bchJ (P26169, Rhodobacter capsulatus SB1003) — curation notes

Provenance of this file

Deep research was unavailable. No deep-research API keys exist in this container, so
just deep-research was not run and no -deep-research-<provider>.md file exists for this
gene. The literature below was assembled directly: from the cached publications, from PubMed
via NCBI E-utilities (esearch/efetch), from Europe PMC availability queries, from PMC full
text where open, and from web search. Every assertion carries a [PMID:n "quote"] or a
UniProt-record pointer.

Papers newly cached for this review with just fetch-pmid: 7876113, 17148453, 23252506,
25978726, 20955518. Of these, only PMID:25978726 has full_text_available: true; the rest,
and the CACAO-cited PMID:8158642, are abstract-only in the cache.

What the protein is

213 aa, no transmembrane segments, no cofactor features, protein existence level
PE 4: Predicted. Encoded in the ~46 kb R. capsulatus photosynthesis gene cluster as
RCAP_rcc00670, immediately downstream of bchE (RCAP_rcc00669) and four genes from
bchF (RCAP_rcc00666). Single V4R domain (Pfam PF02830, SMART SM00989, InterPro IPR004096),
with a dedicated family signature NCBIfam TIGR02019 BchJ and InterPro IPR010249 BchJ.
UniProt curates the pathway as
PATHWAY: Porphyrin-containing compound metabolism; bacteriochlorophyll biosynthesis (light-independent).
UniProt's AltName: Full=4-vinyl reductase; and the domain name V4R ("vinyl-4-reductase") both
preserve an activity assignment that the literature has since withdrawn (see below) — recorded
as a source_defects entry against UniProt in the review.

The CACAO-cited paper: PMID:8158642 (Bollivar, Suzuki, Beatty, Dobrowolski, Bauer 1994)

A cluster-wide mutational survey, not a single-gene paper:
PMID:8158642 Its readouts are exactly loss of
BChl and accumulation of pathway intermediates:
PMID:8158642 and
PMID:8158642

The cache is abstract-only and the abstract does not name the 12 ORFs, so I cannot read off
which category bchJ fell into. The partial-disruption phenotype is, however, precisely what the
same laboratory reported for bchJ one year later (below: reduced BChl plus an accumulating
intermediate, rather than loss of BChl), which makes bchJ the natural candidate for that single
partial mutant. I state this as an inference, not as a quotation from the 1994 paper.

bchJ-specific mutant phenotype

PMID:7876113 and, decisively for the process claim,
PMID:7876113. Note that both an
insertion and a deletion allele were used, which weakens (though does not exclude) a purely
polar explanation.

Confirmed from the R. capsulatus literature by a later group:
PMID:25978726

The contested chemistry: BchJ is not the C8-vinyl (4-vinyl) reductase

The divinyl-protochlorophyllide accumulation above was read as BchJ being the 8-vinyl
(divinyl) reductase, DVR. Three lines retract that:

1. Green sulfur bacteria. Insertional inactivation of both candidate genes in
   Chlorobium tepidum put the activity on CT1063, not bchJ:
   PMID:17148453 The bchJ mutant still made C8-reduced pigments:
   PMID:17148453 Their conclusion:
   PMID:17148453
2. A heterologous complementation test. In R. sphaeroides,
   PMID:23252506, and in a cyanobacterial vinyl-reductase mutant
   PMID:23252506 Hence
   PMID:23252506.
3. Who does it in Rhodobacter. The C8-vinyl reduction in Rhodobacter is carried by
   chlorophyllide a oxidoreductase (COR, BchXYZ) and by BciA where present:
   PMID:25978726 and in the R. sphaeroides ΔbchJ/bciA double mutant
   PMID:25978726

The polar-mutation hypothesis named in this review's brief — that the original 4-vinyl
reductase assignment rested on an insertion that was polar on downstream genes — I could not
substantiate in any source I could read. Chew & Bryant, Canniffe et al. and Tsukatani et al.
all argue from mutant pigment composition and heterologous complementation, not from
polarity. PMID:7876113 used a deletion as well as an insertion. Hahn et al. have argued that
bchE and bchJ lie in one operon whose promoter is upstream of bchE, so polarity is a coherent
worry, but it is not the argument the field actually made, and I record no claim about it in
the review.

What BchJ positively does (still unresolved)

• Binds a Mg-tetrapyrrole, on the R. capsulatus protein itself.
  PMID:20955518 and PMID:20955518 Stoichiometry:
  PMID:20955518 Proposed role:
  PMID:20955518 Note the
  authors' framing: PMID:20955518
• Raises the effective throughput of the intrinsic COR. In R. sphaeroides, an exogenous
  B. viridis BchYZ competes with the intrinsic enzyme only once BchJ is gone:
  PMID:25978726 The mechanism is explicitly
  open: PMID:25978726 and
  PMID:25978726
• The pathway-level membership is not in doubt:
  PMID:25978726

GO term checks (QuickGO, 2026-11-01)

• GO:0030494 bacteriochlorophyll biosynthetic process. Ancestors: GO:0015995, GO:0015994,
  GO:0042440, GO:0046148, GO:0033013, GO:0033014, GO:0006778, GO:0006779, GO:0009058,
  GO:0008152. GO:0015979 photosynthesis is not among them.
• GO:0015979 photosynthesis — "The synthesis by organisms of organic chemical compounds,
  especially carbohydrates, from carbon dioxide (CO2) using energy obtained from light rather
  than from the oxidation of chemical compounds." BchJ performs no step of this; it supplies the
  pigment. The InterPro2GO mapping IPR010249 → GO:0015979 is therefore pigment-supply read as
  pathway participation, and the review MODIFYs it to GO:0030494 (a lateral correction, since
  the replacement is not a descendant).
• GO:0036070 light-independent bacteriochlorophyll biosynthetic process is a descendant of
  GO:0030494, and matches the UniProt pathway line, so it is the most specific correct term.
• No GO molecular-function term exists for a Mg-porphyrin carrier/channelling activity. The
  children of GO:0046906 tetrapyrrole binding are only chlorophyll binding, heme binding and
  cobalamin binding; no "magnesium protoporphyrin IX binding" term exists. GO:0046906 is
  therefore the most specific MF the binding data support, and it is proposed as a NEW row.

CACAO / GONUTS status (checked 2026-11-01)

gowiki.tamu.edu has moved from "temporarily offline" to permanently retired. The site
returns HTTP 410 Gone with the page title "GONUTS has been retired" and the text
"The GO wiki that was at gowiki.tamu.edu is no longer online." EcoliWiki and OMP from the same
lab are still up. The project page's instruction to "re-check the abstract-only rows when GONUTS
returns" is therefore no longer actionable: for CACAO rows whose cached publication is
abstract-only, the paper itself is now the only check available.

Cross-isolate transfer screen (batch-3 screen 2) — result

PMID:8158642 is cited by four CACAO rows, all involved_in GO:0030494 / IMP / dated 20111120:

The two taxon ids are both Rhodobacter capsulatus: 272942 is the strain-level
ATCC BAA-309 / NBRC 16581 / SB1003 reference proteome, 1061 the species node. The split is a
UniProt entry-level artefact, not a curation decision: BCHM_RHOCA is a 1992 Swiss-Prot entry
built on the Youvan et al. 1984 sequence (PMID:6744416) that was never re-filed under the
SB1003 proteome, whereas the other three carry OrderedLocusNames=RCAP_rcc006xx and sit in
UP000002361. All four genes are in the one photosynthesis gene cluster that PMID:8158642
mutated gene by gene, and P26169's own sequence (EMBL Z11165, Burke/Alberti/Armstrong/Hearst)
is from SB1003 — the same strain lineage whose cluster the paper worked in. The predicted
cross-isolate shape is absent, and the neighbour-mis-attribution alternative is absent too:
four genes citing one cluster-wide mutational survey is legitimate, not over-harvesting.