Bioinformatics Identification of Anti-CRISPR Loci by Using Homology, Guilt-by-Association, and CRISPR Self-Targeting Spacer Approaches

Bioinformatics pipeline to identify genomic loci containing Acr homologs and/or Aca homologs. Guilt-by-association (GBA) was used to identify new HTH proteins (Aca proteins) that neighbor homologs of known Acr proteins in the RefSeq bacterial genomes and IMG/VR metagenome-assembled viral contigs (MACs). These Aca proteins were combined with published Aca proteins (Text S1) to search for more Aca homologs. GBA was further used to identify genomic loci (or operons) that contain Aca homologs. More filters were then used to select loci that are from genomes with annotated mobile genetic elements (MGEs), complete CRISPR-Cas systems, and self-targeting CRISPR spacers. For Diamond search (39), the –more-sensitive option was used to improve search sensitivity. For hmmsearch (http://hmmer.org/), the default parameters were used.

Homologs of known Acr families in RefSeq and IMG/VR databases. (A) Breakdown numbers of homologs in each known Acr family (total of 975 for RefSeq and 2,022 for IMG/VR). (B) Breakdown numbers of homologs in each known Acr family that have neighboring Aca homologs (total of 224 for RefSeq and 165 for IMG/VR [Fig. 2]). Note that AcrIIA1 itself is considered an Aca; therefore, the two RefSeq AcrIIA1 genes in panel B coexist with other AcrIIA1 homologs in the same operons.

One hundred seventy-three Acr homologs are from genomes with self-targeting CRISPR-Cas spacers of different subtypes. The y axis shows 26 known Acr families, which have homologs in genomes with self-targeting CRISPR-Cas spacers. The Cas subtypes of the self-targeting spacers (according to the work of Watters et al. [12]) are shown on the x axis. The numbers in the circles are the numbers of Acr homologs.