hydA (P13061, PHNS_DESDA) — periplasmic [NiFe] hydrogenase small subunit

Identity and partner subunit

• P13061 (PHNS_DESDA, Swiss-Prot) is the small (beta) subunit of the periplasmic
  [NiFe] hydrogenase of Desulfovibrio desulfuricans ATCC 27774 / DSM 6949 / MB
  (NCBITaxon:525146), locus Ddes_1039, 315 aa precursor with a Tat signal peptide
  (1–49) and mature chain 50–315 [UniProtKB:P13061 FT SIGNAL 1..49 /note="Tat-type signal"].
• The large (alpha) catalytic subunit reviewed alongside it is
  UniProtKB:Q9L868 (Q9L868_DESDE, TrEMBL, genes/DESDE/Q9L868/).

Are they genuine partners? Yes, despite carrying different UniProt taxon ids.
Checks performed in this review (live UniProt/NCBI queries, 2026-10-28):

• Q9L868 derives from EMBL AF216303 (Saraiva, 1999), whose UniProt reference
  block records STRAIN=ATCC 27774, although the entry's OX line is the
  species-level NCBITaxon:876 rather than the strain taxon 525146. The apparent
  "two strains" split is a record-level artifact, not a biological one.
• AF216303 also encodes Q9L869 (266 aa, "cytochrome-c3 hydrogenase", fragment).
  Its sequence is identical to the P13061 mature chain (residues 50–315) at
  265/266 positions (single difference R→S at mature position 5). So Q9L869 and
  P13061 are the same gene product from the same strain, sequenced twice.
• Q9L868 (543 aa, Flags: Fragment) is identical at 542/543 positions to
  GenBank ACL48945.1 ("nickel-dependent hydrogenase large subunit",
  locus_tag="Ddes_1038", 558 aa) from the ATCC 27774 genome CP001358.
  Ddes_1038 is immediately adjacent to Ddes_1039 (= hydA), i.e. the two
  reviewed accessions are the large and small subunits of one operon in one strain.
• PDB 1E3D is the heterodimeric enzyme from this strain; Q9L868 maps to chains
  B/D (residues 2–543) [UniProtKB:Q9L868 DR PDB; 1E3D; X-ray; 1.80 A; B/D=2-543.].
  The structure paper is PMID:11191224
  (the species is misspelled in the published abstract).

Function

The division of labour in [NiFe] hydrogenases is well established and is visible
in the two records themselves: the bimetallic Ni–Fe catalytic centre is carried
by the large subunit (Q9L868 FT BINDING residues for Ni(2+) and Fe cation),
while the small subunit carries the iron–sulfur relay — two [4Fe-4S] clusters
and one [3Fe-4S] cluster — that conducts electrons between the active site and
the external redox partner
[UniProtKB:P13061 COFACTOR: Name=[4Fe-4S] cluster ... Note=Binds 2 [4Fe-4S] clusters.;
COFACTOR: Name=[3Fe-4S] cluster ... Note=Binds 1 [3Fe-4S] cluster.].
The structural work on this very enzyme shows the cluster chain running up to the
[NiFe] centre PMID:11191224
and identifies the physiological acceptor as the tetrahaem cytochrome c3, docking
at the distal [4Fe-4S] cluster of the small subunit
PMID:11191224.

Therefore, for the small subunit:

• GO:0009055 electron transfer activity is its own molecular function, not a
  merely structural role; it is accepted here as the core MF.
• GO:0047806 cytochrome-c3 hydrogenase activity (EC 1.12.2.1, the whole-enzyme
  reaction) is retained but marked as an over-annotation of an individual chain,
  following the repository's subunit precedent (genes/PSEPK/ureA): the small
  subunit contributes to, rather than independently enables, the holoenzyme
  activity. It is recorded as contributes_to_molecular_function in core_functions.
• The protein is a soluble, Tat-exported periplasmic protein
  [UniProtKB:P13061 "SUBCELLULAR LOCATION: Periplasm."; PTM: Predicted to be exported by the Tat system.];
  it has no transmembrane segment and the holoenzyme was crystallised as a soluble
  heterodimer. The TreeGrafter GO:0016020 membrane row is over-propagation from
  membrane-anchored relatives (e.g. E. coli Hyd-1/Hyd-2 small subunits, which
  carry a C-terminal membrane anchor this protein lacks) and is removed.

InterPro2GO defects (recorded as source_defects)

The live interpro2go file maps the two [NiFe] hydrogenase signatures onto
ferredoxin-specific terms:

InterPro:IPR001821 [NiFe]-hydrogenase, small subunit > GO:ferredoxin hydrogenase activity ; GO:0008901
InterPro:IPR001821 [NiFe]-hydrogenase, small subunit > GO:ferredoxin hydrogenase complex ; GO:0009375
InterPro:IPR018194 Nickel-dependent hydrogenase, large subunit, nickel binding site > GO:ferredoxin hydrogenase activity ; GO:0008901

GO:0008901 is defined as "Catalysis of the reaction: 2 reduced ferredoxin + 2 H+
= 2 oxidized ferredoxin + H2" (the [FeFe]/ferredoxin reaction, EC 1.12.7.2), and
GO:0009375 ferredoxin hydrogenase complex is an intracellular complex term.
Neither fits a periplasmic cytochrome-c3-linked [NiFe] enzyme (EC 1.12.2.1).
Suggested repointing: GO:0033748 hydrogenase (acceptor) activity (acceptor-agnostic,
suitable for a family-level signature) and GO:0044569 [Ni-Fe] hydrogenase complex.

Relation to the CMMMech record (data/records/palladium.yaml)

CMMMech grounds the palladium mechanism in D. desulfuricans to this pair of
accessions. Two points where the CMMMech text is stronger or looser than the
evidence supports (noted here only; CMMMech is not edited by this review):

1. The functional_role for P13061 reads "…small iron-sulfur subunit (EC 1.12.2.1;
   HydA / Ddes_1039) mediating H2 oxidation and periplasmic electron transfer".
   H2 oxidation chemistry happens at the Ni–Fe site of the large subunit; the
   small subunit mediates the electron transfer half of that sentence. The EC number
   properly belongs to the holoenzyme.
2. The Q9L868 entry is given target_organism: NCBITaxon:876 (species level), which
   mirrors UniProt but obscures that the sequence is from ATCC 27774 — the same
   strain as P13061. The two protein examples are operon partners (Ddes_1038/
   Ddes_1039), which is stronger support for the "periplasmic [NiFe] hydrogenase
   system" claim than the record currently states.

The Pd(II)-reduction evidence itself remains indirect for either subunit:
Lloyd et al. used D. desulfuricans ATCC 29577 (a different strain) and inferred
hydrogenase involvement from H2 dependence and Cu2+ pretreatment, explicitly noting
that no deletion mutants were available
PMID:9797331,
and Mikheenko et al. worked in D. fructosivorans
PMID:18689514.
No GO annotation of metal (Pd/U/Tc) reduction is proposed for either subunit here:
no experiment localises Pd(II) reduction to this enzyme, and a hydrogenase reducing
a soluble metal ion non-specifically is not evidence that it executes a
characterised biological process.