This symposium addresses the topic of bioactive components in human milk and examines their role in newborn infants. The milk components examined may play a role in the development of infants (growth factors, hormones) and in protection from infection (cytokines, glycoconjugates). Because of time constraints, many additional bioactive milk components could not be discussed.

I would like therefore to address briefly in this introduction the function of several other bioactive constituents of human milk considered conditionally essential nutrients, i.e., nutrients that are traditionally considered nonessential but may become essential during development or in certain diseases when synthetic capacity does not meet functional demands; such nutrients include nucleotides and long-chain polyunsaturated fatty acids (LC-PUFA). The immaturity of pancreatic digestive function might render the bile salt-dependent lipase and amylase of human milk essential for efficient fat and carbohydrate digestion.

All the major nutrients of human milk have multiple functions, e.g., in addition to providing the building blocks and energy necessary for growth, many have specific functions such as protection against infection and as ligands and carriers (Table 1) (Hamosh 1996).

Table 1. Multiple functions of the major nutrients in human milk1 [View Table]

Although infant formulas provide the nutrients needed for adequate growth, many of the bioactive properties of nutrient and nonnutritive components of human milk cannot be duplicated by formulas. In some cases this is due to quantitative differences between human milk and cow's milk (Jensen 1995), a major component of infant formulas, as well as to qualitative differences (in LC-PUFA, nucleotides, digestive enzymes, immunoglobulins, glycoconjugates, etc.) (Hamosh 1996) (Table 2). Should these components be added to infant formulas? What can we learn from examples where such supplementation has been performed?

Table 2. Conditionally essential nutrients with bioactive functions in human milk1 [View Table]

NUCLEOTIDES

Nucleotide supplementation of infant formula, started in Japan and currently practiced also in some European countries, has been advocated also for the United States. However, close examination of many of the effects of nucleotide supplementation in newborns indicates lack of reproducibility of many findings (Table 3). There is therefore a good case for deferring supplementation until more research can resolve these disparities. Differences in effects of nucleotide supplementation among the various reports might be associated with formula composition, amount of nucleotides added, the process of formula manufacture, gestational and postnatal age of the infants studied and analytical techniques. Careful analysis of the data presented suggests that there is good agreement for a role of nucleotides in immune function (Kulkarni et al. 1994) and in intestinal repair after injury (Brunser et al. 1994, Nunez et al. 1990). The effect on gastrointestinal maturation seems to be either minimal (Uauy et al. 1994) or absent (Carver 1994, Hamosh et al. 1994b). The effects on lecithin:cholesterol acyltransferase, plasma lipoproteins and apoproteins, and plasma and erythrocyte LC-PUFA are weak, of short duration, or absent (Table 3).

Table 3. Conflicting data on nucleotide supplementation of infant formula1 [View Table]

The hypothesis that nucleotides might increase the synthesis of LC-PUFA by inducing intestinal and hepatic desaturase (De Lucchi et al. 1987) remains unsubstantiated in light of the absence of an effect in low-birth-weight (Woltil et al. 1995) and very-low-birth-weight (Henderson et al. 1994) infants.

LONG-CHAIN POLYUNSATURATED FATTY ACIDS

The question of supplementation is still open in the United States, although infant formulas supplemented with DHA are available in Europe and Japan and formula containing arachidonic acid and DHA is available in Europe. There seems to be agreement that preterm infants benefit from this supplementation, although it is not clear whether this is a transient or long-lasting effect. Conflicting reports on the ability of full-term infants to elongate and desaturate the precursors of DHA (linolenic acid) and arachidonic acid (linoleic acid), and thereby to synthesize these fatty acids, have left open the question whether formulas for full-term infants should be supplemented with LC-PUFA at the level present in human milk (Carlson 1997, Hamosh 1994a). Plasma LC-PUFA concentrations in full-term breast-fed infants at 6 and 12 mo of lactation do not depend upon supply from milk (Henderson et al. 1996), suggesting that at this age full-term infants are able to synthesize these fatty acids.

Prenatal LC-PUFA status affects postnatal LC-PUFA independent of postnatal diet (Forman-van Drongelen 1995), suggesting that the importance of placental transfer might extend well beyond the intrauterine period. The LC-PUFA levels in maternal diet and reserves might therefore affect the LC-PUFA level of preterm and full-term infants. Transfer of LC-PUFA to the fetus occurs even under conditions of maternal depletion (Holman et al. 1991).

Because birth levels of LC-PUFA show a strong correlation between arachidonic acid and fetal growth, and between DHA and the length of gestation (Leaf et al. 1992), adequate supply of these fatty acids during pregnancy is essential. Plasma LC-PUFA are similar in mothers of full-term infants from the start of lactation (Spear et al. 1992) to 1 y of breast-feeding (Henderson et al. 1996). The marked decline in milk LC-PUFA after 1-3 mo of exclusive breast-feeding of full-term infants (Bitman et al. 1983, Luukkainen et al. 1994) is therefore not a direct result of a similar decline in serum LC-PUFA (Henderson et al. 1996). Furthermore, there is an increase in plasma LC-PUFA in full-term infants between 6-12 mo of age (Henderson et al. 1996) in spite of a marked fall in the level of these fatty acids in milk (Luukkainen et al. 1994, Henderson et al. 1996). This suggests that LC-PUFA supplementation is not necessary at this age in full-term infants.

One negative effect of supplementation of infant formula with LC-PUFA is their susceptibility to oxidation (Okuda et al. 1994). This might be less of a problem with LC-PUFA in human milk because of the compartmentalization of the fat in highly structured globules where LC-PUFA might be less accessible to oxygen. The high antioxidant activity of human milk (Hamosh 1996) might provide additional protection. The potential negative effects of LC-PUFA supplementation of formulas should be investigated before supplementation becomes routine in the United States.

DIGESTIVE ENZYMES IN HUMAN MILK

There is evidence that this lipase improves fat absorption in newborns (Williamson et al. 1978, Alemi et al. 1981), and in vitro studies show that the enzyme remains active in the infant's gastrointestinal tract (Armand et al. 1996) and therefore could contribute significantly to fat digestion. Stunting in kittens fed formula that does not provide bile salt-dependent lipase has been reported (Wang et al. 1987).

Bile salt-dependent lipase is a constitutive enzyme of the mammary gland; the activity level per milliliter of milk is independent of milk volume and is similar before the onset of lactation, throughout lactation and during weaning (Hamosh 1994b). Furthermore, lipase activity is high in the milk of women who deliver prematurely and is independent of duration of pregnancy. Activity varies among women, but remains constant within each woman (Hamosh 1994b). Human milk BSDL has been cloned, and site-directed mutagenesis has shown that the C-terminal region is not necessary for catalytic activity for certain physiologic functions such as heat stability, stability to low pH and resistance to proteolytic inactivation (Hansson et al. 1994). Its ability to completely hydrolyze triglyceride (because of lack of positional or fatty acid selectivity) and to release of LC-PUFA (which cannot be released by pancreatic lipase because of the proximity of the double bond to the carboxyl end of the fatty acid) make bile salt-dependent lipase highly desirable for neonatal digestion.

In addition to the digestive lipase, human milk contains amylase that could compensate for the very low activity of pancreatic amylase in newborns (Hamosh 1994b). Amylase activity develops very slowly in newborns and amounts to only 0.2-0.5% of adult levels in the first 6 mo after birth (Hamosh 1994b).

Milk amylase is identical to the salivary isozyme, and its activity level is identical in the milk of women who deliver prematurely or at term (Hamosh 1994b). The enzyme has a broad pH optimum of 4.5-7.5 and is relatively stable above pH 3.0; therefore activity could start in the stomach and continue in the intestine. Levels of activity are 10 to 60 times higher in milk than in serum. A combination of in vivo and in vitro studies indicates that amylase activity is maintained for 1-3 h in gastric apirates, that it is unaffected by pepsin and that activity increases in the intestine after milk feeding. Indeed, starch supplements are better tolerated in breast-fed than in formula-fed infants (for recent review see Hamosh 1994b). The latter suggests that milk amylase probably compensates for this deficiency by the hydrolysis of starch supplements or formula oligosaccharides, after supplementation of breast-fed infants or infants who are only partially breast-fed.

In conclusion, we might ask: Should bioactive components of human milk such as the above-described enzymes, LC-PUFA and nucleotides be supplemented to formula-fed infants? Should this supplementation be limited only to preterm infants, who might have the greatest need?

The data available to date suggest that full-term infants might not need to be supplemented. Supplementation with LC-PUFA is probably indicated for preterm infants, although long-term effects are not known and studies to assess possible toxicity have not been performed. Because of conflicting reports on the effects of nucleotide supplementation, additional studies are needed before any recommendation can be made. There are no data on attempts to supplement digestive enzymes. If such supplementation should be tried, one would first have to assess techniques that prevent nutrient hydrolysis before feeding the infant.