Today I had a long Zoom call with someone that we hope to cooperate with for Long COVID. She mentioned that one species of probiotics works very well for some people with long COVID and it’s sibling do not. I explained that my working hypothesis is that a lot of interactions depends on the enzymes.

This resulted in us doing a quick lookup experiment using Microbiome Prescription databases.

• Bifidobacterium adolescentis is what worked well. I looked up what enzymes it had, 469
• Then I removed those that are also found in Bifidobacterium animalis subsp. lactis, that left 51 that was only in Bifidobacterium adolescentis
• Removed those also found in Bifidobacterium breve dropping the count to 28
• Removed those also found in Bifidobacterium longum subsp. infantis dropping the count to 17
• Removed those also found in Bifidobacterium longum dropping the count to 15
• Removed those also found in Bifidobacterium bifidum dropping the count to 13 unique enzymes not in the sibling strains.

I kept to only the probiotics that are retail available.

This hints that the reason that it works is due to the following enzymes (linked to Kyoto Encyclopedia of Genes and Genomes):

• sorbose reductase;Sou1p
• coenzyme F420 hydrogenase;8-hydroxy-5-deazaflavin-reducing hydrogenase;F420-reducing hydrogenase;coenzyme F420-dependent hydrogenase
• opine dehydrogenase;(2S)-2-{[1-(R)-carboxyethyl]amino}pentanoate dehydrogenase (NAD+, L-aminopentanoate-forming)
• homocysteine S-methyltransferase;S-adenosylmethionine homocysteine transmethylase;S-methylmethionine homocysteine transmethylase;adenosylmethionine transmethylase;methylmethionine:homocysteine methyltransferase;adenosylmethionine:homocysteine methyltransferase;homocysteine methylase;homocysteine methyltransferase;homocysteine transmethylase;L-homocysteine S-methyltransferase;S-adenosyl-L-methionine:L-homocysteine methyltransferase;S-adenosylmethionine-homocysteine transmethylase;S-adenosylmethionine:homocysteine methyltransferase
• lysyltransferase;L-lysyl-tRNA:phosphatidylglycerol 3-O-lysyltransferase
• cellobiose phosphorylase
• diacylglycerol kinase (ATP);diglyceride kinase (ambiguous);1,2-diacylglycerol kinase (phosphorylating) (ambiguous);1,2-diacylglycerol kinase (ambiguous);sn-1,2-diacylglycerol kinase (ambiguous);DG kinase (ambiguous);DGK (ambiguous);ATP:diacylglycerol phosphotransferase;arachidonoyl-specific diacylglycerol kinase;diacylglycerol:ATP kinase;ATP:1,2-diacylglycerol 3-phosphotransferase;diacylglycerol kinase (ATP dependent)
• beta-glucoside kinase;beta-D-glucoside kinase (phosphorylating)
• sulfur carrier protein ThiS adenylyltransferase;thiF (gene name)
• allantoinase
• 5-oxoprolinase (ATP-hydrolysing);pyroglutamase (ATP-hydrolysing);oxoprolinase;pyroglutamase;5-oxoprolinase;pyroglutamate hydrolase;pyroglutamic hydrolase;L-pyroglutamate hydrolase;5-oxo-L-prolinase;pyroglutamase
• glutamate decarboxylase;L-glutamic acid decarboxylase;L-glutamic decarboxylase;cysteic acid decarboxylase;L-glutamate alpha-decarboxylase;aspartate 1-decarboxylase;aspartic alpha-decarboxylase;L-aspartate-alpha-decarboxylase;gamma-glutamate decarboxylase;L-glutamate 1-carboxy-lyase
• sirohydrochlorin cobaltochelatase;CbiK;CbiX;CbiXS;anaerobic cobalt chelatase;cobaltochelatase [ambiguous];sirohydrochlorin cobalt-lyase

In theory, the 13 items above may be excellent candidates for novel pharmaceuticals to be trialed for treating long COVID.

This same approach may be done with other conditions and probiotics. By trials identify which probiotics have the most impact, use the same approach to identify possible enzymes causing the difference, then do a clinical trial.