Multienzyme interactions of the de novo purine biosynthetic protein PAICS facilitate purinosome formation and metabolic channeling

Abstract

There is growing evidence that mammalian cells deploy a mitochondria-associated metabolon called the purinosome to perform channeled de novo purine biosynthesis (DNPB). However, the molecular mechanisms of this substrate-channeling pathway are not well defined. Here, we present molecular evidence of protein–protein interactions (PPIs) between the human bifunctional phosphoribosylaminoimidazole carboxylase/succinocarboxamide synthetase (PAICS) and other known DNPB enzymes. We employed two orthogonal approaches: bimolecular fluorescence complementation, to probe PPIs inside live, intact cells, and co-immunoprecipitation using StrepTag-labeled PAICS that was reintegrated into the genome of PAICS-knockout HeLa cells (crPAICS). With the exception of amidophosphoribosyltransferase, the first enzyme of the DNPB pathway, we discovered PAICS interacts with all other known DNPB enzymes and with MTHFD1, an enzyme which supplies the 10-formyltetrahydrofolate cofactor essential for DNPB. We show these interactions are present in cells grown in both purine-depleted and purine-rich conditions, suggesting at least a partial assembly of these enzymes may be present regardless of the activity of the DNPB pathway. We also demonstrate that tagging of PAICS on its C terminus disrupts these interactions and that this disruption is correlated with disturbed DNPB activity. Finally, we show that crPAICS cells with reintegrated N-terminally tagged PAICS regained effective DNPB with metabolic signatures of channeled synthesis, whereas crPAICS cells that reintegrated C-terminally tagged PAICS exhibit reduced DNPB intermediate pools and a perturbed partitioning of inosine monophosphate into AMP and GMP. Our results provide molecular evidence in support of purinosomes and suggest perturbing PPIs between DNPB enzymes negatively impact metabolite flux through this important pathway. There is growing evidence that mammalian cells deploy a mitochondria-associated metabolon called the purinosome to perform channeled de novo purine biosynthesis (DNPB). However, the molecular mechanisms of this substrate-channeling pathway are not well defined. Here, we present molecular evidence of protein–protein interactions (PPIs) between the human bifunctional phosphoribosylaminoimidazole carboxylase/succinocarboxamide synthetase (PAICS) and other known DNPB enzymes. We employed two orthogonal approaches: bimolecular fluorescence complementation, to probe PPIs inside live, intact cells, and co-immunoprecipitation using StrepTag-labeled PAICS that was reintegrated into the genome of PAICS-knockout HeLa cells (crPAICS). With the exception of amidophosphoribosyltransferase, the first enzyme of the DNPB pathway, we discovered PAICS interacts with all other known DNPB enzymes and with MTHFD1, an enzyme which supplies the 10-formyltetrahydrofolate cofactor essential for DNPB. We show these interactions are present in cells grown in both purine-depleted and purine-rich conditions, suggesting at least a partial assembly of these enzymes may be present regardless of the activity of the DNPB pathway. We also demonstrate that tagging of PAICS on its C terminus disrupts these interactions and that this disruption is correlated with disturbed DNPB activity. Finally, we show that crPAICS cells with reintegrated N-terminally tagged PAICS regained effective DNPB with metabolic signatures of channeled synthesis, whereas crPAICS cells that reintegrated C-terminally tagged PAICS exhibit reduced DNPB intermediate pools and a perturbed partitioning of inosine monophosphate into AMP and GMP. Our results provide molecular evidence in support of purinosomes and suggest perturbing PPIs between DNPB enzymes negatively impact metabolite flux through this important pathway. Most metabolic enzymes are part of large metabolic networks where multiple enzymes, of varying substrate affinities and enzymatic activities, must work in concert. Thus, the biochemistry of isolated enzymes, studied in vitro, provides only partial pictures of their behavior in their native context inside living cells. Indeed, in vivo studies have increasingly demonstrated the importance of protein–protein interactions (PPIs) between enzymes, and between enzymes and other proteins and cellular structures, in regulating cell metabolism (1Srere P.A. The metabolon.Trends Biochem. Sci. 1985; 10: 109-110Abstract Full Text PDF Scopus (226) Google Scholar, 2Vélot C. Mixon M.B. Teige M. Srere P.A. Model of a quinary structure between krebs TCA cycle enzymes: A model for the metabolon.Biochemistry. 1997; 36: 14271-14276Crossref PubMed Scopus (131) Google Scholar, 3Wu F. Minteer S. Krebs cycle metabolon: Structural evidence of substrate channeling revealed by cross-linking and mass spectrometry.Angew. Chem. Int. Ed. Engl. 2015; 54: 1851-1854Crossref PubMed Scopus (96) Google Scholar, 4Zhang Y. Beard K.F.M. Swart C. Bergmann S. Krahnert I. Nikoloski Z. Graf A. Ratcliffe R.G. Sweetlove L.J. Fernie A.R. Obata T. Protein-protein interactions and metabolite channelling in the plant tricarboxylic acid cycle.Nat. Commun. 2017; 8: 15212Crossref PubMed Scopus (85) Google Scholar, 5Islam M.M. Wallin R. Wynn R.M. Conway M. Fujii H. Mobley J.A. Chuang D.T. Hutson S.M. A novel branched-chain amino acid metabolon. Protein-protein interactions in a supramolecular complex.J. Biol. Chem. 2007; 282: 11893-11903Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 6Graham J.W. Williams T.C. Morgan M. Fernie A.R. Ratcliffe R.G. Sweetlove L.J. Glycolytic enzymes associate dynamically with mitochondria in response to respiratory demand and support substrate channeling.Plant Cell. 2007; 19: 3723-3738Crossref PubMed Scopus (207) Google Scholar, 7Szecowka M. Heise R. Tohge T. Nunes-Nesi A. Vosloh D. Huege J. Feil R. Lunn J. Nikoloski Z. Stitt M. Fernie A.R. Arrivault S. Metabolic fluxes in an illuminated Arabidopsis rosette.Plant Cell. 2013; 25: 694-714Crossref PubMed Scopus (216) Google Scholar, 8Laursen T. Borch J. Knudsen C. Bavishi K. Torta F. Martens H.J. Silvestro D. Hatzakis N.S. Wenk M.R. Dafforn T.R. Olsen C.E. Motawia M.S. Hamberger B. Møller B.L. Bassard J.-E. Characterization of a dynamic metabolon producing the defense compound dhurrin in sorghum.Science. 2016; 354: 890-893Crossref PubMed Scopus (161) Google Scholar). These interactions can enhance flux by channeling intermediate products between successive enzymes, limit the loss of unstable and/or the accumulation of toxic intermediates, and alter enzymatic activities and/or ligand-binding affinities (9Pareek V. Sha Z. He J. Wingreen N.S. Benkovic S.J. Metabolic channeling: Predictions, deductions, and evidence.Mol. Cell. 2021; 81: 3775-3785Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar). The identification of new multienzyme metabolic complexes, or “metabolons”, revealing their composition and modes of regulation, promises to re-shape our understanding of basic cellular metabolic processes. De novo purine biosynthesis (DNPB) is an ancient and highly conserved pathway, whose products are essential for nearly all aspects of cell biology (10Caetano-Anolles G. Kim H.S. Mittenthal J.E. The origin of modern metabolic networks inferred from phylogenomic analysis of protein architecture.Proc. Natl. Acad. Sci. U. S. A. 2007; 104: 9358-9363Crossref PubMed Scopus (131) Google Scholar, 11Caetano-Anolles G. Yafremava L.S. Gee H. Caetano-Anolles D. Kim H.S. Mittenthal J.E. The origin and evolution of modern metabolism.Int. J. Biochem. Cell Biol. 2009; 41: 285-297Crossref PubMed Scopus (85) Google Scholar). In humans, DNPB generates inosine monophosphate (IMP) from small molecule precursors in 10 reactions catalyzed by six enzymes: synthetase phosphoribosylaminoimidazole carboxylase/succinocarboxamide synthetase and Y. M. Structural biology of the purine Sci. PubMed Scopus Google Scholar). by multiple of this pathway is to a multienzyme the of the first in is highly unstable S. J. Characterization and of an unstable intermediate in the de novo purine PubMed Scopus Google Scholar, J. J. of the of from to PubMed Scopus Google the of from to its are catalyzed by the enzyme is catalyzed by this and to the of the of is an of well an of and can negatively impact cell J. Biochem. PubMed Scopus Google Scholar, S. Y. M. F. J.W. cells are by through of J. PubMed Scopus Google Scholar, F. M. A. of de novo purine biosynthesis with an of Biol. 2015; Full Text Full Text PDF PubMed Scopus Google Scholar, R.G. by metabolite and mammalian of 2009; PubMed Scopus Google Scholar, S. M. S. A. Y. B. C. A. S. J. mechanisms of and cell cycle by of Natl. Acad. Sci. U. S. A. PubMed Scopus Google the enzyme the of and in a the of with the enzymes in the pathway may be important in regulating In to in studies the of to M. Benkovic S.J. of the of in and PubMed Scopus Google Scholar, The of the human bifunctional enzyme A of substrate Biol. Chem. Full Text Full Text PDF PubMed Scopus Google the of into by a DNPB metabolon may be for the to studies support for the purinosome DNPB enzymes, tagged with in HeLa cells, to in S. R. Benkovic S.J. of de novo purine in living PubMed Scopus Google Scholar). These also with and the on H. Kim D. H. H. Y. Y. Benkovic S.J. and of purinosomes with 2016; PubMed Scopus Google Scholar, S. Y. J.W. M. Benkovic S.J. for metabolic Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar, Kim D. C. Benkovic S.J. of purine their to Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar). metabolic of the DNPB pathway, using the of products that is with the intermediates, evidence for intermediate channeling V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google Scholar). In is by metabolism enzymes to DNPB and Metabolic the purine and by channeled suggesting purinosomes can and with the the the of the reduced of the cofactor the for DNPB S.J. D. metabolism in the and mitochondria of mammalian Cell. Full Text Full Text PDF PubMed Scopus Google Scholar, M. Y. of flux for loss of the 2016; Full Text Full Text PDF PubMed Scopus Google the synthetase is to be part of the purinosome to the of and to the effective of into DNPB is have that DNPB enzymes can be suggesting may G. D. H. De novo purine in of the enzymes and of the Biol. Chem. Full Text PDF PubMed Google Scholar, D. D. De novo purine in human of the enzymes and of the Biol. Chem. Full Text PDF PubMed Google Scholar, C. B. S. F. K. G. J. A. K. S. G. Z. of ancient 2015; PubMed Scopus Google evidence for PPIs between DNPB enzymes and interactions between and DNPB enzyme have not Thus, a identification of the purinosomes and its can the assembly of the purinosome through the of or purinosomes only the cells are in purine-depleted or are their present regardless of the of the DNPB Here, we probe the PPIs of the bifunctional enzyme which reactions and of the DNPB pathway. PAICS is in multiple V. S. S. in multiple through of A PubMed Scopus Google Scholar, C. and metabolic of with and Commun. 2016; PubMed Scopus Google Scholar, M. S. S. J. S. and of a de novo purine in 2017; PubMed Scopus Google Scholar, T. S. a human 2017; 8: PubMed Scopus Google Scholar, S. Y. S. J. Y. Z. Z. of highly PAICS in PubMed Scopus Google Scholar, S. B. M. Kim A. C. S. U. a purine metabolic is in and the of Scopus Google suggesting may have a in the flux of the DNPB pathway to the purine demand of cells. We deploy two orthogonal approaches: bimolecular fluorescence to probe PPIs in intact living cells, and co-immunoprecipitation the molecular probe for We that PAICS interacts with all other DNPB enzymes with the exception of the first enzyme in the pathway. PAICS also interacts with MTHFD1, whose activity essential cofactor for DNPB. we these interactions in cells both purine-rich and purine-depleted conditions, the at least a partial is present regardless of the DNPB of the We also that the tagging of PAICS on its C terminus in an enzyme whose activities are whose PPIs are crPAICS cells by the of crPAICS with tagged a of PAICS protein with this metabolic signatures of channeled DNPB to that of reduced DNPB flux and an of between and our results demonstrate the of PPIs between DNPB for the assembly of the purinosome and provide evidence that the disruption of these PPIs a impact on the of this pathway. studies the in HeLa cells with mitochondria and with the We are by the that PPIs between DNPB enzymes may on these and not cell We to probe PPIs between DNPB enzymes in live, intact cells. In are tagged with of a these are into to an intact protein and PubMed Scopus Google Scholar). We two of J. and proteins with protein and PubMed Scopus Google a and A of from by a a in a K. T. K. K. K. of and interactions by bimolecular fluorescence using a new of PubMed Scopus Google Scholar, K. K. A novel of to interactions by in and in vivo bimolecular fluorescence Biol. PubMed Scopus Google Scholar, S.M. of assembly in of J. PubMed Scopus Google Scholar). of for the of PPIs Kim of of proteins in living PubMed Scopus Google Scholar). We by tagging all six DNPB enzymes with by a a and an or We also where the between the and the DNPB enzyme is with a H.J. Kim of a from in human cell and Thus, for we can a to the of that are this can be a of DNPB enzyme for of are with a protein with a large for and Chem. PubMed Scopus Google into which is well from in both and a we cells and the results with We the the fluorescence in cells with of have PAICS tagged on its terminus with we to with and In all the fluorescence is and the of PAICS with is that of PAICS with enzymes and of in the DNPB pathway. the first enzyme in the DNPB pathway, to a We the in both purine-depleted where DNPB be and in HeLa grown in purine-rich where the cellular purine demand is by the in both we We also that the tagging PAICS at its C terminus with in with and whereas the with PAICS with PAICS is known to a this the of PAICS with other DNPB enzymes may be negatively PAICS is on its C Our results that PAICS to with multiple DNPB enzymes. the of to these interactions suggest are not only PPIs whose are or the of can be to in using not evidence for PPIs between enzymes, our results suggest a on may to the of at on DNPB enzymes from cells H. S. S. Y. Benkovic S.J. is in the assembly of the Natl. Acad. Sci. U. S. A. 2013; PubMed Scopus Google Scholar). to the large of enzymes We to cell that and StrepTag-labeled PAICS at or the PAICS in V. M. V. S. M. de novo purine in HeLa cells in accumulation of enzyme and purinosome 2016; PubMed Scopus Google using a PAICS-knockout HeLa cell crPAICS M. V. J. D. D. V. T. M. analysis of purine de novo biosynthesis PubMed Scopus Google Scholar). These cells purine for purine cell cycle and cell We the of and StrepTag-labeled PAICS to these cells in purine-depleted We that PAICS tagged at terminus can crPAICS in purine-depleted that PAICS tagged at terminus its essential enzymatic In both in purine-depleted of cells that on purine for the PAICS into the a DNPB pathway. this we two cell and In our the of PAICS into the crPAICS is by is a and we that the not of the reintegrated a of analysis of these cell PAICS in is only that of HeLa that in is of the DNPB enzymes are the two cell of PAICS in cells using cell PAICS HeLa and is from cell The large of cells PAICS at with to the from only a small to that in the other of cells PAICS to that in In all two cell and was and evidence of in the cells. Thus, for the in these cell PAICS at its PAICS Our results show that PAICS at only a of that in HeLa is to cell and in purine-depleted also suggest that of a DNPB pathway, is in we of cell from these cell by we and We not the of conditions, and we to the of the with and whose are to that of we in the of these enzymes in cells grown purine-depleted purine-rich the other we not the of DNPB enzyme for a PAICS the that the PPIs of may be to the of the PAICS we and in cells in purine-depleted and their to We and of PAICS not in their to These results suggest that tagging of its with our In to the DNPB enzymes, we also PAICS with other enzymes whose activities to the DNPB pathway. In we PAICS interacts with MTHFD1, which supplies the essential for and whose in the purinosome is on the of in channeled DNPB V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google Scholar). We with PAICS in both purine-depleted and purine-rich conditions, was the with DNPB enzymes, we not in from We the disturbed PPIs of C-terminally tagged by our and with disturbed DNPB activity. The a of PAICS that this is to be the We metabolic of these two cell to their DNPB cells grown in and and by with and for and In cells, to the of is by the bifunctional and synthetase to and our conditions, mitochondria are the of and channeled purine synthesis, are to Thus, channeled DNPB flux from purinosomes can be from by DNPB enzymes in the by the in their V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google Scholar). of channeled DNPB is the of to to V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google Scholar). in the native PPIs the purinosome may also alter the partitioning of between AMP and GMP. of both and in AMP their AMP was metabolic cell by and or which disrupts purinosomes and to the of the channeled and pools of DNPB and products We the of and of the AMP and In the composition of AMP was from that of and the on a the of channeled DNPB The between AMP and is with the that this disrupts purinosome Thus, in a DNPB pathway signatures of channeled to that in HeLa However, we a analysis for we by the of DNPB in this cell the of of the DNPB and we of cells. The pools of DNPB and products in to that in purine metabolism is to the we the and in and Our analysis that the of for well in the two PAICS The two cell in their partitioning of to AMP and GMP. Here, a to AMP a of channeled DNPB V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google to We that and with the exception of exhibit of the DNPB enzymes analysis for enzymes that to AMP or that a for to of and Thus, the metabolic in that the of PAICS are to the of PAICS In this we have using orthogonal that the DNPB enzyme PAICS interacts with other DNPB enzymes and well with MTHFD1, which provides the essential cofactor for this pathway. We have this both in live, intact cells and in cells that have not to PAICS from crPAICS interactions be to the of the our results provide molecular evidence in support of the purinosome our results the interactions of PAICS with other DNPB enzymes are present in both purine-rich and purine-depleted the that in the cells the purinosome is from that are present regardless of the DNPB of the be to this is a only in cells or is to human cell is to on our that PAICS may have a in the purinosome through its interactions with the other DNPB enzymes. However, we that PAICS at a of that in HeLa is to support cell and in purine-depleted is that in cell PAICS is at a in large of is to cellular purine demand and that the of PAICS in with other DNPB enzymes we the that the interactions may be multiple DNPB enzymes, with are part of the large PAICS all other enzymes, not with Thus, is to interactions may be revealed we the DNPB enzymes. our with a to we have employed to the DNPB enzymes, A of PPIs between the six DNPB enzymes important the molecular of the The we only the of our a interactions the bimolecular on the of enzymes, and have Y. T.C. D. D. A. evolution of Chem. Biol. PubMed Scopus Google Scholar, J. M. K. J. a to the composition of protein Commun. 2017; 8: PubMed Scopus Google Scholar, T.C. S. T. T. C. in Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar, T.C. in mammalian cells with and PubMed Scopus Google Scholar). a a enzyme is only PAICS and are is to probe PPIs in the to the of between DNPB enzymes, from the assembly of the the of these may important for cells DNPB in response to varying purine We that the tagging of PAICS at its C terminus its Our tagged PAICS our metabolic that to a metabolic In we metabolic signatures of channeled DNPB to that in in the DNPB flux is a PAICS protein and the partitioning of the AMP is also The C-terminally tagged PAICS can crPAICS cells its essential enzymatic activities are is tagging PAICS at its may negatively its enzymatic and/or we that on Y. structure of the human bifunctional enzyme PAICS in purine Biol. 2007; PubMed Scopus Google Scholar, J. J. M. T. of human PAICS substrate and of an Biol. Chem. Full Text Full Text PDF PubMed Scopus Google the of PAICS not to known ligand-binding or The structure of the PAICS a to tagging of PAICS its The PAICS an whereas the of PAICS to be highly and from the of the its on a that on the the of the PAICS a a small may provide to The of cell and on the of the PAICS in both by the into the crPAICS the of reintegrated PAICS on the of in purine-depleted whose PAICS activity can support cell that The that PAICS at a the of the DNPB enzymes are these two cell and the metabolic of and its reduced of is to PAICS and that the in C-terminally tagged PAICS is its can be by enzyme a can for the enzyme In our metabolic we evidence that C-terminally tagged PAICS in we not an accumulation of DNPB of PAICS in to the other for where enzymes their in a channeled from an where only a of enzymes are in and to the an enzyme whose is whose interactions are to by the of the for the reduced The of the purinosome is that the is essential for DNPB is to that the the activities of DNPB enzymes, are We the of the two crPAICS cell provides a of this of the purinosome Our that the enzymatic activities of PAICS are evidence that its interactions with other DNPB enzymes have and that this is correlated with reduced DNPB intermediate pools and a perturbed partitioning of AMP and GMP. our results not only provide molecular in the of in support of the purinosome we also that the disruption of these interactions a impact on the of this important metabolic pathway. HeLa cells on in with and or the was with and the cells are with two tagged and enzymes interactions we to with a a cells are an with a that the cells at in a are using a in both and A of enzyme that these are to to with to 10 cells we the fluorescence of all whose is by HeLa whose is are The is using in in the we all the The in the of all the in crPAICS cells are in at a of and in with and for The cells are to purine-depleted and with of or using the is with purine-depleted with 10 to 10 the cells are and into and in purine-depleted for to this of the crPAICS cells and of cells also the PAICS reintegrated into the the cell have the cell are to be and the cells are in purine-depleted to the loss of reintegrated PAICS to and of and are in at a of in purine-depleted The is with on and and of cells is the cells are in in for at The cells are in of for a of the the cells are in for and using at we of The are in to and cell using a are in and cells are at a of to cells are in in for at by in in for 10 at cells are with and and with in at is also to cellular to the cells are with at in in at with in cells are with 10 and at with in cells are with and with in at and in using to cells, using to of cells. are and The cell is in and with and The cell is by at and the protein is by is at using at a of The are with and by in of are for the proteins are to the The is in in for at with at and at The are on using substrate A of and is in the purine flux and the channeled DNPB flux by mitochondria-associated into DNPB and products and metabolite was in V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google Scholar). in purine-depleted to cells and to for in the was to purine-depleted and with The was with to a of and was for or which cells with by the or by by and the of of is known to of all enzymatic reactions and of the from the purinosome to of the purinosome and of all the in the was in V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google and The in with that the from 10 to of on a using a and acid to the and by mass by from of the and can be in V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google and the and for all the are in The in the mass of and into is in In or and the for all the was for In a we that the PAICS HeLa cell for the cell PAICS activity the of PAICS and crPAICS cells or The purine In the cell by of the all the and products of PAICS and that the reintegrated PAICS We also the of in other that on the metabolism we the in and both at the and and on the of in and using the PAICS cell exhibit metabolic of channeled DNPB by mitochondria-associated the of the and that a model for DNPB V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google The model the of to the in the and The model of the and was with the in and of to the for all the the in in and the and AMP in V. H. Benkovic S.J. and mass channeled de novo purine in PubMed Scopus Google Scholar). the pools are to the in is a is for to in of pools of the between the channeled and pools to and also the the and in this are The that have of with the of this and of for of cell HeLa PAICS cells a from and J. V. and S. J. B. J. Z. and V. J. and V. J. V. and S. J. B. J. V. and S. J. B. and work was by of J. Z. and V. support from the of the The is the of the and not the of the of

Topics & Concepts