Importance of the P4′ Residue in Human Granzyme B Inhibitors and Substrates Revealed by Scanning Mutagenesis of the Proteinase Inhibitor 9 Reactive Center Loop

The cytotoxic lymphocyte serine proteinase granzyme B induces apoptosis of abnormal cells by cleaving intracellular proteins at sites similar to those cleaved by caspases. Understanding the substrate specificity of granzyme B will help to identify natural targets and develop better inhibitors or substrates. Here we have used the interaction of human granzyme B with a cognate serpin, proteinase inhibitor 9 (PI-9), to examine its substrate sequence requirements. Cleavage and sequencing experiments demonstrated that Glu340 is the P1 residue in the PI-9 RCL, consistent with the preference of granzyme B for acidic P1 residues. Ala-scanning mutagenesis demonstrated that the P4-P4′ region of the PI-9 RCL is important for interaction with granzyme B, and that the P4′ residue (Glu344) is required for efficient serpin-proteinase binding. Peptide substrates based on the P4-P4′ PI-9 RCL sequence and containing either P1 Glu or P1 Asp were cleaved by granzyme B (kcat/Km9.5 × 103 and 1.2 × 105s−1m−1, respectively) but were not recognized by caspases. A substrate containing P1 Asp but lacking P4′ Glu was cleaved less efficiently (kcat/Km 5.3 × 104 s−1m−1). An idealized substrate comprising the previously described optimal P4-P1 sequence (Ile-Glu-Pro-Asp) fused to the PI-9 P1′-P4′ sequence was efficiently cleaved by granzyme B (kcat/Km 7.5 × 105 s−1m−1) and was also recognized by caspases. This contrasts with the literature value for a tetrapeptide comprising the same P4-P1 sequence (kcat/Km 6.7 × 104 s−1m−1) and confirms that P′ residues promote efficient interaction of granzyme B with substrates. Finally, molecular modeling predicted that PI-9 Glu344 forms a salt bridge with Lys27 of granzyme B, and we showed that a K27A mutant of granzyme B binds less efficiently to PI-9 and to substrates containing a P4′ Glu. We conclude that granzyme B requires an extended substrate sequence for specific and efficient binding and propose that an acidic P4′ substrate residue allows discrimination between early (high affinity) and late (lower affinity) targets during the induction of apoptosis. The cytotoxic lymphocyte serine proteinase granzyme B induces apoptosis of abnormal cells by cleaving intracellular proteins at sites similar to those cleaved by caspases. Understanding the substrate specificity of granzyme B will help to identify natural targets and develop better inhibitors or substrates. Here we have used the interaction of human granzyme B with a cognate serpin, proteinase inhibitor 9 (PI-9), to examine its substrate sequence requirements. Cleavage and sequencing experiments demonstrated that Glu340 is the P1 residue in the PI-9 RCL, consistent with the preference of granzyme B for acidic P1 residues. Ala-scanning mutagenesis demonstrated that the P4-P4′ region of the PI-9 RCL is important for interaction with granzyme B, and that the P4′ residue (Glu344) is required for efficient serpin-proteinase binding. Peptide substrates based on the P4-P4′ PI-9 RCL sequence and containing either P1 Glu or P1 Asp were cleaved by granzyme B (kcat/Km9.5 × 103 and 1.2 × 105s−1m−1, respectively) but were not recognized by caspases. A substrate containing P1 Asp but lacking P4′ Glu was cleaved less efficiently (kcat/Km 5.3 × 104 s−1m−1). An idealized substrate comprising the previously described optimal P4-P1 sequence (Ile-Glu-Pro-Asp) fused to the PI-9 P1′-P4′ sequence was efficiently cleaved by granzyme B (kcat/Km 7.5 × 105 s−1m−1) and was also recognized by caspases. This contrasts with the literature value for a tetrapeptide comprising the same P4-P1 sequence (kcat/Km 6.7 × 104 s−1m−1) and confirms that P′ residues promote efficient interaction of granzyme B with substrates. Finally, molecular modeling predicted that PI-9 Glu344 forms a salt bridge with Lys27 of granzyme B, and we showed that a K27A mutant of granzyme B binds less efficiently to PI-9 and to substrates containing a P4′ Glu. We conclude that granzyme B requires an extended substrate sequence for specific and efficient binding and propose that an acidic P4′ substrate residue allows discrimination between early (high affinity) and late (lower affinity) targets during the induction of apoptosis. cyotoxic lymphocyte reactive center loop proteinase inhibitor 9 high pressure liquid chromatography stoichiometry of inhibition 3-(3-cholamidopropyl)dimethylammonio-1-propanesulfonic acid N-(9-fluorenyl)methoxycarbonyl ε-dinitrophenyl aminobenzoyl Cytotoxic lymphocytes (CLs)1 play a key role in cell-mediated immunity, destroying foreign, virus-infected, or tumor cells (1Trapani J.A. Bird P.I. Sitkovsky M.V. Henkart P.A. Cytotoxic Cells: Basic Mechanisms and Medical Applications. Lippincott, Williams 18: 255-273Crossref PubMed Google Scholar). The molecular basis of granzyme-induced cell death is not fully understood. The current model is that perforin and granzymes are endocytosed by the target cell and that perforin eventually disrupts the endocytic vesicle, thus releasing granzymes into the cytoplasm (3Froelich C.J. Orth K. Turbov J. Seth P. Gottleib R. Babior B. Shah G.M. Bleackley R.C. Dixit V.M. Hanna W. J. Biol. Chem. 1996; 271: 29073-29079Abstract Full Text Full Text PDF PubMed Scopus (314) Google Scholar,4Pinkoski M.J. Hobman M. Heibein J.A. Tomaselli K. Li F. Seth P. Froelich C.J. Bleackley R.C. Blood. 1998; 92: 1044-1054Crossref PubMed Google Scholar). Caspase activation, loss of mitochondrial membrane potential, proteolysis of key housekeeping proteins, DNA and the of A key in is the serine granzyme B. is an proteinase with a preference for cleaving a that with caspases. the cytoplasm of a granzyme B induces death by at two is caspase and in DNA and The is caspase and of proteins such and F. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). loss of mitochondrial membrane The role of granzyme B in cell death is by lacking granzyme B. that are to DNA and death of target by granzymes Full Text PDF PubMed Scopus Google Scholar). are by granzyme B is to Turbov J. Seth P. Orth K. Froelich C.J. J. PubMed Scopus Google and is that also in cells by M. A.J. Bleackley R.C. J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google R. J. Immunol. PubMed Scopus Google Scholar). the granzyme B substrates within to cleaved Asp but not at sites recognized by caspases. B also in the of the M. Heibein J.A. M.J. Bleackley R.C. Biol. PubMed Scopus Google in of the DNA and in the of F. W. J. PubMed Scopus Google Scholar). of substrates and of the specificity of granzyme B is of substrate that granzyme B an optimal P4-P1 of M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google the activation sites of and of are by granzyme B Turbov J. Seth P. Orth K. Froelich C.J. J. PubMed Scopus Google Scholar). also the optimal of the or that granzyme B an of the caspase This using and extended using to that the residue is and that the for granzyme B is J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). of a comprising an sequence in a that was not cleaved efficiently by granzyme B J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google that residues to substrate binding. A to the interaction of granzyme B with substrates is to its interaction with a natural such a to a and J. J. J. Biol. Chem. Full Text PDF PubMed Google and have a and of a reactive center loop the substrate of its cognate proteinase that is by two proteinase the is cleaved and a to the of the serpin-proteinase Cleavage of the by the proteinase between two residues in the loop P1 and The P1 residue is and the specificity of the serpin-proteinase residues the to the of the interaction J. R. 1998; Full Text PDF PubMed Scopus Google Scholar). The human serpin, proteinase inhibitor 9 (PI-9), is an efficient inhibitor of granzyme B that granzyme apoptosis in cell and is to from J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). that the PI-9 RCL a natural substrate of granzyme B and that residues within the RCL important for with the substrate binding of the Here we by mutagenesis that residues within the P4-P4′ sequence of the PI-9 RCL are for we that the and P4′ residues are important for binding and that the P4′ Glu forms a salt bridge with of granzyme B, serine granzyme B extended substrate and of based on the PI-9 RCL sequence and P1′-P4′ residues confirms and the of specific and granzyme B substrates. PI-9 was using a and described J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). human granzyme B was and using a similar and then by J. Bird Bird P.I. PubMed Scopus Google Scholar). human were described J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). mutagenesis of PI-9 and granzyme B the and J.A. PubMed Scopus Google described in J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google of the and will by the on of the PI-9 mutant was with of granzyme B in of for at of was the was for and then of was for The was to on a liquid with a × and was on a of the at and A acid and B The was with a of B at were and to of acid sequencing on a The interaction of granzyme B with PI-9 and was using described in Scopus Google were by and of granzyme B were not was by with a of of granzyme B in of showed with PI-9 were used in the a of was with of inhibitor and substrate substrate in were at for in a The stoichiometry of inhibition and for the interaction of granzyme B with PI-9 and mutant were described Scopus Google A Scholar). were at and the of the are Cleavage of and granzyme B substrates and of was described J. Bird Bird P.I. PubMed Scopus Google Scholar). The substrate was from Caspase of substrates was previously J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google in containing of the substrate for at An was used to caspase on the substrates and is in the of caspase by in the of The substrates were on a on a are with using B. Peptide A on a The ε-dinitrophenyl residue was into the sequence its and activation, in the of to the of R. W. Scopus Google Scholar). the aminobenzoyl residue was its M. K. PubMed Scopus Google Scholar). of the the were by from the by with a of acid in containing a of the P1 were by two were in This is to the of of a of the is a in the of such M. J. PubMed Scopus Google Scholar). the by with to in the of containing The and of were using and of granzyme B is to granzyme B. The of granzyme B to a of the serine proteinase inhibitor R. Biol. PubMed Scopus Google was used a to model human granzyme B using the J. Biol. PubMed Scopus Google Scholar). The of granzyme B in with to model the region of PI-9 into the of human granzyme B using using similar to those previously described J. R. 1996; PubMed Scopus Google Scholar). we model of human granzyme B the of granzyme using a by and in Scholar). We then the of granzyme B and residues of the to the model of human granzyme B with the region of We the sequence of to that of PI-9 using the in the the P′ region of the inhibitor is in the of the granzyme we used the of to a to the of the RCL of PI-9 into model of human granzyme B. and granzyme B sequence and with a of of the of the granzyme that the of the residues of in consistent with the of the P4-P4′ region of the loop of serine proteinase inhibitors J. Biol. PubMed Scopus Google Scholar). the is a for modeling the P′ region of the PI-9 RCL in the of human granzyme B. We model of human granzyme B in with the region of PI-9 the using a by and in Scholar). We the of and the with the of the We then the sequence of to that of PI-9 We a between the residues and then the model of human granzyme B in with the residues of human PI-9 to with the of the RCL and with was to The of the PI-9 P1 residue Glu340 previously by sequence and by the preference of granzyme B for acidic P1 residues J. Bird J.A. Bird P.I. J. Biol. Chem. 1996; 271: Full Text Full Text PDF PubMed Scopus Google M. J. Biol. Chem. Full Text PDF PubMed Google Scholar). This is by mutagenesis experiments Glu340 with or in with or granzyme B J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). sequence are for P1 residues the RCL from to serpin, and residues in with The is by the serpin, P1 residues for with and M. J. Biol. Chem. 1996; 271: Full Text Full Text PDF PubMed Scopus Google Scholar). the of is acidic residue (Glu344) in the RCL, and granzyme B also M. J. Biol. Chem. Full Text PDF PubMed Google is that either Glu344 or the P1 residue and that Glu340 to inhibitor binding. the of Glu340 the P1 and sequencing experiments were inhibition of the by PI-9 and promote of the by granzyme B, a of PI-9 was This a in the to previously to and the into a granzyme B substrate J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). The was with granzyme B, the and were and by and then the was to acid sequence not of sequence the residues in the and were but the sequence was and to the PI-9 RCL sequence from to This that granzyme B between Glu340 and that for the P1 residue in the PI-9 The serpin-proteinase interaction the of an by to a by of the between the P1 and of the RCL into the A is J. R. 1998; Full Text PDF PubMed Scopus Google P.A. M. PubMed Scopus Google J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). and of the proteinase to the of the from RCL J.A. PubMed Scopus Google Scholar). is RCL the proteinase and the releasing cleaved and P.A. M. PubMed Scopus Google Scholar). the and an or substrate on the of RCL of the interaction of the is or and is substrate the is the is that the is a substrate by the The is a of the of RCL and The the the the and the less is to the substrate of the by the a value for the of of the that the J.A. PubMed Scopus Google P.A. J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). is that the specificity and of the serpin-proteinase interaction is by the P1 residue J. Rev. PubMed Scopus Google Scholar). This is consistent with the inhibition of proteinases that have a substrate and high and and not have sequence the P1 that substrate binding. is also by the of the from to by of its P1 residue from to J. Rev. PubMed Scopus Google Biol. PubMed Scopus Google Scholar). granzyme B from the proteinases in that its and that is a that on a of is and that substrate specificity is also to by residues the This is by that the and residues are important in granzyme B substrates M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google and serine is not of substrates of less residues J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). binding and inhibition of granzyme B by PI-9 is to on the and residues and the We have previously that of the PI-9 P1 residue to Asp in an inhibitor that is less and of a granzyme B substrate and J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). We have also that the PI-9 granzyme B interaction requires a RCL, of the and PI-9 into a substrate identify residues in the PI-9 RCL loop important for granzyme B we the to residues in with the was with were in the P. and to using chromatography were for the to with granzyme B in and then were to and were for the interaction of with granzyme B. the of the with a in the by to with granzyme B not but were in the stoichiometry and of the in and on the of PI-9 were for at and mutagenesis of the PI-9 with granzyme × × × × × × × × × × × × × × × × × × × × × × × × × × × × × × × × × × × a the for the from in in a a the for the from in modeling that not with the of granzyme B residue the of the B interaction and the that is the value for the interaction is to that of on is that proteinase is that or of the RCL into the A the that the will the substrate the PI-9 P4-P1 residues proteinase binding to the to RCL and the are consistent with the substrate and of granzyme B M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). and that a substrate for granzyme B a residue and that the optimal sequence is M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). the of residue in sequence to substrate binding not B will substrates either or at the is and natural substrates have either or the of a in the PI-9 RCL is consistent with residue an important role in granzyme B binding. showed that of the for in the PI-9 RCL in by a in the of in for the binding of granzyme B to substrates and substrates of granzyme × × × × × × × × × × × × × × × × × from in Refs. F. 1998; Full Text Full Text PDF PubMed Scopus Google F. W. J. PubMed Scopus Google PubMed Scopus Google PubMed Scopus Google are experiments by using have by of the P1 Asp for a in a This from in Refs. F. 1998; Full Text Full Text PDF PubMed Scopus Google F. W. J. PubMed Scopus Google PubMed Scopus Google PubMed Scopus Google are experiments by using have by of the P1 Asp for a Peptide substrate a preference for Glu at M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google but of natural substrates that human granzyme B residues at is not important the residue is for granzyme B binding. substrates with at are cleaved by granzyme B at of the of substrates containing Glu J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google substrates containing are cleaved at human granzyme B not substrates containing or efficiently granzyme B M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google in the substrate binding of granzyme B from that the PI-9 from to a on inhibition of human granzyme B a in the the that is important for binding and that important the is to a preference for M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google but substrates containing or are also substrates with are cleaved less efficiently those containing by and human granzyme B, and is in the we for the PI-9 mutant a to the that the residue not to substrate binding. Finally, is that Asp is at P1 for substrate by granzyme B M. J. Biol. Chem. Full Text PDF PubMed Google M. J. PubMed Scopus Google Scholar). we have previously in and of the P1 Glu with Asp in a inhibitor that is cleaved efficiently by granzyme B J. J.A. Bird P.I. Biol. 1998; 18: PubMed Scopus Google Scholar). with in an the of the P1 in substrate and that efficient substrate or inhibitor binding to human granzyme B requires an extended sequence of the and that the of the residues is P1 that granzyme B is by P′ residues in the substrate and that is at J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). is at in a of natural granzyme B substrates a model comprising the optimal P4-P1 residues and the optimal and residues by was not cleaved efficiently by granzyme B J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar). was a substrate a tetrapeptide comprising that at in the P′ residues the granzyme interaction and that a is showed that of the PI-9 and residues to not the of to at the This that the and residues are not for the interaction of granzyme B with substrate of the residue the same the and of the that proteinase binding RCL is of the P4′ residue a the and by The of mutant that in the of an acidic P4′ residue is required for inhibition binds less to the This that substrates an acidic P4′ residue will cleaved less efficiently by granzyme B. a residue is important at the P4′ we Glu344 with The was a inhibitor of granzyme B the the P4′ mutant but with that granzyme B binding is not efficient interaction of human granzyme B with substrates or inhibitors requires an acidic residue at Glu. This is by the that the two to with human granzyme B in and J. Bird J.A. Bird P.I. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Bleackley R.C. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google not have acidic P4′ residues and similar to the P4′ the that the P4-P4′ region of PI-9 a substrate of granzyme B, we substrates based on the we two and of and we used at the and of This was by mutagenesis experiments that of the with either or on The between substrates and is the of the P1 the preference of granzyme B for and the that the PI-9 in a with of a substrate inhibitor we predicted that substrate cleaved in substrates were cleaved by granzyme B, but substrate was cleaved efficiently substrate The with for the of by granzyme the tetrapeptide in a of of J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google with and for substrate of the specificity substrate is better the granzyme B substrates based on of substrates and is described in J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google in a of substrates and is described in J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google substrate binds to granzyme B the tetrapeptide is cleaved at a This to a P4-P1 we an idealized granzyme B substrate comprising the P4-P1 sequence to the P1′-P4′ sequence of PI-9 This substrate to granzyme B the tetrapeptide substrate and was at an a specificity the sequence of substrate the activation for granzyme B on caspase an acidic P4′ residue and is a in target of granzyme B M. A.J. Bleackley R.C. J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google A.J. Bleackley R.C. J. Biol. Chem. 1996; 271: Full Text Full Text PDF PubMed Scopus Google Scholar). the of the P4′ residue in the interaction between granzyme B and we also a substrate to substrate that the P4′ Glu is by in substrate not substrate to granzyme B in and was efficiently are consistent with an important role for the P4′ residue in a interaction with granzyme B in of PI-9 RCL in a we human granzyme B using the of granzyme B R. Biol. PubMed Scopus Google a two proteins are We then the of the PI-9 RCL into the of the model using the of a granzyme R. Biol. PubMed Scopus Google and a J. PubMed Scopus Google in the of human granzyme B is predicted to of a that the binding of an extended substrate sequence and the residues of This is consistent with that the P4-P4′ residues substrate and inhibitor binding. the of the modeling was that the P4′ Glu of the PI-9 RCL is predicted to in to Lys27 of granzyme B, and the two residues are to a salt in the of the model the of the is predicted to a salt bridge with the acid is and from the and This have the of the binding and the of the P4′ residue to in a The of Asp at P4′ is by its a salt bridge with Lys27 to is a in the in but not granzyme B, in the substrate and binding of the granzymes from We also that in granzyme B is to the P4′ Glu but is to a salt the that Lys27 in granzyme B with the P4′ Glu of we two of granzyme Lys27 to and to We predicted that the mutant less efficiently to PI-9 and to substrates containing P′ residues the salt bridge the mutant efficiently granzyme B to inhibitors or substrates is not predicted to in in cleaved a tetrapeptide substrate lacking P′ residues to the same that the of the with or the the K27A mutant showed a in the to and a substrate containing P′ residues. The in for the interaction that the P′ residues play a role in the substrate in the the K27A mutant showed a in its to with PI-9 in the and a the that the P4′ Glu of PI-9 is in a salt bridge with Lys27 of granzyme B and that high of granzyme B with substrates and inhibitors the of P′ residues an acidic of granzyme B with with × × × in a and granzyme B in substrate and that granzyme B and the substrates with similar P4-P1 consistent with of the caspase and apoptosis M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). we the of and to substrates and to caspase and caspase in of the cleaved substrates or two of the and cleaved substrate The of to substrates and P4-P1 with granzyme B. at caspase also Glu at Asp at and Glu at M. V.M. P.A. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). The of caspase and caspase to substrate is not that its P4-P1 sequence is to the of caspase to substrate is and that the P′ residues in substrate efficient interaction with the binding of the that substrates based on the PI-9 RCL and used to and granzyme B in are that the P1′-P4′ residues of granzyme B substrates are important for substrate but the optimal P1′-P4′ sequence for granzyme B to by or also that efficient and granzyme B inhibitors P′ residues and tetrapeptide or granzyme B inhibitors have not to PubMed Scopus Google Scholar). An acidic P4′ residue in the PI-9 RCL is consistent with the that inhibitor to granzyme B binding at the of a that is by the of the Glu at the P1 of the granzyme B sites on substrates that those in caspase and acidic P4′ residues not have acidic P4′ residues in granzyme B This that granzyme B two substrate that are by the or of acidic P4′ that binds with high but and that binds with but also a of targets for granzyme B in the is consistent with its role an of binds to and proteins that early in and then less a of is not to natural substrates into to are early or late and targets binding and have not for with granzyme B that P4-P4′ in natural substrates and inhibitors are required for efficient binding to human granzyme B and that the is an acidic P4′ residue is This is with the substrate of granzyme B and the role in cytotoxic in to the serine proteinases that have a of substrates and sequence is to that serine proteinase to have such an extended sequence for substrate or inhibitor binding. of with its cognate inhibitor inhibitor to the residues of the J. 1996; PubMed Scopus Google Scholar). This is consistent with the role of in activation and its of substrates. We J. for granzyme B used in the of the and and R. for and of the We are also to Bird for with the and to for and a