Page 84 - Read Online

Page 10 of 19 van Beek et al. Microbiome Res Rep 2025;4:13 https://dx.doi.org/10.20517/mrr.2024.45

IAP was positively correlated with Bacteroides at 6 months [Figure 2A], and Actinomyces, Streptococcus, and
genera of the Enterobacteriaceae family at 12 months [Figure 2B]. At 12 months, IAP had negative
associations with members of Clostridia, including Roseburia and Lachnospira [Figure 2B]. BPI was
positively associated with members of Enterobacteriaceae at both time points and negatively with
Lactobacillus at 6 months and members of Clostridia at 12 months [Figure 2].

Indicators of inflammation
Cal had only negative (Prevotella and Akkermansia) associations with bacteria at 6 months [Figure 2A], but
mostly positive ones at 12 months, mainly with Prevotella and Pseudoruminococcus [Figure 2B]. LTF had
negative associations with members of Clostridia at both time points [Figure 2A and B], and with members
of Enterobacteriaceae at 6 months [Figure 2A].

Apart from a strong negative association between ECP and Collinsella, ECP appeared mostly stimulated by

bacteria at 6 months, especially by Streptococcus and Clostridioides [Figure 2A], while it had mostly negative
associations with increasing bacterial populations at 12 months, including Blautia, Lachnospira, and
Haemophilus [Figure 2B].

Strong associations were observed for LCN2. At 6 months, increasing abundances of Streptococcaceae and
Veillonellaceae were associated with increasing LCN2 levels, while members of the class Clostridia and
Roseburia correlated negatively with LCN2 [Figure 2A]. However, these associations were not observable at
12 months [Figure 2B]. At 12 months, LCN2 increase was associated with declining populations of Bacilli
and Enterobacteriaceae [Figure 2B].

Levels of HD5 were negatively associated with several bacterial taxa, including Dialister at both time points,
Bifidobacteriaceae and Ruminococcaceae at 6 months [Figure 2A], and Prevotellaceae at 12 months
[Figure 2B]. HD5 was positively associated with Megasphaera at 6 months [Figure 2A], and with
Actinomyces at 12 months [Figure 2B].

Lysozyme levels were strongly negatively associated with Clostrida members at 6 months [Figure 2A]. Yet,
at 12 months, lysozyme levels were mostly positively correlated with other Clostridia members,
Megasphaera, Pseudoruminococcus, Parabacteroides, and Dorea but negatively with Collinsella and
Bifidobacterium [Figure 2B].

Zonulin was not associated with any bacteria genera at 6 months, but at 12 months, was positively
correlated with several, including Megasphaera, Eggerthella, and several Clostridia members [Figure 2A and
B].

Associations between biomarker concentration and bacterial population growth
We hypothesized that associations between biomarker concentration on day t and the change in bacterial
population size from day t to t + 1 (population growth) would represent bacterial responses to the
biomarkers [Figure 3].

Indicators of gut homeostasis
The association between muc2 and bacterial growth changed from mostly positive at 6 months [Figure 3A]
to mostly negative at 12 months [Figure 3B]. The only exception at 12 months was Bifidobacterium, which
appeared to be stimulated by muc2 [Figure 3B]. At 6 months, higher albumin levels were predictive of
increasing Haemophilus and Lacticaseibacillus [Figure 3A], but no associations were observed at 12 months.