| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1,1,1,2
,
,3
* CSIRO Livestock Industries, Queensland Bioscience Precinct, 306 Carmody Rd., St. Lucia, QLD 4067, Australia
Co-operative Research Centre for Innovative Dairy Products (CRC-IDP), 84 William Street, Melbourne, Victoria 3000, Australia
Deakin University, 225 Burwood Hwy, Burwood, Victoria 3151, Australia
Centre for Advanced Technologies in Animal Genetics and Reproduction (ReproGen), University of Sydney, PMB3 Camden, New South Wales 2570, Australia
3 Corresponding author: ross.tellam{at}csiro.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
2–3)-Gal(β1–3)-GalNAc1] and, to a lesser extent, with N-linked oligosaccharides, which together accounted for approximately 60% of the apparent mass of Muc1. Purified bovine Muc1 directly bound fluorescently labeled Escherichia coli BioParticles (Invitrogen, Mount Waverley, Australia) and inhibited their binding to bovine mammary epithelial cells grown in vitro. It was also demonstrated that the expression of Muc1 mRNA in bovine mammary epithelial cells was markedly upregulated by lipopolysaccharide. Muc1 may be a pattern recognition protein that has the capacity to sequester bacteria and prevent their attachment to epithelial surfaces by immobilizing and subsequently shedding Muc1-bound bacteria from the cell surface. It was concluded that bovine Muc1 is probably an inducible innate immune effector and an important component of the first line of defense against bacterial invasion of epithelial surfaces, particularly mammary epithelial surfaces and the neonatal gut.
Key Words: Muc1 • mucin • polymorphism • innate immunity
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
In addition to providing hydration and lubrication at the epithelial surfaces, mammalian Muc1 may protect epithelial cell surfaces from invasive pathogenic microorganisms (Patton et al., 1995; DeSouza et al., 1999; Lillehoj et al., 2004). The mucin fraction from human MFGM inhibited adhesion of S-fimbriated Escherichia coli to buccal epithelial cells (Schroten et al., 1992), and bovine Muc1 inhibited infection of monkey kidney and human intestinal cell lines with a neuraminidase-sensitive rotavirus strain (Kvistgaard et al., 2004). Perhaps the most compelling evidence is the observation that the intestinal tracts of Muc1-deficient mice were more easily colonized by Helicobacter pylori and were associated with increased gastritis (McGuckin et al., 2007). Despite these investigations, very little is known about the functions of bovine Muc1 isolated from milk or the detailed nature of its gene and protein structures, especially considering the unusually low sequence conservation of the bovine protein.
The bovine Muc1 gene is characterized by 7 exons and the encoded protein consists of 5 structurally distinct domains: 1) the signal sequence; 2) the variable number of tandem repeats (VNTR) domain; 3) the SEA domain (initially identified in sperm protein, enterokinase, and agrin); 4) the transmembrane domain; and 5) the cytoplasmic domain (Figure 1; Gendler, 2001). The VNTR domain is encoded within exon 2 and can be polymorphic in some species. Each VNTR unit (20 AA) is predicted to be extensively glycosylated with O-linked oligosaccharides. The range in the number of VNTR units and the extent of VNTR polymorphism varies substantially across species, with no polymorphism in murine Muc1 and substantial VNTR polymorphisms in humans (21 to 125 VNTR units; Gendler, 2001). For bovine Muc1, up to 5 allelic variants, with VNTR ranging from 10 to 24, have been suggested (DeSouza et al., 2007). The SEA domain in the human ortholog undergoes proteolytic processing, which can potentially lead to shedding of Muc1 from the cell surface (Levitin et al., 2005). However, the processed protein typically forms a stable noncovalently linked heterodimer. It is not known whether bovine Muc1 is similarly processed. The cytoplasmic domain is thought to be involved in signal transduction and protein-protein interactions (Gendler, 2001).
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Detection of Muc1 VNTR Polymorphisms
The VNTR allelic polymorphisms in the Muc1 gene were detected by PCR using primers flanking the VNTR domain in exon 2 (sense primer 5' CATAAACCCCCGCAGAACTA 3'; antisense primer 5' TAATATGGCTGGCAGCAGTG 3'; Figure 1). Polymerase chain reaction was undertaken with 1x PCR buffer, 200 µM deoxynucleotide 5'-triphosphates, 500 nM each primer, 20 to 50 ng of genomic DNA, and 0.6 U of HotStar Taq DNA polymerase (Qiagen) under the following conditions: denaturation at 94°C (15 min), followed by 30 cycles of denaturation at 94°C (30 s), annealing at 60°C (45 s), and extension at 72°C (1.5 min). The amplicons were separated on an agarose gel and visualized with ethidium bromide. A representation of each differently sized PCR product was excised, cloned into pGEM-T Easy vector (Promega, Sydney, Australia), and sequenced using BigDye Terminator v3.1 (Applied Biosystems, Melbourne, Australia). The current annotation of bovine Muc1 in the Btau_4.0 assembly in public databases incorrectly indicates that the gene has 9 exons. This error has arisen from discrepancies between cDNA and genomic sequences caused by VNTR variation of the alleles that were sequenced. Gene symbols are in italics when used in the context of the gene or mRNA and in normal font when referring to a protein. Human gene symbols are in capitals.
Association of Muc1 VNTR Polymorphisms with Dairy Production Traits
Holstein-Friesian sires (157, from Genetics Australia) with extensive EBV data were used for analysis of the association of Muc1 VNTR alleles with dairy production traits. The animals represented high and low extremes of the EBV for SCC; consequently, the statistical model used for trait association analysis required incorporation of a term to account for the introduced bias. A linear mixed model was used to test for association between Muc1 genotype and the EBV for the Australian profit ranking, a composite trait defined by the Australian Dairy Herd Improvement Scheme that includes milk production, milking speed, temperament, survival, SCC, and live weight; the Australian selection index, which accounts for production costs and the value of milk components and is weighted for the EBV protein, fat, and milk; protein (kg); protein percentage; total milk yield (L); fat (kg); fat percentage; and SCC. The model was the following:
 |
is the random error. Only alleles D, E, and G were present in the tested population. The third allele was not included in the model because it is colinear with the first 2 alleles (Allele1 + Allele2 + Allele3 = 2). The term HLSCC was added to each model (other than the SCC trait model) to assess whether extreme tail selection for SCC may have been associated with the trait of interest, and hence may have biased estimates of any association. When found to be nonsignificant, this term was not used in the model. Assessment of the significance of fixed effects was conducted using Wald chi-square (
2) tests. For assessing the significance of the association between the various phenotypes listed above and the Muc1 alleles, the Wald test examined whether β1 = β2 = 0 (2 df), and for HLSCC, the Wald test tested whether the single effect (high vs. low tail) was zero (1 df). The model fitting was implemented using REML procedures in GenStat Release 9 (Payne et al., 2005).
Isolation of RNA and Quantitative Real-Time Reverse Transcription PCR
Total RNA was extracted from bovine MEC grown in cell culture and processed to produce cDNA (Vuocolo et al., 2003). Quantitative real-time reverse transcription PCR (qRT-PCR) using the Sybr Green-based fluorescence detection system and the ABI Prism 7900 Sequence Detection System (Applied Biosystems, Melbourne, Australia) was used to measure mRNA expressed in mammary epithelial cells. The Muc1 sense primer was 5' GAGTGGGTGAAGGCACAGTT 3' and the antisense primer was 5' CAGAACACAGACCAGGAC 3'. Data were analyzed using the Relative Expression Software Tool, which generates a mean normalized expression value of a target transcript relative to a reference gene (Pfaffl et al., 2002). Ten nanograms of reverse-transcribed RNA was used for each qRT-PCR measurement. Four technical replicates and 2 biological replicates were used for each gene. This process allowed quantification of the target gene relative to a constant reference gene in each sample by using threshold cycle data. Three potential reference genes, 18S ribosomal RNA (GenBank accession no. DQ222453), glyceraldehyde 3-phosphate dehydrogenase (GenBank accession no. XM_868165), and ribosomal protein, large, P0 (RPLP0; GenBank accession no. BT021080), were independently assessed. The expression pattern of RPLP0 showed the least variation; therefore, it was used as the reference gene. The RPLP0 sense primer was 5' CAACCCTGAAGTGCTTGACAT 3' and the antisense primer was 5' AGGCAGATGGATCAGCCA 3'.
Each qRT-PCR (5 µL total volume) contained 2.5 µL of 2x Sybr Green Master Mix (Applied Biosystems; 0.25 µL of each primer, giving a final concentration of 450 nM each; 1.0 µL water; and 1.0 µL of a 1:10 dilution of the stock cDNA template. The reaction included an initial denaturation at 95°C (10 min), followed by 40 cycles of 95°C (15 s) and 60°C (1 min). All melt curves showed single peaks consistent with the presence of single amplicons. Each qRT-PCR assay was validated by amplicon size and sequence to authenticate the targeted gene. Amplification efficiency was measured for each amplicon and used in conjunction with the Relative Expression Software Tool gene quantification software (Pfaffl et al., 2002).
Isolation of MFGM
Full-cream milk (80 L, Holstein-Friesian animals, representing Muc1 alleles D, E, and G) was centrifuged at 2,000 x g (15 min) at room temperature (RT). The cream fraction (3.2 L) was isolated and washed (x3) in Tris-buffered saline (TBS; 50 mM Tris, pH 7.5, 150 mM NaCl) containing 1 mM EDTA and 1 mM benzamidine, and centrifuged (2,000 x g, 15 min, RT) after each wash. The cream was resuspended in an equal volume of the buffer and homogenized until it coagulated into butter (approximately 10 min, 4°C). The buttermilk was collected by filtration through cheesecloth and centrifuged (100,000 x g, 1 h, 4°C). The MFGM pellet was washed in PBS, recentrifuged, and stored at –80°C.
Purification of Muc1 Released from MFGM
The MFGM was resuspended in 100 mM Tris, pH 8.0, containing 0.4 mg/mL of trypsin (Promega, Sydney, Australia; 1:20 trypsin to protein) and incubated for 4 h at 37°C. The solution was centrifuged (100,000 x g, 30 min, 4°C) and the supernatant was dialyzed against 20 mM Tris-HCl, pH 7.0, containing 1 mM benzamidine hydrochloride and 1 mM EDTA. This procedure released the extracellular domain of Muc1 from MFGM by exploiting its resistance to tryptic digestion. The sample was then fractionated in the same buffer on a 10-mL DEAE Sepharose column (Amersham Biosciences, Rydalmere, Australia). Muc1 eluted as a broad peak between 200 and 400 mM NaCl. These fractions were pooled, concentrated, and dialyzed against 50 mM sodium phosphate, pH 7.0, containing 0.8 M ammonium sulfate and then applied to a 10-mL Phenyl-Sepharose column (Amersham Biosciences). Muc1 eluted in the break-through fraction and was 90 to 95% pure, as measured by a combination of silver and Alcian Blue-stained SDS-PAGE gels.
SDS-PAGE Analysis
Protein samples representing MFGM and purified bovine Muc1 were analyzed by SDS-PAGE under reducing conditions on 6 to 18% gradient gels. Proteins were stained with Coomassie Blue R250, silver, periodic acid-Schiff reagent, or Alcian Blue.
Biotinylated Lectin Blots
Purified Muc1 (5 µg) was spotted onto a nitrocellulose membranes (Hybond-C, Amersham Biosciences), air-dried, washed in TBS containing 0.05% Tween 20 (TBS-T), and blocked with 2% BSA in TBS-T (30 min, RT). The membranes were incubated with biotinylated lectins (Vector Laboratories, Abacus ALS, Brisbane, Australia) at recommended concentrations (5 to 20 µg/mL) in TBS-T (RT, 2 h), washed in TBS-T (3 x 10 min), incubated with streptavidin-biotinylated horseradish peroxidase complex (1:1,000 dilution; GE Healthcare, Rydalmere, Australia; RT, 2 h), and then washed in TBS-T (3 x 10 min) and TBS (2 x 5 min). Bound horseradish peroxidase-streptavidin complex was visualized on ECL Hyperfilm (Amersham Biosciences) by chemiluminescence after incubation of the membrane for 1 min in 0.04% luminol, 0.01% 4-iodophenyl, and 0.075% hydrogen peroxide in TBS. Binding was visually interpreted.
Amino-Terminal AA Sequencing
To determine the N-terminal AA sequences of proteins present in the MFGM and purified Muc1, proteins were separated by SDS-PAGE and transferred onto a polyvinylidene fluoride membrane (ProBlott, Applied Biosystems). Selected protein bands were sequenced using a 494 HT Procise Sequencer according to a standard protocol (Applied Biosystems).
Peptide Sequencing
Purified Muc1 was dialyzed into 8 M urea and 0.4 M ammonium bicarbonate, reduced (4 mM dithiothreitol, 15 min, 50°C), and alkylated (8 mM iodoacetamide, 15 min, RT). The sample was diluted in water to final concentrations of 2 M urea, 0.1 M ammonium bicarbonate, 1 mM dithiothreitol, 2 mM iodoacetamide, and sequencing-grade modified trypsin was added to a final enzyme-to-protein ratio of 1:20. The reaction mixture was incubated for 24 h at 37°C, and the resulting peptide fragments were analyzed by matrix-assisted laser desorption/ionization-time of flight/time of flight mass spectrometry on a 4700 Proteomics Analyzer according to a standard protocol (Applied Biosystems). Mass spectra were recorded in positive reflector mode at a laser intensity of 4,800. Tandem mass spectrometry (MS/MS) data were acquired using the default MS/MS method (positive ion, 1-kV collision energy, reflectron mode, laser intensity of 5,500). Data were processed using GPS Explorer software, Version 3.5 (Applied Biosystems), which incorporated the Mascot peptide mass fingerprinting and MS/MS ion search software (Version 1.9.05, Matrix Science, London, UK) in conjunction with publicly available bovine protein sequence information.
AA and Monosaccharide Compositional Analyses
Two independent MFGM preparations, as described above, were used to purify Muc1 samples for AA and monosaccharide analyses. Amino acid analysis was carried out by the Australian Proteome Analysis Facility (Sydney, Australia) using standard protocols. Cysteine and Trp contents were not determined. Data are presented as the mean of 2 independent analyses. Monosaccharide composition analysis was performed by the Glycotechnology Core Resource (University of California, San Diego, CA) using standard protocols. The relative amounts of monosaccharides on Muc1 are presented as mole percentage of the total monosaccharide content (mean of 2 independent analyses). Muc1 was also incubated with Clostridium perfringens neuraminidase (Roche Diagnostics, Castle Hill, Australia; 1 mU/µg of glycoprotein) in 50 mM sodium acetate buffer, pH 5.0, containing 150 mM NaCl (37°C, 24 h). Changes in mobility and characteristic staining patterns of Muc1 on SDS-PAGE were monitored, and monosaccharide compositional analysis of a single sample was performed to demonstrate the removal of neuraminic acid (sialic acid; Neu).
Bacterial Culture
A nonpathogenic strain of E. coli (O-, H48) was grown in brain heart infusion (BHI, Oxoid, Adelaide, Australia) medium. Brain heart infusion broth was inoculated with a single bacterial colony-forming unit taken from a BHI agar plate and incubated overnight at 37°C under anaerobic conditions. A log phase culture was prepared by inoculating 10 mL of BHI broth with a 1 x 109 cfu aliquot of the overnight culture, followed by further incubation for 3 h.
Direct Binding of E. coli to Immobilized Muc1
Muc1 was serially diluted (loadings: 0.1 to 100 µg) and adsorbed onto a Whatman Westran S polyvinylidene fluoride membrane (Sigma-Aldrich, Castle Hill, Australia) using a Bio-Dot SF microfiltration apparatus (Bio-Rad, Gladesville, Australia). The membrane was blocked with PBS containing 5% BSA (1 h, RT). Labeling of bacteria with NHS-LC-Biotin (Pierce, Quantum Scientific, Murarrie, Australia) was performed following the instructions of the manufacturer for labeling cell surface proteins. Excess probe was removed (3 washes in 1 mL of PBS containing 100 mM Gly) and the cell pellet was resuspended in 40 mL of PBS containing 5% BSA. An aliquot of this suspension was overlaid onto the membrane (0.3 mL/cm2) and incubated overnight (4°C). The membrane was washed (3 x 5 min; PBS containing 0.1% Tween-20) and adherent bacteria were detected using horseradish peroxidase-conjugated streptavidin complex and standard chemiluminescence techniques (Pierce, Quantum Scientific). Images for densitometric quantification of bacterial binding were captured on a Kodak 440CF imaging station and analyzed using Kodak 1D imaging software (Eastman Kodak Company, Rochester, NY). The assay was independently performed twice with each Muc1 dilution analyzed in triplicate. The difference between the binding to Muc1 and the BSA control at each protein concentration was determined and then expressed as a percentage of the maximal binding. The percentage of specific binding is presented as the mean ± standard error of the mean.
Binding of Fluorescently Labeled E. coli BioParticles to MFGM
Equal volumes of fluorescein-conjugated E. coli (K-12 strain) BioParticles (Invitrogen, Mount Waverley, Australia; 1 mg/mL) and 500 µg/mL of MFGM were mixed and incubated in PBS for 1 h at RT. The BioParticles were visualized using an Axiophot fluorescence microscope (Carl Zeiss Pty. Ltd., North Ryde, Australia).
Culture of Bovine MEC
The isolation, culture, and characterization of primary MEC have been described previously (Strandberg et al., 2005). Bovine MEC were seeded at 4 x 104 cells per well in a 6-well plate coated with collagen type I from calf skin (Sigma-Aldrich) and cultured until 80% confluent. They were then washed in PBS before addition of fresh media with or without LPS (E. coli strain 055:B5, Sigma-Aldrich) or lipoteichoic acid (LTA; Staphylococcus aureus; Sigma-Aldrich) at final concentrations of 50 and 20 µg/mL, respectively. After incubation for 0, 2, 4, 8, 16, and 24 h, the cells were washed with PBS before harvesting for subsequent RNA extraction. For the LPS dose-response study, bovine MEC were stimulated with LPS (0.01 to 50 µg/mL) for 24 h.
Binding of Bacteria to Bovine MEC
The effect of Muc1 on the binding of the fluorescein-conjugated E. coli (K-12 strain) BioParticles to MEC was measured as follows. Bovine MEC (1 x 105 cells/well) were cultured in a collagen type 1 (calf skin)-coated black 96-well plate (Greiner Bio-One GmbH, Frickenhausen, Germany) overnight (37°C). The BioParticles (1 mg/mL) were added to MFGM (500 µg/mL) at 20°C with agitation for 2 h, before addition of 100 µL of the mixture to the MEC (in triplicate). After further incubation (2 h, 37°C), the cells were washed (x2) with PBS and residual cell-bound florescence (emission 480 nm, excitation 520 nm) measured on a Polarstar fluorometer (BMG Labtech Pty. Ltd., Mornington, Australia). Bovine serum albumin was used as the control.
Statistical Analysis
In addition to the statistical methods outlined for the detection of Muc1-trait associations, data on other measures (E. coli binding, Muc1 expression fold changes) are presented in the text and figures as the mean ± standard error of the mean. Significance for these measures was ascertained using ANOVA. P-values 
0.05 were considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
The allele frequencies and observed heterozygosity values obtained using GenePop software (Raymond and Rousset, 1995) for the Muc1 gene in 2 cattle herds representing Brahman and Holstein-Friesian animals are shown in Table 1. The predominant alleles were D (16 VNTR units), E (14 VNTR units), and G (11 VNTR units). The shorter (F, G, H, I) and longer (A, B, C) alleles were relatively rare or absent in this population. Alleles B (20 VNTR units) and C (17 VNTR units) were found only in Brahman animals.
|
|
|
The identity of the purified protein as Muc1 was demonstrated by direct N-terminal AA sequencing, which revealed the sequence NVPTLTTSD. This sequence began at residue 23 immediately following the predicted signal sequence (European Molecular Biology Laboratory accession no. AJ400824; Figure 4). Two Muc1 peptides were also identified from the purified protein after additional in-solution tryptic digestion and detection of peptides by using matrix-assisted laser desorption/ionization-time of flight/time of flight mass spectrometry peptide sequencing. These peptides, VSLYFLSFR and SIWGLILQIYKQR, are located within the Muc1 SEA domain. Very few peptides were produced by the action of trypsin on the purified Muc1 released from MFGM. The lack of tryptic peptides from the cytoplasmic domain and the proteolytic release of the intact glycosylated extracellular domain indicate the absence of the transmembrane and cytoplasmic domains in the protein released from MFGM. There are only 3 tryptic cleavage sites between the end of the most C-terminally identified peptide and the transmembrane domain; thus, the C-terminus of the purified protein must reside within this 25-AA segment, which includes the consensus sequence (G^SVVV) associated with natural proteolytic processing of Muc1 in other species (Figure 4). Muc1 polymorphic protein variants were also independently isolated and purified from MFGM derived from cows that were homozygous for allelic variants D or G (Figure 3c). The size difference estimated by SDS-PAGE (approximately 20%) for the 2 alleles agrees with the predicted value (18%), assuming a putative C-terminus near the conserved natural proteolytic processing site (G^SVVV).
|
2–3)-Gal(β1–3)-GalNAc1] bound to Ser and Thr residues (Burchell et al., 2001). The relative abundance of Neu is also consistent with the unusual staining properties of Muc1 subjected to SDS-PAGE. Present in substantially lower quantities were N-acetylglucosamine and mannose (5.2 and 7.1 mol %, respectively), and lesser extents of glucose and fucose (both <1 mol %), which are generally characteristic of N-linked oligosaccharides (Lakhtin, 1995). Muc1 has 3 predicted N-linked glycosylation sites outside its VNTR region, and 1 of the more prevalent VNTR units also has a potential N-linked site. It is estimated that Muc1 contains approximately 60% (wt/wt) carbohydrate. The abundance of Neu on bovine Muc1 was underscored by its reduced mobility, less diffuse band, and loss of Acian Blue staining in SDS-PAGE after treatment with C. perfringens neuraminidase (Figure 3d). Monosaccharide compositional analysis of Muc1 after neuraminidase treatment showed that greater than 95% of Neu was efficiently removed (result not shown).
Binding of Biotinylated Lectins to Muc1 Released from MFGM
Muc1 was screened for binding by 18 biotinylated lectins by using immunoblots to provide information on its oligosaccharide structures (Table 3). Muc1 reacted with peanut lectin (PNA) but only after desialylation with neuraminidase, and with Jacalin lectin before and after desialylation. Both PNA and Jacalin have high affinities for T-antigen [Gal(β1–3)-GalNAc1; Lakhtin, 1995]. However, unlike PNA, Jacalin also binds this structure when Gal is substituted with an additional monosaccharide, such as Neu (Tachibana et al., 2006). Evidence for Neu substitution at the C-3 position of Gal was obtained from the positive reaction of Muc1 with Maackia amurensis lectin II (hemagglutinin), which binds Neu in an 2–3 linkage. This reactivity was lost upon neuraminidase treatment. Muc1 reacted only weakly with Sambucus nigra lectin, which preferentially binds 2–6-linked Neu (Lakhtin, 1995). Collectively, these lectin-binding properties demonstrated the presence of O-linked sialyl-T-antigen, as suggested from the monosaccharide compositional analysis. The latter analysis also implied the presence of some N-linked glycans, and this was supported by reactivity of desialylated Muc1 with concanavalin A, which preferentially binds N-linked oligosaccharides with terminal -linked mannose structures (Lakhtin, 1995).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
The substantial VNTR allelic variation in bovine Muc1 may underlie functional differences in the protein. There was suggestive evidence for an association of the Muc1 variants with the milk production traits protein percentage and fat percentage, although further research is required to validate these associations and to rule out the possibility of a more distantly placed polymorphism in linkage disequilibrium with specific Muc1 VNTR alleles.
Muc1 allelic variants of very different sizes and with consequential glycosylation differences may play a physical role in dictating the size of the milk fat globules. This, in turn, could affect the triglyceride-carrying capacity of milk and hence lead to an association between Muc1 alleles and the fat content of milk. Further, measurement of β-CN and s1-CN mRNA in bovine mammary tissue showed highly significant correlations with Muc1 mRNA (data not shown), thereby potentially supporting a relationship between the Muc1 alleles and milk protein content. There was no association between Muc1 alleles and SCC, which is a measure of susceptibility to mastitis in dairy cattle, even though Muc1 binds E. coli. One possibility is that the bacterial-binding capacities of the allelic variants are equivalent despite their considerable structural differences. The abilities of the D and G alleles to prevent adherence of bacteria to cells grown in culture were also compared. Although there was no significant difference, there was a trend for the longer allele to be more efficacious (result not shown). These 2 alleles may not be sufficiently different to highlight functional differences. It is noteworthy that human MUC1 shows an association between VNTR allele size and susceptibility to gastritis (Vinall et al., 2002), and a variety of human mucins have shown associations between their VNTR alleles and disease susceptibility (Thornton et al., 2008). Thus, functional differences among the bovine Muc1 allelic variants are anticipated. The binding of bacteria by Muc1 may help prevent mammary tissue infections (mastitis) and reflect an early biological response to infection, which is not associated with changes in neutrophil populations (i.e., SCC).
The functions of Muc1 in milk are probably closely associated with its major architectural feature, extensive glycosylation. Bovine Muc1 is heavily glycosylated, primarily with simple sialylated O-linked oligosaccharides, reflecting a predominance of O-linked sialyl-T-antigen oligosaccharide. A similar conclusion was reached in another study, but unlike the results of the current analyses, it also reported indirect evidence for a substantial quantity of N-linked glycans (Liu et al., 2005). The reasons for this latter discrepancy are unclear. Each of the 4 different major VNTR units has 7 or 8 potential O-linked glycosylation sites. The most prevalent of the 4 VNTR units also has a potential N-linked glycosylation site. Thus, allelic variation not only defines the probable extent of O-linked glycosylation, which depends on the number of VNTR units, but also the extent and density of N-linked glycosylation, which structurally probably extends well beyond the background of the shorter O-linked oligosaccharides and may be functionally relevant.
The VNTR domain of the Muc1 protein is probably a filamentous or rod-like structure caused by several factors: 1) the abundance of Pro, which favors an extended rod-like conformation; 2) the abundance of Neu residues, which create strong repulsive charges on the surface of the protein; and 3) the immobilization by the oligosaccharides of a significant layer of water (Shi and Caldwell, 2000). The AA sequences flanking the hydrophilic VNTR domain are predicted to have a greater secondary structure and greater hydrophobicity (PepPlot; Gribskov et al., 1986). This alternating hydrophobic and hydrophilic structure of the extracellular domain may provide a nonspecific binding scaffold characterized by an array of low-affinity hydrophobic and hydrogen bonds that provide a pattern recognition function with flexibility to adapt to the surface shape of macromolecules, particles, and bacteria. It has been proposed that one of the primary functions of mucins is to protect epithelial cell surfaces from bacterial invasion. This could be achieved by the immobilization of bacteria at those surfaces through such nonspecific binding interactions with mucins (Patton et al., 1995). Subsequent shedding of the mucin by specific proteolytic processing or, in the case of the mammary gland, the formation of MFGM could allow elimination of bound bacteria in the normal fluid efflux.
The current study has added further evidence to the emerging role of Muc1 as a defensive antibacterial agent by demonstrating that it has the ability to directly bind biotinylated E. coli and inhibit the binding of fluorescently labeled E. coli BioParticles to bovine MEC. The binding and consequent immobilization of bacteria by Muc1 may play a key role in protecting epithelial cells from bacterial infection. Indeed, the gastrointestinal tract of Muc1-deficient mice was more easily colonized by H. pylori, thereby providing strong support for the role of Muc1 in protection of epithelial tissues from bacterial infection (McGuckin et al., 2007). The extensive sialylation of bovine Muc1 is probably important for its ability to bind bacteria because 2–3-linked Neu present on many cellular glycoproteins plays a crucial role in adherence of several bacterial species, particularly gram-negative species. Because the extent of O-linked glycosylation, including sialylation, is proportional to the number of VNTR units in the allelic variants, the marked differences in sizes between the smallest and largest Muc1 alleles (e.g., 7 vs. 23 VNTR units) may significantly affect the bacterial-sequestering activity of Muc1. Indeed, comparison of the abilities of the G (11 VNTR) and D (16VNTR) Muc1 protein alleles to inhibit bacterial binding to MEC was consistent with this proposal, although the results were not significant (result not shown). Comparison of more divergent but rarer alleles may be informative. The ability of bovine Muc1 to sequester bacteria suggests that it is a component of the innate immune defense system. Consistent with this proposal, stimulation of bovine MEC with LPS caused rapid upregulation of the expression of Muc1 mRNA, a characteristic of many innate immune genes.
Not only does milk have great importance in providing nutrition to newborn mammals during a period of intense growth and development, but it also contains a plethora of bioactive proteins, lipids, and sugars. Milk contains several antimicrobial proteins, which presumably protect the mammary gland and also the gut of suckling young from bacterial infections (Peterson et al., 1998). This strategy has to cope with 2 biological issues. First, any antibacterial agent will be severely diluted by the volume of milk and then shed at each milking. Thus, such agents must be synthesized and maintained at high levels in milk to exert their defensive capability. To minimize energy expenditure, the agent would preferably have pleotropic functions, particularly in relation to the structural composition of milk. Second, the exposure of mammary tissue and the neonatal gut to diverse infectious agents suggests that any defensive strategies require broad specificity. Bovine Muc1 fits both criteria because it is relatively abundant in milk (but concentrated on MFGM), it may have multiple functions, and it binds a variety of bacterial species (P. Parker, L. Sando, R. Pearson, K. Kongsuwan, S. Smith, and R. L. Tellam; unpublished data).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
2 Current address: School of Veterinary Science, University of Queensland, St. Lucia 4067, Australia.
Received for publication March 16, 2009. Accepted for publication June 16, 2009.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSFOOTNOTESACKNOWLEDGMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or LTA (
). Bovine MEC were stimulated with 50 µg/mL of LPS or 20 µg/mL of LTA, and Muc1 expression was monitored over 24 h by qRT-PCR. Data are expressed as mean fold change ± 1 SEM (n = 3) compared with expression in unstimulated cells. *P < 0.05.