Page 15 - Read Online

Page 8 of 15 Lai et al. Microbiome Res Rep 2024;3:21 https://dx.doi.org/10.20517/mrr.2023.76

MR. In addition, MF-C was classified into a single cluster as compared to the other suppliers produced by
the same MF substrate. Deoxyguanosine and N-gamma-L-glutamyl-D-alanine showed a high abundance,
whereas L-proline, erinapyrone B, p-anisic acid, and (Z)-3-methyl-2-(2-pentenyl)-2-cyclopenten-1-one
were relatively low in MF-C. Therefore, it was worth noting that the microbial activity and flavor profile
were different because of discrepancies in suppliers.

According to the cluster analysis, MF-A and MLS-A were classified into an identical group, both collected
from the same supplier with discrepancies in substrates. Therefore, to determine the markers for
differentiating the differences between the metabolites of the two samples, the Euclidean algorithm based on
OPLS-DA was applied to mark the differential metabolites (P < 0.05) whose abundance matched top thirty.
As shown in Figure 2A, the differential metabolites were determined by calculating VIP (variable important
in projection) value. Taxifolin, dactilin, quercitrin, lampranthin II, and some other glucoside/rhamnoside/
rutinoside compounds were highlighted in MLS-A, which were annotated to the flavonoid glycoside
subclass. The native Lactiplantibacillus plantarum fermentation could enrich its antioxidant activity through
the accumulation of flavonoids . The regulation of dietary flavonoid intake can improve health status by
[33]
[34]
reconstructing gut microbiota . The majority of gut microbiota are associated with flavonoid
transformation. For example, Lactococcus is involved in the C-deglycosylation of flavonoids . Most
[35]
flavonoids contribute to the color change of yellow food [36,37] . Thus, it was quite reasonable to presume that
the flavonoid was related to the golden yellow Suancai color. In addition, the differential metabolites in
MF-A compared to MLS-A were involved in most pathways, among which aesculetin was found in the class
[38]
of coumarins and derivatives, which exhibit anti-inflammatory activity . As shown in Figure 2B, the
differential metabolites with VIP > 1 were used to reveal the discrepancies between the different suppliers
(MF-A and MF-E). Luteone shows highest antifungal activity against food spoilers and retains its activity at
low acidic pH . Thus, more luteone in MF-A might inhibit unwanted fungal growth and help in normal
[39]
fermentation process of Suancai. Taken as a whole, the metabolites of non-salt Suancai varied with
substrates of ingredients and their suppliers.

Metagenomic analysis
Microbial composition
A total of 40,780,878 to 51,208,942 raw reads were generated from metagenomic sequencing analysis. After
quality processing and elimination of sample host gene, the effective percentage of clean reads was 99.19%
to 99.46%. Sequences were assembled using Megahit, resulting in contigs that ranged from 13,850 to 82,205.
All contigs were then subjected to ORF prediction and 25,981 to 110,566 ORFs were obtained. Alpha
diversity was an effective index to characterize within-habitat diversity, which includes Chao1 and Simpson
index. As shown in Supplementary Table 4, the microbial richness (Chao 1) was highlighted in MF-A, while
the highest microbial diversity (Simpson) was in MR-A when compared between different substrates of
ingredients. In addition, among different suppliers, MF-B and MF-C showed highest richness and diversity,
respectively. Lactobacillaceae was the predominant family in all samples (58.78%~96.17%). Through species
annotation [Supplementary Figure 5], 566 species were co-detected in all samples, accounting for
66.04%~94.08%, which indicated that the basic species composition was roughly the same. However, the
36~291 differential species related to the discrepancies of samples.

The microbial composition at the genus level is reported in Figure 3A. Lactiplantibacillus, Leuconostoc, and
Lactococcus were the dominant microbes across all the samples but with differences in abundance. The
microbial compositions differed significantly depending on the substrates/suppliers of ingredients. The
relative abundance of Lactiplantibacillus was higher in MR-A, which was derived from food and showed
good resistance and adhesion in the gastrointestinal tract, and exhibited antioxidant and antimicrobial
properties . Among the suppliers, Lactiplantibacillus was predominant in MF-A, MF-D, and MF-E
[40]