The Pseudoalteromonas multipartite genome: distribution and expression of pangene categories, and a hypothesis for the origin and evolution of the chromid

Abstract

Bacterial genomes typically consist of one large chromosome, but can also include secondary replicons. These so-called multipartite genomes are scattered on the bacterial tree of life with the majority of cases belonging to Proteobacteria. Within the class gamma-proteobacteria, multipartite genomes are restricted to the two families Vibrionaceae and Pseudoalteromonadaceae. Whereas the genome of vibrios is well studied, information on the Pseudoalteromonadaceae genome is much scarcer. We have studied Pseudoalteromonadaceae with respect to the origin of the chromid, how pangene categories are distributed, how genes are expressed relative to their genomic location, and identified chromid hallmark genes. We calculated the Pseudoalteromonadaceae pangenome based on 25 complete genomes and found that core/softcore are significantly overrepresented in late replicating sectors of the chromid, regardless of how the chromid is replicated. On the chromosome, core/softcore and shell/cloud genes are only weakly overrepresented at the chromosomal replication origin and termination sequences, respectively. Gene expression is trending downwards with increasing distance from the chromosomal oriC, whereas the chromidal expression pattern is more complex. Moreover, we identified 78 chromid hallmark genes, and BLASTp searches suggest that the majority of them were acquired from the ancestral gene pool of Alteromonadales. Finally, our data strongly suggest that the chromid originates from a plasmid that was acquired in a relatively recent event. In summary, this study extends our knowledge on multipartite genomes, and helps us understand how and why secondary replicons are acquired, why they are maintained, and how they are shaped by evolution.

Keywords: Pseudoaltermonas; Alteromonadales; chromid; multipartite; pangenome; secondary replicons.

Figure 1

Summary of an ML-phylogenetic tree showing evolutionary relationships between Alteromonadales families. See Supplementary Figure S1 for complete tree. Multipartite genomes are restricted to Pseudoalteromonadaceae, which is placed at the crown of the tree and branches off from families containing monopartite genomes. The arrow highlights the LCA of Pseudoalteromonadaceae, and the likely point of origin of the Pseudoalteromonadaceae chromid. The color scheme shows family affiliation of genera. Numbers of strains included in collapsed nodes are shown in parentheses. Bootstrap values at the nodes were calculated using the ML method, and the GTR+G + I model, with 1000 replicates.

Figure 2

Heatmaps of distribution of core, softcore, shell, and cloud genes in Pseudoalteromonas genomes. Genes were placed into one of six equally sized sectors of chromids (A) or chromosomes (B), with sector 1 starting at the origin of replication (12 o’clock). Unidirectionally replicated chromids are found in species that belong to Clade 1, as shown in the ML-phylogeny (GTR+G + I model, 1000 replicates), whereas bidirectionally replicated chromids belong to representatives of Clade 2. Heatmaps are based on log ₁₀ ratio values of the probability of a gene belonging to a sector on average divided by the probability given an equal distribution of genes among all sectors. Positive values (shades of orange) suggest that gene categories are overrepresented, whereas negative values (shades of blue) suggest underrepresentation. Asterisks indicate significant over- or under-representation of gene categories using Dunńs test P-value ≤ 0.05 (see Supplementary Files S2 and S3 for more details). The phylogenomic tree is based on 469 single marker genes identified by EzTree (Wu 2018). Bootstrap values at the internal nodes were inferred from a ML−G + I analysis.

Figure 3

Global expression maps of P. fuliginea BSW20306 chromosomal and chromid genes centered on the median. Data points (log₂ ratio RPKM CDS: RPKM median) for each CDS are shown, as well as a trend line averaged over a sliding window of 100 data points. The temperatures 4° and 32° corresponds to sub-optimal growth conditions and 15° corresponds to optimal growth conditions.