Breast milk consists of proteins, carbohydrates – especially lactose -, fat (97% triglycerides), vitamins, minerals, and viable cells, for example, leukocytes and macrophages
[18–20]. The protein fraction comprises multiple bioactive proteins with diverse functions such as lactoferrin, Mac-2–binding protein, secretory IgA (sIgA), but many of these are related to defense functions. Earlier studies pointed to variable DMBT1 expression in normal breast tissue and a strong expression was detected in some cases with inflammation
[13–15]. According to these studies, DMBT1 is expressed in the epithelium of the mammary ducts and glands, i.e. to structures, which are also responsible for the production and secretion of breast milk. We therefore considered it possible that DMBT1 could be secreted to breast milk after delivery. While Danielsson Niemi and co-workers could not detect DMBT1 in breast milk performing Western blotting with four different antibodies (mAb143, mAb303, Hyb213-01, Hyb213-06) against DMBT1
, our data point to the presence of DMBT1 in breast milk as confirmed by using two different methods (ELISA and Western blotting) and two different antibodies (Hyb213-06, anti-DMBT1p84). These findings strongly suggest that DMBT1 is one of the various breast milk components with functions in innate immunity, where it may function as a pattern-recognition molecule for pathogens and would co-localize with various of its known binding partners such as sIgA or lactoferrin.
The DMBT1 concentrations in our study ranged between 0.112 μg/mL and 17.984 μg/mL. The highest DMBT1 concentration was found in the first week after delivery, which then declines to an apparent steady-state level of approximately 2 μg/mL in week 3 and 4 after birth. This resembles the dynamics of other protective proteins such as lactoferrin, Mac-2 binding protein and sIgA in breast milk.
Lactoferrin, an interaction partner of DMBT1, has functions in the defense against bacterial and viral infections and was found in concentrations up to 3.3 mg/mL
[19, 25, 26]. Montagne et al. found the highest lactoferrin concentrations directly after birth (5.8 mg/mL), decreasing in the first consecutive days and then again increasing after day 28
. The Mac-2 binding protein (Mac-2 BP) is - like DMBT1 - a member of the SRCR superfamiliy. Mac-2 BP is expressed in tissues containing cavity-lining secretory epithelia (stomach, gut). It can function as an immune-stimulatory and anti-infective agent. D’Ostilio
 detected Mac-2 BP in human breast milk and showed an increase until days 2-3 postpartum (13.4 μg/mL to 79.2 μg/mL) followed by a decrease of the Mac-2 BP concentration to <10 μg/mL on day 6 postpartum. They additionally performed one measurement 4 weeks after delivery and found concentrations of 5.3 ± 4.8 μg/mL. The mothers and infants in this study were healthy and had no infections. Likewise, the highest levels of secretory IgA (sIgA) were found directly after birth (19.0 mg/mL) and a decrease was reported throughout the lactation period (mature milk: 1.1 mg/mL)
. Similarly, cytokine levels significantly decrease in mature milk in comparison to the first milk samples postpartum, but in patients with preeclampsia the high proinflammatory cytokine levels were persistent throughout lactation
. Thus, the DMBT1 concentrations in our study follow a similar course, but were lower than the detected concentrations of lactoferrin, Mac-2 BP, and sIgA. Assuming an average amount of breast milk of 120 mL per day in the first days after delivery for term neonates, these infants receive 385 μg/day DMBT1 (3.21 μg/mL x 120 mL/day). In extreme premature infants, the amount of breast milk in the first days of life was about 12 x 1 mL to 12 x 3 mL per day (depending on the gestational age of the infants) corresponding to 39 μg and 116 μg DMBT1 per day.
Interestingly, our results revealed that breast milk from mothers with newborns suffering from neonatal infections showed higher DMBT1 concentrations in comparison to the breast milk from mothers with healthy newborns. In contrast, no correlation was found between the DMBT1 concentration in the breast milk and a maternal bacterial infection or a risk factor for chorioamnionitis (premature rupture of the membranes). It is, however, documented that sIgA levels in breast milk increase in response to the maternal environment
[30, 31]. In this case, antigen exposure in the gut or the respiratory tract activates B cell trafficking to the mammary glands, which then results in IgA secretion and translocation of sIgA to the breast milk
[30–32]. While the precise mechanisms behind the up-regulation of DMBT1 levels in breast milk remain to be determined, it resembles the observations made for the Mac-2 BP. Fornarini et al. showed a significant association between the Mac-2 BP concentration in breast milk and an acute respiratory infection of the newborn in the first 12 months of life. The infants were breast feeded for 4-5 months. Newborns with high Mac-2 BP levels in the breast milk of their mothers had a lower rate of respiratory infections indicating a protective effect
. Because we analyzed only neonatal infections in our study group and the breast feeding time considered in our study is substantially shorter, it is not possible with the present data sets to compare the effects of DMBT1 on the infections rate as was performed in the earlier study of Mac2-BP.
Because of the blood volume using for blood culture (0.5-1 mL) only one third of the infants with neonatal infections showed positive blood cultures. To compensate for possibly low sensitivity we used additional criteria for the diagnosis infection (C-reactive protein >10 mg/L, clinical signs of infection as fever, and/or detection of bacteria in smears e. g. omphalitis), but we cannot rule out that a very strong inflammation was interpreted as neonatal infection.
According to the present data, the use of the DMBT1 concentration in the breast milk as a marker for neonatal infection is limited. The DMBT1 concentration in the breast milk of newborns with neonatal infection has a range between 1.838 and 17.984 μg/mL. This large range and the crossover of this range with the range of the DMBT1 concentration of healthy neonates would result in a high rate of false positive predictions. One reason could be basal differences of the DMBT1 expression levels in the breast tissues of individual women
[13–15], which may depend on variable promotor activity
. Thus, the DMBT1 concentration of the breast milk can potentially contribute to diagnose a neonatal infection, but it is not able to replace established methods such as measurement of C-reactive protein levels to determine a neonatal infection.