The protein encoded by PP_2928 (UniProt Q88IS1), generically annotated in databases as a
"saccharopine dehydrogenase," is more precisely a carboxyspermidine dehydrogenase
(equivalently carboxynorspermidine dehydrogenase/synthase; CASDH/CANSDH), EC 1.5.1.43. It is a
soluble, cytoplasmic, NADPH‑dependent oxidoreductase that catalyzes the reductive condensation
of L‑aspartate‑4‑semialdehyde with a diamine (putrescine → carboxyspermidine, or
1,3‑diaminopropane → carboxynorspermidine). This is the penultimate step of the alternative
("carboxyspermidine") biosynthetic pathway for the polyamine spermidine (and sym‑norspermidine).
Its product is decarboxylated by the immediately adjacent enzyme PP_2929 (carboxyspermidine/
carboxynorspermidine decarboxylase, CASDC, EC 4.1.1.96) to yield spermidine (or norspermidine).
Because P. putida KT2440 lacks the canonical spermidine synthase (SpeE) and S‑adenosylmethionine
decarboxylase (SpeD), PP_2928 is expected to be the organism's principal route to spermidine.
| Attribute | Value |
|---|---|
| UniProt | Q88IS1 (Q88IS1_PSEPK), 414 aa, 45.4 kDa, evidence level 4 (Predicted) |
| Locus | PP_2928 (OrderedLocusName); RefSeq WP_003248123.1; GenBank AAN68536 |
| Organism | Pseudomonas putida KT2440 (ATCC 47054 / DSM 6125), TaxID 160488 |
| Domains | PF03435 Sacchrp_dh_NADP (res 5–146); PF16653 Sacchrp_dh_C (res 150–396) |
| Superfamily | Gene3D 3.40.50.720 NAD(P)-binding Rossmann fold; eggNOG COG1748 (saccharopine dehydrogenase & related) |
| PANTHER | PTHR43796:SF2 "Carboxynorspermidine synthase" |
| KEGG | ppu:PP_2928 — name "Saccharopine dehydrogenase", no KO, no pathway assigned |
The gene symbol and organism match the UniProt record. The database label "saccharopine dehydrogenase"
is a family‑level automatic annotation, not an experimentally determined activity; this report
resolves the specific function using domain architecture, orthology, genomic context, and pathway logic.
Four independent lines of evidence converge:
IIGAGGVAK (residues 7–14). CASDH is a member of this saccharopine‑dehydrogenaseQuantitative orthology (added Iteration 2). By global (Needleman–Wunsch) alignment, PP_2928
(414 aa) is 45.2% identical over its full length to the reviewed, experimentally characterized
carboxynorspermidine synthase/dehydrogenase of Vibrio cholerae (UniProt Q9KRL3, VC_1624, 414 aa;
Lee et al. 2009), 46.2% to V. alginolyticus CANSDH (P0DPE4, 414 aa), 44.5% to Idiomarina
loihiensis (Q5QY81) and 42.3% to Geobacter sulfurreducens (Q74A52) — all annotated CANSDH/CASDH
(EC 1.5.1.43). By contrast PP_2928 is only 21.1% identical to its own operon partner CASDC
(PP_2929), confirming they are distinct enzymes of the same pathway. ~45% identity over the entire
length, to a reviewed, biochemically validated enzyme, is well above the threshold for confident
functional transfer and moves the assignment beyond family-level inference.
Reactions catalyzed (from the 45%-identical reviewed ortholog Q9KRL3; the enzyme is dual-substrate):
(i) L‑aspartate‑4‑semialdehyde + 1,3‑diaminopropane + NADPH → carboxynorspermidine + NADP⁺ + H₂O (Rhea 34115)
(ii) L‑aspartate‑4‑semialdehyde + putrescine + NADPH → carboxyspermidine + NADP⁺ + H₂O (Rhea 34111)
Aspartate‑β‑semialdehyde is the aminopropyl donor, replacing decarboxylated S‑adenosylmethionine used
by the canonical pathway. The V. cholerae ortholog is classified in UniProt as "saccharopine
dehydrogenase family, Carboxynorspermidine synthase subfamily" — the exact subfamily to which
PP_2928's domain architecture belongs. Deletion of this enzyme in V. cholerae abolishes both
norspermidine and spermidine and causes accumulation of 1,3‑diaminopropane (Lee et al. 2009),
demonstrating the dual‑substrate physiology.
Cofactor/active-site conservation (Iteration 3). The PP_2928 ↔ V. cholerae CANSDH alignment is
colinear over 417 columns with only 6 gaps (no large insertions/deletions), indicating identical
two-domain 3D architecture. The N‑terminal NAD(P)H‑binding Rossmann fingerprint is conserved
(PP_2928 …IIGAGGVAK…IAIASRNISKC… vs CANSDH …LQIGAGGVGW…ITIASRSIAKC…; the βαβ glycine‑rich
GAGGV motif and the downstream IASR dinucleotide‑binding motif are retained), confirming that
PP_2928 binds an NADP(H) cofactor exactly as the characterized enzyme does.
P. putida KT2440 lacks the canonical spermidine‑biosynthesis genes — spermidine synthase
speE (K00797) and SAM decarboxylase speD (K01611) are absent from the genome (KEGG). It does
encode the alternative‑pathway decarboxylase CASDC (PP_2929). Therefore the two‑step
CASDH (PP_2928) → CASDC (PP_2929) route is the only genomically encoded path to spermidine.
This mirrors the situation described by Hanfrey et al. (2011): species that make spermidine but lack
SpeD/SpeE instead carry carboxynorspermidine dehydrogenase and decarboxylase orthologues. The reaction
maps to KEGG map00330 (Arginine and proline metabolism).
Complete pathway reconstruction from primary metabolites (Iteration 3). P. putida KT2440 encodes
every step needed to feed PP_2928 and to finish the pathway:
| Step | Enzyme | Locus (KEGG KO) |
|---|---|---|
| L‑aspartate → L‑aspartate‑4‑semialdehyde | aspartate‑semialdehyde dehydrogenase asd | PP_1989 (K00133) |
| L‑arginine → agmatine | arginine decarboxylase speA | PP_0567 (K01585) |
| agmatine → putrescine | agmatinase speB | PP_2196 (K01480) |
| L‑ornithine → putrescine | ornithine decarboxylase speC | PP_0864 (K01581) |
| ASA + putrescine + NADPH → carboxyspermidine | CASDH | PP_2928 (this study; K13746 unassigned in KEGG) |
| carboxyspermidine → spermidine | CASDC | PP_2929 (K13747) |
Both PP_2928 substrates are therefore demonstrably produced by the cell. The 1,3‑diaminopropane branch
is incomplete (DABA decarboxylase ddc/K13745 is absent in the KEGG annotation, although DABA
aminotransferase dat is present at PP_2800/PP_4223), so putrescine is the more likely in vivo amine
substrate and spermidine the more likely product, even though the enzyme is intrinsically
dual‑specificity. The canonical genes speD (K01611) and speE (K00797) are absent from KEGG, so this route is
the sole genomically encoded path to spermidine.
Ruling out a canonical bypass (Iteration 4). P. putida does encode one aminopropyltransferase‑
domain protein, PP_1867 (Q88LR2; Pfam PF01564 Spermine_synth; PANTHER PTHR43317 thermospermine
synthase). However, (i) its "spermidine synthase" name derives only from automated ProtNLM
text‑mining (ECO:0008006), and KEGG lists it as a "conserved protein of unknown function" with no KO;
and (ii) more decisively, canonical spermidine/spermine synthases require decarboxylated
S‑adenosylmethionine (dcSAM) as the aminopropyl donor, produced by SAM decarboxylase (SpeD) —
which P. putida does not encode (no KEGG K01611; zero hits in a UniProt proteome search for
adenosylmethionine decarboxylase). Without dcSAM, no aminopropyltransferase can synthesize spermidine.
The CASDH→CASDC route instead uses L‑aspartate‑4‑semialdehyde as the aminopropyl donor and is fully
encoded, so PP_2928/PP_2929 constitute the only complete de novo route to spermidine in P. putida.
(By contrast, P. aeruginosa makes norspermidine through a genuinely SAM‑dependent aminopropyltransferase
system, speD2/speE2 = PA4773–PA4775 (Bolard et al. 2019) — a route absent from P. putida, underscoring
that the two pseudomonads use different chemistries.)
PP_2928 is a cytoplasmic enzyme: 414 aa, soluble, with no predicted signal peptide or
transmembrane segment (the N‑terminus is a Rossmann dinucleotide‑binding start, MKKNVLIIGAGGVAK).
NADPH‑dependent polyamine‑biosynthetic reactions occur in the cytosol.
The AlphaFold model of Q88IS1 (AF‑Q88IS1‑F1, v6) is very high confidence: mean pLDDT = 95.9, with
92% of residues ≥90 and none <50. The model is well‑ordered across both the N‑terminal Rossmann
NAD(P)‑binding domain and the C‑terminal substrate‑binding domain, consistent with the two‑domain
carboxynorspermidine‑synthase/saccharopine‑dehydrogenase fold and with the colinear (gap‑poor)
alignment to the characterized V. cholerae CANSDH.
Polyamines (spermidine/norspermidine) are essential polycations supporting translation, growth, and
biofilm formation. In Vibrio cholerae, deletion of CANSDH or CANSDC abolishes spermidine/norspermidine,
reduces planktonic growth by 50–60%, and severely impairs biofilm formation (Lee et al. 2009). By
analogy — and given the absence of a bypass route — PP_2928 is expected to be important for optimal
growth and biofilm physiology in P. putida (a specific P. putida knockout has not, to my knowledge,
been reported).