Nicotinate Metabolism (R-HSA-196807)

[ukemi]

• Make mouse models : http://noctua.geneontology.org/editor/graph/gomodel:678073a900002053? (copied from fly with instances cleaned up);
• Make fly model

In fly:

• De novo synthesis from tryptophan appears not to be functional (no ortholog for human HAAO (3-hydroxyanthranilate 3,4-dioxygenase) and human QPRT (Nicotinate-nucleotide pyrophosphorylase) (PMID:38481042) and no quinolinic acid detected in fly)
• No ortholog in fly for human NAMPT (NAM to NMN reaction)
• fly has nicotinamide amidase (Naam) (Nam to NA), no ortholog in mammals

For the moment I made 3 separate models for NA, Nam and NR but we are thinking to merge them into one unique model.
In the end, I made a unique model for production that have all the various inputs. The 3 separate models were left in development.

• the mitochondrial NAD transporter is missing. I was wondering if it would be possible to add it to the human model (SLC25A51)Q9H1U9 the uniprot entry has all the publication links). Upstream of R-HSA-8955030? Yes - see 20250130 nicotinate metabolism R-HSA-196807 notes
• R-ALL-197220 (ChEBI:15763) should be replaced with the conjugate base CHEBI:57502.
• Should the output (ChEBI18304) of R-HSA-197235 be deamido-NAD zwitterion (CHEBI:14105) or deamido-NAD(2−) (CHEBI:58437). Fly model has the latter, but is the zwitterion more correct for Ph7.3? ChEBI:58437 is correct, per RHEA
• Should NAD+ be CHEBI:57540? Yes (but already fixed)

I used Rhea reactions RHEA 22860 and RHEA 24384 to build the model and they both use CHEBI:58437. Do you think they made a mistake?

• @rozaru for one of your models (gomodel:678073a900000085 ), you have 2 activities occurring in the cytosol and one occurring in the cytoplasm. I assume that is because the enzyme is attached to the Golgi membrane. This is one of those cases where the reaction occurs in a different location that the localization of the protein and is one of the weaknesses of the GO-CAM. Just a heads up that in my model, I changed everything to cytosol.

This is an issue that I kept having problems with. Here the correct GO term is cytosol but in fly the experimental evidence is for cytoplasm. While looking for better evidence I kept the cytoplasm GO term. When I will make the model into production I will change it to cytosol. Sorry for the confusion.

• There are two Rheas associated with R-HSA-76496. One is in all caps and the other is not.

• For R-HSA-200474, the blurb says that the annotation is for the cytosolic form of NMNAT3, but it is the mitochondrial form in the pathway browser. The function of the enzyme in mitochondria has been called into question in mice, PMID:26756334. But this seems to be an exception and is based on a phenotype. Lots of papers show it localized to mitochondria,but it's not clear that its function there has to do with its enzymatic activity in NAD metabolism, see PMID:29046881.

• I think PMID:27796300 allows you to elevate PARP14 from candidate status.

• I'm not sure about the representation of NMA being a direct regulator of R-HSA-2309773. In most of the papers I have looked at, it is clear that administration of it causes an increase in PGI2, but most of them say that the mechanism is unclear and likely works through COX2. In the mouse model I have left the exact mechanism of regulating PTGS2 activity out and made the NNMT activity part of the regulation of prostaglandin biosynthesis. This leaves the opportunity to still regulate indirectly.

[rozaru]

I am just adding the link to the ticket that we made with @sjm41 about the NAD metabolic process that we are trying to sort out.
geneontology/go-ontology#29050

[ukemi]

@deustp01 I think there might be a mistake in this pathway in Reactome? The blurb of R-HSA-197250 states that the reaction produces nicotinamide D-ribonucleotide (NMN), which would be CHEBI:14649 or maybe CHEBI:16171. However, the output of the reaction is nicotinate D-ribonucleotide (15763). If the output really is the amide, this would mean that the salvage pathways would not need feed into the 'de novo' pathway via nicotinate D-ribonucleotide with further processing to NAD via the activity of the synthase. Instead, the nicotinamide D-ribonucleotide (NMN) would be directly converted to NAD by the NMNAT paralogs. The MF for this reaction would be 'nicotinamide-nucleotide adenylyltransferase activity' (GO:0000309) and not 'nicotinate-nucleotide adenylyltransferase activity' (GO:0004515). This is consistent with the figure 2 in PMID:33028824](https://pubmed.ncbi.nlm.nih.gov/33028824/) and the studies presented in PMID: 34290089. There is also some evidence in PMID:38357931 that at least NADSYN1 is not important for salvage, you can take a look. Note this reaction, R-HSA-8939959, is the one from the amide. I think this should be injected into the salvage pathway after R-HSA-197250. The question now is whether the nuclear and mitochondrial paralogs can catalyze both reactions. At any rate, I think the amide one needs to be added to the pathway for NMNAT1 and NMNAT3.

I am still puzzled about the 'de novo' pathway in PMID:33028824](https://pubmed.ncbi.nlm.nih.gov/33028824/) as it shows the synthase acting on NAMN rather than NAAD. According to GO:0008795 it does act on the latter. So that leaves the question of how we go from NAMN to NAAD. Can the NMNAT paralogs also accept nicotinate D-ribonucleotide and catalyze the reaction shown in Reactome to NAAD?

I'd love to have you sanity check all of this. Note also @rozaru 's pathway (http://noctua.geneontology.org/editor/graph/gomodel:678073a900000085?)which can go through the acid/base because of the Naam enzyme that's not present in mammals. And her pathway has the fly NMNAT enzyme working on the acid. The authors of PMID:15310905 and PMID:14704851 state that the enzyme can work on the amide and the acid. https://enzyme.expasy.org/EC/2.7.7.1

[ukemi]

The definition of GO:0034354 'de novo' NAD biosynthetic process from tryptophan

"The chemical reactions and pathways resulting in the formation of nicotinamide adenine dinucleotide (NAD), beginning with the synthesis of tryptophan from simpler precursors; biosynthesis may be of either the oxidized form, NAD, or the reduced form, NADH."

is not correct. this pathway is downstream of the CATABOLISM of tryptophan through [CHEBI:77803] 2-ammonio-3-(3-oxoprop-1-enyl)but-2-enedioate(1-), which spontaneously forms quinolinate(2-)

[ukemi]

If I create a construct of 'biosynthetic process' has_primary_output some NADP+, it is not inferred to be an NADP biosynthetic process. Based on the definition this term refers to both the biosynthesis of NADP+ and NADPH. The ontology group need to check the differentia of the logical definition.
geneontology/go-ontology#29666

[ukemi]

@deustp01 There might be a pathway missing in Reactome. The authors of PMID:19001417 show evidence that in mammals nicotinamide riboside can be acted on by Pnp1 to form nicotinamide, which is then worked on by Nampt (R-HSA-197250) and fed into the salvage pathway (figure 1). I think I am going to add this mouse pathway unless this was later debunked. I don't see it in later pathways. Might be real mentioned in PMID:17914902 (already a Reactome ref) as well.

The evidence may be weak (assay of NR cleavage in a crude extract, blocked by a specific inhibitor) but see 20250130 nicotinate metabolism R-HSA-196807 notes for a possible new reaction

[ukemi]

@rozaru and @deustp01
I am looking at the chain of two reaction in the Reactome nicotinamide salvage pathway, R-HSA-8938073 and R-HSA-5359451. These two reactions convert NAD to NMA the latter of which is a regulator of prostaglandin biosynthesis, R-HSA-2309773 and R-HSA-76496. The first reaction is just one of the examples that @rozaru pointed out a few weeks ago where enzymes use NAD and convert it to NAM. At that time, @rozaru proposed that since this was a one step event, by current GO standards it wasn't part of a process (something I am not entirely comfortable with, but that ship has left). As a result, IIRC we had proposed to obsolete NAD catabolic process.

I've decided to model this area of biology as prostaglandin biosynthesis and its regulation. http://noctua.geneontology.org/editor/graph/gomodel:67b1629100003590?
My dilemma is assigning R-HSA-8938073 and R-HSA-5359451 to processes. I have assigned R-HSA-8938073 to protein modification since that is one of the outcomes of the PARP reaction, but I am unsatisfied with that since the important part of the reaction for this model is the conversion of the NAD. What I would like to do instead is make R-HSA-8938073 and R-HSA-5359451 part of NAD catabolic process and then make the NAM output the regulator of the prostaglandin biosynthesis.
What do you think of this? We can discuss it on a future weeds call.

[rozaru]

At that time, @rozaru proposed that since this was a one step event, by current GO standards it wasn't part of a process (something I am not entirely comfortable with, but that ship has left). As a result, IIRC we had proposed to obsolete NAD catabolic process.

Sorry it wasn't clear, there was no intention to obsolete the GO term "NAD catabolic process”; the term is going to stay. My comment was just that the products of enzymes that catabolise NAD (1 step reaction) are usually re-cycled to remake NAD. In that case I found it difficult to decide where the boundaries between catabolism and NAD salvage lie. However, there may be pathways that don’t convert these products back into NAD and in that case the catabolic term will be useful.