| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Departamento de Bioquímica y Biología Molecular, Facultad de Biología, Universidad de Salamanca, 37007 Salamanca, Spain
Departamento de Fisiología, Facultad de Veterinaria, Universidad de León, 24071 León, Spain
Corresponding author:
P. Hueso; e-mail:
phueso{at}usal.es.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: Infant formula • human • bovine lactation • milk oligosaccharide
Abbreviation key: DSL = disialyllactose, DSLNT = disialyllacto-N-tetraose, LSTa = sialyllacto-N-tetraose a, LSTb = sialyllacto-N-tetraose b, LSTc = sialyllacto-N-tetraose c, OBSA = oligosaccharide-bound sialic acid, RT= retention time, 3'S3FL = 3'-sialyl-3-fucosyllactose, 3'SL = 3'-sialyllactose, 3'SLN = 3'-sialyllactosamine, 6'SL = 6'-sialyllactose, 6'SLN = 6'-sialyllactosamine
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
It has been described that certain oligosaccharides inhibit specific pathogen-binding to host cell receptors in the intestinal tract. Fucosyloligosaccharides from human milk inhibit the diarrheagenic effect of the heat-stable toxin of Escherichia coli and Campylobacter jejuni. The binding of Streptococcus pneumoniae, enteropathogenic E. coli and Haemophilus influenza to their receptors is also inhibited by human milk oligosaccharides (for a review, see Kunz and Rudloff, 1993; McVeagh and Miller, 1997; Newburg, 1997, 1999). Other oligosaccharides would protect against other pathogens. In in vitro assays, human milk oligosaccharides are minimally digested (Engfer et al., 2000; Gnoth et al., 2000), a prerequisite for them to be effective in the small intestine in preventing infections. However, little is known about the digestion of milk oligosaccharides during their passage through the calf gastrointestinal tract.
The milk contents of each species are specifically tailored to meet the special needs of the newborns. Therefore, the aim of this work was to carry out a comparative study of the oligosaccharide contents of human and bovine milk from different stages of lactation. Sialyloligosaccharides were chosen since they are the most abundant oligosaccharides in bovine milk (about 80%) and are found in high amounts in human milk (about 1 g/L). Breast-fed infants appear to have fewer or less severe gastrointestinal, respiratory and urinary infections than formula-fed infants (Kovar et al., 1984; Marild et al., 1990). Infant formula manufacturers have tried to emulate human milk in macronutrient composition and occasionally with respect to the actual molecular components. In addition, various substances, including oligosaccharides, have been attributed functional effects in certain special situations. Because the infant formulas designed for nutrition during the first weeks after birth are currently produced from different bovine milk derivatives, their sialyloligosaccharide contents were also compared to those of bovine and human milks.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Samples from fresh cow milk and colostrum were obtained from six individual multiparous Spanish-Brown cows at Huergas de Babia (León). Each cow provided four samples from four different lactation stages: Colostrum (d 2), transitional milk (d 7), mature milk (mo 3), and late lactation milk (mo 10) samples were obtained from the morning milking, immediately frozen at -20°C, lyophilized, and homogenized.
Five infant formulas (chosen from those most frequently consumed in Spain) from three different baby food companies were analyzed. These are recommended for use from birth to 3 to 4 mo of age. All formulas are bovine-milk based.
Oligosaccharide Standards
Several sialyloligosaccharide standards from human milk and bovine colostrum of known concentration were injected to obtain calibration curves on an HPLC apparatus. Disialyllacto-N-tetraose (DSLNT) 3'-sialyllactose (3'SL), 6'-sialyllactose (6'SL), 3'-sialyllactosamine (3'SLN), 6'-sialyllactosamine (6'SLN), 3'-sialyl-3-fucosyllactose (3'S3FL), sialyllacto-N-tetraose a (LSTa), sialyllacto-N-tetraose b (LSTb), and sialyllacto-N-tetraose c (LSTc) were purchased from Glyko (Upper Heyford, UK). Disialyllactose (DSL) was purchased from Sigma (St. Louis, MO).
Oligosaccharide-Bound Sialic Acid Quantitation
For quantitation of oligosaccharide-bound sialic acid (OBSA) contents, lyophilized milk samples and formulas were reconstituted in water according to their original water content or the manufacturers’ instructions, respectively; defatted by centrifugation (3000 x g, 30 min at 4°C), and filtered through glass wool. The filtrates were mixed with an equal volume of a 24% TCA (Sigma) solution, allowed to precipitate for 10 min on ice after vortexing, and centrifuged at 3000 x g for 30 min at 4°C. Protein precipitates were washed twice with cold 12% TCA to remove any traces of carbohydrates. After this, the supernatant solution and washings were combined and hydrolyzed in 0.05 M H2SO4 at 80°C for 1 h. The hydrolysates were cooled to room temperature and purified by ion-exchange chromatography on a Dowex 2 x 8 column (2 ml), formic form. Sialic acids were eluted with 0.1 M formic acid, lyophilized, dissolved in distilled water, and used to determine OBSA contents (Martín et al., 2001). They were then quantitated by the resorcinol procedure of Svennerholm (1957). Assays were made in triplicate; i.e., each individual sample was analyzed three times.
Isolation of the Oligosaccharide Fraction
Isolation of total oligosaccharides was performed as described by Kobata (1972), with several modifications. Briefly, samples were defatted by centrifugation (3000 x g, 30 min, 4°C) and filtered through glass wool. The filtrates were mixed with two volumes of ethanol and allowed to stand overnight at 4°C. Most proteins and lactose precipitated and were removed by centrifugation at 0°C and evaporating off the ethanol. The crude oligosaccharide fraction was further purified by molecular exclusion chromatography (Sephadex G-25) to eliminate residual proteins and peptides. Ninhydrin (0.3 g of ninhydrin in 95 ml of butanol) -positive fractions were discarded. Orcinol (0.1% orcinol in 20% H2SO4 ) -positive fractions were collected together and kept at -20°C until use.
HPLC Analysis
Sialylated oligosaccharide contents were analyzed by HPLC according to the method previously described by Michalski (1995). Analyses were performed on a Waters apparatus (Waters, Milford, MA), using an NH2-bound silica column (Carbohydrate Analysis, 300 x 3.9 mm, Waters). HPLC elution of bovine milk samples was performed with a gradient of acetonitrile and 15 mM KH2PO4 adjusted to pH 5.2: isocratic elution with 25% 15 mM KH2PO4 for 15 min, linear gradient up to 50% 15 mM KH2PO4 for 45 min, maintaining these conditions for 15 min. Flow rate was 1 ml/min. Samples of human milk and infant formulas were eluted under almost identical conditions, except that isocratic elution at 25% KH2PO4 lasted 10 min, and the linear gradient to 50% lasted 50 min for better separation. Oligosaccharides were detected at 206 nm and were identified by comparison of their retention times (RT) with those of standard oligosaccharides injected previously. Oligosaccharide quantification was carried out from the calibration curves of standard oligosaccharides using Millenium software (Waters) coupled to the HPLC system. All assays were carried out in triplicate.
Neuraminidase Treatment
Neuraminidase (E.C.3.2.1.18) from Clostridium perfringens was purchased from Sigma. Milk oligosaccharide samples were treated with neuraminidase in 50 mM sodium acetate buffer, pH 5.5, for 16 h at 37°C. Digested oligosaccharides were later analyzed by HPLC.
Statistical Analysis
Results are expressed as means ± SD. Oligosaccharide differences among stages were tested by one-way ANOVA using the SPSS 9.0.1 program for Windows (SPSS Inc., Chicago, IL). The Scheffé test was used for comparisons between groups. Differences were considered significant at a probability level of P < 0.01.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
, the OBSA content decreased along lactation, colostrum being the significantly different stage. In bovine milk, the slight increase in OBSA contents in the late lactation stage was not statistically significant. Infant formulas proved to be quite heterogeneous in their OBSA content (14.6 to 68.6 mg/L). In all cases, the average value obtained (38.1 mg/L) was quite low with respect to human milk and similar to that of midlactation bovine milk.
|
Human milk.
Along human lactation, a very similar chromatographic profile was observed (Figures 1a and 1b) from one stage to another. The chromatograms displayed 11 major peaks, but only seven of them could be identified using the standards available. These peaks corresponded to the following oligosaccharides: 3'SL, 6'SL, 3'S3FL, LSTa (NeuAc
2-3Gal ß1-3 GlcNAc ß1-3 Gal ß1-4 Glc), LSTb [Gal ß1-3 (NeuAc2-6) GlcNAc ß1-3 Gal ß1-4 Glc], LSTc (NeuAc2-6Gal ß1-4 GlcNAc ß1-3 Gal ß1-4 Glc), and DSLNT. LSTb and LSTc had very similar RT, and were, therefore, considered as LSTb+c, although it was not possible to know what percentage of the area corresponded to each of them. 3'SLN, 6'SLN, and DSL were not found in any of the human samples. The first peak in the chromatogram (RT:12.5 min) appeared slightly later than the large neutral oligosaccharide peak (not shown). The material in this peak was isolated and proved to be ninhydrin-positive, orcinol-positive, and resorcinol-negative. Most of the acidic oligosaccharides identified decreased significantly along lactation (Figure 2). 3'SL did not vary significantly along lactation. The most representative constituents were 3'S3FL, LSTa, and LSTb+c from among all the sialyloligosaccharides identified in this study. Peaks with an RT of >45 min could not be identified. The sialyloligosaccharide composition of individual women varied mainly with respect to one of the component, i.e., 3'-sialyl-3-fucosyllactose. Three of the women had high amounts of this sialyloligosaccharide (Figure 1), while the rest of the women had more moderate amounts.
|
|
shows the quantitative variations along lactation. Most of the significant differences were seen with respect to the colostrum, which was the stage with the highest values in most of the oligosaccharides identified. Nevertheless, 3'SLN was significantly higher in transitional milk.
|
) were generally characterized by the absence of peaks. Only one unidentified peak was present in all the chromatograms. This peak had the same RT as peak "a" in bovine milk. In some chromatograms, there were trace peaks of 3'SL, 6'SL, or unidentified compounds.
Neuraminidase Treatment
When oligosaccharides from human milk were treated with neuraminidase, all the peaks disappeared from the chromatographic profile; only those with an RT <8 min (neutral oligosaccharides) remained. The peak with an RT of 12.5 min did not disappear. After enzymatic treatment of bovine samples, only peak "a" remained unaltered in the chromatograms; the rest of the peaks disappeared.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Because treatment of human milk oligosaccharide samples with neuraminidase yielded a chromatogram lacking peaks (except for the peak with an RT of 12.5 min, which was assumed to be a retarded neutral oligosaccharide), all of the peaks in Figure 1a and 1b can be said to contain a sialic acid residue. No important qualitative changes in the sialyloligosaccharide chromatographic profile were observed from one stage to another. Nevertheless, a significant quantitative difference was observed between colostrum and mature milk, consistent with the overall tendency for OBSA to decrease. Considering the sialyloligosaccharides individually, in general, we obtained slightly higher values than those previously reported (Kunz and Rudloff, 1993; Coppa et al., 1999). Nonetheless, the significant differences among the stages reported in previous works were essentially in agreement with our own findings.
The data reported in the literature (Coppa et al., 1999) refer to a fairly constant value along lactation for LSTb, while LSTc apparently declines sharply in the mature milk. This prompts us to speculate that the significant decrease in our LSTb+c peak would be mainly due to the LSTc fraction.
A major sialyloligosaccharide species in human milk, DSLNT, has not previously been reported to decrease along lactation. Although we found the same amount of DSLNT in our mature milk samples as in previous works (Kunz and Rudloff, 1993; Coppa et al., 1999), our much higher values for colostrum could explain the significant descending trend observed by us.
We found large amounts of 3'S3FL in all three stages of lactation of three of the donors. Although its presence in human milk has been described previously (Grönd-berg et al., 1989), it has not been reported to be an important sialylated oligosaccharide in other quantitative studies (Kunz and Rudloff, 1993; Coppa et al., 1999). We are unaware of whether the presence of 3'S3FL in our samples is a genetic-dependent phenomenon. Human milk samples were collected from private donors, and no information about their blood groups was available.
Studies on individual variations in milk samples have been published (Newburg and Neubauer, 1995; Kunz et al., 1999). The relative proportions of 3'SL, 6'SL, and sialyltetraoses have been found to vary among individuals. Because diurnal differences in oligosaccharide contents have also been reported (Viverge et al., 1986), the level of variation seems to be extremely high. Nonetheless, the biological relevance of such variations remains obscure.
Parallel experiments carried out in our laboratory with human milk from Cuban donors showed a very similar composition in the sialyloligosaccharide content. However, we found a statistically significant difference in the LSTb+c fraction (data not shown). While in colostrum, Cuban women seemed to have lower amounts of LSTb+c than Spanish mothers, in transitional milk, the proportions of these tetraoses switched. Variations among different populations have been reported previously (Erney et al., 2000), but further experiments in this field need to be done.
Bovine milk contains acidic rather than neutral oligosaccharides (Saito et al.,1984). To date, 10 sialylated oligosaccharides have been described (Gopal and Gill, 2000). The main species identified in our work were 3'SL, 6'SLN, 6'SL, and DSL, which in bovine colostrum have already been reported to be predominant (Newburg et al., 1996). No data about sialylated oligosaccharides at other stages of lactation have been found.
We found a small amount of 3'SLN in the four stages considered, transitional milk being the richest in this compound. This species has not been previously reported in bovine colostrum or milk.
As in previous reports (Veh et al., 1981; Parkkinen and Finne, 1987), we found 3'SL to be the most representative species in bovine colostrum, although 6'SLN remained at constantly high values along lactation. Regarding the rest of the species identified, except for DSL, our values are appreciably higher than those reported in the literature. The presence of 2-3 isomers diminished along bovine lactation.
For all the sialyloligosaccharides identified, except 3'SLN, a significant difference was seen between the colostrum and other lactation stages. From a nutritional point of view, we consider the sialyloligosaccharide composition of mature stages to be very interesting, as it is commonly used in the manufacture of infant formulas.
In some chromatograms, we found trace peaks with RT values similar to those of the sialyltetraose standards. Sialyltetraoses have not been previously reported in cow’s milk or colostrum, so we cannot assign any structure without further studies. Urashima and coworkers (1991) described a neutral lacto-N-neo-tetraose. It is clear that individual breeds, diet, the length of the dry period of cows, and time postparturition make it difficult to assume a typical composition profile (Gopal and Gill, 2000).
With respect to the first peak, "a", in the bovine milk chromatograms, this was assumed to be a retarded neutral oligosaccharide, since it mostly remained unaltered after digestion with neuraminidase.
As shown in Figure 1d, the chromatographic profile of infant formulas was different from that of bovine milk or human milk. Nevertheless, the chromatograms of the individual infant formulas differed from one to another, as some of them were totally flat (total absence of peaks). By contrast, others had very low amounts of 3'SL and 6'SL or showed unidentified trace peaks. In any case, they resembled those from bovine mature milk more than those from human colostrum and milk.
Since several studies have reported that most human milk oligosaccharides are not digested nor absorbed into the small intestine of infants (Engfer et al., 2000; Gnoth et al., 2000), it is reasonable to speculate that they could play a defensive role, among others. In this sense, several reports about the inhibition of microbial adhesion to host cells by fucosylated and sialylated oligosaccharides have been published (Kunz and Rudloff, 1993; McVeagh and Miller, 1997; Newburg, 1997, 1999). In calves, sialylated oligosaccharides could play the same role before they become ruminants. Indeed, experiments carried out at our laboratory have demonstrated that bovine milk sialylated oligosaccharides bind to several enterotoxigenic E. coli strains isolated from calves (unpublished data).
The bioactivity of oligosaccharides from bovine and human milk would be similar and, therefore, bovine oligosaccharides could be used in milk products as bioactive components in human nutrition (Gopal et al., 2000). Nonetheless, breast-fed infants are better protected against several types of infections than formula-fed infants (Newburg, 1997). Infant formulas are designed specifically to mimic the composition of human milk or the functional aspects of human breast-feeding. The idea of supplementing formula milk with oligosaccharides similar to those found in human milk has been discussed previously (McVeagh and Miller, 1997; Motil, 2000). Large-scale production of oligosaccharides is currently being investigated, and its application in infant nutrition has been suggested (Endo and Koizumi, 2000). However, further research is needed as regards the structural identity, form, amounts, stability, and biological role of human milk oligosaccharides before this step can be taken.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The results point to the general conclusion that, regarding both qualitative and quantitative aspects, milk from humans and cows and infant formulas have different oligosaccharide contents. The manufacturing process for infant formulas has virtually eliminated OBSA. In this sense, bottle-fed infants are subject to reduced sialyloligosaccharide intake as compared to breast-fed infants.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
We acknowledge the generous collaboration of the private donors, as well as the great help given by Mrs. Esmeralda Gómez de Miguel, midwife from the Health Center "La Alamedilla" (Salamanca, Spain), who kindly provided us with a large part of the samples. We thank Mr. Gerardo Álvarez, who carefully collected all the bovine samples. We also wish to thank Dr. Gilda Marquina from the Center of Molecular Immunology (Havana, Cuba) for providing the Cuban samples. We are also grateful to Mr. N. Skinner (from the Servicio de Idiomas, Universidad de Salamanca) for revising the English version of the manuscript.
Received for publication June 14, 2002. Accepted for publication August 5, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
1-3 Gal 1-4 Glc. Biochim. Biophys. Acta 1073: 225–229.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
