FIGURE 5.

Ordination plots comparing (I) resistome composition and (II–IV) mobilome composition using non-metric multidimensional scaling (NMDS) for (A) human and (B) cattle metagenomic datasets. All ordinations were based on Euclidean distances derived from Hellinger-transformed normalized counts of positive alignments. Results of resistome and mobilome ordinations are reported at the class level of ontology, though these results remained consistent when analyzed at the mechanism level (data not shown). For the human dataset (A) where patients were parenterally administered a cocktail of antimicrobial drugs for 4 days, the resistome and all studied components of the mobilome (i.e., ICE, plasmids, and prophages) demonstrated a concomitant shift in composition after peak antimicrobial administration (Day 4) which persisted 4 days after peak exposure (Day 8) (ANOSIM P < 0.001) and reverted closer to the original composition seen at baseline by Days 42 and 180. In the U.S. beef cattle dataset (B) where cohorts of cattle were either given metaphylactic tulathromycin upon arrival or remained untreated, there was no detectable difference in resistome composition between treatment and control groups at baseline and after 11 days of monitoring (ANOSIM P > 0.05). On the other hand, the ICE (II) and plasmid (III) components of the mobilome differed between treated and untreated animals (ANOSIM P < 0.05). However, all cattle demonstrated a significant alteration in resistome composition over time, regardless of treatment status (ANOSIM P < 0.01). While there was a shift in the plasmidome among all groups over time (ANOSIM P < 0.05), the remaining components of the mobilome remained unchanged.