Plastid ancestors lacked a complete Entner-Doudoroff pathway, limiting plants to glycolysis and the pentose phosphate pathway

The Entner–Doudoroff (ED) pathway provides an alternative to glycolysis. It converts 6-phosphogluconate (6-PG) to glyceraldehyde-3-phosphate and pyruvate in two steps consisting of a dehydratase (EDD) and an aldolase (EDA). Here, we investigate its distribution and significance in higher plants and determine the ED pathway is restricted to prokaryotes due to the absence of EDD genes in eukaryotes. EDDs share a common origin with dihydroxy-acid dehydratases (DHADs) of the branched chain amino acid pathway (BCAA). Each dehydratase features strict substrate specificity. E. coli EDD dehydrates 6-PG to 2-keto-3-deoxy-6-phosphogluconate, while DHAD only dehydrates substrates from the BCAA pathway. Structural modeling identifies two divergent domains which account for their non-overlapping substrate affinities. Coupled enzyme assays confirm only EDD participates in the ED pathway. Plastid ancestors lacked EDD but transferred metabolically promiscuous EDA, which explains the absence of the ED pathway from the Viridiplantae and sporadic persistence of EDA genes across the plant kingdom.

This manuscript reports the absence of a fully functional ED pathway in plants and potentially their plastid ancestor.To support this claim, authors showed the absence of EDD enzyme in plants through phylogenetics and conducted substrate docking to confirm the substrate specificity for EDD and DHAD, two structurally similar dehydratase.The in vitro biochemical assays for EDD activities using either genuine EDDs or structurally similar DHAD candidates further validated the substrate docking simulation.In addition, the metabolite analysis of KDPG indicates its absence in many eukaryotes.This result is in contradiction with a previous report that KDPG could be detected in at least barley roots.Overall, I found the evidence convincing to illustrate the absence of the EDD gene in the plant domain, thus supporting the notion that the ED pathway is absent in plants.
The prevalence of EDA enzyme in all life forms thus is intriguing to understand it potential contribution to cell metabolism.The authors further showed that EDA has moonlighting aldol cleavage activities toward several phosphate sugars.The confocal microscopy is also a nice touch to show the subcellular localization of EDA in plants.
In summary, I believe the main result here is significant in that it provides a right path for future studies to understand the potential role of EDA in contributing cell metabolism, rather than through the ED pathway activity.I also like some of the discussion authors provided in terms of how ED activity could divert fluxes out of CBB or cancel G6P shunt fluxes if it was to exist in either chloroplast or cytosol.The method description is also thorough in describing different procedures.I have some additional minor comments listed below: -The previous research in cyanobacteria labeled DHAD as a potential candidate for EDD activities.In the substrate docking simulation, did authors try docking 6PG into DHAD for potential interactions or did it fail to dock? -Please correct the mislabeling in the Figure 3  -Please provide the MS/MS spectrum of 6PG, KDPG, and GAP either in the main figures or in supplemental figures to support Figure 3  -Line 319-: "promoter"?Do you mean monomer?-Line 324: here should be Fig.2C and 2D -Line 461-462: Please double check the accuracy of this statement regarding non-phosphorylating ED in Synechocystis.
Reviewer #2 (Remarks to the Author): In this interesting manuscript, the authors use different methodologies to shed light on the evolutionary history and distribution of the Entner-Doudoroff (ED) pathway.
The manuscript text and figures are very clear.I am very pleased by the complementary bioinformatics and experimental approaches used by the authors, which I find adequate and elegant.
The general conclusions of the study are convincing and important.The results support the notion that the ED pathway is absent from eukaryotic lineages and allow proposing that the plastid ancestor likely lacked 6-PG dehydratase (EDD), and hence a complete ED pathway.
The authors also discuss relevant points, including an alternative hypothesis for the lack of EDD in photosynthetic eukaryotes, the persistence of KDPG aldolase (EDA) in plants, and the potential exploitation of the ED pathway in plant synthetic biology approaches.

I recommend publication.
Minor comments: 1.It would have been interesting to explore the physiological and metabolic impact of localizing EDD to the plant cytosol, plastids, and mitochondria.

Line 353:
There is a typo; please change assaus to assays.
3. Line 361: Perhaps "not detected" is better than "blocked" in this context.4. Figure 1C: "Bacteria" could be centered like the other biological lineages names.The curved blue line could be separated from "Cyanobacteria". 5. Figure 3A: Retention times 12 and 4 for 6-PG, KDPG, and GAP should be separated.In their article "Plastid ancestors lacked a complete Entner-Doudoroff pathway, limiting plants to glycolysis and the pentose phosphate pathway", Evans et al. describe a well-conducted study into a big question in the evolution of plant biochemistry: the distribution of a functioning ED pathway.Overall, the study is of good quality.I however have remarks regarding the interpretation / framing.

6.
Rerding data availability: have the raw spectra data from the mass spec been uploaded?E.g., to MetaboLights or another adequate database?I think that the study is of great value and the analyses carried out are an important contribution to the scientific discourse.Yet, I am not convinced by the evidence for the "plastid ancestor" (highlighted in the title and throughout).As fars as I can judge, the entire statements rests on the following: "Although the true sister group to plastids is uncertain, proposed candidates for the most recent common ancestor to plastids and free-living cyanobacteria coalesce around Gleomargarita lithospora and several closely related taxa which include Synechococcus sp., Synechocystis sp., Prochlorococcus marinus, Trichodesmium sp., Oscillatoria sp., and Arthrospira sp."When we consider features of the cyanobacterial progenitor of plastids, we are talking about an orgnanism that likely lived around 2 billion years ago.Screening a few cyanobacteria is not enough.Prokaryotes are known to loose and gain genes with ease, the latter propelled by LGT.The problems that result from this are expertly explained in: Endosymbiotic gene transfer from prokaryotic pangenomes: Inherited chimerism in eukaryotes by Ku C, Nelson-Sathi S, …, and Martin WF.PNAS 2015 112 (33) 10139-10146 https://doi.org/10.1073/pnas.1421385112When talking about these extant organisms (leaves on the tree from this year and not the 2billion-year-old phylogenetic entitity that woudl branch off a deep node on the tree), please do not talk about them them as if they would represent 1to1 the yanobacterial plastid progenitor (which lived about 2 billion years ago!).Talk about the groups that they belong to.Talk about the last common ancestors that they might share with the cyanobacterial plastid progenitor.Talk in trees.
"none of which is considered a candidate for the ancestor of plastids."-> Again, these are all extant organisms and not an organism that lived 2 billion years ago.See above.And when it comes to the groups of filamentous nitrogen fixing cyanos (like Nostoc), this is also not true.See, e.g., Dagan et al 2013 GBE --DOI: 10.1093/gbe/evs117 --for a discussion on this shifting target, see: https://doi.org/10.1016/j.cub.2016.12.006 Figure 1: It would help a lot to reconcile the gene phylogeny here with a species tree to show the distribution.Also the choice of organisms is unclear.A lot of Chlorella/Chlorella-like algae have been picked, but only one streptophyte alga, not including the closest algal relatives to land plants, see, cite and add: https://www.biorxiv.org/content/10.1101/2023.01.31.526407v1 & https://www.nature.com/articles/s41477-023-01491-0#Abs1& https://www.sciencedirect.com/science/article/pii/S0092867420304827& https://doi.org/10.1016/j.cub.2022.08.022Furthermore, please cite all the genome data that have been used--I do not find them in the reference list or material and methods e.g. the recently sequenced Chlorelloids or the key species Chara braunii.
The biochemical data and figures are, as far as I can judge, all sound and of adequate quality.See comment about data availability above.The only aspect that I would like the authors to consider here is to also provide in the figures 3 to 5 the phylogenetic context from which the enzymes were derived would be very helpful (writing e.g.embryophytes, Chlorophyceae etc. below the species abbreviation would help already a lot) legends.Fig. 3A, B, C should be B, C, D.
and Figure 4 conclusions.-Line 52-53: rephrase "and the availability of non-glycolytic sources of ATP…" -Line 100: I think authors meant "in vitro" rather than "in vivo" for plant EDA activity in Ref 11? -Line 312: Fig 2C is DHAD substrate domain.Here should be Fig 2B to reference E. coli EDD domain 1.Again, Fig. 2D should be 2B as well in line 315.