Structural and functional studies of Arabidopsis thaliana legumain beta reveal isoform specific mechanisms of activation and substrate recognition

Abstract

The vacuolar cysteine protease legumain plays important functions in seed maturation and plant programmed cell death. Because of their dual protease and ligase activity, plant legumains have become of particular biotechnological interest, e.g. for the synthesis of cyclic peptides for drug design or for protein engineering. However, the molecular mechanisms behind their dual protease and ligase activities are still poorly understood, limiting their applications. Here, we present the crystal structure of Arabidopsis thaliana legumain isoform β (AtLEGβ) in its zymogen state. Combining structural and biochemical experiments, we show for the first time that plant legumains encode distinct, isoform-specific activation mechanisms. Whereas the autocatalytic activation of isoform γ (AtLEGγ) is controlled by the latency-conferring dimer state, the activation of the monomeric AtLEGβ is concentration independent. Additionally, in AtLEGβ the plant-characteristic two-chain intermediate state is stabilized by hydrophobic rather than ionic interactions, as in AtLEGγ, resulting in significantly different pH stability profiles. The crystal structure of AtLEGβ revealed unrestricted nonprime substrate binding pockets, consistent with the broad substrate specificity, as determined by degradomic assays. Further to its protease activity, we show that AtLEGβ exhibits a true peptide ligase activity. Whereas cleavage-dependent transpeptidase activity has been reported for other plant legumains, AtLEGβ is the first example of a plant legumain capable of linking free termini. The discovery of these isoform-specific differences will allow us to identify and rationally design efficient ligases with application in biotechnology and drug development. The vacuolar cysteine protease legumain plays important functions in seed maturation and plant programmed cell death. Because of their dual protease and ligase activity, plant legumains have become of particular biotechnological interest, e.g. for the synthesis of cyclic peptides for drug design or for protein engineering. However, the molecular mechanisms behind their dual protease and ligase activities are still poorly understood, limiting their applications. Here, we present the crystal structure of Arabidopsis thaliana legumain isoform β (AtLEGβ) in its zymogen state. Combining structural and biochemical experiments, we show for the first time that plant legumains encode distinct, isoform-specific activation mechanisms. Whereas the autocatalytic activation of isoform γ (AtLEGγ) is controlled by the latency-conferring dimer state, the activation of the monomeric AtLEGβ is concentration independent. Additionally, in AtLEGβ the plant-characteristic two-chain intermediate state is stabilized by hydrophobic rather than ionic interactions, as in AtLEGγ, resulting in significantly different pH stability profiles. The crystal structure of AtLEGβ revealed unrestricted nonprime substrate binding pockets, consistent with the broad substrate specificity, as determined by degradomic assays. Further to its protease activity, we show that AtLEGβ exhibits a true peptide ligase activity. Whereas cleavage-dependent transpeptidase activity has been reported for other plant legumains, AtLEGβ is the first example of a plant legumain capable of linking free termini. The discovery of these isoform-specific differences will allow us to identify and rationally design efficient ligases with application in biotechnology and drug development. The plant cysteine proteases of the legumain family (C13 family, EC 3.4.22.34) have an important role in processing and maturation of seed storage proteins within the vacuole and, therefore, are also referred to as vacuolar processing enzymes (VPEs) (1Hara-Nishimura I. Takeuchi Y. Nishimura M. Molecular characterization of a vacuolar processing enzyme related to a putative cysteine proteinase of Schistosoma mansoni.Plant Cell. 1993; 5 (8312744): 1651-165910.2307/3869746Crossref PubMed Scopus (135) Google Scholar). Plant legumains are structurally related to the mammalian caspases and exhibit a strong substrate sequence preference for cleavage after asparagine and, to a lesser extent, aspartate residues (2Abe Y. Shirane K. Yokosawa H. Matsushita H. Mitta M. Kato I. Ishii of and in protein sequence 1993; PubMed Google K. and characterization of proteinase an PubMed Scopus Google Scholar). are also referred to as the to legumain isoform is Arabidopsis thaliana for legumains and and other to I. I. M. legumains have a role in seed and PubMed Scopus Google Scholar). Plant legumains are in and consistent with their the seed and or and K. and their functions in Plant PubMed Scopus Google K. M. Kato Nishimura M. I. vacuolar processing is in seed the of seed Cell. PubMed Scopus Google M. M. The vacuolar processing enzyme Plant PubMed Scopus Google K. K. I. processing enzymes in the plant PubMed Scopus Google Scholar). legumains are in and have been to in plant programmed cell and for the are in K. I. processing enzyme in plant programmed cell Plant PubMed Scopus Google Scholar). legumains Arabidopsis thaliana legumain isoform β (AtLEGβ) important functions in the processing and maturation of seed storage proteins within storage K. M. Y. M. Kato K. M. M. M. Nishimura M. I. processing enzymes are for processing of seed storage proteins in Arabidopsis PubMed Scopus Google Nishimura M. I. processing enzyme of that to the PubMed Scopus Google Scholar). The of legumains is in Arabidopsis legumain and to seed storage proteins protein in the of the vacuolar processing enzyme family of cysteine Cell. PubMed Scopus Google Scholar). AtLEGβ for and γ that AtLEGβ is the in protein processing in K. M. Y. M. Kato K. M. M. M. Nishimura M. I. processing enzymes are for processing of seed storage proteins in Arabidopsis PubMed Scopus Google Scholar). of AtLEGβ the and proteins K. and their functions in Plant PubMed Scopus Google K. M. Y. M. Kato K. M. M. M. Nishimura M. I. processing enzymes are for processing of seed storage proteins in Arabidopsis PubMed Scopus Google protein in the of the vacuolar processing enzyme family of cysteine Cell. PubMed Scopus Google K. of cysteine the for the of storage and after in PubMed Scopus Google Scholar). of plant legumains peptide ligase and activity proteins and mechanisms of PubMed Scopus Google PubMed Scopus Google Y. Y. is an ligase peptide and PubMed Scopus Google K. by a PubMed Scopus Google I. of peptides by a PubMed Scopus Google I. I. in protein PubMed Scopus Google a efficient protein PubMed Scopus Google Scholar). we that the ligase activity H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google Scholar). However, is still is also true for the other thaliana legumain the seed peptides are important for plant Y. Y. is an ligase peptide and PubMed Scopus Google K. by a PubMed Scopus Google and their processing are for cyclic peptides in PubMed Scopus Google activities of PubMed Scopus Google Scholar). the in has and and the and their processing are for cyclic peptides in PubMed Scopus Google I. of proteins in PubMed Scopus Google Scholar). peptides are to in pH and for biotechnological and drug design Y. a for drug PubMed Scopus Google M. K. as a molecular for drug PubMed Scopus Google as PubMed Scopus Google Scholar). in is by is a in the specificity, and of different plant legumain to the discovery of a of ligase activity and a that allow us to ligase activity sequence a efficient protein PubMed Scopus Google K. H. Molecular for the of cyclic peptides by plant PubMed Scopus Google Scholar). However, to these ligase activity of different legumain is of plant legumains that are as of a with activity and a legumain and activity that are by an activation peptide the a efficient protein PubMed Scopus Google H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google H. of Arabidopsis thaliana legumain and processing of and PubMed Scopus Google K. Y. for by Google Scholar). is in mammalian and plant legumains, the activation of thaliana isoform significantly that of legumain H. of legumain and resulting in a and activity PubMed Scopus Google H. and structural legumain its activation and PubMed Scopus Google Scholar). is present in an dimer state that is by and pH and protein concentration H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google Scholar). we have that to the monomeric of the a two-chain intermediate state. cleavage the of the within the and is by protein is after an cleavage within the an two-chain intermediate pH the dimer to monomeric two-chain the and to the The of the dimer and two-chain the of a activation of plant legumains, and However, the of these isoform-specific differences in activation are to with still Here, we present the crystal structure of to the discovery of a activation in to Combining structural and biochemical we for the first that plant legumains isoform-specific autocatalytic activation mechanisms and of activity and we that seed AtLEGβ is an ligase capable of peptide AtLEGβ ligase activity is to peptide cleavage the efficient of free and termini. AtLEGβ is the first example of a plant legumain for we the of free peptide termini. of isoform-specific differences in plant legumains and their in plant the to rationally design peptide ligases with in biotechnology and drug development. isoform-specific differences different we determined the crystal structure of to a of The of the isoform an and a legumain and activity with and The and are by an activation peptide that the the structure of the structure of the two-chain by a of However, the Whereas the of AtLEGβ and γ with an of we differences in the with a of with is also in with a sequence of the β and γ with the we an isoform-specific the of the is also in legumain and and structure determined a The by the for a is in of of is in The structure determined a The by the K. and PubMed Scopus Google Scholar). in a is monomeric in crystal is in and is with a in The of the is with an of dimer in the with two-chain to different of the the as in two-chain is by a hydrophobic that is by the in is by a the and hydrophobic the the and the to the in and are by the of the as in and of an in is of of and monomeric state of as a important of is that in a dimer state in is by dimer the activation and activity of H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google Scholar). in the crystal structure of we in the to in dimer by However, these by different and to an of the within the and and of the β and γ dimer revealed isoform-specific The dimer is by and and a cyclic protein the The and a that is stabilized a hydrophobic by as as and and hydrophobic is stabilized by a of the and and and of the the in the crystal structure by the the and stabilized by as as by a hydrophobic of the with in the and The hydrophobic of the as by the residues important for dimer in However, a dimer of residues in AtLEGβ and that will of and these that the is and in we pH the of a to legumain a in the crystal structure of with AtLEGγ, we that the to the nonprime substrate binding in a to we in mammalian and and the is substrate the in a state. Additionally, we a the of the is binding the in an and cleavage of the and substrate to the and in the crystal structure of thaliana legumain isoform is the plant that the binding a in plant legumain activity Additionally, is by the ionic with to the to two-chain and mammalian the is stabilized by and the of the AtLEGβ a vacuolar is and, therefore, in the an to the of we the of and we that the has a hydrophobic with by and are also in and is in to the has a is by and a hydrophobic to the substrate binding the and in in of a of the The residues in other are in and with the differences in state, these to the that will differences in the activation and pH stability of the thaliana legumain Because the the and the in is hydrophobic in we that its activation rather of an activation that the activation of AtLEGβ that of mammalian with activation pH we that autocatalytic activation that will the to to the we of the pH that the is pH pH and However, the to us that AtLEGβ a two-chain state, cleavage and the to we pH we a of the two-chain state and two-chain intermediate state of AtLEGβ in after we identify autocatalytic cleavage and the in and to to pH we cleavage the and Because of the of the cleavage after is to in with the cleavage is the and and in with processing allow the of the as in mammalian legumain H. of legumain and resulting in a and activity PubMed Scopus Google Scholar). and are within the the and is within plant legumains, and cleavage within sequence to for the autocatalytic activation of legumain Nishimura M. I. processing enzyme is by of the and PubMed Scopus Google Scholar). activation of AtLEGβ pH with cleavage the and in the of the the and the the for to cleavage the and we processing the of the Here, is important to that an by a that AtLEGβ capable of after the within the and the as by a an and experiments, we that the cleavage the cleavage by the two-chain However, processing within the pH we that the cleavage is by the cleavage is a as the sequence is present in the we the of and we that differences of their pH stability profiles. we the stability of we a stability pH is different and mammalian the stability of the is pH H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google H. and structural legumain its activation and PubMed Scopus Google Scholar). we that AtLEGβ pH and, the a in pH stability pH is in to and also mammalian the is pH However, the hydrophobic and in legumain and a stability the the the H. and structural legumain its activation and PubMed Scopus Google Scholar). pH the of the of the is for by the legumain and AtLEGγ, the stabilized by of the the stability pH The in AtLEGβ the a strong pH is for in with the pH The of the with the and the of these in and is the pH that is also for the that the in is present in an H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google K. The protease autocatalytic and the structural for ligase Google Scholar). the of the by the and we the crystal structure of to the AtLEGβ state. we the of AtLEGβ and we that their is AtLEGβ exhibits a a that is by 5 and The protein that activity, specificity, activation and PubMed Scopus Google Scholar). AtLEGβ the and the nonprime substrate binding The a plant that is the that is the with mammalian legumain of or in a of protein that the is also for we peptide and with for and and role of peptide in protein PubMed Scopus Google Scholar). are in the substrate binding The is the and the as the for the is of the of the the we that the residues and also with the related the residues the and to in The of the substrate and activity of AtLEGβ and However, we the activity the we a activity for AtLEGβ with γ Because the of the residues in β and we in activity to a rather differences in substrate the we differences the and the nonprime and in sequence and in a in a rather in AtLEGβ and K. M. Kato Nishimura M. I. vacuolar processing is in seed the of seed Cell. PubMed Scopus Google and these differences a of the of the we the crystal of two-chain and and their we that the of the substrate and of the state of we that a of substrate However, the different in AtLEGγ, we of the the and the a the crystal we in to Whereas AtLEGβ an broad to the the as in and and the and the as the for to us that with to AtLEGβ with of we determined for AtLEGβ and γ the we binding to for AtLEGβ we AtLEGγ, and AtLEGγ, for that the in activity AtLEGβ and γ by differences in substrate we a substrate of we a in activity for AtLEGβ and γ the is the the in we an activity for pH is also in with the pH stability of the these us that the and as a we plant legumain in the Whereas the structural in we differences the is in and sequence and a and these that the is a of the activity of legumains, to The of the for legumain activity is by a as a of ligase activity K. H. Molecular for the of cyclic peptides by plant PubMed Scopus Google Scholar). the substrate of we experiments, peptides as substrate of the and of proteases by peptide and PubMed Scopus Google M. K. of protease by peptide and Cell. PubMed Scopus Google Scholar). Here, we a peptide that an by with for different pH we a strong preference for in the pH we also an of cleavage residues time that with in are The substrate preference also pH the of with pH with the of the and with for legumain H. and structural legumain its activation and PubMed Scopus Google Scholar). we a preference for residues in the the that is Additionally, we a preference for in the has been to the in legumain ligase Y. Y. is an ligase peptide and PubMed Scopus Google Scholar). these are in with the of the and binding Whereas the binding is and with the the a H. of Arabidopsis thaliana legumain and processing of and PubMed Scopus Google Scholar). residues in the will the binding of the and residues the and binding still to allow efficient cleavage of the peptide The of the is by and the by is in plant legumains that have been structurally The is and with mammalian legumain a the and, M. M. cleavage by of PubMed Scopus Google Scholar). has the by that also will have a the the of for substrate specificity, we the AtLEGγ, has a the we the nonprime and substrate binding that is the the the we the substrate of AtLEGβ protein of the thaliana of as a substrate with AtLEGγ, or free with different and cleavage determined the and of in Cell. PubMed Scopus Google Scholar). we peptides significantly after with AtLEGβ pH and and to cleavage in proteins we a preference for the we a preference for and residues in the that the of the peptide is is by a Additionally, we also a preference for the and and in the we a preference for hydrophobic in the AtLEGγ, we significantly peptides cleavage in proteins that a cleavage in with peptide the of of the cleavage in proteins by in proteins of AtLEGβ and of proteins by and the activity of we with different peptides and the of the cyclic we that AtLEGβ the peptide to the and to cyclic pH is in with the reported pH of legumain ligase activity H. of Arabidopsis thaliana legumain and processing of and PubMed Scopus Google H. and of an peptide ligase in PubMed Scopus Google Scholar). the an the we a of is than that with AtLEGγ, in H. of Arabidopsis thaliana legumain and processing of and PubMed Scopus Google Scholar). the by as is the in AtLEGβ still to peptide in to the of we the with is the sequence for we a Y. Y. is an ligase peptide and PubMed Scopus Google us that for the is peptide in these we also AtLEGβ to and and of the protease AtLEGβ to the free and the cyclic that AtLEGβ is a transpeptidase also a ligase and the peptides with or the cleavage of the of the to free pH pH a of a legumain capable of linking free peptide termini. is a for the of the is by the the and is with structural that the also in plant The crystal of and a dimer state that by the and the However, in these with activation rather with is as protein the will to the Additionally, is a two-chain intermediate state, is to we show for the first time that are isoform-specific differences in the activation and activity of thaliana we that is monomeric in autocatalytic activation of the mammalian also a latency-conferring dimer state that in the crystal we dimer is and and, in we that the is rather hydrophobic and in the stability of AtLEGβ and mammalian legumains AtLEGβ autocatalytic cleavage of the and in the of the in mammalian legumain H. and structural legumain its activation and PubMed Scopus Google Scholar). Whereas cleavage pH cleavage the of the is to pH is in with the of the and Additionally, pH the ionic that is linking the of the to the will will the of the an a intermediate state, the of the by However, as we a in experiments, will a that of will AtLEGβ to activation will the two-chain state in the other the in the AtLEGβ crystal a role in substrate these that AtLEGβ and γ of of plant legumains, their and also with to different mechanisms of and plant legumains are for after However, we show that differences in the nonprime substrate binding different legumain will substrate by the and time of substrate that the encode The and significantly in plant legumains, the within the plant legumain in substrate with peptide are to differences of the and concentration of peptide However, the substrate is become time experiments, we show that is a legumain substrate pH a as a to substrate to a cleavage in a and example the autocatalytic activation of cleavage after residues the and is to that the differences in the the plant legumains will have an cleavage rather than sequence with these we AtLEGβ and γ cleavage in protein in the as a of ligase activity K. H. Molecular for the of cyclic peptides by plant PubMed Scopus Google Scholar). a in that with an in ligase activity. However, AtLEGβ and γ encode are we show that the sequence also the of in we that the is a of protease activity. Because protease and ligase activities are the an of ligase activity. the of the nonprime substrate is the of the ligase substrate in a substrate of nonprime also in of cyclic and also with that has a nonprime binding for is a ligase with broad substrate we that peptides are ligase resulting in efficient of cyclic with the that nonprime are to a ligase substrate the in the the the nonprime substrate of substrate by the that is the is true is Additionally, the time of the is of the cyclic and the cyclic However, and in of cyclic that will as a substrate pH to as a free in a substrate pH The will also by a a substrate with to will have a than free the substrate the for binding and the substrate binding we that AtLEGβ and γ encode substrate binding differences in AtLEGβ and γ by their different nonprime substrate binding for and to the of ligase activity, as a for ligase activity a efficient protein PubMed Scopus Google Scholar). to in an enzyme with ligase activity. Because a the AtLEGβ the isoform differences in ligase activity. the sequence is to the and to for ligase activity, is in thaliana legumains K. Y. for by Google Scholar). However, to is is a of the nonprime binding and is different in and the differences in the we that is rather the of the nonprime substrate binding that substrate and ligase activity. The of AtLEGβ to free is also a biotechnological of as will allow us to the of cleavage still with the of or also that free is a in plant legumain that a of different legumain is that is an enzyme present in we show that the two-chain state in is with to ligase activity, as is pH ligase activity is Because two-chain AtLEGβ has the pH stability as with a pH stability two-chain AtLEGβ will a However, differences in substrate and we that two-chain legumain with the still present exhibits activity rather than activity H. and structural legumain its activation and PubMed Scopus Google Scholar). The activity is structurally by the the we a preference for aspartate and residues in the of protein with a of residues and activity of two-chain However, has to with as the in for and the or two-chain AtLEGβ activity will and the of The Arabidopsis thaliana vacuolar processing enzyme isoform β (AtLEGβ) the Arabidopsis as a we an the the and and The an sequence for in the and an by a we a the is the of by a PubMed Scopus Google Scholar). that the of the cyclic the to a the of the of the and an the and The for of by The the and cell as H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google Scholar). and as H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google H. of legumain and resulting in a and activity PubMed Scopus Google Scholar). The protein in a of pH and by the the a of in a of and within a protein concentration of autocatalytic we a in a and in and to the The with a a of and an of and K. and the PubMed Scopus Google of the and PubMed Scopus Google Scholar). by molecular PubMed Scopus Google the crystal structure of two-chain with the sequence of of in K. for molecular PubMed Scopus Google and in for structure PubMed Scopus Google a and and structure to the the with of application to and the PubMed Scopus Google after pH an for the of PubMed Scopus Google Scholar). the pH of of we a concentration of in of pH pH and for by the of to AtLEGβ a we the in a of pH and for of by AtLEGβ a in a of pH and by the in H. structure of plant legumain a two-chain state with activity Cell. PubMed Scopus Google The activity of AtLEGβ the and in of pH and of the substrate after the enzyme in an in determined of AtLEGβ or γ with of the substrate in and the state of of a in a of pH and the state of pH we an in of pH and as a the stability of and pH we the as H. K. H. and of dimer and PubMed Scopus Google Scholar). The of pH and as for the of protein by Scholar). to an proteins an a cell The with with the with with of protein by the with an the substrate of AtLEGβ and we of protease cleavage peptide of the and of proteases by peptide and PubMed Scopus Google M. K. of protease by peptide and Cell. PubMed Scopus Google Scholar). The peptide as K. H. of the PubMed Scopus Google Scholar). The with a of in pH The peptide with proteases in of and pH and after and of with and and with and for and with pH for and and thaliana M. K. M. Nishimura M. I. processing enzyme is for cell in Arabidopsis PubMed Scopus Google the Arabidopsis The with a in pH and protease The by for 5 The protein concentration determined by the as a AtLEGγ, or to the a of in the and in for The by of The by for the of protein in in the of and PubMed Scopus Google and in pH The protein determined the The with 5 for by with for and by of for The with and and or and of the of and for The with pH and for of protein by and in pH The with in a for of peptides to the of in Cell. PubMed Scopus Google Scholar). with a of in with for for the The with a concentration of and for 5 the a The peptides peptides the with with the first and in the to a for to The in a with of of a of and of with a for and a of to a as M. M. M. a processing the PubMed Scopus Google Scholar). with the in in a an of The for with of for the with the 5 for and for with of the time with and a of by and a of to peptide an of The for PubMed Scopus Google experiments, the with and peptide or and as and as peptides that a in after protease with the or present in the putative cleavage to putative peptides of the and to the cleavage the cleavage as M. K. of protease by peptide and Cell. PubMed Scopus Google and as K. K. of protein by PubMed Scopus Google Scholar). experiments, the thaliana with of for that cleavage by as Further by or of peptide or as and or or as Further and with the peptides and as H. of Arabidopsis thaliana legumain and processing of and PubMed Scopus Google Scholar). of the peptide and AtLEGβ in a of and pH or pH for the and by The and structure in have been with the with the have been with the I. Y. H. of the and its related PubMed Scopus Google with the for the and for the are within the for peptide with Arabidopsis thaliana isoform Arabidopsis thaliana legumain isoform and activity activation peptide vacuolar processing enzyme cyclic protein stability

Topics & Concepts