Tabular analysis of polyketide synthase domain active site residues for the PKZILLAs

In brief, the presence or absence of expected conserved enzyme active site residues within domains of the PKZILLAs, were matched to the polyketide ketosynthase (KS), acyl carrier protein (ACP), dehydratase (DH), ketoreductase (KR), flavin oxygenase (FLX), and enoyl (ER) domains, using either multiple sequence alignment liftover of active site location & residue as defined by InterProScan annotations, or, locations & residues defined as active site residues from literature sources.

./KS_ALL represents the scripts and source files used to generate the analyses. See run.sh , within.

Post_process_KS-ALL_analysis.ipynb represents the Jupyter notebook used to process the raw analyses into the presented tabular (.xlsx) form.

See Manuscript Materials and Methods for further details.

Summary of the active side residue results:

KS:All scored as active on the basis of catalytic cysteine. The below KS domains are scored as KS0, on the basis of variation of the histidine in the HGTGT motif:

PKZILLA-1__KS14 QGLGS

PKZILLA-1__KS19 QGLGS

PKZILLA-1__KS22 QGLGS

PKZILLA-1__KS25 QGLGS

PKZILLA-1__KS28 QGLGS

PKZILLA-1__KS31 GANGS

PKZILLA-1__KS34 QGLGS

PKZILLA-2__KS3 QGLGS

PKZILLA-2__KS6 QGLGS

PKZILLA-2__KS9 QGLGS

PKZILLA-2__KS11 QGLGS

PKZILLA-2__KS12 QGLGS

PKZILLA-3__KS6 AANGS

ACP:All scored as active (PPant prosthetic group attachment residue present)

DH:All scored as active, except possibly:PKZILLA-3__DH9

ER:Active site residues not defined in InterProScan or literature sources.

FLX:Active site residues not defined in InterProScan or literature sources.

KR:

KR domains come in two flavors: the canonical "full-length" contiguous domain, and the non-canonical split domain with two subdomains.KRs are 2 reciprocally similar subdomains - where the N-terminal subdomain "KRn" has a structural role, and the C-terminal subdomain "KRc" has the catalytic role.

Typically, the two KR subdomains are contiguous. 

However, it's precedendented (Porcine FAS) that these two KRn and KRc subdomains can become split with an ER in the middle. In the PKZILLAs, this can be observed for the KR-ER-KR in Module 28, it's actually KRn-ER-KRc. The first KRn-subdomain (KR26) is missing the catalytic residues (as are all of the KRn domains), while the second KRc-subdomain (KR27) has all of them. This split KR domain structure seems to be not that uncommon in the PKZILLAs, and the intervening domain is not limited to being an ER.

The KR.xlsx excel summarizes the status of the KRs, as active or inactive, and full-length (f), n-terminal subdomain (n), or c-terminal subdomain (c). The below KR domains are scored as inactive:

PKZILLA-1__KR26PKZILLA-1__KR29PKZILLA-1__KR31PKZILLA-2__KR6PKZILLA-2__KR11PKZILLA-2__KR20

Summary of the domain phylogenetics results:

Present in separate Zenodo item: https://doi.org/10.5281/zenodo.10152639