INTRODUCTION

Many studies of the modification of mucins associated with diseases and infection have used histological (Corfield et al., 1992b; Ferri et al., 2001) or immunological (Aebischer et al., 2000; Sutton, 2001; Backhed et al., 2003; Cohen et al., 2003), lectin histochemical (Alroy et al., 1984; Fischer et al., 1984; Abel et al., 1987; Chan and Wong, 1991; Fujino and Fried, 1993; Baczako et al., 1995; Choi et al., 2003; Lueth et al., 2005) or immunocytochemical (Burchell et al., 1987; Reis et al., 1999; Kubota et al., 2007; Tsubokawa et al., 2009) techniques rather than chemical methods. Mucins collected from tissues, especially from the gastrointestinal tract, are contaminated with non-mucin substances such as proteins, acid, pepsin, bile, DNA and RNA which must be removed for chemical analysis. (Ferri et al., 2001)

2.1.1. Purification of mucins

Various methods have been used to purify mucins including: (1) gel filtration chromatography; (2) affinity chromatography; (3) cesium chloride (CsCl) isopycnic density gradient centrifugation; (4) equilibrium density gradient centrifugation and (5) ion exchange chromatography (reviewed by(Allen, 1981). These methods use differences in MW, density and charge of mucin molecules. Using a single method or a combination of two or three of the above methods, mucins have been isolated and characterised in pigs (Hashimoto et al., 1964; Neiderhiser et al., 1971; Pearson

et al., 1981; Minkiewicz-Radziejewska et al., 2000), rats (Keryer et al., 1973; Clark and Marchok, 1979; LaMont and Ventola, 1980; Malinowski and Herp, 1981; Bodner and Baum, 1984; Tabak et al., 1985; Ohara et al., 1997), monkeys (Herzberg et al., 1979; Devaraj et al., 1993), humans (Chao et al., 1988; Parker et al., 1993; Thornton et al., 1995), sheep and cattle (Tettamanti and Pigman, 1968) from tissues including the stomach, salivary, submandibular and submaxillary glands, the colon and intestine.

A frequently used protocol is a combination of gel filtration followed by CsCl density gradient centrifugation (Feste et al., 1990; Paszkiewicz-Gadek et al., 1995; Minkiewicz-Radziejewska et al., 2000). The advantage of this technique is that it has been used for the purification of mucus glycoprotein in a range of species from many different organs such as the gastrointestinal tract, submaxiliary, saliva gland and lung. In all instances, the gel elution profile has been entirely predictable, with mucin eluting in the void volume at a low flow rate. Gel filtration should be used before CsCl density gradient centrifugation, as Marshall and Allen (1978), who used them in the reverse order, found that a single ultracentrifugation did not succeed in separating glycoproteins from nucleic acids. (Marshall and Allen, 1978)(Feste et al., 1990; Parker et al., 1993; Paszkiewicz-Gadek et al., 1995)

The removal of proteins and DNA/RNA using these methods can be validated at each step by monitoring the absorbance at 280 nm, assaying fractions with Periodic Acid Schiff (PAS) to detect neutral carbohydrates and an Enzyme- Linked Lectin Assay (ELLA) to detect sugars on glycoconjugates. The peak fractions can then be subjected to two SDS-PAGE, one stained with PAS and the other transferred by Western blot and probed with lectins to locate the glycoprotein peak. Monitoring by A280 alone is not adequate because mucin glycoproteins

contain only a small amount of aromatic amino acids. Although PAS is a sensitive detection method for neutral sugars, these may also be present on non-mucin glycoconjugates. Thus, using only one assay to detect mucin-containing fractions is not reliable. The sensitivity and reliability of the ELLA method has been investigated and shown to be specific for detecting mucins (Kodaira et al., 2000; Gull et al., 2007). Western blot followed by lectin probing is also effective in detecting the carbohydrate epitopes.

2.1.2. Molecular weight of mucins

The MW of native mucin molecules are very high (approximately 106 _or

greater) because of subunit assembly (Caspar, 1966). Mucins are oligomeric structures consisting of many monomers linked by non-covalent bonds i.e., physical forces such as hydrophobic or ionic interactions; or covalent bonds i.e., linked end-to-end via disulfide bonds in some mucins (Gum et al., 1992; Meerzaman et al., 1994; Desseyn et al., 1997; Toribara et al., 1997). The MW of mucin depends on the source of the sample, method used for preparation and experimental conditions employed (Holden et al., 1971a). For example, freshly- prepared pig gastric mucin glycoproteins have MW of 2x106 _{Da (Carlstedt and}

Sheehan, 1984; Carlstedt et al., 1985) whereas the MW of commercial pig gastric mucins is 1.25x106 _{Da (Jumel et al., 1996).}

Molecular weights of sheep mucins have been reported. Those were investigated including colonic mucin (Allen and Kent, 1968; Kent and Draper, 1968), submaxillary mucin (Gottschalk et al., 1972; Hill et al., 1977a, b), and small intestinal mucin (Mukkur et al., 1985). Ovine submaxillary mucins have a MW as low as 394,000 Da using 0.2M sodium chloride extraction or as high as 1.3x106 _Da

(Gottschalk et al., 1972). Hill et al. (1977a) reported the MW of sheep submaxillary mucins to be between 550,000 and 650,000 Da. Mukkur et al. (1985) reported the MW of sheep intestinal mucins to be 5.0±0.1x106 _{Da. Differences in MW of mucins}

can arise from the size and number of subunits per one native molecule (Bennett et al., 1998). (Gottschalk et al., 1972; Hill et al., 1977a; Hill et al., 1977b),

Native mucins have a very high MW and maintain their structure as secreted while reduced mucins exist in subunit form (the smallest unit is a monomer). During reduction, bonds or interactions that link subunits together to make up the native molecule are broken down, releasing reduced products. The size of mucin subunits varies according to the reducing conditions. Allen et al. (1989) studied the reduction of pig gastric mucin glycoprotein in different conditions. The glycoprotein had a MW of 2x106 _{Da only when the extraction and isolation of mucins were implemented in}

the presence of guanidinium hydrochloride (GuHCl) and treated with dithiothreitol (DTT), while a 5x105 _{Da subunit was observed when mucins were further treated}

with 0.2 M β-mercaptoethanol. This suggested that mild reduction would result in

larger MW glycoproteins and complete reduction would result in smaller MW subunits (Allen et al., 1988)

2.1.3. Characterisation of bonding and linkage in mucins

Protease inhibitors such as phenylmethylsulphonyl fluoride (PMSF) and N- ethylmaleimide (NEM) are generally required to protect the native mucin molecule from being attacked and broken down into small fragments by proteolytic enzymes. Urea acts as a reducing or physical deaggregating agent (Maxfield and Davis, 1963) by disrupting non-covalent bonds, increasing the solubility of proteins; SDS also disrupts non-covalent bonds in proteins and denatures and causes the molecules to lose their conformation (unfolding into a rod-like shape). Using urea and SDS, mucin complexes linked by non-covalent bonds would be broken. The presence of covalent bonds can be investigated by using DTT, which reduces the mucin molecule by cleaving -S-S- groups, unfolding the peptide chains or by β- mercaptoethanol, which is used to completely reduce disulphide bonds.

The sheep fundic and duodenal mucus glycoproteins (mucins) were purified from the solubilised mucus by the procedure of size exclusion chromatography on a Sepharose 4B (gel filtration) column, followed by density gradient centrifugation in GuHCl/CsCl (Figure 2.1). This method was shown to be effective in isolating mucin glycoproteins without contamination. During the gel filtration process of mucin purification, the MW of native and reduced mucins was estimated.