Production of methane in bacteria about gapseq HOT 18 CLOSED

There is a proposed alternative that was suggested in some published papers. One paper is "https://www.nature.com/articles/ismej2014256" and we also tried to describe this alternative pathway here "https://www.sciencedirect.com/science/article/pii/S0960852417320722" (Fig. 3). Since in some genomes, it was not possible to find genes such as EC2.1.1.258, a possible alternative pathway was proposed which is converting Acetyl-CoA to Pyruvate, Serine, Glycine, NH2CH2-DHLP, CH2THF. From my knowledge, this pathway was not clearly demonstrated, but it could be a possible alternative.

This is the description reported by Nobu and coll.: "Thermotogae, a phylum often found in wastewater treatment ecosystems (Narihiro et al., 2014), may be responsible for syntrophic acetate degradation, as observed for Pseudothermotoga lettingae strain TMO (Balk et al., 2002); however, the metabolic pathway remains unclear because of absence of acetyl-CoA synthase/CO dehydrogenase required for conventional syntrophic acetate catabolism through the Wood–Ljungdahl pathway (Muller et al., 2013). Interestingly, the TA-degrading community Mesotoga (bin TAMoga) encodes a potential alternative pathway mediated by the glycine cleavage system and tetrahydrofolate pathway (Supplementary Note, Supplementary Figure S2 and Supplementary Table S6), and H2-generating energy conservation complementary to syntrophic carboxylate degradation (that is, FixABCX, Rnf and ECHyd) (Figure 2)."

from gapseq.

Hi Stefano,
[updated]
indeed, this is odd. I quickly checked the db and there seems to be in issue in our reaction db in this specific pathway. I will try to fix this in the next days.
Best
Silvio

from gapseq.

Thanks Silvio,
these are the command lines I used:
gapseq find-transport Moorella_thermoacetica_ATCC39073.fna.gz
gapseq find -p all -b 200 -m Bacteria Moorella_thermoacetica_ATCC39073.fna.gz
gapseq draft -r Moorella_thermoacetica_ATCC39073-all-Reactions.tbl -t Moorella_thermoacetica_ATCC39073-Transporter.tbl -p Moorella_thermoacetica_ATCC39073-all-Pathways.tbl -c Moorella_thermoacetica_ATCC39073.fna.gz -u 200 -l 100
gapseq fill -m Moorella_thermoacetica_ATCC39073-draft.RDS -n MM_anaerobic_Acetate_H2.csv -c Moorella_thermoacetica_ATCC39073-rxnWeights.RDS -g Moorella_thermoacetica_ATCC39073-rxnXgenes.RDS -b 100
Then I tried also to "draft" and fill with the default parameters (without using -u 200, -l 100, -b 100) but nothing changed.

The same happened with other species having the reductive acetyl coenzyme A pathway and the result seems the same. I will send you the files via email.
Thanks a lot!

Stefano

from gapseq.

I'm having a similar issue, but will keep an eye out for the fix you mentioned above! Thanks for follow up on this so quickly.

from gapseq.

Thank you for your patience – we are currently working on a fix but might need a few days more.

from gapseq.

No worries, thanks for your work on it.

from gapseq.

We committed an update to address this issue. There were two underlying problems. First, some reactions in the Ljungdahl-Wood pathway were missing/blocked, which we corrected with the update. The second cause was, that gapseq's reaction database mainly reflects reaction directions assuming a lower H2-partial pressure, which favors hydrogen production by hydrogenases but impairs the assimilation. The growth environments of homoacetogenic bacteria often have a high H2-pressure, shifting the respective reaction directions towards H2 oxidation. We have added an option for the gapseq fill command to adjust the reactions directions according to a high-H2-pressure environment: -e highH2. Please note that this is, at the moment, an experimental solution to the issue and we plan to extent this option to adjust environmental conditions (e.g. pH) when constructing models.

Here is an example workflow to get a reconstruction for the acetogenic oganism Moorella thermoacetica:

# the setup
gapseq=~/path/to/gapseq/./gapseq
model="Moorella_thermoacetica_ATCC39073"

# Reconstruction
$gapseq find -p all -b 200 -m Bacteria $model.fna.gz
$gapseq find-transport -b 200 $model.fna.gz

$gapseq draft -r $model-all-Reactions.tbl \
-t $model-Transporter.tbl \
-p $model-all-Pathways.tbl \
-c $model.fna.gz \
-u 200 -l 100

$gapseq fill -m $model-draft.RDS \
-n ../dat/media/MM_anaerobic_CO2_H2.csv \
-c $model-rxnWeights.RDS \
-g $model-rxnXgenes.RDS \
-b 100 -e highH2

This is probably almost the workflow you have used. But please note the new additional option -e highH2. This specifies that you are running the gapfilling in an environment with high partial pressure for hydrogen gas (i.e. > 100 mbar). By this specification, certain H2-forming reactions are blocked for the direction of H2-evolution, while certain H2-consuming reactions are enabled (e.g. https://doi.org/10.1074/jbc.M112.395038).
(You might need to adjust the path to your gapseq installation and to the medium)

With this example workflow, I observed the expected acetate formation from CO2 and H2 with a high flux through the LWP.

However, I also tried to achieve synthrophic acetate oxidation (SAO) by Clostridium ultunense on anaerobic minimal medium with acetate but no H2, which did not yet work (still methane production from acetate). We will need to have a closer look at this and, admittedly, also need to read more about the pathways involved in SAO.

So in summary, it is not a complete fix but we hope it is a step closer to a better reconstruction for anaerobic bacteria, which heavily rely on C1 metabolism.

Genome for the above mentioned example:
Moorella_thermoacetica_ATCC39073.fna.gz

from gapseq.

Thanks for your work on this. Maybe it could be something related to inappropriate detection of methyl coenzyme M reductase activity-could be that the sequence of that enzyme + another in the homoacetogenic strains is similar enough to give you a false positive.

from gapseq.

@jdwinkler-lanzatech yes, thanks for pointing this out. That is definitely one part of the problem. I found also two other aspects that might contribute and where I'm currently running some tests. I'll hope to be able to commit a fix this week. Step by step :)
In general, predicting flux distribution that involve the LWP are a difficult task as it operates at the edge of what is thermodynamically feasible.

from gapseq.

Agreed. Let me know if I can weigh in on any proposed solutions.

from gapseq.

Many thanks for your work on this. I agree that the main problem of WL pathway is its ability to work at the limit of thermodynamical feasibility, and this makes predictions very difficult. Additionally, WL pathway exists both in Bacteria and Archaea but they are different. I understand that there are some species particularly difficult to manage because of their high distance from well-known species or because of their peculiar pathways.
I hope you will manage to solve the problem!

from gapseq.

You might be able to special case the directionality of the pathway based on the presence or absence of Coenzyme B-related enzymes, since that cofactor is only (currently) known to be in methanogenic Archaea. I'm generally leery of such assumptions though, so probably better to make it a flag rather than a default fixed behaviour. Thanks again for your work on this though!

from gapseq.

Hi,

First of all, thank you both for the input! I made a few changes to the db, although they were minor. More important was that we needed to add Protons to the gapfilling medium alongside acetate. The reason is that acetate (charge -1) is imported into the cell via a proton (charge +1) symport. I added an anaerobic minimal media with acetate and protons to gapseq (dat/media/MM_anaerobic_Acetate.csv) that can be used for gapfilling of putative SAOB.

I tested four organisms:

• (CO2 + H2 -> Acetate) Moorella thermoacetica. Pred. phenotype as expected: acetogenic
• (CO2 + H2 -> Acetate) Blautia hydrogenotrophica DSM 10507. Pred. phenotype as expected: acetogenic
• (Acetate -> CO2 + H2) [Clostridium] ultunense DSM 10521. False phenotype prediction: methane production
• (Acetate -> CO2 + H2) Syntrophaceticus schinkii. Pred. phenotype as expected: acetate oxidation

Of course I hoped for a better prediction for C. ultunense. When I checked the Wood-Ljungdahl pathway I noticed that no genes for the WLP from acetyl-CoA to 5-Methyl-THF were found. There's not so much literature on this organism, but I found this thesis by Shahid Manzoor, were he describes exactly the same missing part of the WLP (page 86). Thus, since gapseq doesn't identify the full pathway, gapseq's gapfilling adds the wrong reactions that enables anaerobic growth on acetate. Maybe you two know more about an alternative route in this organism or genes for the respective reactions in this organism? From screening literature, I have the feeling that knowledge about reactions and genes involved in SAO is rather scarce. Besides that, we'll think of a way for gapseq to exclude archaea-specific reactions for gapfilling of bacterial models.

In constrast, the predictions for S. schinkii is as expected.

The genomes and a file with the command I used are here:
Blautia_hydrogenotrophica_DSM_10507.fna.gz
Clostridium_ultunense_DSM_10521.fna.gz
Syntrophaceticus_schinkii.fna.gz
LWP.pdf

Again, not a full fix, but maybe something that will already help you to get more out of gapseq for your questions :)

Silvio

from gapseq.

I'm not too familiar with that specific organism unfortunately, so I can't be of much help. Could be that the missing enzyme is encoded by a more divergent protein that could be picked up with a HMM scan, but that is probably beyond the scope of reasonable modifications. I'll try with the canonical acetogens to see what happens though!

from gapseq.

Yes, I agree.
In general, there are three possibilities: First, as you said, the genes for the enzymes in the LWP are to divergent and therefore are "under the radar" of gapseq's homology detection. Second, the genome is incomplete, and the respective genes are missing in the assembly. Or third, the organism uses a different metabolic route that gapseq does not (yet) cover.
I found that the C. ultunence has homologous genes for the the NQR-complex, S. schinkii seems not to have this complx. The NQR-complex might be important for energy metabolism for C. ultunense. At the moment, gapseq does not have a pathway or a reaction for this sodium pump complex. We'll include this and see if it changes predictions. Not sure if it's part of the issue, but it is good to include this reaction anyway.

from gapseq.

Thanks, Stefano! I will check whether something in the gapseq reaction database is blocking possible alternative routes that are proposed in the articles you mentioned.

from gapseq.

Hi,

as one part of a fix (commit 8777d33), methane production should not occur anymore in bacterial reconstructions. However, it does not solve the problem with SAOB, yet. Instead of methane, the reconstructions for SOAB that I tested produce butyrate or propionate instead. We need to further analyze this.

from gapseq.

please reopen if there is still an issue with this

from gapseq.