Main

During the first months of life, maternal milk is the best source of nutrition for the newborn(1–7). It contains nutrients, enzymes, hormones, growth factors, and host defense agents. The latter are a combination of various specific and nonspecific factors such as oligosaccharides, hormones, LF, vitamin A and C, the B complex, binding proteins, lysozyme, and antibodies (IgA and low concentrations of IgG and IgM)(8–14).

Careful consideration must be given to low-birth-weight or preterm neonates. During the last few years, advances in their care have allowed the survival of a large number of extremely precarious infants. This has shifted the attention of neonatologists to the identification of the most suitable nutritional support for such infants. It is well known that early nutritional management plays an important role in the immediate survival, subsequent growth, and long-term outcome of extremely immature infants. These subjects, because of their unlimited nutritional reserves and gastrointestinal immaturity, risk being fed with nutrients inappropriate in composition or quantity. Nutritional problems are complicated by a number of additional stress, deficit, and metabolic factors specific to these infants.

For the above reasons, identification of the best dietetic program is one of the most important topics in neonatology. The requirements of preterm neonates seem to be higher than those of term infants(15–18), and different authors have suggested a fortification of human milk to satisfy the preterm infant's needs. According to this strategy, the benefits of human milk (for example, immunologic protection and long-term influence on neurodevelopment)(19) should be associated with the increase of protein content. In fact, preterm newborns require 2.25-3.1 g of protein per 100 kcal, whereas term infants need only 1.8-2.8 g/100 kcal(20).

Some researchers have shown that "early" preterm milk has higher nitrogen and total fat concentrations(21–28). In the article by Anderson(24), nitrogen derivatives, protein and nonprotein components such as urea, ammonia, and free amino acids, are 15-20% higher in preterm versus term milk whereas lactose concentration is lower (approximately - 15%) in preterm milk(23,24). These data are not confirmed in other studies; in fact, nitrogen content varies greatly from study to study, and the differences between term and preterm milk were not statistically significant in any study. However, although the differences in nutrient composition of preterm and term milk remain uncertain, all authors agree that these differences disappeared after the first month of lactation and that the variability in nutrient content (protein, fat, and minerals) is higher in preterm than in term milk(9,21–23,28).

The purpose of the present study was to evaluate differences in protein concentration between preterm and term milk; in particular, we studied the LF, ALA, β-casein, and lysozyme contents at different days during lactation. Knowledge of the composition of preterm newborn mother's milk could be an important step for formulation of the most suitable diet for premature infants.

METHODS

Milk samples were collected from two different groups of healthy mothers. Among them, eight women delivered prematurely between the 27th and 30th week of gestation (the selection was made at the Department of Pediatrics, San Gerardo Hospital, Monza) and eight women (volunteer donors, with an uneventful gestation) delivered normally between the 37th and 41st week of gestation. The mothers filled out a diet form in which they indicated the foods consumed day by day.

Collection and analysis of milk samples. TM collected milk samples by manual pressure of the breast during the first 72 h at the hospital and on the other days at home. To standardize the collection, milk samples were taken between 0700 and 1000 before the baby was fed. Mothers of the premature hospitalized infants were asked to collect the samples at home, using a pump and with the same time schedule. All samples were stored at -20°C until analysis. Samples of the first 24 h after the lactation onset were called d 0 and, thereafter, every 24 h as d 1, 2, 3, etc. The mothers collected one milk sample per day during the first week postpartum and then one sample every 5 d.

The milk samples were mixed with distilled water 1:4 (vol/vol); the samples obtained were treated with sample buffer (containing 0.25 M Tris-HCl pH 6.8, 7.5% glycerol, 2% SDS, 5% β-mercaptoethanol) in the ratio 1:1 (vol/vol) and heated in boiling water for 10 min at 100°C.

Purified milk proteins. Purified human milk proteins were purchased from Sigma Chemical Co. (St. Louis, MO) at the highest available purification level. These were LF (lyophilized powder, minimum 90% protein by electrophoresis); human serum albumin (lyophilized powder, minimum 99% protein by agarose electrophoresis, essentially globulin-free); IgA (lyophilized powder, minimum 95% protein by electrophoresis); IgG (lyophilized powder, minimum 95% protein by electrophoresis); human casein (lyophilized powder, approximately 85% protein by Biuret); lysozyme (lyophilized powder, approximately 10% protein, activity minimum 100 000-200 000 U/mg protein); and ALA (lyophilized powder, minimum 90% protein by electrophoresis).

The proteins were suspended in sample buffer at a final concentration of 1 mg/mL.

SDS-PAGE. To identify the different proteins in human milk sample, we used a gradient polyacrylamide gel with the following characteristics: a) gradient running gel: 9 - 19% acrylamide, 0.08-0.17% bis-acrylamide, 0.36 M Tris-HCl buffer pH 8.8, 35% glycerol, 0.1% SDS, 0.02% ammonium persulfate, and 0.15% TEMED; b) stacking gel: 3.5% acrylamide, 0.09% bis-acrylamide, 0.125 M Tris-HCl buffer pH 6.8, 0.1% SDS, 0.02% ammonium persulfate, and 0.15% TEMED; and c) running buffer: 25 mM Tris, 0.19 M glycine, and 0.1% SDS (wt/vol) pH 8.8.

After the electrophoretic run (90 V at room temperature for approximately 6 h), gels were dyed with Coomassie Brilliant Blue G-250 according to the method of Neuhoff et al.(29). All materials and instruments were purchased from Bio-Rad (Richmond, CA).

Protein bands were identified by comparing their electrophoretic mobility to those of purified human milk proteins. Quantitative analyses were performed using a gel scanner (Sharp JX-330, Pharmacia Biotech) and the Image Master 1 D software, which allows the quantification of proteins, calculating the average density of pixels across the band length and integrating over the bandwidth.

To calculate the amount of protein in each band, we used a calibration curve generated by plotting the known value of purified protein loaded onto the gel versus the corresponding area obtained by integration. Every sample was analyzed at least four times, and the data are reported as mean ± SE. The coefficient of variation for the values of each sample was always <5%.

Statistical analysis. The significance of the differences between the mean values was calculated by multivariate analysis of variance (MANOVA) and then by the Fisher multiple range test.

RESULTS

In Table 1, the clinical data of lactating women included in the study are shown. Every mother was in good nutritional condition and consumed a balanced diet as determined by the standard form on which the mothers reported their daily diet (Table 2). Height, weight (before pregnancy), and socioeconomic conditions were comparable. PM were older (34 ± 0.86 y) than TM (26.88 ± 0.74 y). There was a different increase in weight during pregnancy between PM and TM, both for the different gestational age (low increases) and for maternal pathologies (increase of 27 kg for gestosis). Table 2 lists the weeks of gestation and infant birth weight; body weight was obviously lower in preterm neonates (0.95 ± 0.31 versus 3.32 ± 0.53 kg, mean ± SD).

Table 1 Clinical data of lactating mothers
Full size table
Table 2 Data on gestation, newborn weight, and mother's diet in relation to each mother
Full size table

The SDS-PAGE profiles of purified human milk proteins and milk samples at different days of lactation from TM (0, 11, 31, 90, 171) and PM (0, 12, 31, 91, 182, 241) are shown in Figures 1 and 2, respectively.

Figure 1
figure 1

SDS-PAGE of purified milk proteins and milk samples at different days of lactation from mothers with term parturition. H and L indicate heavy and light chains, respectively, of IgA and IgG; CAS, casein; LYS, lysozyme; HSA, human serum albumin.

Full size image
Figure 2
figure 2

SDS-PAGE of purified milk proteins and milk samples at different days of lactation from mothers with preterm parturition.

Full size image

The human milk proteins appear well separated and easily identifiable. LF, human albumin, lysozyme, and ALA are single bands. Because samples were prepared under denaturing conditions, IgA and IgG are split into light and heavy chains. Human casein contains two major components, described as β and κ caseins, according to their specific electrophoretic mobility(13).

In the milk samples from both groups of mothers, the total quantity of the protein associated with different bands is higher at d 0; then, from the second day of lactation, protein quantity decreases progressively. This pattern is markedly evident for immunoglobulins (IgA and IgG). In Figure 3A, the SDS-PAGE pattern of some milk samples (TM) and known amounts of purified human milk proteins are shown; the amount of LF and ALA in the samples was calculated considering the linear regression deriving from the values of purified proteins (Fig. 3B). β-casein and lysozyme concentrations were determined similarly.

Figure 3
figure 3

A, Gel electrophoresis of purified standard proteins (LF and ALA) and milk samples from mothers with term parturition. MK1 = 0.4 µg LF + 3 µg ALA; MK2 = 1.5 µg LF + 1.5 µg ALA; MK3 = 3 µg LF + 0.4 µg ALA. B, Example of calibration curve for the determination of LF.

Full size image

The variation of LF protein concentration (expressed as g/L) in the milk of TM and PM is shown in Figure 4. In TM, LF concentration falls rapidly from d 0 (9.75 ± 0.25 g/L) to d 10 (2.43 ± 0.18 g/L), at which time a plateau is reached (values ranging between 1.4 and 2.6 g/L). According to the data obtained from the only mother who lactated for 9 mo (data in the Figure stop at d 140), LF concentration also remains in the same range after the sixth month. The values are approximately 1.58 g/L at d 175 and 2.97 g/L at d 235. A higher variation is observed in the milk from PM; moreover, at d 0, the LF concentration (4.30 ± 0.75 g/L) was lower than that of TM samples, gradually reaching the TM value at d 10 (2.29 ± 0.34 g/L). After d 60, when only two mothers were able to continue lactation, LF concentration remained fairly constant (1.8-4.0 g/L). The only significantly difference between the two groups was observed at d 0 (4.30 ± 0.75 versus 9.75 ± 0.25 g/L; p = 0.04).

Figure 4
figure 4

LF concentrations (mean ± SE) in milk from mothers delivering term and preterm babies during 5 mo of lactation. Each value is the average of at least four independent determinations. Because only two preterm mothers were able to continue the lactation after d 60, no SE is reported thereafter for the preterm group.

Full size image

In all samples, ALA concentration shows a higher variability compared with LF profiles both in TM and PM milk samples, as shown by SE in Figure 5. In the TM milk, ALA concentration is similar during all the lactation period: 3.56 ± 0.45 g/L at d 0, 3.73 ± 0.38 g/L at d 10, 3.68 ± 0.66 g/L at d 34, 3.23 ± 0.69 g/L at d 66, and 2.83 ± 0.60 g/L at d 115.

Figure 5
figure 5

ALA concentrations (mean ± SE) in milk from mothers delivering term and preterm babies during 5 mo of lactation. Each value is the average of at least four independent determinations. Because only two preterm mothers were able to continue the lactation after d 60, no SE is reported thereafter for the preterm group.

Full size image

Similarly, in PM milk, ALA concentration is 4.57 ± 0.66 g/L at d 0, 3.84 ± 0.38 g/L at d 10, and 2.95 ± 0.42 g/L at d 34. Samples from the only two mothers who continued lactation for >2 mo show a similar trend. No statistical difference was evident between ALA values of TM and PM samples.

β-casein shows a variable trend: the average values go up and down cyclically and SE is high (Fig. 6). The mean values vary between 2.7 and 5 g/L for TM and between 1.8 and 5.7 g/L for PM. No statistically significant difference was observed.

Figure 6
figure 6

β-CAS concentrations (mean ± SE) in milk from mothers delivering term and preterm babies during 5 mo of lactation. Each value is the average of at least four independent determinations. Because only two preterm mothers were able to continue the lactation after d 60, no SE is reported thereafter for the preterm group.

Full size image

Lysozyme concentration is constant during all the lactation period, as shown in Figure 7; in TM, mean values range between 0.16 and 0.46 g/L whereas in PM the range is between 0.17 and 1.10 g/L. Although the mean values of PM samples are always higher than those of TM, no statistically significant difference can be found.

Figure 7
figure 7

LYS concentrations (mean ± SE) in milk from mothers delivering term and preterm babies during 5 mo of lactation. Each value is the average of at least four independent determinations. Because only two preterm mothers were able to continue the lactation after d 60, no SE is reported thereafter for the preterm group.

Full size image

DISCUSSION

Although breast-feeding has been recommended for all neonates, some doubts still exist concerning the suitability of human milk for premature infants. The advantages are that breast milk contains a number of trophic factors important for stimulating the development of the immature organisms (digestive, hepatic, and renal systems) and that it contributes to protection against infections and to prevention of food allergy(17). Moreover, maternal lactation is also advantageous for the psychologic benefit on the mother who is actively involved in the care of her infant.

However, there is now increasing evidence that human milk cannot completely satisfy the needs of these vulnerable babies(16–19).

Because many of the physiologic advantages of human milk are provided by proteins, we evaluated, using an electrophoretic technique, the profiles of four human proteins (LF, ALA, β-casein, and lysozyme) during lactation in the milk of PM and TM.

The mothers included in our study were in good nutritional condition; moreover, they were not receiving any medication that could interfere with lactation. It is important to underline the general difficulty in finding a sufficient number of PM who were able to feed their babies for a long period of time. In fact, 50% of PM interrupted breast-feeding during the first 45 d, whereas most of the TM fed their babies for at least 5 mo. Many psychologic, dietetic, and hormonal factors can be implicated in the disappearance of milk in PM.

LF is an iron-chelating protein with high antibacterial activity against a wide variety of microorganisms that require iron for growth. It has also been suggested that LF can be involved in the process of iron absorption(8,30).

LF concentration in the milk of TM presented a common trend; in particular, the highest concentration of LF was shown in the very first days, when it plays its role as a potent antibacterial agent. LF content declines very slightly from d 0 to 10 when a plateau is reached. The decreasing trend of LF concentration in TM during the first day of lactation confirms the data by Hirai et al.(31) who reported for colostrum, transitional milk (d 4-7), and mature milk (d 20-60) values of 6.7 ± 0.7, 3.7 ± 0.1, and 2.6 ± 0.4, respectively.

In the milk of PM, the variability is higher; moreover, the LF content does not present the high value in the first days as observed in the milk of TM. This means that the PM milk has a lower antibacterial activity in the first days of life, and this LF shortage could be critical for preterm subjects, considering their particular susceptibility.

The presence of ALA in human milk is essential because of its high content of some essential amino acids (tryptophan, phenylalanine, lysine, and cysteine). It is involved in supporting rapid growth and some developmental processes. It also plays an important role in the lactose synthesis of the mammocyte(14).

ALA trend is variable both in TM and PM milks, with considerable differences among the mothers and for the same mothers during the lactating period. ALA concentrations for TM are comparable to those reported by Cuilliere et al.(32): colostrum 4.9 g/L, transitional milk 5.2 g/L, and mature milk 3.4 g/L.

β-casein is another important protein nutritionally; it represents 15% of the total protein content of human milk, presents a balanced amino-acidic pattern, and supplies calcium and phosphates(33). The trend of this protein during the lactation period is similar for TM and PM; both β-casein concentrations fluctuate around 3.5 g/L. These data correlate well with those by Kunz and Lönnerdal(13) who reported TM values ranging between 3.3 and 5.9 g/L.

Lysozyme is a basic polypeptide with antibacterial activity that contributes to the immunologic system; the concentration in human milk is about 3000 times higher than in cow's milk(34). It has been suggested that lysozyme in human milk has a bacteriostatic function in the gastrointestinal tract of breast-fed infants.

The lysozyme concentrations show a constant trend, with values ranging between 0.2 and 0.5 g/L in TM and between 0.2 and 1 g/L in PM. Although the mean values of PM samples are always higher than those of TM, no statistically significant difference can be found. Similar values for TM lysozyme concentrations (0.1-0.4 g/L) were reported by Hennart et al.(35).

Although the number of lactating mothers included in term and preterm groups is not very high, we can conclude that the concentration of the four proteins analyzed in the milk of PM is very similar to that of TM. A higher interindividual variability of protein content was observed in PM.

These findings confirm the data of some authors(36–41) but are in conflict with those reported by others(21–28).

To explain the variability of protein profiles during lactation, we verified whether maternal diet and nutritional status could influence the quality of the milk. No correlation was found between maternal diet (a quali-quantitative registration of foods consumed was made by mothers every day) and milk concentration of proteins here considered.

It is necessary to reaffirm that since PM samples are not richer in protein than those of TM, breast milk requires fortification for preterm neonates who have a higher protein requirement. Although our study is surely limited, we hope that the research presented in the present study may contribute to improving knowledge in this field with the final aim of identifying the best nutritional solution for assuring better growth of preterm infants.