Page 85 - Read Online
P. 85

van Beek et al. Microbiome Res Rep 2025;4:13  https://dx.doi.org/10.20517/mrr.2024.45  Page 11 of 19





























                Figure 3. Biomarker levels predicting gut microbe changes. Associations between faecal biomarker concentrations and daily changes in
                gut microbe abundances at the family and genus levels at (A) 6 months and (B) 12 months of age. The colour represents the strength of
                                                                                                         **
                                                                                                   ***
                association from a linear mixed model adjusting for total protein concentration. P-values are indicated as asterisks: P < 0.001 ; P < 0.1 ;
                    *
                P < 0.5 . The microbial class is presented as the sidebar colour.
               IgA at 6 months was predictive of increasing Veillonella and Citrobacter [Figure 3A], while at 12 months,
               Lacticaseibacillus declined in association with high IgA levels [Figure 3B].

               High IAP levels were generally associated with bacterial declines, especially among Clostridia. BPI was
               negatively associated with the growth of Bilophila, Lachnospira, and members of Negativicutes at 12 months
               [Figure 3B], but not at 6 months.

               Indicators of inflammation
               Cal was mostly negatively associated at 6 months, notably with Akkermansia, Klebsiella, and Prevotella, and
               was only positively correlated with Erysipelatoclostridium [Figure 3A]. Cal was positively correlated with
               Lactobacillus and Pseudoruminococcus at 12 months [Figure 3B]. LTF and ECP concentrations had no
               strong associations with microbial changes, apart from a notable strong negative association between ECP
               and Akkermansia at 12 months. High levels of LCN2 were associated with increasing abundances of many
               Clostridia genera at 6 months [Figure 3A], but these were mostly non-significant at 12 months. High levels
               of HD5 were negatively associated with Akkermansia (significant only at 6 months) and Haemophilus (12
               months), and positively with Lactobacillus, Enterococcus, and Enterocloster (12 months) [Figure 3A and B].
               High lysozyme levels predicted a decline in Anaerostipes at 6 months. At 12 months, it correlated with
               increasing levels of Pseudoruminococcus and decreasing levels of Prevotella and Senegalimassilia. A high
               zonulin level was associated with increasing abundance of Blautia at 6 and 12 months[Figure 3A and B]. At
               6 months, it was also associated positively with Bilophila and Flavonifractor [Figure 3A], and at 12 months
               with Clostridioides [Figure 3B].


               DISCUSSION
               Utilising densely sampled absolute abundance time series of infant gut microbiota and immune-related
               biomarkers, we were able to identify potential host-microbe interactions occurring in vivo in healthy
               infants. Although microbial programming of the immune system is considered important during early life,
   80   81   82   83   84   85   86   87   88   89   90