Results

Blocking primers development

We discovered three unique regions in M. festivus rRNA 18S V4-V5 gene consensus sequence following its comparison with seven potential fish parasites [Fig. 1]. One region was located 5’ of the consensus sequence and had a length of 24 pb, a second one was located at the middle of the sequence and had a length of 40 pb, and the third was located 3’and had a length of 28 pb. From these regions, two combinations of blocking primers were developed, the F-primers[635F-C3 and 1062R-C3] and the M-primers [816F-C3 and 846R-C3], respectively located at the extremities and at the middle of the target gene.

Blocking primer evaluation

The relative abundance of host DNA in samples amplified using theF-Primers was not significantly different from the control group [Fig. 2]. Indeed, none of the three concentrations ofF-primers significantly reduced the number of host amplicons in samples according to the paired samples Wilcoxon test (p-values > 0.05). Also, using an increasing concentration ofF-primers , from 2X to 10X, did not improve its effectiveness and led to inconsistent results in samples with a low host DNA to gut content ratio [Table 1].
According to the paired sample Wilcoxon test, the M-primers at a concentration of 2X allowed a significant reduction of the mean relative abundance of host sequences present in samples when compared to the control group (p-value = 0.004). The mean relative abundance of the class Actinopterygii, the taxonomic class of M. festivus , in theM-primers group represented 34.01 % of the abundance of the control group [Fig. 2]. Also, the M-Primers efficacy at inhibiting host’s DNA amplification was significative in samples starting with a low and a high host DNA to gut content ratio [Table 1]. For example, the mean relative abundance of the class Actinopterygii in gut IDs 1 to 4 of the M-primers group represents 14.58 % of the relative abundance of the control group. Furthermore, host DNA relative abundance in gut IDs 7 and 8 respectively reduced from 72 and 77.9 % to 24 and 21.5 %. However, using a concentration of more than 2X of theM-primers completely inhibited the PCR amplification [Table 1].

Alpha diversity

The Faith phylogenetic alpha diversity index was significantly lower in the M-Primers group than in the control group in the Tukey HSD test (p-value = 0.02) [Fig. S2]. However, the total number of ASVs (observed diversity), the Shannon and the Simpson alpha diversity indexes were not significantly lower when comparing the same groups (p-value > 0.05). Furthermore, when omitting the class Actinopterygii, there was no significant difference in alpha diversity between the M-primers group and the control group in the Tukey HSD test (p-values > 0.05) [Fig. S3]. For theF-Primers , their effect was inconsistent at different concentrations and did not significantly influence the alpha diversity in samples. Overall, we observed a mean of 30 ASVs per sample, with 20 % of these from the Phylum Vertebrata. Also, two genera of plants,Cyperus sp. and Oryza sp. , were the most abundant ASVs for the samples analysed. These two genera, from the order Poales, averaged 48 % of the total relative abundance in samples.

Beta diversity

According to the two-way PERMANOVA comparing Bray-Curtis dissimilarity indexes and considering the eight gut IDs and the four blocking primer combinations as factors, most of the variance was explained by differences between gut IDs (R2 = 0.85; p-value = 3.0 x 10-5) and a moderate part of the variance was explained by differences between blocking primer groups (R2 = 0.07; p-value = 1.0 x 10-4). Furthermore, the M-primers was by far the combination of blocking primers with the most influence on the sample’s beta diversity in the Pairwise Adonis test (For M-primers , R2 = 0.09; p-value = 0.007).
In a PCoA based on Bray-Curtis’s dissimilarity indexes, there was a significant shift in beta diversity in the four samples which contained a high ratio of host DNA to gut content (Gut IDs: 5 to 8), but no important variation in beta diversity for the four gut samples with a low abundance of host DNA contamination (Gut IDs: 1 to 4) [Fig. 3]. This result was very similar when producing a PCoA based on weighted Unifrac distances [Fig. S4]. For samples 5 to 8, the usage ofM-primers led to a shift in beta diversity, reducing the dissimilarity between gut IDs 5 to 8 and 1 to 4 [Fig. 3]. Consequently, when computing a pairwise Adonis test using a dataset omitting the class Actinopterygii, M-Primers did not have a significant effect on the beta diversity of samples when compared to the control group (R2 = 0.03; p-value = 0.07). Again, theF-Primers had little influence on the beta-diversity in samples and showed inconsistent patterns that did not support their effectiveness at blocking host DNA amplification.

Blocking primers for parasite detection

The usage of M. primers enhanced the detection of low abundance parasitic taxonomic classes in gut IDs 5 to 8 [Fig. 4]. For instance, we detected 12 different taxonomic classes in theM-primers group against 9 in the group without blocking primers in the top 10% ASVs barplots [Fig. 4]. From these, M-Primersenhanced the detectability of Trematode in gut ID 6 and Arcellinida in gut ID 7. Also, these classes were detected in higher relative abundance in the M-Primers group. Also, a parasitic Ciliophora from the genus Nyctotherus sp. was detected in high relative abundance in two gut samples (Gut IDs 2 and 7) and in lower abundance in one (Gut ID 4) [Fig. 4]. Finally, we detected the presence of a Tubilinea in gut ID 7 and a very low relative abundance of microsporidia in gut ID 4.