Main

Human milk is a unique and well-balanced source of nutrition for the newborn. Milk contains not only major nutrients (proteins, carbohydrates, and lipids) but also a variety of components such as minerals, vitamins, enzymes, hormones, growth factors, and immunoglobulins important for growth and healthy development during the neonatal period (1–3). As advances in neonatal care have significantly increased, the survival of extremely prematurely born neonates have increased (4). Yet these infants have the highest incidence of morbid events, such as NEC (5). Questions have arisen about the most suitable and relevant nutritional support for these immature infants (6, 7).

Within the last three decades, a large number of biologically active peptides have been identified in human milk (8–10), and the list is continuously growing (11–13). Among these, EGF is one of the major peptide growth factors present both in colostrum and human milk (14–17). Human milk EGF levels are highest in the first days after parturition and then gradually decrease during the first 2 wk of lactation (17). Another structural homolog of EGF, TGF-α is also present in human colostrum and milk (15, 18), but at much lower concentrations than is EGF. However, neither peptide is found in commercial infant formulas.

Although the roles of milk-borne EGF and TGF-α on the developing neonate are not clearly understood, recent studies indicate the importance of these peptides in repair processes in injured intestinal mucosa (19). The biologic actions of EGF and TGF-α are mediated via binding to the EGF-R, and the expression of EGF-R in fetal and neonatal gut (20) suggests their role in intestinal development (21). With decreasing gestational age, critical gastrointestinal functions are compromised and extremely premature infants are at highest risk of developing gastrointestinal disorders, such as NEC. Further, the feeding of human milk to premature infants is associated with a lower incidence of gastroenteritis (22) and NEC than formula (5, 23).

To determine whether the concentration of growth factors in human milk varies with gestation, we evaluated the concentrations of EGF and TGF-α in human milk collected during the first month of lactation from mothers with EPT, PT, and FT infants. Because neonatal gastrointestinal diseases typically develop within the first postpartum weeks, milk EGF and TGF-α concentrations and protein content were evaluated longitudinally during the first month of lactation.

METHODS

Sample collection.

Mothers of infants in the neonatal nursery at Texas Children's Hospital were recruited. Informed written agreement was obtained. The protocol was approved by the Institutional Review Board for Human Subject Research at the Baylor College of Medicine Hospital, Houston, TX, U.S.A. Human milk samples were obtained from mothers of infants born EPT (23-27 wk gestational age; n = 16), PT (32-36 wk gestational age; n = 16), and FT (38-42 wk gestational age; n = 15). Milk samples were collected in midmorning by a single expression from one breast and brought to the hospital unfrozen at the end of the first, second, and fourth weeks postpartum. Each mother supplied three samples, one for each time point. All samples were stored at -70°C before analysis.

Milk sample preparation.

Frozen milk samples were quickly thawed, mixed (1:4 by volume) with 50 mM Tris buffer (pH 7.4), and centrifuged at 16,000 × g at 4°C for 15 min. Defatted milk whey was carefully aspirated and centrifuged again (as above). The clear milk whey volume was recorded, and samples were lyophilized and stored at -20°C before RIA analysis.

Protein measurements.

Milk protein concentrations were measured as described previously (24). Assays for total protein content (25) were determined by spectrophotometry (SPEC-TRAmax PLUS, Molecular Devices, Sunnyvale, CA, U.S.A.).

EGF and TGF-α RIA.

Before RIA analysis, each sample was rehydrated with double-distilled water. Human milk EGF content was measured using human EGF RIA kit, (Biomedical Technologies, Inc, Stoughton, MA, U.S.A.) as described in the manufacturer's protocol. Milk TGF-α content was measured using rat TGF-α RIA kit (Peninsula Laboratories, Inc, Belmont, CA, U.S.A.) as described in the manufacturer's protocol. The specificity of anti-human TGF-α antibody guaranteed by the manufacturer is 100% cross-reactivity with human TGF-α and 0% cross-reactivity with human EGF. Interassay variation for both EGF and TGF-α was 10% or less.

Statistics.

Statistical analyses were performed by ANOVA followed by Fisher's protected least significant difference using the statistical program StatView for Macintosh computers (Abacus Concepts Inc, Berkeley, CA, U.S.A.). A value of p < 0.05 was considered significant at the 95% confidence level. All data are expressed as the mean ± SEM.

RESULTS

Protein concentrations in human milk.

Milk protein concentrations were evaluated to verify that any changes in peptide growth factor concentrations were not a result of differences in total protein content. Total protein concentration in all experimental groups was highest on d 7 of lactation and then gradually decreased during the first month of lactation (Fig. 1). There were no statistically significant differences in milk protein concentrations among experimental groups on d 7. However, on d 14 the EPT group exhibited significantly higher milk protein concentration compared with the PT and FT groups (p < 0.05). On d 28, statistically significant reduction in milk protein concentration was observed between the EPT and PT groups (p < 0.05).

Figure 1
figure 1

Total protein content in human milk from mothers of infants born EPT (23-27 wk gestational age, n = 16), PT (32-36 wk gestational age, n = 16), and FT (38-42 wk gestational age, n = 15) during the first month of lactation. Columns are mean values, vertical lines are SEM. *p < 0.05 EPT vs PT or FT group.

Full size image

Milk EGF levels.

Human milk from the EPT group (23-27 wk gestational age) had significantly higher concentrations of EGF compared with values in milk from the PT and FT groups (Fig. 2). The same trend was observed when milk EGF data were expressed as the content of EGF per gram of milk protein (EPT, 7.9 ± 0.9; PT, 4.8 ± 0.5; FT, 4.9 ± 0.5 μg/g protein; p < 0.01). Milk concentrations of EGF between the PT and FT groups were not significantly different. Longitudinally, there was no statistical difference in EGF concentration from the EPT group. However, the EGF concentration from the PT and FT groups did show statistically significant decrease with time (PT d 7 versus d 28, p < 0.001; FT d 7 versus d 28; p < 0.05).

Figure 2
figure 2

EGF concentrations in human milk from mothers of infants born EPT (23-27 wk gestational age, n = 16), PT (32-36 wk gestational age, n = 16), and FT (38-42 wk gestational age, n = 15) during the first month of lactation. Milk EGF concentrations were measured using a homolog RIA. Columns are mean values, vertical lines are SEM. **p < 0.01 and *p < 0.05 EPT vs PT or FT group.

Full size image

Milk TGF-α levels.

Human milk TGF-α concentrations showed a similar pattern as observed for EGF values (Fig. 3), with significantly higher TGF-α concentrations in the EPT group compared with the PT or FT groups (p < 0.05). Moreover, the absolute values of milk TGF-α were 100-1000 times less (ranging from 0.2 to 0.6 μg/L) compared with the corresponding EGF data (ranging from 40 to 560 μg/L). Longitudinally, TGF-α concentration from the EPT and PT groups were statistically decreased between d 7 and d 28 (p < 0.01). There was no statistically significant difference in the FT group longitudinally.

Figure 3
figure 3

TGF-α concentrations in human milk from mothers of infants born EPT (23-27 wk gestational age, n = 16), PT (32-36 wk gestational age, n = 16), and FT (38-42 wk gestational age, n = 15) during the first month of lactation. Milk TGF-α concentrations were measured using a homolog RIA. Columns are mean values, vertical lines are SEM. **p < 0.01 and *p < 0.05 EPT vs PT or FT group.

Full size image

DISCUSSION

The present study describes EGF and TGF-α levels in human milk from mothers with EPT (23-27 wk gestation) babies and compares them to the EGF and TGF-α levels in the milk from mothers with PT (32-36 wk) or FT (38-42 wk) babies. Human milk EGF concentrations in the EPT group were significantly higher (60-80%) compared with the PT and FT groups. This trend was observed throughout the first month of lactation, suggesting that increased secretion of EGF into the EPT milk is a long-lasting phenomenon. A similar pattern was detected in the milk TGF-α levels. This is in contrast to other milk nutrients whose concentrations decrease with time of lactation (2). Milk of mothers from the EPT group had significantly higher concentrations of TGF-α.

Results from our studies clearly indicate elevated EGF levels in human milk from mothers with EPT babies throughout the first month of lactation. Human milk EGF levels reported in previous studies have been contradictory and not clearly defined. Whereas Read et al. (17) found markedly higher EGF levels in colostrum compared with mature milk, Connolly and Rose (15) detected similar concentrations of EGF in colostrum and milk. Moran et al. (16) reported no difference in the concentration of EGF in milk from women delivering preterm versus full-term babies. However, the latter study compared preterm infants of 27 to 32 wk of gestation. In our study significant differences in milk EGF levels were detected in EPT babies (23-27 week gestation) only. Read et al. (26) reported elevated EGF levels in colostrum and human milk of mothers with very premature babies compared with less premature or term babies. However, they used a radioreceptor technique, which is based on a competitive binding of ligands to the EGF-R. This method simultaneously detects several members of the EGF-like peptide family, such as EGF, TGF-α, heparinbinding EGF, and betacellulin, all reported to be found in milk (12, 14, 18, 27). In the present study, we used a peptidespecific and quantitative RIA, with no cross-reactivity among members of the EGF-like family of peptides. In addition, the milk samples were divided into three experimental groups based on the gestational age of the newborns, and the results were consistent during the three periods.

The presence of TGF-α in colostrum and human milk is even less understood than EGF. Originally, TGF-α-like activity was detected in human colostrum; however, the exact concentration was not described (28). Later, Connolly and Rose (15) detected the presence of TGF-α in human milk using a radioreceptor technique. Okada et al. (18) verified TGF-α concentration in human milk using a specific RIA assay. Wagner et al. (29) have found that human milk contains different biochemical forms of TGF-α that can originate from human milk macrophage secretion. Results from these studies have shown that TGF-α concentration in mature milk is markedly lower compared with EGF levels. To our knowledge, this is the first report of TGF-α levels in the milk of mothers with extremely prematurely born babies. In the present study, TGF-α levels were evaluated in human milk samples at different gestational ages and lactation times. Similar to EGF, TGF-α levels were the highest in the EPT group compared with PT and FT groups. Although statistically significant, the differences in TGF-α levels in milk from three gestational age groups were less pronounced compared with the differences in EGF levels from the same experimental groups. However, TGF-α levels were measured in the aqueous portion of human milk only. Recent studies indicate that TGF-α activity can also be detected in the fat portion of human milk (30). Therefore, milk fat globules can perhaps serve as an additional source of milk TGF-α (31).

What is the physiologic relevance of these results? Britton et al. (32) have shown that EGF is resistant to degradation by gastric juices of human premature infants. The presence of EGF-R in fetal and neonatal gut is well established (20), and developing intestine is considered to be a target organ for milk-borne EGF and TGF-α (19, 33). With decreasing gestational age of newborns, the critical functions of the gastrointestinal tract are compromised, and neonates are at high risk of developing intestinal disorders. A typical example is neonatal NEC, in which 90 to 95% of the cases occur in prematurely born infants (34, 35). A multicenter study from the United Kingdom has shown that the incidence of NEC among premature babies fed human milk is 6-10 times lower compared with formula-fed babies (5). However, the component(s) of human milk associated with these protective effects were not identified. Amniotic fluid contains significant concentrations of EGF that gradually increase during pregnancy, with the highest level achieved at the end of the normal gestation period (36). In contrast to human milk, EGF is absent in commercially available infant formulas (37, 38). Clinical studies have shown reduced serum and salivary EGF levels in neonates with NEC when compared with healthy babies (39, 40). In an animal model of NEC, we have recently shown that milk-borne EGF reduces the incidence and severity of NEC-like injury in neonatal rats (41). In suckling animals, supplementation of EGF into formula enhances the growth of stomach and the small intestine (42), induces precocious maturation of intestinal brush-border disaccharidase activities (43), and modulates intestinal nutrient transport (44). Based on these clinical and experimental studies, we speculate that elevated levels of EGF in human milk of extremely prematurely born babies can be potentially responsible for the protective effect of maternal milk against neonatal diseases.

CONCLUSIONS

In the present study we have shown that human milk EGF and TGF-α concentrations are markedly higher in mothers with EPT babies, compared with those with PT or FT babies. Because EPT infants are often maintained for extended periods on total parenteral nutrition and enteral formula is devoid of peptide growth factors, a better understanding of the physiologic functions of milk-borne growth factors on the development of neonates is critical for future clinical use of these growth factors.