deprecate crpP - strong evidence this is marker-by-neighborhood only and NOT a ciprofloxacin phosphotransferase.

[danielhhaft]

NCBI's Pathogen group originally treated crpP as the ciprofloxacin phosphotransferase it was reported to be, but now instead agrees with rebuttals to that publication, including

PMID: 34001507] (http://www.ncbi.nlm.nih.gov/pubmed/34001507)) "CrpP is not a Fluoroquinolone Inactivating Enzyme."

NCBI's RefSeq and PGAP group describes CrpP this way, in the BlastRule https://www.ncbi.nlm.nih.gov/genome/annotation_prok/evidence/NBR016116/

ICE-associated protein CrpP

"CrpP, described originally as a protein encoded by Pseudomonas aeruginosa plasmid pUM505, was reported to confer elevated resistance to ciprofloxacin, but not to four other fluoroquinoline-type antibiotics tested. The mechanism proposed was ATP-dependent phosphorylation. However, the original report is controversial. CrpP appears more closely related to the ribosome modulation factor Rmf than to any known aminoglycoside phosphotransferase. Some recent reports question the proposed mechanism, link CrpP to mobile integrative and conjugative elements (ICE), and conclude CrpP does not directly confer resistance to ciprofloxacin. CrpP remains interesting as a marker because of its association with metal resistance proteins and virulence factors coharbored in the same mobile elements."

Recently, Ana Rita Rebelo wrote to us at NCBI and apaprently also to card@mcmaster.ca . Perhaps there is a ticket already that I missed.
I see PMID:34001507 is cited by the CARD entry, but the entry lacks the banner that tet(34) has, saying the node in the hierarchy is now "private", and no longer used by CARD's RGI.

This ticket is a suggestion by NCBI to harmonize on the removal of CrpP as an antibiotic resistance protein, as Dr. Rebelo suggested. We see that over 100 isolates are ciprofloxacin-sensitive despite the presence of this marker - still a minority, but a significant observation - and we plan to explore this further. Please interpret this GitHub ticket as a request to open discussion of the topic, and see if CARD has a quick answer.

Many thanks!

Dan

Daniel Haft
Staff Scientist
National Center for Biotechnology Information (NCBI)
National Library of Medicine (NLM)
National Institutes of Health (NIH)
USA

[danielhhaft]

I'd like to update my comments.

We completed our analysis. Among isolates tested, isolates with crpP were actually more likely to be sensitive to cipro than isolates without. Furthermore, for the majority of isolates resistance to ciprofloxacin, resistance could be explained by the expected point mutations.

I have now made crpP non-reportable in AMRFInderPlus, effectively killing it from our database, effective with our next release. I encourage CARD to do the same.