452 Btochtmwa etBwphyswaActa, 742 (1983) 452-462 ElsewerBmmedlcal Press BBA31462 THE CONFORMATION AROUND THE PEPTIDE BOND BETWEEN THE P|- AND P2-POSITIONS IS IMPORTANT FOR CATALYTIC ACTIVITY OF SOME PROLINE-SPECIFIC PROTEASES GUNTER FISCHER, JOCHEN HEINS and ALFRED BARTH Martin - Luther-Umversttat Halle, BIB Bwchemw (Abtedungen fur Enzymologte und Wtrkstoffbtochemw), Domplatz 1, 402 Halle (G D R ) (Recewed June 29th, 1982) (Rewsed manuscript received October 25th, 1982) Key words Conformattonal spectfictty, Prohne peptMase, Substrate conformatton, DlpeptMyl peptMase IV Proline-containing dipeptidyl-4-nitroanilides have been synthesised and subjected to dipeptidyl peptidase IV-catalysed hydrolysis at high enzyme concentrations to collect information on the conformational specific- ity of the enzyme active site for a nonscissile bond. Descriptions of the biphasic kinetics were carried out in terms of cis/transinterconversion of the substrates. The results show that the enzyme can cleave only the trans-conformation of the substrate. The competitive inhibition by Gly-Pro-OH and Ala-Pro-OH is also specific for the trans form of the dipeptides. The interpretation of the results obtained from these kinetic studies has led to proposals for the stepwise cleavage of biologically active peptides like substance P and fl-casomorphine by dipeptidyl peptidase IV. Introduction Prohne-contalnlng peptldes are widely distrib- uted among biological actwe sequences The pro- teolytlc cleavage of pept~de bonds m this type of compound ~s catalysed by vinous types of prohne-speclfic endo- or exopeptldases [1]. The specificity of these enzymes can be described, in part, by the posmon of the prohne residue with respect to the susceptible pepUde bond. If the prohne is located in a peptlde chain, as In. Ps-P2-Pro~P;-P~- the peptlde bond C-tern'anal to prohne is hydro- lysed by prollne-speclfic endopepudases (EC 3.4.21 26) like postprohne-cleavmg enzyme [2] and the prohne-speclfic endopeptldase from Flavobac- tertum menmgoseptwum [3]. Some of these enzymes have further structural reqmrements. For example dlpeptldyl peptldase IV (&peptidyl peptldase IV, 0167-4838/83/0000-0000/$03 00 © 1983 Elsevier Blome&cal Press EC 3 4 14.-) [4,5] requires a protonated a-armno group m the P2- position for successful cleavage. + ~ t p H 3 N-P2 Pro-PrP ~ When an llnldlc group is present in the P~- posi- tion, neither proline-speclfiC endopeptldase nor &- peptldyl peptldase IV is active. It is, however, a fact that amino acids like prohne, hydroxyprolme or sarcosme wlthm a peptlde chain might intro- duce structural heterogeneity because of the two possible conformations about the Px-Pro- bond Although the -CO-NH- peptlde unit generally shows a large preference for the trans conforma- tion, muno aods are known to introduce consider- / C C~
`
`t r a n s
`
`c~
`
`\
`
`H
`
`0
`
`~C
`
`H
`
`cis
`
`SAXA-DEF-00237
`
`Page 1 of 11
`
`AstraZeneca Exhibit 2031
` Mylan v. AstraZeneca
` IPR2015-01340
`
`N~
`N\\
`Scheme 1
`

`
`able amounts of cts-conformaaon in solution [6-8]. In contrast to the situation with the -Px-Pro umt, the cls population of the Pro-P x- unit is surely lower than 1% [9]. In fact, lnvestlgaUons of the conformatlonal speoficlty of amlnopeptldase P (EC 3 4.11.4) [10,11] and prohdase (EC 3.4.13.9) [12] have dem- onstrated a reqmrement for the trans conforma- tion about the SClSSlle bond. For this type of proteolytlc enzyme, the secondary amlde bond ~tself ~s the center of catalytic action: -PI~-Pro-P~ On the other hand, the cytosohc leuclne ammo- pepudase (EC 3 4 11 1) shows no preference for a specific conformaUon about a nonsclssde bond. When Leu-Phe-Pro-OH is used as a substrate, it is possible to find hydrolysis of the Leu-Phe bond in the conformation with a trans Phe-Pro umt as well as with the cls Phe-Pro peptlde bond [11]. In accord with these results, it is clear that the interpretation of kinetic constants of proteolytlc enzymes working on proline-contalnmg peptldes will reqmre m every case a knowledge of confor- matlonal specificity of the enzyme and, if neces- sary, the cts/trans isomer ratio for the substrate. In fact, because the free energy, AG ° , for the cts//trans isomensatlon of ollgopeptldes contain- mg lmlno acids is intrinsically rather small [13,14], minor changes of external conditions for enzyme catalysis (state of protonauon, solvent composi- tion) might produce large differences in the amount of enzymatlcally hydrolysable isomer m soluuon The present paper reports the conformatlonal speclfloty of &peptldyl peptldase IV and, to a smaller extent, that of prohne-speoflc endopeptl- dase. For th~s purpose, the reaction of Gly-Pro-4- mtroandide (Gly-Pro-NHNp) and Ala-Pro-4- nltroamhde (Ala-Pro-NHNp) was investigated at very high enzyme concentrations. Conformat~onal speoflclty in catalysis could be also important m relation to product mlubmon by the product P2- Pro-OH. The prohne-speclfic-endopeptidase-sub- strate Ala-Pro-Pro-4-mtroandlde (Ala-Pro-Pro- NHNp) has four possible isomers [6], any or all of which might serve as a substrate for prohne-speoflc endopeptldase. This could be detected by the ob- 453 servatlon of different kinetic phases at high pro- hne-speclfiC endopeptldase concentrations. Materials and Methods DlpepUdyl pepudase IV was purified from pig kidney as described in a previous paper [4] Pro- line-specific endopeptldase was gift from T. Yoshlmoto, Faculty of Pharmaceutical Sciences, Nagasaki, Japan. The lyophyhsed enzyme was dis- solved in buffer and centrifuged before the kinetic runs. Ala-Pro-4-mtroandlde and Gly-Pro-4- nltroamhde were syntheslsed by using the nuxed-anhydride techmque and were characterlsed as described previously [18]. The chemical punty of the substrates was checked additionally by fol- lowing the alkaline hydrolysis to the endpomt. The optical purity was assayed using dlpeptldyl peptl- dase IV in long incubation (4-8 h) expenments. Dlpeptidyl peptldase IV can hydrolyse only the L-:somer m both positions of the peptlde chain. The concentrations gwen m the results were calculated from these determinations The purity of the substrates was approx. 93% Lys+(4- NO2Z)-Pro-OH and Gly-Pro-OH HC1 were gifts from K. Neubert Enzyme and protem assay Dipeptidyl peptldase IV was assayed by actwe- site t~tratlon using Pro-Pro-4-nltroanllide as Utrant. The molecular weight of dxpeptldyl peptidase IV was assumed to be 115 000 per catalytic subunlt [2,4,20] The fraction of actxve enzyme was de- termined by measuring the 'burst' reglon in the pre-steady state phase of the enzyme kinetics at various substrate concentrations. A typical purity of the enzyme used in thas work was 60-70% of total protein concentration. Prohne-speclflC endopeptldase activity was as- sayed by a spectrophotometric method using Ala- Pro-Pro-NHNp as substrate. Protein was mea- sured by the method of Lowry et al. [22] using bovine serum albumin as standard. Esttmatton of pK ~ values pK a values of substrates were deterrmned by potentiometnc titration with Radiometer titration equipment (PHM 26, TTT 1, ABU 13, G 2222 C, K 4112) at 30.0°C and 0.1 # KC1 In order to
`
`SAXA-DEF-00238
`
`Page 2 of 11
`
`

`
`454 avoid incorrect results from the hydrolysis of the nltroanllldes m more basic solutions, only the first part of the titration curve was used to calculate pK~ values 13 C NMR spectra The 13C NMR spectra were recorded by pulsed Fourier transform methods on a Bruker WP-200 spectrometer Carbon-13 chemical shifts are ppm downfleld from an external reference of TMS. The p2H of the peptlde samples was adjusted with 2HCI and was measured by a combination electrode (Radiometer) at 25°C. The p2H was calculated from p2H = pHreadm$ + 0 4 [19] gmettc expertmenls Fast kinetic experiments were performed on an Dlonex model I10 stopped-flow spectrophotom- eter with a dead time of 3.5 ms One syringe contained enzyme m 0.1 M phosphate buffer and the other contained the substrate m the same buffer solution. Reaction progress was monitored at 390 and 450 nm with a 20 mm cell The spectrophotometer was used only m the transmit- tance mode. The time-dependent transmittance was dlgltlsed and stored by a transient recorder (Bio- matlon, model 2805). The stored 2048 data points were transmitted on-line to a 9825 desk-top com- puter (Hewlett-Packard) and converted to the time-dependent absorbances Different sets of data points were taken for calculation of the first order kinetic parameters and for time-course analysis Slow kinetic measurements were done on a Perkm-Elmer 356 spectrophotometer and Specord 40 M (VEB Carl Zelss, Jena). Results a) The substrate hydrolysts The enzymatic rate of hydrolysis of GIy-Pro- NHNp was determined at a dlpepudyl peptldase IV concentration of 2.1 • 10-7-4.7 • 10 -7 mol • 1 -l Since crystal structure data for the N-termmal- protonated substrate show exclusively the trans conformation [23], the dissolved compound was preeqmhbrated m the final incubation buffer (pH 7.2; 0.1 tool 1 -I phosphate) for at least 30 mln After the enzymatic reaction was initiated by adding dlpeptldyl pepudase IV, two kmetlcally different processes were momtored by means of the mcrease of absorbance at 390 nm or 450 nm, produced by the reaction product 4-nltroandme (Fig 1 a and b) The kinetic pattern of the prohne-speclfiC endo- peptldase (0 184-0 329 mg/ml) catalysed hydroly- sis of Ala-Pro-Pro-NHNp looks very similar to that of the dlpeptxdyl peptldase IV experiments (Fig. lc and d) In all three cases the rate of the slower kinetic phase, as measured by ksiow, is nearly independent of the enzyme concentration (Table I). In contrast to this, the rate of the fast kinetic process increases with increasing enzyme con- centraUon. For GIy-Pro-NHNp, the inmal con- centrat~on of the substrate S O ~s much smaller than the K m value, so that the relationship v = (kcat/gm) [Eo] [So] 1s approximately valid. This led to a nearly first order decrease of the substrate concentration for the fast, enzyme-catalysed pro- cess This allows a rough companslon of kslow and the apparent rate constant for the enzymatlcally catalysed reaction The slow kinetic phase ~tself follows first-order kinetics for more than six t~/2 values In order better to estimate the amphtude ASs~ow, related to the substrate concentration involved in the slow phase, and the rate constant kstow, the ume-dependent absorbance signal was dlgmsed as 100 data points at equal intervals. By non-linear regression analysis [24], the first order rate con- TABLE I INFLUENCE OF ENZYME CONCENTRATION ON RATE CONSTANTS kslow AND AMPLITUDE RATIOS ~sf~, /ASs~ow Substrate E o kslow AStast / (mol 1 -I) (s -l) ASslow (x 10 -2) Gly-Pro-NHNp b 1 55 10 -6 626 90 237 10 -6 645 906 Ala-Pro-NHNp c 155 10 -6 741 159 237 10 -6 664 169 Ala-Pro-Pro-NHNp 0 329 a 1 66 10 6 0 185 a 2 03 9 3 " g l -t, substrate concentration 258 10 -5 mol 1 -I b 6 45 10 -5 tool 1-1 substrate concentration c 46 10 -5 tool 1-1 substrate concentration
`
`SAXA-DEF-00239
`
`Page 3 of 11
`
`

`
`E 0.8331 0.581 0.329 0 077
`
`0.761
`
`;
`
`;
`
`0.272
`
`;
`
`;
`
`;
`
`;
`
`'
`0
`
`:"
`
`:
`
`:
`
`''"
`
`f:
`5
`
`i;
`10
`
`•
`
`; "i:
`15
`
`:;
`
`:
`
`:
`i
`20 t rsecl
`L
`.l
`
`A
`
`0 772
`
`E3
`
`:
`
`;
`
`;
`
`;
`
`!
`
`;
`
`:
`
`|
`
`;
`
`:
`
`;
`
`:
`
`;
`
`,
`
`0
`
`50
`
`100
`
`150
`
`t Fsecl
`
`0.72, 0 67( 0.621
`0 1 2 3 4 tLsec j C 0.715 0 655 0.595 0 53'
`
`455 o 5o 1oo 15o t [.c] Fig 1 A Rate of dlpepadyl peptldase IV-catalysed hydrolysis of Gly-Pro-NHNp at 25 0°C and pH 7 2 m phosphate buffer (0 1 mol 1-1) The concentration of GI~'-Pro-NHNp was 3 09 10-5 mol l-l, the concentration of enzyme was 1 08 10-6 mol 1-t B Same reacUon momtored on a slowbr tzme base The first order rate constant was calculated from the data points collected after terrmnat~on of the rapid change m absorbance The calculation gives kstow = 3 82 10 -2 s- = and ASslow = 0 065 The theoretical curve is drawn as a hne, with some of the experimental values shown as points C Rate of prohne-spectfic endopeptldase-catalysed hydrolysis of Ala-Pro-NHNp at 30°C and pH 7 5 in phosphate buffer The concentration of the substrate was 2 58 10-Smol 1- i, the concentration of the enzyme was 0 185 g 1- l D Conditions aS" m b, kslo. = 2 03 10-2s - t and ASslow = 0 073 stant kslow and the amphtude ASslow were obtamed. The theoretical curve and some of the experimen- tal data points used In the regression analysis are TABLE II INFLUENCE OF SUBSTRATE CONCENTRATION ON THE KINETIC CONSTANTS OF DIPEPTIDYL PEPTI- DASE IV CATALYSIS AIa-Pro-NHNp kslow % cts ASrast/ASstow (mol l-l) a (s-l)(×10 -2) 2 83 10 -4 5 98 5 88 16 0 6 90 10 -5 4 36 7 15 13 0 1 38 10 -5 6 64 5 58 16 9 275 10 -6 255 846 108 Enzyme concentratzon 2 37 10 -6 mol l-I shown m Fig lb and c to permit a check of the quahty of curve fitting. Both fast and slow kinetic phases are different in their rates by a factor of 31, when the experi- mental condmons of Fig la and b are used. The amphtude ASfast of the fast reaction was obtamed from the knowledge of the slow phase value ASslo.,,, and the total change of absorbance after long incubation experiments (ASslow + ASfast ). For some cases, especmlly at low temperatures, the rate con- stants are different enough (kfast/k~low >> 50) to permit &rect measurements of absorbance of the fast process as a plateau region. The ratio of amphtudes ASf~,st/ASslow is mde- pendent of the enzyme concentration and of the substrate concentration as well (Table II)
`
`SAXA-DEF-00240
`
`Page 4 of 11
`
`E
`: :
`E
`0.51E
`0 02~
`

`
`456 For the experimental conditions of Fig 1 a and b, for Gly-Pro-4-NHNp the ratio 9 1 is measured. The sum of both amphtudes ASfast + ASs,ow is di- rectly proportional to the 4-nltroamhne concentra- Uon, winch can be calculated from the substrate concentratmn, after correction for optical and chelmcal purity of the dlpeptldyl-4-mtroanflldes The rate constant kslow is also independent of the substrate concentrauon (Table II) It has been demonstrated that the acuwty of dlpepUdyl peptldase IV can be competitively ln- Inblted by Lys+(4-NOEZ)-Pro-OH in a powerful manner If the slow phase of the kinetics pattern reflected any property of &peptxdyl peptldase IV, tins process would be influenced by the lninbltor. In the presence of 3.24 10-Smol • 1-1 Lys-~-(4- NO2Z)-Pro-OH and 4.24.10 -5 mol.1 -I Gly- Pro-NHNp k~low is unaffected, while the fast kinetic process is slowed down to one third of this original velocity The apparent small influence of the minbltor on ASslow (Table IV) seems to be an artefact It appears from the difficulty of evaluat- ing tins quanUty, that the rates of the two kinetic phases approach each other. The hrmts of error m the nonlinear regression analysis also become greater, which ~s consistent w~th tins explanaUon In summary, these results suggest that the slow phase is an intrinsic property of the substrate, whale the rapid phase results from enzyme cataly- sis Blphasic kinetics were found not only at 30°C, but throughout the temperature range from 9 to 30°C. For Gly-Pro-NHNp, the dependency of the fracUonal concentraUon of the slow phase and the rate constants of the slow phase from the tempera- ture are summansed m Table III. The actlvatmn enthalpy AH¢. for the slow phase TABLE III TEMPERATURE DEPENDENCE OF CIS-TRANS INTER- CONVERSION OF Gly-Pro-NHNp t kslow % cts (°C) (s- 1) 9 2 8 15 10 -3 7 84 150 143 10 -2 839 199 237 10 -2 852 250 382 10 -2 909 304 645 10 -2 994 -
`
`In (T k---) sec -1
`
`44
`
`40
`
`36
`
`32
`
`314
`103 x T -1
`
`36
`
`Fig 2 Eyrlng-plot for kslow of Gly-Pro-NHNp For experimen- tal condmons, see Fig la The hnear regression analys~s gives y = -3 50 103 x +7 87 (r 2 = 0 99997, n = 5) of Gly-Pro-NHNp obtained from the Eynng plot given m Fig. 2, is + 66.9 kJ mol-i. For the reac- Uon enthalpy AH ° of the same process (Kea = [fast]/[slow]) a value of -8.2 kJ.mol -t (r ~= 0.952; n = 5 for hnear regression) was esUmated from a van't-Hoff I~lot. The population and the rate constant of the slow-fast mterconversion do not depend markedly on the pH m the range 5.5-7.5 (Table IV). TABLE IV pH DEPENDENCE OF RATE CONSTANTS AND FRAC- TIONAL CONCENTRATION OF CIS ISOMER, AT 30°C IN PHOSPHATE BUFFER kslow values are in s- i pH b GIy-Pro-NHNp AIa-Pro-NHNp kslow ~ cts kslow % cls (XIO -2) (xlO -2) 55 388 64 6 8 5 82 10 8 7 64 6 5 72 645 99 741 59 75 581 105 734 66 75 a 664 153 a In presence of 3 24 10 -5 mol 1-1 Lys-(c-4NO2-Z)-Pro-OH HC1 as a compeutlve mlubltor (K, = 2 8 10 -6 mol 1- t) The mmal rate of the hydrolysis m 4 25 10 -5 mol 1-i GIy-Pro- NHNp is reduced to 38% b Enzyme concentration 3 10 10 -6 mol 1- z
`
`SAXA-DEF-00241
`
`Page 5 of 11
`
`48
`i I
`

`
`o2.- A ! 457
`,c3
`
`c
`
`t
`
`/
`--CONH--
`CON
`
`~ L
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`!
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`
`[
`
`200
`
`180
`
`160
`
`140
`
`120
`
`100
`
`80
`
`0
`
`40
`
`20
`
`0
`
`lapin
`
`Fig 3 The 13C NMR spectrum of GIy-Pro-NHNp (3 3 10 -1 moll -1) m 2H20 at 25°C and p2H 6 8 All these values should be considered m the hght of the fact that the fracttonal concentration of N-terminal unprotonated substrate, at pH 7.5, is 21.6% for Gly-Pro-NHNp and 16.3% for Ala- Pro-NHNp (pK a values of 8.06 and 8.21, respec- hvely). Keq and ks]ow must both be considered as mixed constants at pH > 6.5. b) Carbon-13 NMR spectra The laC NMR spectrum of GIy-Pro-NHNp, studied at p2H 6 8 exhabits a doubhng of the resonance for some of the C-atoms of the pyrroh- dine ring. (Fig. 3) This shows that more than one species of the substrate exasts. From the relatwe mtensiUes of these resonances, ~t follows that the minor population is only 7.3% of the total. The minor resonance of the CV-atom of the pyrrohdine ring is located at 22.4 ppm, 2 37 ppm upfield from the more intense stgnal In contrast for the C& atom, a downfield shift of the minor signal is observed. Thel3C chemical shifts for atoms of the pyrrohdlne ring have been employed prewously to dffferentmte between the cts and trans forms of prohne-containing peptldes [25]. The behaviour of the chemtcal shifts m our case indtcates that the minor resonance arises from the cts isomer of Gly-Pro-NHNp The cts/trans isomerism is also clearly confirmed in the ~3C-NMR spectrum of Ala-Pro-NHNp. As a consequence of the lower cts-isomer concentration with this substance, a re- liable integration of the spectrum was not ob- tamed The observed chenucal shifts were Ctrra,s 23.09 ppm, CcV, s 21.50 ppm, C/~,r,,s 27.70 ppm and C~, 30.35 ppm These results are sufficient evidence that the slow kineUc process observed m the dlpeptldyl pepudase IV-catalysed hydrolysis corresponds to the rate of cls/trans lnterconverslon of proline- contamlng substrates c) Product mhtbmon The apparent conformatlonal speoflcxty of dl- pepUdyl peptldase IV, indicated above, could lead, under appropriate circumstances, to pure isomers as hydrolysis products, as a result of the small residence time of the substrate on the enzyme matrix, compared to relaxatmn Ume for cts/trans ~somerlsm. Thus another question is raised: the conformatlonal speclfioty of dlpeptidyl peptldase IV m compeUUve mhibiuon by the product d~- peptldes P2-Pro-OH These compounds have been employed previ- ously to inhibit dlpepudyl peptldase IV [2,4] We can estimate the approximate relaxatmn hme for the lsomensatmn of the zwltterlomc d~pepudes as follows For HaN+-Gly-Pro-O -, we take tl/2 =
`
`SAXA-DEF-00242
`
`Page 6 of 11
`
`CNC--3
`/.N C
`! I I I I I I I 1 I I I I I I I I 1 I I I
`

`
`458 25. 17 so% * 8 __~_. _- -_ -_ - 87.8% 75~o 96% l | , • , i . , , , , . , , , ; ; . : : , • , • : ; 1.000 1 627 2 254 2.881 1/c x In(l/(1-c)) Flg 4 Trine-course analysLs using Eqn l for GIy-Pro-NHNp (3 25 10 -5 mol 1 -]) with d]pept]dyl pept]dase l 07 10-6mol l-u Arrows indicate fractmn of substrate conversmn 0.693/(k_, + kl), k]/k_n = Keq = [trans]/[ClS] with k n =9.6.10 -3 S -1 and k_~-60- 10 -3 s -I from the data of Lln and Brandts [12] at 30°C, this gives 40 s for the q/2 value. Time-course experiments of dlpeptidyl peptl- dase IV-catalysed hydrolysis, at low enzyme con- centratlon with Gly-Pro-NHNp as a substrate, yield the same K,-value for the product Gly-Pro-
`
`(n
`,,.,,
`
`K3
`4)
`
`~ 22-
`
`r"
`m
`
`~ 18- 14 1.0 I I
`I I 1 3 SoXlO 4 Fig 5 Replot of intercepts and slopes from the time-course analysis against the concentration of Gly-Pro-NHNp Dnpeptldyl pepudase IV concentration, 1 07 l0 -6 mol l-u
`
`1 3 Sox 10 4
`
`5
`
`5
`
`SAXA-DEF-00243
`
`Page 7 of 11
`
`

`
`TABLEV KINETIC CONSTANTS OF DIPEPTIDYL PEPTIDASEIV CATALYSIS AIa-Pro-NHNp Gly-Pro- NHNp K~ (mol l-l) a 1 40 10 -5 1 25 10 -4 kca t (s- i ) a 63 7 75 6 Km (mol 1-1) b - 82 10 -5 ken t (s- I ) b _ 74 1 K,(mol.l-l) a 133 10 -40 105 10 -3 K,(mol.l-I) b - 468 10 -4 K~(mol.l-t) c 480 10 -4 1 12 10 -3 Dlpeptldyl peptldase IV concentratmn 2 14 10 -7 mol i-i b Dlpeptldyl peptldase IV concentrauon 1.07 106 mol 1- l, K, at tins enzyme concentratmn corresponds to the mmnsm K, value c Found with added eqmhbrated lrdUbltor d Nearly the mtnns~c K,, the catalytm reactmn ~s much faster than for Gly-Pro-NHNp OH, formed in the catalytic reaction, as is formed for the chemically synthesized dipeptide, when it is preeqmhbrated in buffer solution and subjected to a conventional initial-velocity inhibition experi- ment At higher dipeptldyl peptldase IV levels, the whole reaction takes place on a faster time scale (tl/2 << 40 s) It is then possible that in the case of conformaUonal speclfioty, a new 'mtnnslc' inhtbi- tory constant K, of Gly-Pro-OH might be ob- tained. The time-course dependencies were calcu- lated using Eqn. 1 [26]: ( ) (,o,,l t 1 Km Km 1 + -~-~, ) cln 1 (I) Vm~, 1+-~- so+ v-~. ~ -c S o = initial concentraUon of substrate, S = substrate concentration at the time t. A plot of t/c vs. (l/c) In [1/(l - c)] with c = (S o - S)/S o leads to a strmght hne, as shown m Fig. 4. Additional studies of the same type, with differ- ent substrate concentrations S o and the same en- zyme concentration m all cases, give a series of straight lines with different intercepts and slopes. Replots of the intercepts vs. S O (Fig. 5) then gwe the 'mtnnsic' K,, 'intrinsic' K m and Vma x. The intrinsic lonetm constants result when the tl/2 of the fast kinetic phase is short compared with the 459 relaxation times of all isomerisatlon processes of the substrate and product. If a certain high level of chemically syntheslsed dlpeptide (I o >> So) is ad- ded initially, the inhibitor concentration can be considered constant throughout the time-course In tins case Eqn. 2 is vahd [26] t Kin/ 4~ltn i c (2) The slopes of the linear plot of t/c vs. (l/c) In [1/(1 - c)] at a constant tugh lntubitor concentra- tion and different substrate concentrations, to- gether with a value of Km//Vmax, estimated in an experiment without inhibitor, give a set of nearly constant K,-values. The mean value of K 1 pro- duced by six different substrate concentrations is gwen m Table V as k,L All data of lntnnsic and normal klnetm constants are collected m Table V Discussion It is quite clear from the results employing high concentrations of enzyme that only 90% of Gly- Pro-NHNp and 95% of Ala-Pro-NHNp can act as a substrate for dlpeptldyl peptldase IV. We note that only the fast kinetic process is sensitive to the enzyme level, the substrate concentration and also to lnlubltion expenments (Tables I, II and IV). The ratio of amplitudes of the slow and fast phases is influenced by temperature variation Therefore, for the slow part of the kinetic pattern, a rate- limmng structural change occurring in the sub- strate molecule is a reasonable model. As is evident from the 13C-NMR spectra for TABLE VI FRACTIONAL CONCENTRATION OF CIS ISOMER FOR VARIOUS STRUCTURES ' Substrate cls Refs H3N +-Gly-Pro-Phe-O- 0 27 28 H3N+-GIy-Pro-AIa-O - 0 17, 0 15 1 I, 27 H3N+.Gly.Pro.OH a 0 16, 0 18 7, 12 H3N +-Ala-Pro-NHNp 0 06 this work H3N +-GIy-Pro-NHNp 0 09 tins work H3N +_Ala.Pro_OR a 0 11 7 a Products of substrate cleavage
`
`SAXA-DEF-00244
`
`Page 8 of 11
`
`

`
`460 solut]ons of Ala-Pro-NHNp and Gly-Pro-NHNp, there are two conformations present, with the trans isomer strongly dominating The fraction of cts Isomer, obtamed by NMR spectroscopy, agrees very well with the amplitude ASslow of the slow process of the substrate cleavage. Cts/trans ratios found m the present case are consistent w~th data for related compounds, estimated by independent methods (Table VI) The reacuon enthalpy AH ° obtained from the temperature dependence of AS~low also indicates a typical cts/trans lnterconverslon. One explanation for such kinetic behaviour is an absolute confor- matlonal speohcity of d]peptldyl pepudase IV for the trans conformation at the nonsc~sslle bond of P2-Pro-peptldes Studies of Lln and Brandts [9-12] indicate that the sosslle pepude umt must be m the trans conformauon for successful action of some other proline-speclfiC peptidases This then differs from the conformauonal specificity shown by dlpepudyl peptldase IV, where a wrong peptlde bond conformaUon, remote from the reacuon center, totally prevents the catalytic action. The situation also differs from that m the chymotryp- sln-catalysed hydrolysis of P3-Pro-Phe-NHNp and related substrates, where both ~somers were found to react, but at different rates (Fischer, G. and Bang, H., unpubhshed results) dlpepudyl peptl- dase IV can, of course, cleave substrates that are initially m the cts form, but only after conversion to the trans conformation, a process which takes place with the rate constant kslow. O
`
`~L____~ 0
`N
`
`÷ Enzyme - OH \
`
`/
`
`O"
`
`k, O- Enzyme k~ N k'.7
`
`"
`
`+ H2 N---.o
`* Enzyme
`
`Scheme II explains the acuon of dlpepudyl peptldase IV on prohne-containlng peptldes: kslow must be assigned to the single rate constant k 1 of the isomerisation equilibrium, since the backward reaction k_ 1 will not occur at the l'ugh dlpepudyl peptldase IV level used m deternunmg kslow The temperature dependence of ksjow for Gly-Pro-4- NHNp ggves an activation enthalpy AH*, exactly m the range predicted for tlus type of lsomerlsa- tlon [12]. It is further evident from the nearly pH-mdependent rate of lsomensatlon that ttus ln- terconversion is not catalysed strongly by aods or bases. Furthermore, uncharged H2N-P2-Pro-NHN p seems to have the same fractional concentrations of isomers as does the protonated compound, be- cause the cts/trans ratio is essenUally independent of pH This contrasts strongly to the behawour of dlpeptldes, where the introduction of negatwe charge m H2N-P2-Pro-O- favours the cts isomer. The loss of posmve charge upon deprotonatlon of H~ N-P2-Pro-NHN p, however, does not have a slrmlar effect [6]. The mechanism m scheme II includes the suggestaon that the hydrolysis will produce trans d~peptldes as products. If this Is true, then m a transient phase, the length of which would depend on the relaxation time of the conformatxonal equi- librium, a different picture from the normally ob- served competltwe inhlb]tlon by P2-Pro-OH would be produced. If the enzymaUc reaction time ~s short compared to the relaxaUon time, the mh~bl- tion constant K, will be different from that mea- sured with longer reaction times or preeqmhbra- tlon. In fact, the time-course studies of dlpeptldyl pepudase IV-catalysed hydrolysis suggest that the competitive lntubmon by P2-Pro-OH is specific for the trans xsomer (Table V). When the hydrolysis is completed m an very short period of time (/I/2 -- 2 5 s) the K, value ~s 2 4-Umes smaller than the K, value measured with chemacally syntheslsed, pre- eqmhbrated dlpepUde To exclude effect of en- zyme concentration on the constant, both types of experiments were performed at the same dlpepu- dyl peptldase IV level For lower dlpeptldyl peptldase IV concentraUons and thus slower en- zymic hydrolysis without added Gly-Pro-OH, the K,-value determined from time-course experiments reaches the K,-value for preeqmhbrated &pepude
`
`SAXA-DEF-00245
`
`Page 9 of 11
`
`Scheme II
`

`
`From the fractional concentrations of cts ~somers in soluUons of H~-N-Gly-Pro-O- (0.55) and H~--N-Ala-Pro-O- (0.55), estimated from NMR experiments at 25°C in 2H20, the predicted 'in- trinsic' K~ is 1.8-times smaller than the normal K, This is, within the limats of error, in good agree- ment with our value of 2.4. K m and kca t constants obtained from time-course studies agree very well with those esttmated from initial velocmes. The constants of Table V have been corrected for the small extent of cts isomer in the substrates. The results dtscussed above lead to interesting conclusions m relation to the stepw~se degradation of some peptlde hormones hke substance P. [15,16], fl-casomorphin [17] and promehttin [29]. Prolyl residues, located within a peptide chain, might influence the isomer ratio in an different way than is the case when they are found adjacent to the N-terminal amino acid m P2-Pro [27]. The step- by-step cleavage at the bonds Indicated by arrows will then give peptlde fragments during a transient phase of the enzymatic process which retain their original isomer ratios. This ratio could strongly differ from the equilibrium ratio normally present m peptlde solutions p2- Pro-~ PI -Pro-~ P~- Pro - The duration of this transient phase will depend on the structure of the peptide chain, and on other variables and events, but may be long enough to have biological significance. Thus It seems im- portant to investigate the conformatlonal specific- ity of dipeptldyl peptidase IV m relation to the P~-Pro- bond. Prolln-speofic endopeptidase, the other enzyme investigated, shows a slmdar picture (Fig. lc and d) for the hydrolysis of Ala-Pro-Pro- NHNp at very high enzyme concentraUons. In contrast to the SltUatton with AIa-Pro-NHNp, the presence of a second prolyl residue in this com- pound makes four tsomers posstble. In Ser-Pro- Pro-OH all predicted isomers, trans-trans, trans-cts, cls-trans and c,s-cts, have been found [6]. From the biphaslc kinetics with prohne-specific endopeptidase it ~s evident that not all of the isomers serve as substrates. From (a) the trans-trans isomer population of the closely related H~- N-Ser- Pro-Pro-OH (91% [6]) and (b) the high ratio ASfast/ASs~ow (Table I), it appears that the all-trans form is one of the correct substrates A more 461 detailed analysis of the conformatlonal require- ments of this enzyme will require additional infor- mation about the Isomer ratios in Ala-Pro-Pro- NHNp. Acknowledgements We wish to thank Dr. K. Neubert for samples of mhlbitors, Mrs. S. Flatau for the 13C-NMR spectra, Mrs. C. Metz for techmcal assistance, Mrs. B. Schneewelss for the computer programs and Prof. Dr, Yoshlmoto for samples of prohne- specific endopeptidase. The authors thank Prof. Dr. R.L. Schowen for reading of the manuscript. References 1 Walter, R, Simmons, W H and Yoslumoto, T (1980) Mol Cell Blochem 30, I 11-127 2 Yoslumoto, T, Flschl, M, Orlowska, R C and Walter, R (1978) J Blol Chem 253, 3708-3716 3 Yostumoto, T. and Tsuru, D (1978) Agnc Blol Chem 42, 2417-2419 4 Wolf, B, Fischer, G and Barth, A (1978) Acta Blol Med Ger 37, 409-420 5 Fischer, G and Barth, A (1981) Beltr Wtrkstoffbtochem 11, 105-133 6 London, R E, Matwloff, N A, Stewart, J M and Cann, J R (1978) Blochenustry 17, 2277-2283 7 Evans, C A and Rabensteln, D L (1974) J Am Chem Soc 96, 7312-7317 8 Wuthnch, K, (1976) NMR m Biological Research Pepudes and Proteins, pp 184-188, North Holland, Amsterdam 9 Brandts, J F, Halvorson, H R and Brennan, M (1975) B~ochenustry 14, 4953-4958 10 Lm, LN and Brandts, JF (1980) Blochermstry 19, 3055-3059 11 Lm, LN a,ld Bran&s, JF (1979) Btochermstry 18, 5037-5042 12 Lm, L N and Brandts, J F (1979) Biochermstry 18, 43-47 13 Mata, M L, Oreil, K G and Rydon, HN (1971) Chem Commun 1209-1211 14 La Planche, LA and Rogers, MT (1964) J Am Chem Soc 86, 337-342 15 Kato, T Nagatsu, T, Fukasawa, K, Harada, M Nagatsu, I and Sakaklbara, S (1978) Biochlm Blophys Acta 525, 417-422 16 Heymann, E and Mentlem, R (1978) FEBS Lett 91, 360-364 17 Hartrodt, B, Neubert, K, Fischer, G, Schulz, H and Barth, A (1982) Pharmazae 37. 165-169 18 Barth, A, Mager. H, Fischer. G. Neubert, K and Schwarz, G (1980)Acta B~ol Med Ger 39, 1129-1142 19 Glasoe, PK and Long, FA (1960)J Phys Chem 64, 188-190
`
`SAXA-DEF-00246
`
`Page 10 of 11
`
`

`
`462 20 Svensson, B, Damelsen, M, Staun, M, Jeppesen, L, Noren, O and SjOstroem (1978) Eur J Blochem 90, 489-498 21 Bender, M L (1971) Mechanism of Homogenous Catalys~s from Protons to Protein, pp 414-418, Wdey-Intersclence, New York 22 Lowry, O H, Rosebrough, N J, Farr, A L and Randall, RJ (1951)J Blol Chem 193, 265-275 23 Reck, G and Barth, A (1981) Cryst Struct Commun 1001-1005 24 Zuberbuhler, A D and Kaden, T A (1977) Ctumla, 442-444 25 Deslauner, R and Smith, I C P (1980) in Blolo~cal Mag- netlc Resonance (Berhner, L J and