Fig. 4: Phage–antibiotic pressures drive divergent Pseudomonas subpopulation evolution, coupling resistance adaptation with constrained fitness.

a Phylogenetic tree of 22 nonmucoid accessory genomes. Isolates were outgrouped by ancestral isolate –538a and grouped into four clades. b Circular plots of variants mapped to the –538a reference genome, ordered chronologically and colored by clade in (a). Red bars indicate nonsynonymous mutations (single-nucleotide polymorphisms [SNP]. insertions [INS], and deletions [DEL] in lipopolysaccharide (LPS) biosynthesis, envelope, and metabolic genes. The five outermost rings denote P1A-resistant isolates. c Boxplots comparing planktonic growth of Clade 4 nonmucoids in phage-free media. Boxes represent interquartile ranges of area under the growth curve (AUC), lines mark medians, whiskers denote minima and maxima, and individual replicates (n = 4) are shown as points. One-way ANOVA with Tukey multiple comparisons of AUC were performed. ****denotes adjusted p < 0.0001. d Phylogenetic tree of nine mucoid accessory genomes, outgrouped by ancestral strain –440b and separated into three clades. e Circular plots of variants mapped to the –440b genome, arranged chronologically and color-coded by clade from (d). Mutations predominantly occurred in efflux transporters, outer membrane, and alginate synthesis genes. f Schematic summarizing lineage trajectories across relative resistance and fitness space. Each bubble represents a clade, scaled by isolate density and colored as in panels (a, d). Phage-driven evolution of nonmucoids (Clades 2–4) increased resistance at a cost to fitness, whereas mucoid lineages remained stable within the low-risk, antibiotic-susceptible range. Together, these pressures shifted the population toward a lower-risk, chronic configuration. For both trees, scale bar = 0.1 nucleotide substitutions per site; bootstrap values > 0.2 are indicated at nodes.