Script: analyze_plsA.py (uv run python analyze_plsA.py). It fetches the
sequences live (UniProt, UniParc, NCBI protein), writes them to
sequences.fasta, and writes every number below to results.json. Nothing
here is hard-coded.
Questions
- How close is Q9CDI2 to the PlsA proteins whose plasmalogen-forming activity
was shown experimentally?
- Does Q9CDI2 carry the full PlsAR domain set (Act1, Act2, Red1, Red2), or
only the activator half implied by its name ("2-hydroxyglutaryl-CoA
dehydratase"; KBase/eggNOG "hgdC")?
Methods
- Global (Needleman-Wunsch) alignment, BLOSUM62, gap open -11 / extend -1
(Biopython PairwiseAligner), against single-chain homologs; local
(Smith-Waterman) alignment against the two separate C. perfringens
proteins. Identity = identical columns / all alignment columns, gaps
included.
- Pfam HMMs PF01869 (BcrAD_BadFG), PF09989 (DUF2229) and PF06050 (HGD-D),
downloaded from the InterPro API and searched with pyhmmer (E <= 10, so
that weak hits are reported) against the full chain, against residues
901-1436 alone, and against controls.
Results
Sequence identity of Q9CDI2
| Comparator |
What it is |
Identity |
Q9CDI2 span |
| UPI00072AC48C (1441 aa) |
KBase NZ_VXKC01000008.1_19 / WP_081041604 |
98.3% |
1-1436 |
| ABJ73971.1 (1436 aa) |
L. cremoris ATCC BAA-493 LACR_2543, the PlsA expressed in PMID:41427725 |
96.6% |
1-1436 |
| A2RP33 (1436 aa) |
L. cremoris MG1363 llmg_2516 |
96.7% |
1-1436 |
| AAO81118.1 (1415 aa) |
E. faecalis EF1327 PlsA (PMID:33350306) |
67.3% |
1-1436 |
| G0VRE7 (1431 aa) |
M. elsdenii MeHAD, MELS_0169 (PMID:37348645) |
48.7% |
1-1436 |
| BAB80901.1 (1004 aa) |
C. perfringens CPE1195 PlsA (Act1-Act2-Red1) |
51.1% |
1-989 |
| BAB80900.1 (420 aa) |
C. perfringens CPE1194 PlsR (Red2) |
41.3% |
1011-1411 |
PMID:41427725 names the BAA-493 gene "plsABAA493 (4,311 bp)" and lists its
locus tag as LACR_2543. 4,311 bp is 1,436 codons plus a stop codon, which
matches ABJ73971.1.
Domain coverage
- Pfam on the full chain: BcrAD_BadFG at 7-257 (i-E 1.3e-24) and 322-576
(i-E 1.5e-40), so Act1 and Act2 are present. DUF2229 is present at 676-897
(i-E 4e-80; Red1).
- C-terminal region: PF06050 (HGD-D) gives no hit on Q9CDI2, either on the
full chain or on residues 901-1436 alone. A weak partial DUF2229 match
(HMM 17-63, i-E ~1e-6) lies at 1023-1089. C. perfringens PlsR, the
experimentally required Red2 subunit, behaves the same way: it also gives
only weak, partial DUF2229 matches (i-E 1.2e-7 for HMM 14-66). In EF1327,
HGD-D gives only fragmentary hits (best i-E 2.6e-4).
- The local alignment of Q9CDI2 residues 1011-1411 to PlsR (41.3% identity
over 402 columns) is the direct evidence that the C-terminal ~430 residues
are the Red2 domain, fused in one chain as in EF1327. The Pfam HGD-D model
is too divergent to detect this domain.
Act-domain cysteines
Of the nine cysteines in Q9CDI2 residues 1-620, eight align to cysteine in
both EF1327 and C. perfringens PlsA (C126, C165, C445, C449, C484, C606,
C609, C613). C572 is Ala in EF1327. PMID:33350306 reports that Act1 and Act2
share conserved cysteines for iron-sulfur cluster ligation with HAD and BCR
activators. This analysis shows only that the cysteines are conserved among
PlsA proteins. It does not map them onto HgdC cluster ligands.
MeHAD cysteine mutant mapped onto Q9CDI2
PMID:37348645 reports that a combined mutant of MeHAD Cys125, Cys164, Cys445
and Cys484, the predicted [4Fe-4S] ligands numbered on the MeHAD sequence
G0VRE7, cannot make plasmalogens. In the global alignment of Q9CDI2 with
G0VRE7 (48.7% identity), these residues align to:
| MeHAD (G0VRE7) |
Q9CDI2 |
| C125 |
C126 |
| C164 |
C165 |
| C445 |
C445 |
| C484 |
C484 |
All four Q9CDI2 positions are among the eight Act-region cysteines conserved
in EF1327 and C. perfringens PlsA (above). The mutant result is known only
from the abstract (the cached record has no full text), so it is unknown
whether the four mutations were also tested singly.
PANTHER subfamily of the characterized PlsA proteins
The UniProt PANTHER cross-references, fetched by the script, place Q9CDI2,
A2RP33 (L. cremoris MG1363), Q8DVN8 (S. mutans SMU_438c), Q835P9 (E. faecalis
EF1327) and G0VRE7 (M. elsdenii MeHAD) all in PTHR32329:SF4.
Conclusions
- Q9CDI2 is a full-length, single-chain plasmalogen synthase (PlsA/PlsAR
architecture Act1-Act2-Red1-Red2). It is essentially the same protein
(96.6% identity) as the L. cremoris PlsA shown to confer plasmalogen
synthesis on E. coli.
- The ProtNLM name "2-hydroxyglutaryl-CoA dehydratase" and the eggNOG name
"hgdC" describe only one family relationship. Neither describes this
protein's function.
- Limits: identity is not proof of identical activity. No L. lactis subsp.
lactis knockout or in vitro enzyme assay exists. Cysteine conservation is
not cluster-ligand assignment.