To our knowledge, this is the first study describing the effect of a probiotic mixture, orally consumed during the last trimester of pregnancy, on the vaginal microbiota and immune response. Although several health-promoting activities of probiotics have been described in relation to the gut homeostasis [16, 32], less information is available regarding the interactions between orally administered probiotic bacteria and the vaginal microbial habitat.
The first step in ascertaining the influence of the dietary supplementation with the probiotic VSL#3 on the vaginal microbiota of pregnant women was the characterization of vaginal bacterial communities by using an integrated approach based on PCR-DGGE and qPCR.
DGGE population profiling, conducted with universal primers for bacteria and Lactobacillus-specific primers, allowed us to investigate the variations of the predominant vaginal bacterial communities and Lactobacillus species occurring both physiologically in the last trimester of pregnancy and potentially associated with VSL#3 intake. The influence of the probiotic intake in modulating the predominant bacterial populations and Lactobacillus species could be hypothesized since significant differences between DGGE profiles at W33 and W37 were found only in women belonging to P group. Notably, the lower percentage of women belonging to P group who displayed significant differences in Lactobacillus-specific DGGE profiles between W33 and W37, compared to the universal bacterial DGGE patterns, suggested a major stability of lactobacilli population and a more extended impact of the probiotic VSL#3 on total bacteria than lactobacilli. However, no significant changes in single species abundances were found between W33- and W37-related universal DGGE profiles. Differently, the statistical analysis of the peak heights of the Lactobacillus-specific DGGE densitometric curves allowed us to identify a band, corresponding to L. helveticus, which significantly decreased after probiotic supplementation. Strains belonging to L. helveticus are used as starter cultures in the manufacturing of a variety of fermented dairy products, to modulate flavor. The presence of L. helveticus in vagina, likely due to the migration from the gut, can be related to a diet rich in yogurt and cheese. This work is not the first describing L. helveticus in vaginal samples. Stoyancheva et al.  identified this species among several Lactobacillus isolates from vaginal fluids of healthy Bulgarian women in childbearing age by using three different molecular techniques, amplified ribosomal DNA restriction analysis, ribotyping and PCR with species-specific primers. The decrease of L. helveticus observed in our study could be due to a competition between the Lactobacillus strains present in VSL#3 formula and dairy L. helveticus strains in colonizing vaginal environment.
Cluster analysis showed that universal and Lactobacillus-specific DGGE profiles related to the time points W33 and W37 of the control women were closely related. Also the DGGE patterns of the majority of women administered with VSL#3 grouped according to the subject and not to the time point, revealing that the inter-individual variability was higher than variability induced by the probiotic supplementation.
The hypothesis of a positive action of VSL#3 on the vaginal microbiota of pregnant women was further supported by qPCR results, which suggested a role of the probiotic product in counteracting the decrease of the health-promoting Bifidobacterium genus and the increase of the BV-related Atopobium genus, that occurred in control women during late pregnancy. Notably, group B Streptococcus, which was found in two women (N.1 and 10) before the probiotic intake, was no longer found after the dietary supplementation (data not shown).
The second step of the present research was the investigation of the vaginal immunological profiles of the pregnant women in order to search for correlations between the VSL#3 intake and changes in vaginal immune response. Pregnancy has been referred to as a state of relative immune compromise. This notion has been related to both demonstration of depression of certain aspects of cell-mediated immunity and clinical observations of an increased severity of numerous infectious conditions in pregnant women . On the other hand, preterm cervical ripening can be likened to an inflammatory process with cytokines as important mediators .
Bioplex immunoassay was used in the present work to measure levels of 27 cytokines, chemokines and growth factors in the vaginal samples of the pregnant women belonging to P and C groups. In group C a significant reduction at W37 was found for IL-4, IL-7, IL-9, IL-10 and RANTES. IL-4 is a key regulator in humoral and adaptive immunity. It has many biological roles, including the stimulation of activated B-cells and T-cell proliferation, and the differentiation of CD4+ T-cells into Th2 cells. A regulatory role is also exerted by IL-10. In relation to pregnancy, IL-10 decreases the production of pro-inflammatory cytokines, such as IL-8, IL-6, TNFα, IL-1β and prostaglandin E2 in lipopolysaccharide-stimulated fetal membranes [35, 36]. Both IL-4 and IL-10 are produced by Th2 cells. IL-7 and IL-9 are hematopoietic growth factors that act as regulators of cell survival, proliferation and homeostasis of a variety of hematopoietic cells. RANTES is a potent and versatile chemokine, capable of attracting monocytes, lymphocytes, basophils and eosinophils. This cytokine has been implicated in the regulation of the inflammatory response and recruitment of macrophages to the implantation site in early pregnancy . However, no variations in RANTES levels have been associated with preterm cervical ripening and labor . Immunological profiles related to women belonging to C group indicated that some fluctuations in vaginal immune-modulators occurred physiologically during the last trimester of pregnancy. In particular, it is noteworthy the decrease of IL-10 and IL-4, important regulatory cytokines controlling the inflammatory reaction responsible for uterine contractions and cervical ripening at the labor time . In P group a significant variation was registered only for the chemokine Eotaxin, which decreased after probiotic supplementation. Eotaxin selectively recruits eosinophils, and for this reason is implicated in allergic responses . By comparing the data related to the two study groups, the following hypotheses could be formulated regarding the possible impact of the probiotic intake on cytokine secretion during late pregnancy: (i) probiotics counteracted the decrease of anti-inflammatory cytokine levels occurring in C group; (ii) probiotics induced the decrease of a pro-inflammatory cytokine in P group, showing a global anti-inflammatory effect on the vaginal immunity. In addition, a stabilization effect on the vaginal immunity during late pregnancy could be attributed to the probiotic intake, since the percentage of women with modified amounts of immune-mediators decreased from 67% to 31% in relation to the dietary supplementation.