, devoid of carbohydrate, has a molecular weight of 58,300, and serine and threonine account for about one-third of all residues. The glycosylated molecule with its full complement of oligosaccharides has amolecular weight of 154,000. These glycosylated monomers self-associate as a function of ionic strength and protein concentration to form oligomers with molecular weights ranging from 0.5 to 1.0 X lo6. Native porcine submaxillary mucin has also been reported to consist of subunits ( M , = 900,000) that are linked by disulfide bonds to form dimers (Mr = 2,000,000) and higher aggregates (19). However, the properties of porcine submaxillary apomucin have not been described because of the difficulty of completely removing the carbohydrate. Several types of 0-linked carbohydrates are found in submaxillary mucin, and themost complex is a pentasaccharide with the following structure (8). GalNAcal,3
\ /
Mucins are the major glycoproteins in mucous secretions of epithelial cells in the salivary, gastrointestinal, respiratory, and urogenital tracts (1-4). They form highly viscous solutions and serve to lubricate and to protect these tissues from dehydration and the external environment (5, 6). The unique physical and biological properties of mucins are determined by the oligosaccharides in 0-linkage to the hydroxyl groups of serine and threonine in the protein core. These oligosaccharides, which account for as much as 60-70% of the weight of the molecule, have been characterized for mucins from bovine (7), porcine (8),and ovine (9) submaxillary gland, human stomach (10) and colon (11, 12), porcine trachea (13) and Cowper’s gland (14), and bonnet monkey cervix (15). All
* This work was supported by Research Grants GM25766 (to R. L. H.) and GM21479 (to W. C. J.) from the National Institute of General Medical Sciences, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I Supported by National Research Service Award GM07184 from the National Institute of General Medical Sciences, National Institutes of Health.
Fucal,2
Ga1@1,3
\
GalNAcaSer/Thr
/
NeuNGca2,B
This structure (where Fuc is fucose and NeuNGc is N-glycolylneuraminic acid) endows the mucin with blood group A activity, and on removal of the nonreducing terminal N acetylgalactosamine residue, blood group A activity is lost. We wish to report here the preparation of porcine submaxillary apomucin and adescription of some of its chemical and physical properties. Several lines of evidence are presented indicating that theapomucin exists in solution as anextended nonglobular structure. A preliminary account of this work was reported earlier (20). EXPERIMENTALPROCEDURES
Preparation of Apomucin-Porcine submaxillary glands were typed for blood group A activity by assaying Triton X-100 extracts for blood group A al,3-N-acetylgalactosaminyltransferaseactivity (21). Porcine submaxillary mucin was prepared from A-negative salivary glands by the method of De Salegui and Plonska (22) with omission of the hydroxylapatite step. Mucin (59.8 mg, 6 ml)was deglycosylated by treatment with protease-free neuraminidase (0.1 unit), fucosidase (0.06 unit), a-N-acetylgalactosaminidase (0.75 unit), and endo-a-N-
11339
11340
Submaxillary Apomucin
acetylgalactosaminidase (0.2 unit) for 92 h a t 37 “C in 0.01 M sodium cacodylate, pH 6, containing 0.04% sodium azide, followed by exhaustive dialysis against 0.01 M Tris-HC1 buffer, pH 8. Apomucin was obtained by chromatography of 1-ml aliquots of the reaction mixture at a flow rate of 9 ml/h on a column (0.7 X 5.2 cm) of Whatman DE52 equilibrated with 0.01 M Tris-HC1 buffer, pH 8. Apomucin emerged from the column unretarded upon elution with 0.01 M Tris-HC1 buffer, pH 8. Preparation of Anti-apomucin Antibodies-Rabbit anti-porcine submaxillary gland apomucin antibodies were prepared in amale New Zealand white rabbit by multiple subcutaneous injections in the hindquarters with 200 pg of apomucin emulsified in Freund’s complete adjuvant. Injection of apomucin (200 pg) was repeated in the same manner in Freund‘s incomplete adjuvant 8 and 15 days later. One week after the third injection, antiserum was collected. IgG was purified by chromatography on Protein A-Sepharose (23). Molecular Weight Determination-The molecular weight of apomucin was determined by gelfiltration in 6 M guanidine hydrochloride as described (24). Reduced and alkylated standard proteins and apomucin were chromatographed at a flow rate of4.6 ml/h on a column of Sepharose 6B (1.5 X 85 cm) equilibrated with 6 M guanidine hydrochloride, pH 5.Void and included volumes were determined with calf thymus DNA and tritiated water, respectively. Fractions (approximately 1.2 g) were collected and weighed to determine elution volumes. Standard proteins and their molecular weights were: phosphorylase b, 97,400; transferrin, 76,600; bovine serum albumin, 67,000; ovalbumin, 43,000; malate dehydrogenase, 33,000; cytochrome c , 12,400. Standard proteins were detected spectrophotometrically at 280 nm. Nitrocellulose blotting was used to detect apomucin. Aliquots (5 pl) of each fraction were applied to a nitrocellulose filter and dried. The filter was then washed, blocked with bovine serum albumin, and incubated with rabbit anti-porcine apomucin IgG and lZ5I-goatantirabbit IgG as described (25, 26). The fractions were cutout and counted to detect the apomucin. Determination of Stokes Radius-The Stokes radius of apomucin was determined by column chromatography on Sephadex G-200 (fine, 1.6 X 50 cm) equilibrated with 10 mM sodium cacodylate, pH 6.5,250 mM NaCl as described (27, 28). Fractions (approximately 1.0 g) were collected and weighed to determine elution volumes. Void and included volumes were determined with blue dextran 2000 (Pharmacia P-L Biochemicals) and sodium chloride, respectively. Standard proteins and theirStokes radii were: 0-galactosidase, 68.6 8, 129); lactate dehydrogenase, 42.0 8, (29); malate dehydrogenase, 35.1 A (30);ovalbumin, 27.5 A (31); and myoglobin, 18.9 A (29). Standard proteins were assayed as described (27). Apomucin was detected by nitrocellulose immunoblotting as described above. Circular Dichroism of Apomucin-Apomucin was dissolved in 0.01 M sodium phosphate, pH 7.5, at 0.3-0.6 mg/ml. Concentration was determined by analysis with ninhydrin of a hydrolyzed aliquot (32). Normal absorption spectra were measured to 178 nm on a Cary 15 purged with nitrogen, and theextinction coefficient of the maximum at 191 nm was determined to be 10,300. Circular dichroism spectra were measured to 178 nm on vacuum UV instrumentation described elsewhere (33). The instrument was calibrated with (+)-10-camphorsulfonic acid assuming A (290.5) is +2.37. Measurements were made with the samples in standard cylindrical quartz cells of nominal 50and 100-pm path length, and the spectral slitwidth was a constant 1.6 nm. Spectra were terminated when the total optical density of cell, solvent, and sample exceeded 1.0. CD spectra digitized every 2nm were analyzed for secondary structure according to the earlier procedure (32), except that BenceJonesprotein, carboxypeptidase A, concanavalin A, hemerythrin, rubredoxin, and thermolysin were added to the original set of 16 proteins. Following the Variable Selection Method (34), proteinswere systematically removed from the set of 22 to reduce the number of variables. The idea is to eliminate proteins that have contributions to the CD from sources not found in apomucin. Of course it is not known a priori which proteins should be eliminated, so the criteria used were: 1)improving the fit of the calculated CD to themeasured CD to below 2.0 A squared units, for a sum total of the squares of the residuals; 2) having the sum of fractions of secondary structure between 0.90 and 1.10; and 3) not having any secondary structures with fractions more negative than -0.05. It was found to be mandatory to eliminate a-chymotrypsin,elastase, and papain. With a fourth protein removed, all the criteria were met in 10 out of the 18 possibilities. With a fourthand fifth protein removed, an uncountable number of combinations met all of the criteria. The fractions of secondary structure presented under “ h s u l t s ” are the average of all
these runs that met the criteria, and therepeatability gives the range of these runs. Clearly the method is extremely stable. Isolation and Sequence Analysis of Apomucin Tryptic PeptideApomucin (1.5 mg in 2 ml) was hydrolyzed with 0.05mgof tosylphenylalanyl chloromethyl ketone-trypsin (Worthington) in 0.2 M ammonium bicarbonate, pH 8.4, for 10 h at 37°C. Additional trypsin (0.05 mg) was then added, and hydrolysis continued for another 10 h. The tryptic hydrolysate was lyophilized, acidified with 10%trifluoroacetic acid, relyophilized, and dissolved in 0.225 ml of 0.1% trifluoroacetic acid. The tryptic hydrolysate (0.1 ml) was chromatographed in a Laboratory Data Control high performance liquid chromatography system with a large pore Vydac C18 reverse phase column (0.46 X 25 cm) and the peptides in the eluate detected at 225 nm. The column was developed with 0.1% trifluoroacetic acid (eluant A) at a flow rate of 0.6 ml/min, and after the initial breakthrough peak of absorbing material emerged, the column was developed for 120 min with a linear gradient of 0-62.5% of eluant B (0.1% trifluoroacetic acid in 80% 2-propanol). An incompletely resolved complex of peptides emerged a t and immediately after the breakthrough of the 2propanol gradient. One of these peaks (29 min after starting the gradient) was rechromatographed on the same column with a gradient of 0-87.5% eluant B (120 min) and gave a symmetrical peak 38 min after that of the gradient. The peptide in this peak was sequenced by automated Edman degradation in an Applied Biosystems 470A gas phase sequenator. Phenylthiohydantoin derivatives were analyzed on a Hewlett-Packard1090 high performance liquid chromatograph with a Du Pont phenylthiohydantoin column and the isocratic eluant suggested by the manufacturer (6 p M phosphoric acid, pH 3.2, containing 18%acetonitrile and 16%tetrahydrofuran). Analytical Procedures-Protein concentrations were determined by the method of Lowry et al. (35) with bovine serum albumin as standard unless stated otherwise. Determination of the protein concentration of apomucin by the ninhydrin method (32) showed that the concentration given by the ninhydrin method was 0.61 times the concentration given by the Lowry method. Polyacrylamide gel electrophoresis in sodium dodecyl sulfate (36), silver nitrate staining of polyacrylamide gels (37), iodination of goat anti-rabbit IgG (38), and Western blotting (25, 26)were performed according to published methods. Acid hydrolysates of apomucin were prepared by hydrolysis for 24,48, and 96 h in 6 N HCl containing 0.1% phenol at 110°C. Cystine and cysteine were determined after oxidation to cysteic acid (39). Tryptophan was determined after acid hydrolysis in the presence of 4% thioglycolic acid (40). Amino acid compositions were determined with a Beckman 630 high performance amino acid analyzer with buffers provided by the manufacturer. N-Acetylgalactosamine was determined after acid hydrolysis (4 N HCl, lOO”C, 5 h) with a Beckman amino acid analyzer. Neutral sugar was determined by the method of Dubois et al. (41) with methyl a-D-galactopyranoside as standard. Sialic acid was measured as described (42) after hydrolysis in 0.1 N HzS04 at 80°C for 1 h with N-acetylneuraminic acid as standard. Protease activity was detected as described (43) with glycinated [methyl-14C] hemoglobinas substrate. The following enzymes were prepared as previously described Clostridium perfringens a-Nacetylgalactosaminidase, fucosidase, and neuraminidase ( X ) , and Streptococcus pneumonioe endo-a-N-acetylgalactosaminidase (44). Amino acid sequences were analyzed for secondary structure according to theprogram of Chou and Fasman (45). RESULTS
Preparation of Apomucin-Hydrolysis of blood group Anegative submaxillary mucin (59.8 mg) with a mixture of protease-free glycosidases, followed by chromatography of the hydrolysate to remove the glycosidases and thecontaminating proteins, gave apomucin (19.8 mg). No protease activity (43) was detected in the hydrolysis mixture throughout the course of hydrolysis; thus, the mucin contained no protease. The purified apomucin was devoid of sialic acid, N-acetylglucosamine, and neutral sugars and contained only 8 residues of Nacetylgalactosamine/molecule, assuming a molecular weight of 96,500 (see below). Incontrast, before deglycosylation mucin contained 263 residues of N-acetylgalactosamine/molecule. The nature of the proteins in native mucin and apomucin was assessed by polyacrylamide gel electrophoresis in sodium dodecyl sulfate under reducing conditions (Fig. 1).
Submaxillary Apomucin
A
B
11341
TABLE I The amino acid composition of porcine submaxillary mucin and apomucin Amino acid
2 0 5 4
FIG.1. Polyacrylamide gel electrophoresis of native mucin and apomucin in sodium dodecyl sulfate under reducing conditions. Lane A, 10 pg of porcine submaxillary gland mucin; lone B , 13.6 pg of porcine submaxillary gland apomucin. The gel was stained with silver nitrate. The molecular weights of standard proteins (not shown) are indicated on the ordinate: myosin, 205,000;8-galactosidase, 116,000,phosphorylase b, 97,400; and bovine serum albumin, 67,000.
Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Cystine' Valined Methionine Isoleucine Leucine Tyrosine Phenylalanine Histidine Lysine Arginine
Mucin" Apomucin" residueslmolecule
147 294* 54 61 252 179
32 124 233 76 70 239 154
85' 28'
Total number of residues
3 28 26 10 12 7 21 33
28
1068 1128
Apomucin samples were hydrolyzed for 24,48,and 96 h, and the compositions were averaged to give the values shown. Mucin was hydrolyzed for 24 h only. Tryptophan was not found in apomucin. Value listed was extrapolated to zero time of hydrolysis. Cystine and tryptophan were not determined in mucin. Valine is masked by galactosamine in mucin. e Values listed were those obtained after 96 h of hydrolysis.
The proteinswere detected withsilver nitrate, because mucin Phosphorylase b and apomucin do not stain with Coomassie Blue. The electroApomucin phoretic pattern of native mucin revealed several proteins Transferrin (lane A), although the mucin itself failed to enter thegel and 0 remained in the stacking gel. Purified apomucin was free of these proteins; however, it behaved anomalously, migrating as a broad band with an apparentmolecular weight in excess of 205,000. The molecular weight of anomucin, obtained as described below, is 96,500. This would suggest the abnormal migration of apomucin is mostlikely due to abberrant binding of sodium dodecyl sulfate to the protein. The broad band in lane B (Fig. l), however, is apomucin since it reacted with anti-apomucin IgG onimmunoblots (25, 26). None of the contaminating proteins of mucin (Fig. 1, lane A) reacted on immunoblots with anti-apomucinantibodies. Amino Acid Composition and Amino-terminal End Group 1 Analysis-The amino acid composition of apomucin is given 1.2 1.4 1.6 1.8 2.0 2.2 in Table I. The number of amino acid residues/molecule was Ve / Vo calculated assuming a M , = 96,500, a value obtained as FIG.2. Determination of the molecular weight of apomucin described below. Acid hydrolysates of apomucin contain nine by gel filtration in 6 M guanidine hydrochloride. amino acids insignificant amountsof which threonine, serine, 77% of the composition. Thirty- were performed in an automated gas-phase sequenator (46). glycine, and alanine make up nine percentof the residues are potential sites for 0-glycosyl- No single identifiable residue was found after eight cycles, ation. The molecule appears to be devoid of aspartic acid, suggesting that the amino terminus blocked. is leucine, cystine, methionine, tyrosine, phenylalanine, tryptoDetermination of Molecular Weight-The molecular weight phan, lysine, and histidine. Each of these amino acids was was determined by gel filtration of apomucin and reduced and present in less than 0.1 mol percent of the total amino acids alkylated proteins of known molecular weight in 6 M guaniin the molecule. dine HCl. Underthesedenaturingconditionsthe elution The composition of native porcine submaxillary mucin is position of a protein molecule is a function of chain length good and, therefore, a measure of molecular weight. The elution also given in Table I for comparison,anditisin agreement with thatpreviously reported (22). Clearly, mucin position of apomucin, as shown Fig. in 2,corresponded to that contains small amounts of amino acids not foundapomucin. in of a molecule with a molecular weight of 96,500. Apomucin These amino acids, which include aspartic acid, lysine, histi- emerged as a single peak from this column. dine, leucine, and the aromatic residues, are thought to be Determination of Stokes Radius-The Stokes radiusof apoderived from contaminating proteins in the mucin prepara- mucinwas alsodetermined bygel filtrationusingnative tion. globular proteins of known molecular weight as standards Several cycles of Edman amino-terminal end group analysis undernondenaturing conditions.Apomucinemerged as a
11342
Apomucin
Submaxillary
single peak, and its Stokes radius was found to be greater terminus of the peptide were clearly identified and are shown than 68.6 A (Fig. 3). If apomucin were a globular protein with in Fig. 5 along with the predicted secondary structure. The this Stokes radius, it would have a molecular weight greater sequence itself is consistent with the amino acid composition than 515,300. Therefore, the observed Stokes radius is far of apomucin and contains 9 of the 10 amino acids present in greater than expected for a globular protein with a molecular apomucin. The predicted secondary structure is dominated by weight of 96,500. A globular protein of molecular w$ight forms that would preclude a globular structure. Helices are 96,500 should have a Stokes radius of approximately 38 A. not predicted, and the region that could form p-structures is Circular Dichroism Spectrum-The nature and extent of shorter than usually found in globular proteins. the secondary structure in apomucin was determined by CD spectroscopy. The CD spectrum of apomucin is characterized DISCUSSION by a large negative minimum at 197 nm (Fig. 4).Analysis of Enzymatic deglycosylation of porcine submaxillary mucin the spectrum predicts that apomucin contains 42% aperiodic gives apomucin, which bas a M , = 96,500 (Fig.2) and a Stokes or “other” structure, 40% @-turns, 10% anti-parallel pleated sheet, and 8%helical structure, with a repeatability of +2%. radius greater than 68 A (Fig. 3). If porcine apomucin were a Parallel p-sheet was predicted as -4 k 1%,which is taken to globular protein as suggested earlier for ovine apomu5in (48), be zero. These results indicate that a considerable portion of it wouldhave a smaller Stokes radius (about 38 A). The the apomucin molecule is lacking traditional secondary struc- molecular weight and Stokes radius measurements are comture and that the conformation of the native molecule is pletely inconsistent with the view that porcine apomucin has a globular structure. Indeed they indicate that apomucin is consistent with an extended nonglobular structure. Prediction of the Secondary Structure of Portions of Porcine more likelyan extended structure. If the apomucin aggregated and Ovine Apomucins-A tryptic peptide was obtained from under conditions used to obtain the Stokes radius, then a porcine apomucin andsubmitted to sequence analysis by misleadingly large valuewouldbeobserved.Aggregation automated Edman degradation (46) as described under “EX- seems highly unlikely, however, in viewof the very low perimental Procedures.” Thirty-six residues from the amino concentration (50 pg/ml) of apomucin used in thesemeasurements. Moreover, the circular dichroism spectrum of apomucin is quite unlike that expected for a globular protein but has the general shape that is considered characteristic of the Apomucin “random coil” conformation. However, so-called “random coil” CD spectra of denatured proteins and model polypeptides have significant differences in shape and vary in magnitude A.
:!/actoridore Lactote
Dehydrogenore
i
H
h
i
Ovine Submaxillary Gland Apomucin
w
40 Malate Dohydrogenore
w
w
w
w
w
w
w
w
40 30 Thr-Gln-Leu-Pro-Gly-Val-Thr-Gly-Thr-Ser-Ala-~al-Thr-Gly-Ser-Glu-Pro-Gly-Leu-Pro h h i 1 i B bb b B B i B B B B t T t TT TT Ti TT TT T T T
Ovalbumin
20-
20
10
1
Ser-Ser-Val-Pro-Gly-Glu-Ser-Ala-Thr-Pro-Gln-Gln-Pro-Gly-Ala-Leu-Ser-Glu-Ser-Thr i h H i i b t T t t T T t T T T t T T T T t t
Myoglobin
w
4
H
w
w
50
0.4 0.8
erf
1.2
1.6 2.0 2.4
-’ ( 1 -
Ser-Thr-Gly-Val-Ser-Gly-Leu-Pro-Gly-Thr
KD)
T
FIG. 3. Determination of the Stokes radius of apomucin by gel filtration. The Stokes radius is plotted uerslls err’ (1 - KD) where e r f ’ is the inverse error function and KD isthe partition coefficient of a protein (28).
B.
B T
B T
i T 9
B t
T 9
T
T
Porcine Submaxillary Gland Apomucin 10
1
I
i
H
I
i
h
20
Ile-Ser-Val-Ala-Gly-Ser-Ser-Gly-Ala-Pro-Ala-Val-Ser-Ser-Gly-Ala-Ser-Gln-Ala-Ala i H H H h h i i i i b i B b b i i t t T T T T tT t T T T T 9
9
9
30 36 Gly-Thr-Ser-Gly-Ala-Gly-Pro-Gly-Thr-Thr-Ala-Ser-Ser-Val-Gly-Val h
1
T
WAVELENGTH (nm)
FIG. 4. The circular dichroism spectrum of apomucin.
T
T
T
T
T
T
T
i
i
t
t
T
b t
i T
B i
9
9
B
FIG. 5. The amino acid sequence and the predicted secondary structure of tryptic peptides from ovine (A) and porcine (B) apomucin. The porcine peptide (6.9 nmol) was submitted to automated sequence analysis (46) with a yield of 3.3 nmol at the first cycle and 33 pmol at the thirty-sixthcycle. Serine and threonine were identified qualitatively. The sequence of the ovine peptide was determined as described earlier (18). The predicted secondary structure was determined by the methods of Chouand Fasman (45). The letters indicate the following type of secondary structures. H, strong helix former; h, weak helix former; E , strong @-sheetformer; b, weak @sheet former; T,strong turn; t , weak turn; i, indifferent; w ,aqueous; and g, hydrophobic.
Submaxillary Apornucin over a factor of more than 2 (49). Furthermore, some native proteins, such as a-chymotrypsinand soybean trypsin inhibitor, exhibit CD spectra with “random coil” characteristics. These facts indicate that there arereal differences in secondary structure among these “random coils.” Indeed, there is no agreement in the literature as to what these “random coils” are, andsome workers believe them to be ordered but extended structures (50). When we analyze “random coil” CD using basis spectra derived from globular proteins, we find that these CD spectra correspond to large amounts of /3-turn and aperiodic (“other”) structure together with some antiparallel p-sheet. The amounts of these three structures in the analysis change with the shape and magnitude of the “random coil” CD. Here, analysis of the spectrum suggests that apomucin contains about 42% aperiodic structure, 40% ,&turns, 10% antiparallel pleated sheet, and 8% a-helix. The nonglobular structure of porcine apomucin with a large amount of @-turnis consistent with the predicted secondary structure (Fig. 5) of a 50-residue tryptic peptide (18) from ovine apomucin. The predicted secondary structure of 36 residues in a tryptic peptide from porcine apomucin (Fig. 5 ) reveals the same structural features that as from ovine mucin. The predicted secondary structure is dominated by turns that could preclude a globular structure, and regions that may potentially form ordered secondary structures as in globular proteins are too short to doso. Others (48)have used electron microscopy and ultracentrifugation for analysis of ovine apomucin structure and concluded thatit was a globular molecule. It is difficult to reconcile the results reported here with those reported for ovine apomucin, but clearly the electron microscopic methods used could not detect long extended molecules with the molecular dimensions of apomucin. The amino acid composition of porcine apomucin (Table I) is rather unusual since it contains only nine different amino acids. The molecule probably contains 10 amino acids, since the glutamine that isobserved in the partial sequence (Fig. 5 ) is found as glutamic acid in the acid hydrolysates used to determine the composition. Few proteins are as deficient as apomucin in the 20 amino acids usually found in proteins. This composition may reflect the need of the protein to serve simply as ascaffold upon which the oligosaccharides of native mucin are attached. It is not known which of the threonine and serine residues in the partial sequence of porcine apomucin are glycosylated. Nevertheless, several of these residues are adjacent to glycine residues and in regions dominated by turns. Serine and threonine residues in such sequences have been shown (51) to have a high probability of being glycosylated. Two other proteins with stretches of sequence rich in serine, threonine, glycine, and alanine are the proteoglycan core protein to which glycosaminoglycans are attached ( 5 2 ) and theproduct of the per locus in Drosophila (53). The structure of porcine submaxillary mucin cannot be ascertained directly by the studies reported here. Nevertheless, they are consistent with previous suggestions (19) that native mucin, containing about two-thirds carbohydrate by weight, is a long extended structure composed of aggregates of glycosylated subunits. The aggregates have been estimated to have a M , = 2-3 x IO6 and may be composed of smaller subunits with a M, = 9 x lo5 (19). Ovine submaxillary mucin also aggregates, but the size of the native mucin aggregates and the subunits seems to be somewhat smaller. Hill et al. (16) reported that the subunits of ovine mucin have a M , = 154,000. The studies reported here indicate that the fully glycosylated porcine subunit may have a molecular weight of about 275,000, based on the molecular weight of apomucin
11343
( M , = 96,500) and assuming mucin contains 65% carbohydrate by weight. It is difficult to reconcile the results reported here with those of Shogren et al. (19), who used light scattering to examine the structure of porcine submaxillary mucin. These workers found a decrease in the molecular weight of mucin on treatment with P-mercaptoethanol in 6 M guanidine hydrochloride, suggesting that disulfide bonds are involved in forming mucin aggregates. Mucin so treated has an M , = 0.9 x lo6. Moreover, they concluded that peptide bonds formed by aspartic acid were cleaved in mucin exposed to 6 M guanidine hydrochloride at pH 2. Since the apomucin isolated here has no detectable half-cystine nor aspartic acid, it is evident thatthe protein backbone (apomucin) of mucin cannot undergo such types of reaction. They also suggest that there are aspartic acid-rich regions of sequence in mucin, but this seems unlikely in view of the amino acid composition (Table I). One possible way to explain the findings of Shogren et al. (19) is to invoke the need of a protein containing aspartic acid and half-cystine in the aggregation of subunits. Such a protein would be analogous to thelink proteins of proteoglycan aggregates that aid in binding proteoglycan subunits to hyaluronic acid (47). If such a protein acts in aggregation of mucin subunits,then reduction of its disulfide bonds by mercaptoethanol or cleavage of its aspartyl residues would disrupt its structure and perhaps lead to disaggregation of mucin aggregates. That such a protein is present in mucin preparations is suggested by the fact that as isolated mucin contains small amounts of several amino acids not present in apomucin (Table I) anda number of protein species are found on gel electrophoresis of mucin (Fig. 1). We have assumed that these proteins are nonspecific contaminants in mucin, but one or more could be serving as a link protein. This can readily be tested experimentally, and such studiesare in progress. REFERENCES 1. Pigman, W. (1977) in The Glycoconjugates (Horowitz, M. I., and Pigman, W., eds) pp. 137-152, Academic Press, Orlando, FL 2. Horowitz, M. I. (1977) in The Glycoconjugates (Horowitz, M. I., and Pigman, W., eds) pp. 189-213, Academic Press, Orlando, FL 3. Holden, K. G., and Griggs, L. J. (1977) in The Glycoconjugates (Horowitz, M. I., and Pigman, W., eds) pp. 215-237, Academic Press, Orlando, FL 4. Doehr, S. A. (1977) in The Glycoconjugates (Horowitz, M. I., and Pigman, W., eds) pp. 239-257, Academic Press, Orlando, FL 5. Gottschalk, A., and McKenzie, H. A. (1961) Biochim. Biophys. Acta 54,226-235 6. Gottschalk, A., Bhargava, A. S., and Murty, V. L. N. (1972) in Glycoproteins: TheirComposition,Structure and Function (Gottschalk, A., ed) pp. 810-829, Elsevier Scientific Publishing CO.,Inc., New York 7. Gottschalk, A., and Graham, E. R. B. (1959) Biochim. Biophys. Acta 34,380-391 8. Carlson, D. M. (1968) J. Biol. Chem. 2 4 3 , 616-626 9. Graham, E. R. B., and Gottschalk, A. (1960) Biochim. Biophys. Acta. 3 8 , 513-524 10. Oates, M. D., Rosbottom, A. C., and Schrager, J. (1974) Carbohydr. Res 34, 115-137 11. Podolsky, D. K. (1985) J. Biol. Chem. 260,8262-8271 12. Podolsky, D. K. (1985) J. Biol. Chem. 2 6 0 , 15510-15515 13. Rana, S. S., Chandrasekaran, E. V., Kennedy, J., and Mendicino, J. (1984) J . Biol. Chem. 259, 12899-12907 14. Chandrasekaran, E. V., Rana, S. S., Davila, M., and Mendicino, J. (1984) J. Biol. Chem. 259,12908-12914 15. Hatcher, V. B., Schwartman, G. 0. H., Jeanloz, R. W., and McArthur, J. W. (1977) Biochemistry 16, 1518-1524 16. Hill, Jr., H. D., Reynolds, J. A., and Hill, R. L. (1977) J. Bioi. Chem. 252. 3791-3798
11344
Submaxillary Apomucin
17. Allen, A., and Snary, D. (1972) Gut 13, 666-672 (1951) J. Biol. Chem. 193, 265-275 18. Hill, Jr., H. D., Schwyzer, M., Steinman, H. M., and Hill, R. L. 36. Laemmli, U. K. (1970) Nature 227,680-685 (1977) J. Biol. Chem. 252, 3799-3804 37. Morrissey, J. H. (1981) Anal. Bwchem. 117,307-310 19. Shogren, R. L., Jamieson, A.M., Blackwell, J., and Jentoft, N. 38. Greenwood, F, C., and Hunter, W. M. (1963) Biochem. J. 8 9 , 114-123 (1984) J. Biol. Chem. 2 5 9 , 14657-14662 20. Eckhardt, A. E., Timpte, C. S., Abernethy, J. L., and Hill, R. L.39. Hirs, C. H. W. (1967) Methods Enzymol. 1 1 , 5 9 4 2 (1986) Fed. Proc. 45, 1819 40. Matsubara, H., and Susaki, R. M. (1969) Biochem. Biophys. Res. 21. Schwyzer, M., and Hill, R. L. (1977) J. Biol. Chem. 252, 2338Commun. 36, 175-181 2345 41. Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., and Smith, F. (1956) Anal.Chem. 28, 350-356 22. DeSalegui, M., and Plonska, H. (1969) Arch. Biochem. Biophys. 129,49-56 42. Warren, L. (1959) J. Biol. Chem. 234, 1971-1975 23. Goding, J. W. (1976) J. Zmmunol. Methods 13, 215-22643. Williams, H. R., and Lin, T. (1971) Biochim. Biophys. Acta 250, 24. Fish, W.W., Mann, K. G., and Tanford, C. (1969) J. Biol. Chem. 603-607 244,4989-4994 R., Glasgow, L. 44. Paulson, J. C., and Hill, R. L. (1977) J. Biol. Chem. 252,8615-8623 25. Yen, T. S. B., and Webster, R. E. (1981) J. Biol. Chem. 256, 11259-11265 45. Chou, P. Y., and Fasman, G . D. (1978) Annu. Reu. Biochem. 47, Biochem. 112, 195-203 251-276 26. Burnette, W. N.(1981) Anal. 27. Sadler, J. E., Rearick, J. I., Paulson, J. C., and Hill, R. L. (1979) 46. Hewick, R. M., Hunkapiller, M. W., Hood, L. E., and Dreyer, W. J. Bwl. Chem. 254,4434-4443 J. (1981) J.Biol. Chem. 256, 7990-7997 28. Ackers, G . K. (1967) J.Biol. Chem. 242,3237-3238 47. Doege, K., Hassell, J. R., Caterson, B., and Yamada, Y. (1986) Proc.Natl.Acad.Sci. U. S. A . 83, 3761-3765 29. Sober, H. A. (ed) (1968) in Handbook of Biochemistry, 2nd Ed., pp. C-10-11, Chemical Rubber Publishing Co., Cleveland, OH 48. Rose, M. C., Voter, W. A., Sage, H., Brown, C. F., and Kaufman, 30. Clarke, S. (1975) J. Bwl. Chem. 250,5459-5469 B. (1984) J. Biol. Chem. 259,3167-3172 31. Castellino, F. J., and Barker, R. (1968) Biochemistry 7, 2207-49. Johnson, W.C., Jr. (1985) Methods Biochem. Anal. 31, 61-163 2217 50. L., Tiffany, M. and Krimm, S. (1969) Biopolymers 8,347-359 32. Hennessey, J. P., Jr., and Johnson,W. C., Jr. (1981)Biochemistry, 51. Aubert, J. P., Biserte, G., and Loucheux-Lefebvre, M. H. (1976) Biophys. Biochem. Arch. 175,410-418 20,1085-1094 52. Bourdon, M. A., Oldberg, A., Pierschbacher, M., and Ruoslatti, 33. Johnson, W. C., Jr. (1971) Reo. Sci. Znstrum. 42, 1283-1286 E. (1985) Proc. Natl. Acad. Sci. U. S. A . 82, 1321-1325 34. Mosteller, F., and Tukey, J. W. (1977) Data Analysis and Regres53. Jackson, F. R., Bargiello, T. A., Yun, S. H., and Young, M. W. sion, Addison-Wesley, Publishing Co., Reading, MA 35. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1986) Nature 320, 185-188
