Various Problems With Lipoic Acid Biosynthesis

[Devlin-Moyer]

Background:

Depends on changes proposed in #683, and, to a lesser extent, #682 and #684

Lipoic acid biosynthesis from octanoyl-ACP is currently represented by MAR06402:

• apoA1 [c] + octanoyl-[ACP] [c] --> [ACP] [c] + [protein]-N6-(octanoyl)lysine [c] (MAM01350c + MAM02643c --> MAM00184c + MAM00210c)
• GPR: LIPT2 (ENSG00000175536)

and MAR06403:

• [protein]-N6-(octanoyl)lysine [c] + 5 H+ [c] + 7 NADPH [c] + 2 PAPS [c] + 2 SAM [c] --> [protein]-N6-(lipoyl)lysine [c] + 2 5-deoxyadenosine [c] + 6 H2O [c] + 2 methionine [c] + 7 NADP+ [c] + 2 PAP [c] (MAM00210c + 5 MAM02039c + 7 MAM02555c + 2 MAM02682c + 2 MAM02877c --> MAM00209c + 2 MAM01098c + 6 MAM02040c + 2 MAM02471c + 7 MAM02554c + 2 MAM02681c)
• GPR: LIAS (ENSG00000121897)

These papers (1, 2, 3, 4, 5, 6, 7, 8) collectively indicate that these two reactions should instead look like:
MAR06402:

• octanoyl-[ACP] [m] --> [protein]-N6-(octanoyl)lysine [m] + mitoACP [m] (MAM02643m --> MAM00210m + MAM02484m)
• GPR: NDUFAB1 and GCSH and LIPT2 (ENSG00000004779 and ENSG00000140905 and ENSG00000175536)

and MAR06403:

• [protein]-N6-(octanoyl)lysine [m] + 3 H+ [m] + NADPH [m] + 2 SAM [m] + 4Fe4S [m] --> [protein]-N6-(dihydrolipoyl)lysine [m] + 2 5-deoxyadenosine [m] + 4 Fe²⁺ [m] + 2 HS^- [m] + 2 methionine [m] + NADP⁺ [m] (MAM00210m + 3 MAM02039m + MAM02555m + 2 MAM02877m + <4Fe4S> --> MAM00208m + 2 MAM01098m + 4 MAM01821m + 2 MAM02042m + 2 MAM02471m + MAM02554m)
• GPR: GCSH and LIAS and FDX1 and FDXR and (NFU1 or ISCA1 or ISCA2) (ENSG00000140905 and ENSG00000121897 and ENSG00000137714 and ENSG00000161513 and (ENSG00000169599 or ENSG00000135070 or ENSG00000165898))

Explanation:

• Lipoic acid is synthesized from octanoyl-ACP in the mitochondria by strictly mitochondrial enzymes and is only attached to strictly mitochondrial enzymes (sources: 1, 2))
• LIPT2 transfers an octanoyl from NDUFAB1 (ENSG00000004779) (which functions as the mitochondrial Acyl-Carrier Protein (ACP)) to GCSH (ENSG00000140905), not apoA1 (MAM01350c), (same sources as above)
• Since all lipoylated proteins appear to be strictly mitochondrial, the current cytosolic metabolites for [protein]-N6-(lipoyl)lysine and [protein]-N6-(dihydrolipoyl)lysine should just be completely replaced with mitochondrial ones. These metabolites also participate in MAR00022, MAR06404, MAR08433, and MAR08434, so those would also need to be edited. Since the other metabolites in MAR06404 are cytosolic NAD(H), they can also just be replaced with mitochondrial NAD(H). See Misrepresentation of Catalytic Activity of LIPT1 #683 re: MAR08433 and MAR08434 and why this would not make MAR06404 redundant with MAR06409.
• The sulfur atoms used by LIAS come from a 4Fe4S iron-sulfur cluster bound to LIAS, and the sulfur atoms in that iron-sulfur cluster ultimately come from cysteine, not PAPS (MAM02682c; sources: 1, 2, 3, 4). New 2Fe2S iron-sulfur clusters are produced by a complex formed by ISCU2, NFS1, ISD11 (LYRM4), NDUFAB1, FXN, FDX2, and FDXR (sources: 1, 2).
  • That complex only produces 2Fe2S clusters, and combining two of those into a 4Fe4S cluster involves HSC20 (HSCB), HSPA9, GLRX5, GRPEL1, ISCA1, ISCA2, IBA57, FDX2, FXR, hydrolysis of two ATP, and oxidation of one NADPH. While LIAS can only use 4Fe4S clusters, some proteins apparently use 2Fe2S clusters, so I’m proposing two separate reactions for each of these steps.
  • Newly synthesized 4Fe4S clusters are primarily delivered to LIAS by NFU1, but can also be delivered directly by ISCA1 and ISCA2 to a lesser extent (source), and since these delivery proteins tend to specific for a few particular 4Fe4S-using proteins, I’m proposing that (NFU1 or ISCA1 or ISCA2) be added to the GPR of MAR06403.
• As far as I can tell, each turnover of LIAS only requires 1 NADPH, not 7, since the NADPH only seems to be used to reduce FDX1 (source), which reduces the iron-sulfur cluster that LIAS uses to generate 5’-deoxyadenosyl radicals. FDX1 does a one-electron reduction of that iron-sulfur cluster twice (once for each of the two sulfur atoms attached to the octanoyl residue), and NADPH does two-electron reductions, so each turnover of LIAS requires oxidizing one NADPH. Relatedly, FDX1 and FDXR should also be added to the GPR of MAR06403.
• After each turnover of LIAS, the remaining 2 sulfur atoms and 4 iron atoms from the 4Fe4S cluster that gets partially dismantled are released into solution as Fe²⁺ and S^2- (sources: 1, 2). I used HS^- instead of S^2- in the new version of MAR06403 above because S^2- is a very strong base and rapidly becomes HS^- in physiological pH ranges, and there was already a metabolite for HS^- but not S^2- in Human-GEM. Two more reactions also need to be added to connect this new mitochondrial HS^- with the existing sulfur metabolism reactions:
  • HS^- + H⁺ + GSH (glutathione) + FAD -> GSSH (glutathione persulfide AKA S-sulfanylglutathione) + FADH₂, GPR: SQOR (ENSG00000137767), references: PMID:25225291, RHEA:62608
    • Uniprot lists the electron acceptor as ubiquinone, but also lists FAD as a cofactor, so as we decided in Proposed fixes to DHAP <=> sn-glycerol-3-phosphate reactions #599 and FAD/Ubiquinone Duplicate Reactions #607, I'm using FAD as the electron acceptor here and also proposing to add SQOR to the GPR of MAR06911 (redox between FAD and ubiquinone)
  • GSSH + O₂ + H₂O -> GSH + sulfite + 2 H⁺, GPR: ETHE1 (ENSG00000105755), references: PMID:25225291, E.C.: 1.13.11.18, RHEA:12981
• Human-GEM does not currently have a mitochondrial metabolite for 5’-deoxyadenosine. Apparently, nobody really knows what happens to 5’-deoxyadenosine in human cells (source), so I’m just proposing adding a gene-less [m] -> [c] transport reaction to keep the new mitochondrial 5’-deoxyadenosine in MAR06403 connected to the existing [c] <-> [e] transport reaction MAR09637.

Proposed Changes:

• Create new metabolite MAM02643m to represent mitochondrial octanoyl-ACP (only necessary if changes proposed in Mitochondrial Fatty Acid Synthesis #684 haven’t already been implemented)
• Create new metabolite MAM02042m to represent mitochondrial HS^-
• Create a new metabolite MAM01098m to represent mitochondrial 5’-deoxyadenosine
• Create a new metabolite to represent a mitochondrial 4Fe4S iron-sulfur cluster:
  • formula: S4Fe4
  • charge +2
  • InChI: InChI=1S/4Fe.4S/q;;2*+1;;;;
  • CHEBI:33722
• Create a new metabolite to represent a mitochondrial 2Fe2S iron-sulfur cluster:
  • formula: S2Fe2
  • charge: +2
  • InChI: InChI=1S/2Fe.2S/q2*+1;;
  • CHEBI:33737
• Create a new metabolite to represent mitochondrial GSSH:
  • formula: C10H17N3O6S2
  • charge: 0
  • InChI: InChI=1S/C10H17N3O6S2/c11-5(10(18)19)1-2-7(14)13-6(4-21-20)9(17)12-3-8(15)16/h5-6,20H,1-4,11H2,(H,12,17)(H,13,14)(H,15,16)(H,18,19)/t5-,6-/m0/s1
  • KEGG: C17267, CHEBI:52857
• Create new entries for all genes mentioned above that are not already present in Human-GEM:
  • ENSG00000100209, HSCB
  • ENSG00000105755, ETHE1
  • ENSG00000109519, GRPEL1
  • ENSG00000113013, HSPA9
  • ENSG00000135070, ISCA1
  • ENSG00000136003, ISCU2
  • ENSG00000137714, FDX1
  • ENSG00000137767, SQOR
  • ENSG00000165898, ISCA2
  • ENSG00000169599, NFU1
  • ENSG00000181873, IBA57
  • ENSG00000182512, GLRX5
  • ENSG00000214113, LRYM4
  • ENSG00000244005, NFS1
  • ENSG00000267673, FDX2
• Replace MAM00210c with MAM00210m
  • Edit rather than copy the existing entry for MAM00210c to create the entry for MAM00210m
• Replace MAM00209c with MAM00209m in MAR00022, MAR06404, MAR08433, and MAR20168
  • Edit rather than copy the existing entry for MAM00209c to create the entry for MAM00209m
• Replace MAM00208c with MAM00208m in MAR06404 and MAR08434
  • Edit rather than copy the existing entry for MAM00208c to create the entry for MAM00208m
• Change MAR06402 to MAM02643m -> MAM00210m + MAM02484m, GPR: ENSG00000004779 and ENSG00000140905 and ENSG00000175536, references: PMID:24777537;PMID:28757203
• Change MAR06403 to MAM00210m + 3 MAM02039m + MAM02555m + 2 MAM02877m + <4Fe4S> --> MAM00208m + 2 MAM01098m + 4 MAM01821m + 2 MAM02042m + 2 MAM02471m + MAM02554m
  • GPR: ENSG00000121897 and ENSG00000137714 and ENSG00000140905 and ENSG00000161513 and (ENSG00000135070 or ENSG00000165898 or ENSG00000169599)
  • references: PMID:24777537;PMID:31493409;PMID:36281303;PMID:37453661
• Replace MAM02553c with MAM02553m, MAM02552c with MAM02552m, and MAM02039c with MAM02039m in MAR06404
• Add ENSG00000137767 to the GPR of MAR06911
• Create a new reaction to represent synthesis of 4Fe4S iron-sulfur clusters in mitochondria:
  • 2 MAM01371m + 2 MAM02040m + MAM02555m + 2 <2Fe2S> -> 2 MAM01285m + 3 MAM02039m + MAM02554m + 2 MAM02751m + <4Fe4S>
  • GPR: ENSG00000100209 and ENSG00000109519 and ENSG00000113013 and ENSG00000135070 and ENSG00000161513 and ENSG00000165898 and ENSG00000181873 and ENSG00000182512 and ENSG00000267673
  • references: PMID:31918395;PMID:31935115
• Create a new reaction to represent synthesis of 2Fe2S iron-sulfur clusters in mitochondria:
  • 2 MAM01628m + 2 MAM01821m + MAM02555m -> 2 MAM01307m + MAM02039m + MAM02554m + <2Fe2S>
  • GPR: ENSG00000004779 and ENSG00000136003 and ENSG00000161513 and ENSG00000165060 and ENSG00000214113 and ENSG00000244005 and ENSG00000267673
  • references: PMID:31918395;PMID:31935115
• Create a new reaction to represent formation of GSSH from GSH and HS^- by SQOR:
  • MAM01802m + MAM02026m + MAM02042m -> MAM01803m + <GSSH>
  • GPR: ENSG00000137767
  • references: PMID:25225291
  • E.C.: 1.8.5.8
  • KEGG: R11929, RHEA:62608
• Create a new reaction to represent conversion of GSSH into GSH and sulfite by ETHE1:
  • MAM02040m + MAM02630m + <GSSH> -> MAM02026m + 2 MAM02039m + MAM02949m
  • GPR: ENSG00000105755
  • references: PMID:25225291
  • E.C.: 1.13.11.18
  • KEGG: R08678, RHEA:12981
• Create a new reaction MAM01098m -> MAM01098c, references: PMID:34126623