The primary protection of the egg after being laid relies firstly on a physical defence (the eggshell and the eggshell membranes) and secondly on chemical defences mainly present in the egg white, but also in other compartments. IgY, IgM and IgA  participate with numerous major proteins  and newly identified minor proteins and peptides  in the innate defences of the egg. While IgY concentration have been shown to vary in egg yolk depending of the nature and degree of antigen exposure of hen , no evidence in the literature explored whether the antimicrobial peptides/proteins of the egg are modulated by the microbial environment of the hen. The present study demonstrates the regulation of some egg white antimicrobial activities by microbial environment of the hen.
This investigation used an experimental design based on the comparison of three extreme conditions of rearing laying hens: germ-free (GF), specific pathogen-free (SPF) and conventional (C) conditions. GF hens are characterized by the absence of microbiota at the intestinal level. This influences their metabolism and intestinal morphological parameters . SPF hens are raised in strictly hygienic conditions and are not vaccinated. Due to the absence of any interactions with other pathogenic microorganisms, the SPF model is frequently used to explore immunological responses to pathogenic or vaccine antigens [21, 22]. In contrast, C laying hens are bred under commercial conditions and might occasionally be exposed to pathogens. These contrasting breeding conditions provide extremely wide qualitative and quantitative variations in terms of bacterial populations in contact with the hens: the absence or presence of surrounding microbes and gut microbiota, for the GF or C groups respectively, and an intermediate group, the SPF hens, hosting a controlled microbiota in a pathogen-free environment. The maintenance of GF hens until they are sexually mature (4–5 months) and beyond requires efficient isolators, sterilized food and water, and qualified animal handlers. These constraints could partly explain why such an animal model has never been used before. In our attempt, the non-contamination of GF hens was not successfully achieved. An agent, Penicilium, was detected at month four, in two independent isolators, but more importantly, in spite of this fungal contamination, the hens remained free of bacteria relevant to our initial objective.
The GF group definitively showed different immunological statuses compared to the C and SPF groups, as reflected by higher expressions of IL-1β, IL-8 and TLR4 genes in the jejunum and cæcum of these groups, compared to the GF group. IL-1β and IL-8 are two pro-inflammatory cytokines which are often used as markers of inflammation . TLR4 is a host cell membrane receptor that detects lipopolysaccharide from Gram-negative bacteria and elicits innate immune response following bacterial infection. The difference in expression levels of IL-1β, IL-8 among the three groups was larger in the cæcum (2- to 64-fold) than in the jejunum (2- to 4-fold) in the SPF and C groups as compared to the GF group. Such expected differences are probably due to the bacterial load, which is much higher in the cæcum than in the jejunum . In contrast, no differences in IL-1β, IL-8 and TLR4 gene expression were observed in the oviduct (magnum) between the experimental groups. Under normal non-pathogenic conditions, the magnum and the other segments of the hen oviduct (infundibulum, isthmus and uterus) constitute an aseptic environment in which the egg is formed in a 24 hour period . These aseptic conditions are probably responsible for the absence of pro-inflammatory gene induction. Altogether, these observations suggest that the presence or absence of microflora and associated stimuli, at the intestinal or oviduct levels respectively, directly influences the local inflammatory state and the tissue expression of IL-1β, IL-8 and TLR4 genes.
The egg white is the largest compartment of the egg in terms of variety and concentration of antimicrobial proteins. Among the major egg white antimicrobial proteins are ovotransferrin and lysozyme, which are active against Gram-negative and Gram-positive bacteria [4, 25]. Apart from these major egg white compounds, a number of minor molecules with potent antimicrobial activities have recently been identified and further characterized. Of these, we characterized the antibacterial activities of two peptides of the beta-defensin family, namely gallin and the avian beta-defensin [26, 27]. While gallin is active against E. coli, AvBD11 possesses a broad spectrum of antibacterial activities against both Gram-positive and Gram-negative bacteria, The ability of the hen to modulate these compounds in response to microbial environments has not been explored. Egg whites of the C and SPF groups had greater inhibitory activities on the growth of S. aureus and S. uberis (Figure 2A, B, P < 0.01) than those of the GF hens. In contrast, anti-Salmonella (S. Enteritidis and S. Gallinarum), anti-E. coli and anti-L. monocytogenes activities were similar in the egg whites of all three experimental groups. Our results demonstrated that the breeding conditions of hens have an impact on some of the antibacterial properties of their eggs, according to the degree of bacterial contamination of their environment. However, the response seemed specific to certain bacterial strains, suggesting that it might result from change in some antimicrobial egg molecules with a particular spectrum of activity, predominantly toward Gram-positive bacteria in our study. In order to give some insight into the putative mechanisms at the origin of the increased egg white antibacterial activity against S. aureus and S. uberis observed in SPF and C groups, we further analysed the level and/or activity of a panel of proteins representative of the main modes of action of egg antimicrobials (chelating, antiprotease and lytic effects). That was carried out by quantifying egg white activities or magnum gene expression of proteins representative of this diversity of antibacterial actions.
The main bacteriolytic molecule of the egg white is the lysozyme. This well-studied cationic protein is an enzyme catalysing the cleavage of peptidoglycan, a major compound of Gram positive bacterial cell walls. No variation between GF, SPF and C was observed for the lysozyme-mediated lytic activity of egg whites. This is in agreement with the lysozyme amounts, which appeared to be constant within all experimental groups, as assessed by western blot (data not shown) and with previous findings showing the genetic stability of lysozyme concentration in the egg white . In the opposite, hen age and acute administration of different immunostimulatory substances to hens modulate its activity [9, 10]. However, our results were coherent with unmodified anti-L. monocytogenes activity. Egg white exerts a potent bactericidal activity against L. monocytogenes and the main egg component possessing anti-Listeria activities is the lysozyme. In contrast, L. monocytogenes, S. aureus and S. uberis seemed to be less sensitive to the egg white antimicrobial activities and grew in less diluted egg white. A number of S. aureus strains are known to develop resistance to lysozyme, whereas the activity of egg white lysozyme on S. uberis strains requires further study. The fact that no variation between GF, SPF and C was observed for the lysozyme-mediated lytic activity of egg whites supports the hypothesis that enhanced anti-S. aureus and anti-S. uberis activities in SPF and C egg white are not related to lysozyme, but most probably to additional compound(s). Egg white contains numerous bactericidal molecules including the avian defensins. These cationic peptides can disrupt the bacterial membrane, resulting in the cell lysis [7, 28]. Thus, gallin and avian beta-defensins (AvBDs) 10, 11 and 12 which have been detected in the egg white by proteomic analysis  and/or in the magnum at transcriptional level  are alternative candidates to explain a change in antimicrobial activities. The quantification of these peptides was not possible because neither specific antibodies nor quantitative ELISA kits are available. Variation at the transcriptional level was therefore analysed by RT-qPCR in the magnum as a potential marker for relative protein synthesis between experimental groups. Previous studies showed that hens intravenously injected with lipopolysaccharide showed a transitory increased expression of AvBD10, AvBD11 and AvBD12 in the vagina [30, 31]. In our steady-state experimental conditions, even if C and SPF hens were more challenged immunologically than GF hens, their magnum showed no stimulation of AvBD10, AvBD11, AvBD12 and gallin expression, suggesting that these molecules are not responsible for the increased antimicrobial activity observed in the egg white. Therefore, the higher anti-S. aureus and anti-S. uberis activities in the egg white of C hens did not appear to rely on AvBD10, AvBD11, AvBD12 and gallin.
Egg white contains large amounts of chelating molecules with antimicrobial activities, the most representative being ovotransferrin and avidin. Ovotransferrin was quantified both at the protein (western blot, data not shown) and transcriptional levels, while avidin was assessed only at the transcriptional level. No modifications in any of the three hen groups were revealed for these molecules. It is believed that the most efficient antimicrobial molecule against Gram-negative bacteria E. coli and S. Enteritidis in the egg is ovotransferrin via its iron depriving mechanism [25, 32]. S. Enteritidis is of major concern in public health as it is considered as the first foodborne disease agent in eggs and egg products . This bacterium is capable of invading the intact egg when laid and, via different mechanisms, of withstanding the antibacterial molecules as well as the harsh pH conditions in the egg white during its storage . The absence of variation in S. Enteritidis growth in any of the three conditions was consistent with our observations showing that ovotransferrin was not modified, either at protein or transcriptional levels.
Egg white antiproteases might play a role in egg innate immunity by exhibiting antimicrobial activities. Cystatin is a potent antimicrobial, active against a variety of bacteria including Escherichia coli and S. aureus. Two other egg antiproteases, ovomucoid and ovoinhibitor, are known to inhibit bacterial peptidases [35, 36] in spite of limited data regarding their antimicrobial properties. In particular, their effect on S. aureus is yet unknown. Likewise, there is no data in the literature demonstrating anti-S. uberis properties for ovomucoid, ovoinhibitor and cystatin. In our study, the analysis of egg white antiprotease activities and magnum gene expression of these molecules was of interest as staphylococci and streptococci are bacteria known to secrete extracellular peptidases that presumably play some role in virulence. In particular, S. aureus produces and releases to the extracellular milieu several enzymes belonging to distinct classes of proteases, such as serine- (Protease V8 or SspA), cysteine- (Staphopains A and B, also known as ScpA and SspB) and metallo- (Aureolysin Aur) proteases . S. uberis produces extracellular proteases that are involved in the regulation of biofilm formation . Our results showed that global anti-trypsin, anti-chymotrypsin and anti-papain-like protease activities were not influenced by the microbial environment of hens. Moreover, gene expression analyses of ovoinhibitor, cystatin and ovomucoid in the magnum did not show any differences among the three experimental groups. These observations suggest that increased egg white activities against S. aureus and S. uberis do not rely on these egg antiproteases.
The egg white pH affects global egg white antimicrobial activity. High pH values are bactericidal for S. aureus and are correlated with anti-S. Enteritidis activity . Egg white pH was slightly higher in C (+0.19) and SPF (+0.13) groups as compared to GF (pH = 8.41). However, for this magnitude of changes, there was no correlation between pH and anti-S. aureus or anti-S. uberis activities (correlation coefficients were respectively −0.16 and −0.50; p > 0.1) so this parameter is unlikely to explain the bacterial growth inhibition.
Our observation that only two out of the six bacteria studied responded to the treatment, suggests that the effect results from some specific egg molecules. However, our attempt to identify the molecular origin of the change in the egg white antimicrobial activities observed for S. aureus and S. uberis was not fruitful. It strongly suggests that additional egg components, not investigated in the present study, are involved in this regulation. The sequencing of the hen’s genome and the development of proteomic [29, 41, 42] and transcriptomic  approaches reveal hundreds of minor peptides and proteins expressing a large range of biological functions including protection against diverse pathogens (bacteria, viruses, fungi)  in the different egg compartments. An alternative explanation for the difficulty in identifying the minor egg molecules responsible for the increased antibacterial effect towards S. aureus and S. uberis is that we explored the gene expression of candidate proteins, and not the egg protein or peptide levels or activities in the eggs. However, by using such extreme experimental situations (GF, SPF, C), this strategy should be valid and this was confirmed by the dramatic changes observed for interleukins at the intestinal level. It is obvious, however, that numerous alternative candidates amongst the newly identified molecules may be at the origin of the observed changes, including histone-like proteins or lipolysaccharide-binding proteins .