Tài liệu Báo cáo khoa học: Different modes of dipeptidyl peptidase IV (CD26) inhibition by oligopeptides derived from the N-terminus of HIV-1 Tat indicate at least two inhibitor binding sites doc

Different modes of dipeptidyl peptidase IV (CD26) inhibition by oligopeptides derived from the N-terminus of HIV-1 Tat indicate at least two inhibitor binding sites Susan Lorey 1 , Angela Sto¨ ckel-Maschek 1 ,Ju¨ rgen Faust 1 , Wolfgang Brandt 2 , Beate Stiebitz 1 , Mark D. Gorrell 3 , Thilo Ka¨ hne 4 , Carmen Mrestani-Klaus 1 , Sabine Wrenger 5 , Dirk Reinhold 5 , Siegfried Ansorge 6 and Klaus Neubert 1 1 Department of Biochemistry/Biotechnology, Institute of Biochemistry, Martin-Luther-University Halle-Wittenberg, Halle, Germany; 2 Institute of Plant Biochemistry, Leibniz Institute Halle, Germany; 3 AW Morrow Gastroenterology and Liver Center, Royal Prince Alfred Hospital, Newtown NSW, Australia; 4 Department of Internal Medicine, Institute of Experimental Internal Medicine and 5 Institute of Immunology, Otto-von-Guericke-University Magdeburg, Germany; 6 IMTM Magdeburg, Germany Dipeptidyl peptidase IV (DP IV, CD26) plays an essential role in the activation and proliferation of lymphocytes, which is shown by the immunosuppressive effects of syn- thetic DP IV inhibitors. Similarly, both human immuno- deficiency virus-1 (HIV-1) Tat protein and the N-terminal peptide Tat(1–9) inhibit DP IV activity and T cell prolifer- ation. Therefore, the N-terminal amino acid sequence of HIV-1 Tat is important for the inhibition of DP IV. Recently, we characterized the thromboxane A2 receptor peptide TXA2-R(1–9), bearing the N-terminal MWP seq- uence motif, as a potent DP IV inhibitor possibly playing a functional role during antigen presentation by inhibiting T cell-expressed DP IV [Wrenger, S., Faust, J., Mrestani- Klaus, C., Fengler, A., Sto ¨ ckel-Maschek, A., Lorey, S., Ka ¨ hne, T., Brandt, W., Neubert, K., Ansorge, S. & Rein- hold, D. (2000) J. Biol. Chem. 275, 22180–22186]. Here, we demonstrate that amino acid substitutions at different positions of Tat(1–9) can result in a change of the inhibition type. Certain Tat(1–9)-related peptides are found to be competitive, and others linear mixed-type or parabolic mixed-type inhibitors indicating different inhibitor binding sitesonDPIV,attheactivesiteandoutoftheactivesite. The parabolic mixed-type mechanism, attributed to both non-mutually exclusive inhibitor binding sites of the enzyme, is described in detail. From the kinetic investigations and molecular modeling experiments, possible interactions of the oligopeptides with specified amino acids of DP IV are sug- gested. These findings give new insights for the development of more potent and specific peptide-based DP IV inhibitors. Such inhibitors could be useful for the treatment of auto- immune and inflammatory diseases. Keywords: DP IV; CD26; HIV-1 Tat; parabolic inhibition; mixed-type inhibition. Dipeptidyl peptidase IV (DP IV, CD26, EC 3.4.14.5) is a membrane-bound serine protease first identified in rat kidney [1]. The enzyme occurs in most mammalian epithelial tissues, such as kidney, liver and intestine [2,3]. DP IV catalyzes the cleavage of dipeptides from the N-terminus of oligopeptides and polypeptides provided the penultimate residue is proline [4]. In the immune system DP IV is an activation marker of T lymphocytes and is also expressed on B lymphocytes and NK cells [5–7]. A contribution to signal transduction processes is ascribed to DP IV by various authors [8–11]. Furthermore, the enzyme functions as a binding molecule for adenosine deaminase [12]. The DP IV-catalyzed hydrolysis of the N-terminus of different chemokines resulting in changed receptor binding potentials reflects the importance of the enzymatic activity of DP IV in humans [13,14]. DP IV inhibitors are currently tested by different laboratories and companies as therapeutics in diseases such as diabetes and multiple sclerosis [15,16]. The human immunodeficiency virus-1 transactivator Tat (HIV-1 Tat, 86 amino acids) is a protein encoded by the HIV-1 genome. Tat is an intracellular protein playing an essential role in transactivation of viral genes and in viral replication [17]. HIV-infected T cells release Tat into the culture supernatant [18]. Addition of Tat to the cell culture medium induces a number of immunosuppressive effects, such as the inhibition of antigen-, anti-CD3- and mitogen- induced lymphocyte proliferation [19,20]. The mediation of these inhibition effects may be achieved via the interaction of Tat with cell surface proteins, for instance DP IV. In concordance with this finding, Gutheil et al.[21]showed that Tat binds with high affinity to DP IV and functions as a potent inhibitor of the enzyme, indicating the possible role Correspondence to K. Neubert, Department of Biochemistry/ Biotechnology, Institute of Biochemistry, Martin-Luther-University Halle-Wittenberg, Kurt-Mothes-Str. 3, Halle, Germany. Fax: + 49 345 5527011, Tel.: + 49 345 5524800/5524849, E-mail: neubert@biochemtech.uni-halle.de Abbreviations: G-CSF, granulocyte colony stimulating factor; HIV-1, human immunodeficiency virus-1; IL, interleukin; pNA, p-nitroani- lide; R110, rhodamine 110; DP IV, dipeptidyl peptidase IV; TXA2-R, thromboxane A2 receptor. Enzyme: Dipeptidyl peptidase IV (DP IV, CD26, EC 3.4.14.5). (Received 9 September 2002, revised 10 February 2003, accepted 13 March 2003) Eur. J. Biochem. 270, 2147–2156 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03568.x of Tat–DP IV interactions in AIDS. The immunosuppres- sive effects of specific DP IV inhibitors and Tat are similar [20]. We found that the MXP motif of the N-terminal region of Tat is an important sequence for inhibitory activity, showing the inhibition of DP IV-catalyzed hydrolysis of IL-2(1–12) and the inhibition of mitogen-induced prolifer- ation of human T cells by Tat(1–9) [22]. Amino acid substitutions at positions 5 and 6 of Tat(1–9) resulted in a weakening of the immunosuppressive effects reflecting the importance of the amino acid sequence out of the N-terminal MXP motif [23]. Recently, we characterized the N-terminus of thethromboxane A2 receptor, TXA2-R(1–9), as a possible endogenous inhibitory ligand of DP IV with N-terminal MWP sequence [24]. To gain further insight into the molecular mechanisms of Tat–DP IV interactions and thereby contributing to the understanding of the functional effects mediated by signal transduction processes induced by Tat, kinetic investiga- tions mainly of Tat(1–9)-derived peptides as inhibitors of DP IV were carried out. It is shown that the type of inhibition is determined not only by the N-terminal amino acid motif XXP as we discussed earlier [25] but also by the subsequent amino acids. The inhibition of DP IV by the peptides Tat(1–9), Trp1-Tat(1–9), Gly3-Tat(1–9) and Ile3- Tat(1–9) that follows a rarely described parabolic mixed- type mechanism is presented in detail. Furthermore, the improvement of the inhibitory potency of oligopeptides containing a tryptophan residue in position 2 is demonstra- ted, and we made an effort to explain this by docking studies based on a model of the C-terminal domain of DP IV. We present here the first evidence for the existence of at least two different inhibitor binding sites on DP IV, one in the catalytic site and the other outside the catalytic site. This could be important for the development of new, more effective DP IV inhibitors. Experimental procedures Synthesis of oligopeptides All nonapeptides and Met-IL-2(1–12) (Table 1) were syn- thesized by solid-phase synthesis with Fmoc technique using a peptide synthesizer 433A (Applied Biosystems). The tripeptides MWP and MWV were prepared by solution synthesis. All peptides were purified by reversed-phase HPLC and analyzed by mass spectrometry, 1 HNMR spectroscopy and elemental analysis. The chromogenic DP IV substrates Ala-Pro-pNA [4] and Gly-Pro-R110- CO-(CH 2 ) 4 Cl [26,27] were synthesized according to standard procedures of peptide synthesis and were purified by HPLC. The synthesis and characterization of the DP IV inhibitor Pro-Pro (P) [OPh-4 CL] 2 has been described earlier [28]. Enzyme purification Human soluble DP IV was produced recombinantly in CHO cells [13]. The cell culture supernatant of the transfected cells was applied on a FPLC POROS HQ ion exchange column and eluted with an increasing gradient of NaCl. DP IV-containing fractions were subsequently ana- lyzed by PAGE (silver stained) and the fractions without any contaminations were pooled for further use. Enzymatic assay All enzymatic assays were performed in 0.04 M Tris/HCl buffer (pH 7.6, I ¼ 0.125 M withKCl)at30 °C. The number of the active sites of DP IV was determined by incubating DP IV with different concentrations (10 )9 M to 10 )8 M )of the irreversible DP IV inhibitor Pro-Pro (P) [OPh-4 CL] 2 for 12 h at 30 °C. After completion of inactivation the Table 1. Kinetic constants of the inhibition of DP IV-catalysed hydrolysis of Ala-Pro-pNA and Gly-Pro-R110-CO-(CH 2 ) 4 Cl. Kinetic constants were determined by coincubation of at least six different inhibitor concentrations and six different substrate concentrations. The enzymatic assays contained 0.04 M Tris/HCl buffer (pH 7.6, I ¼ 0.125), 4.04 · 10 )8 M DP IV and were incubated at 30 °C. The hydrolysis of Ala-Pro-pNA was measured by monitoring the released p-nitroaniline at 390 nm. The hydrolysis of Gly-Pro-R110-CO-(CH 2 ) 4 Cl was measured by monitoring the released R110-CO-(CH 2 ) 4 Cl at 494 nm. The kinetic constants were evaluated using slope and y-axis-intercept replots of the Dixon plot and/or Lineweaver–Burk plot. Compound Amino acid sequence K i ( M ) acd Type of inhibition Tat(1–9) MDPVDPNIE 2.67 · 10 )4 8.9 0.3 6.5 Parabolic mixed-type Tat(1–9) a MDPVDPNIE 2.30 · 10 )4 0.8 0.8 2.2 Parabolic mixed-type Trp1-Tat(1–9) WDPVDPNIE 1.50 · 10 )4 46 1.5 15 Parabolic mixed-type Gly3-Tat(1–9) MDGVDPNIE 4.87 · 10 )4 3.7 0.3 2.2 Parabolic mixed-type Ile3-Tat(1–9) MDIVDPNIE 1.75 · 10 )3 1.7 9.2 0.01 Parabolic mixed-type Lys2-Tat(1–9) MKPVDPNIE 4.27 · 10 )5 10 Linear mixed-type Trp2-Tat(1–9) MWPVDPNIE 2.12 · 10 )6 16 Linear mixed-type Trp2-Tat(1–9)* MWPVDPNIE 1.70 · 10 )6 4.8 Linear mixed-type Met-Trp1-G-CSF(1–8) MWPLGPASS 1.24 · 10 )5 16 Linear mixed-type Met-IL-2(1–12) MAPTSSSTKKTQL 2.69 · 10 )4 9.4 Linear mixed-type Met-Trp-Val MWV 2.00 · 10 )4 15 Linear mixed-type Trp2,Ile3-Tat(1–9) MWIVDPNIE 4.36 · 10 )5 Competitive TXA2-R(1–9) MWPNGSSLG 5.02 · 10 )6 Competitive Met-Trp1-IL-2(1–8) MWPTSSSTK 1.59 · 10 )5 Competitive Met-Trp-Pro MWP 2.45 · 10 )5 Competitive a Gly-Pro-R110-CO-(CH 2 ) 4 Cl was used as substrate. 2148 S. Lorey et al. (Eur. J. Biochem. 270) Ó FEBS 2003 hydrolysis of 10 )4 M Gly-Pro-pNA was measured by moni- toring the released p-nitroaniline (pNA) at 390 nm over 120 s. The DP IV concentration in the assay was obtained from the graph of initial velocity vs. the inhibitor concentra- tion as the intersection of the regression line with the x-axis. The DP IV activity was determined using Ala-Pro-pNA or Gly-Pro-R110-CO-(CH 2 ) 4 Cl as substrates. The inhibi- tion of the hydrolysis of the substrates in at least five different concentrations (10 )5 M to 8 · 10 )5 M )inthe absence and presence of different inhibitor concentrations around the expected K i values was analyzed by detecting the enzymatically released pNA (e 390 nm ¼ 11 500 M )1 Æcm )1 [4]) or R110-CO(CH 2 ) 4 Cl (e 494nm ¼ 29 653 M )1 Æcm )1 [27]), respectively. The measurements were carried out on a Beckmann DU-650 UV/VIS spectrophotometer. The reac- tion was started by adding the enzyme (4.04 · 10 )8 M )and was run in duplicates over 90 s. Evaluation of kinetic constants The kinetic data were calculated using the software MICRO- CAL ORIGIN 4.10 and SIGMAPLOT 5.0. First, steady state kinetics were analyzed using Eqn (1) for the Dixon plot where K i is the binding constant of the inhibitor to the noncompetitive site on the enzyme, whilst a is the factor relating the difference in affinity of the inhibitor for the same site in the enzyme-substrate complex [29]. 1 v ¼ 1 þ aÁK m S½  aÁK i ÁV max Á I½þ 1 V max 1 þ K m ½S  ð1Þ In the case of linear behavior the distinction between competitive and linear mixed-type inhibition as well as the determination of the K i valueandthefactora was carried out using the replot of slopes vs. 1/[S] (Eqn 2). For competitive inhibition a straight line goes through the origin whereas for linear mixed-type inhibition the y-axis intercept is greater than zero [29]. slope ¼ K m K i ÁV max Á 1 ½S þ 1 aÁK i ÁV max ð2Þ In the case of parabolic behavior of the Dixon plot the data were plotted according to Lineweaver–Burk (Eqn 3), yielding straight lines without a common point of intersec- tion. For the calculation of the K i valueandthefactorsa, c and d the slopes and intercepts were replotted vs. [I] according to Eqns (4) and (5), respectively. Here, cÆK i represents the competitive inhibition constant. 1 v ¼ K m V max 1 þ ½I c 1þc  ÁK i þ ½I 2 cÁdÁK 2 i 0 @ 1 A Á 1 ½S þ 1 V max 1 þ ½I aÁK i  ð3Þ slope LineweaverÀBurk ¼ K m V max 1 ½I þ 1 c 1þc  ÁK i þ ½I cÁdÁK 2 i 0 @ 1 A Á½I ð4Þ intercept LineweaverÀBurk ¼ 1 V max þ 1 V max ÁaÁK i Á½Ið5Þ Molecular modeling A model of the C-terminal region containing the catalyti- cally active domain of DP IV has been developed by us and was described previously [30]. Based on this structural model we intended to investigate possible docking arrange- ments of Tat(1–9) and Trp2-Tat(1–9) with DP IV in the presence of the substrate Ala-Pro-pNA located at the active site. The molecular graphics program SYBYL (TRIPOS Associates Inc.) with a slightly modified TRIPOS force field [31] was used. The parameters e of the van der Waals force field term of all carbon atoms were increased by 0.2 kcalÆ mol )1 . The nonbonded cut-off was set to 16 A ˚ . This allows the application of simulated annealing techniques without applying a huge water box surrounding the whole enzyme– ligand complex and periodic boundary conditions. Two independent simulated annealing runs were carried out. The first run was started with the solution conformations of trans Tat(1–9) [23] as well as trans Trp2-Tat(1–9) [24] both determined by NMR investigations. In the second run a random-coil conformation was used as starting structure. Performing 30 cycles of simulated annealing for each run by heating the system to 700 K within 2000 fs and cooling to 100 K in 2000 fs the ligands do not move far away from the enzyme at the high temperature, only about 10 A ˚ on average. During the annealing phase a multitude of stable docking conformations were obtained. The backbone atoms of the enzyme were kept fixed. Constraints were applied between one N-terminal hydrogen atom and one oxygen atom of the side chain carboxylic group of Glu668 and between the carbonyl carbon atom of Pro of the substrate Ala-Pro-pNA and the Ser630 oxygen atom of the enzyme to hold the substrate inside the active site of DP IV. The resulting 30 low-temperature docking arrangements of each run of Tat(1–9) and Trp2-Tat(1–9) were saved in a database and subsequently minimized with the standard TRIPOS force field using Gasteiger charges [32] and a distance dependent dielectric function of e ¼ 4r. Results Kinetic analysis of the inhibition of DP IV Previous investigations have shown that peptides with the N-terminal MXP sequence inhibit DP IV and suppress DNA synthesis of peripheral blood mononuclear cells [22–24]. Our aim was to obtain more information about the interactions of peptidergic inhibitors with DP IV and the kinetic mechanisms of inhibition. For that purpose we investigated the inhibitory effects of Tat(1–9)-derived pep- tides obtained by amino acid substitutions at positions 1, 2 or 3 of Tat(1–9). Furthermore, a number of other oligopeptides with the XXP motif were investigated. The oligopeptides were stable under assay conditions and were not cleaved enzymatically by DP IV as proved by HPLC. DP IV retained its full enzymatic activity in the presence of 10 )3 M oligopeptide solutions as analyzed by dilution experiments. Therefore, putative loss of DP IV activity caused by precipitation or inactivation can be excluded. The K i values of the inhibition of DP IV by these compounds were determined in the range between 10 )6 M and 10 )3 M Ó FEBS 2003 DP IV inhibition by XXP containing oligopeptides (Eur. J. Biochem. 270) 2149 (Table 1). Unexpectedly, some structurally highly related peptides were found to inhibit DP IV according to three different mechanisms. The inhibition of DP IV by the N-terminal nonapeptide of HIV-1 Tat, Tat(1–9), was characterized as parabolic mixed-type inhibition where two non-mutually exclusive inhibitor binding sites exist at the enzyme (Fig. 1) [29]. The inhibitor interacts with the enzyme both at the active site and at an additional binding site. Binding of one inhibitor molecule out of the active site, here defined by the K i value, decreases the affinity for binding of the substrate. The resulting IES complex is catalytically inactive. Binding of one inhibitor molecule at the competitive site, here defined by the cÆK i value, completely excludes binding of the substrate. The interaction of a second inhibitor molecule with the IE complex yielding the IEI complex is character- ized by the dÆK i value. So far, this type of a parabolic inhibition mechanism has been documented only in a few publications [33,34]. For DP IV, this inhibition type is described here for the first time. Using Ala-Pro-pNA as substrate, a K i value of 2.67 · 10 )4 M and an a value of 8.9, reflecting the increased affinity of the substrate or the inhibitor to the free enzyme compared to the EI or ES complex, respectively, were determined (Table 1). The binding affinities of the inhibitor to both binding sites, yielding the IE or EI complex, were only slightly different. On the other hand, the binding affinity of the second inhibitor molecule to the EI complex was decreased by a factor of 6.5. The interaction of the inhibitor with two non-mutually exclusive binding sites at the enzyme was reflected in the parabolic behavior of the Dixon plot (Fig. 2A), which therefore was not suitable for the determination of the K i value. From the Hill-plot, binding of the inhibitor to different binding sites was reflected by the change of the slope in dependence of the inhibitor concen- tration. At high inhibitor concentrations a hill coefficient of )1.6 was determined. The Lineweaver–Burk plot generated straight lines at different fixed inhibitor concentrations (1 · 10 )4 M to 8 · 10 )4 M ). All lines intersected in the second quadrant but without a common point of intersection (Fig. 2B). The replot of slopes (slope Lineweaver–Burk vs. [I]) produced a parabolic curve (Fig. 2C), the replot of y-axis intercepts (y-axis intercept Lineweaver–Burk vs. [I]) provided a straight line not going through the origin. By using the larger substrate Gly-Pro-R110-CO- (CH 2 ) 4 Cl a similar K i value for the inhibition of DP IV by Tat(1–9) was determined (2.30 · 10 )4 M )whereasthea Fig. 1. Kinetic model of a parabolic mixed-type inhibition. Fig. 2. Dixon plot, Lineweaver–Burk plot and slope replot of a parabolic mixed-type inhibition. (A) Influence of eight fixed concentrations of Tat(1–9) (1 · 10 )4 M to 8 · 10 )4 M ) on the hydrolysis of different substrate concentrations Ala-Pro-pNA (s,1· 10 )5 M ; h,1.5· 10 )5 M ; n,2· 10 )5 M ; ,,3· 10 )5 M ; e,4· 10 )5 M ; d, 8 · 10 )5 M ) represented as a Dixon plot (1/v vs. [I]). (B) Lineweaver– Burk plot (1/v vs.1/[S],[I]¼ s,8· 10 )4 M ; h,6· 10 )4 M ; n, 5 · 10 )4 M ; ,,4· 10 )4 M ; e,3· 10 )4 M ; d,2· 10 )4 M ; j, 1 · 10 )4 M ; m, no inhibitor). (C) Slope replot of Lineweaver–Burk plot (slopes vs. [I]). The reaction mixture contained 0.04 M Tris/HCl buffer (pH 7.6, I ¼ 0.125), 4.04 · 10 )8 M DP IV, and was incubated at 30 °C. The hydrolysis of Ala-Pro-pNA was measured by detecting the released pNA at 390 nm over 120 s. 2150 S. Lorey et al. (Eur. J. Biochem. 270) Ó FEBS 2003 value and the d value were reduced (Table 1). The K m value for the DP IV-catalyzed hydrolysis of Gly-Pro-R110-CO- (CH 2 ) 4 Cl was estimated as 4.02 · 10 )5 M (S. Lorey, unpublished results) indicating a fourfold lower apparent affinity of the substrate to the active site of DP IV in comparison to the smaller substrate Ala-Pro-pNA (K m 1.13 · 10 )5 M , S. Lorey, unpublished results). The substitutions of Met1 [Trp1-Tat(1–9)] or Pro3 of Tat(1–9) [Gly3-Tat(1–9), Ile3-Tat(1–9)] did not change the inhibition type. The K i values of the inhibition of DP IV by these peptides were in the high micromolar up to the millimolar range, and using the substrate Ala-Pro-pNA the a values were a >1 (Table 1). As shown for Tat(1–9), the binding affinity of a second inhibitor molecule of Gly3- Tat(1–9) or Trp1-Tat(1–9) to the EI complex was decreased. On the other hand, in the case of Ile3-Tat(1–9) the formation of the IEI complex was favored in comparison to the formation of the EI complex. Moreover, for this peptide binding to the competitive binding site of DP IV is diminished (c ¼ 9.2) in comparison to the other peptides. The peptide containing Trp at position 2, Trp2-Tat(1–9), turned out to be a stronger DP IV inhibitor than the parent peptide but inhibited DP IV following a different inhibition type. Trp2-Tat(1–9) as well as the oligopeptides Lys2-Tat (1–9), Met-IL-2(1–12), Met-Trp1-G-CSF(1–8) and MWV were characterized as inhibitors according to the model of linear mixed-type inhibition [29]. In this case, the inhibitor and the substrate combine independently and reversibly to the enzyme, not competing for a common site, forming ES, EI and IES complexes. The inhibitor is not able to bind to the competitive binding site. The IES complex is catalyti- cally inactive. The binding affinity of substrate and inhibitor to the free enzyme and to the EI or ES complex, respectively, differs by the factor a.TheK i values of the inhibition of DP IV by these oligopeptides were in the micromolar range and the a values were in a range between 9.4 and 16 (Table 1) indicating a greater affinity of the substrate and the inhibitor to the free enzyme compared to the EI and ES complex, respectively. Figure 3 illustrates the kinetics of DP IV inhibition by Lys2-Tat(1–9). The Dixon plot was characterized by straight lines at different fixed substrate concentrations intersecting in the second quadrant (Fig. 3A). The replot of slopes represented a straight line not going through the origin reflecting the linear mixed-type inhibition (Fig. 3B). In the Lineweaver–Burk plot straight lines at different fixed inhibitor concentrations (1 · 10 )5 M to 3 · 10 )4 M ) intersected with a common intersection point in the second quadrant (not shown). As shown above for Tat(1–9), the use of the larger substrate Gly-Pro-R110-CO-(CH 2 ) 4 Cl did not affect the K i value of Trp2-Tat(1–9) but resulted in a decreased a value. Interestingly, whereas Trp2-Tat(1–9) bound to the non- competitive binding site, its N-terminal tripeptide MWP and Trp2,Ile3-Tat(1–9) exclusively bound at the competitive binding site. These peptides and the N-terminal peptide TXA2-R(1–9) of the thromboxane A2 receptor and the oligopeptide Met-Trp1-IL-2(1–8), all bearing Trp in posi- tion 2 similar to Trp2-Tat(1–9), were characterized as competitive inhibitors of DP IV with K i values between 5.02 · 10 )6 M and 4.36 · 10 )5 M (Table 1). This inhibition type is characterized by the formation of EI and ES complexes resulting from a direct competition of the substrate and the inhibitor molecules for binding at the active site [29]. The kinetics of DP IV inhibition by TXA2- R(1–9) is depicted in Fig. 4. The Dixon plot (1/v vs. [I]) provided straight lines at different fixed substrate concen- trations intersecting in the second quadrant (Fig. 4A). The replot of slopes (slope Dixon vs. 1/[S]) represented a straight line through the origin characterizing the competitive inhibition mechanism (Fig. 4B). The Lineweaver–Burk plot (1/v vs. 1/[S]) yielded straight lines at different fixed inhibitor concentrations (10 )6 M to 2 · 10 )5 M )withacommon point of intersection on the y-axis at 1/V max (not shown). Docking of Tat(1–9) and Trp2-Tat(1–9) to DP IV Based on the X-ray structure of the related enzyme prolyl oligopeptidase [35], which together with DP IV belongs to Fig. 3. Dixon plot and slope replot of a linear mixed-type inhibition. (A) Influence of nine fixed concentrations of Lys2-Tat(1–9) (1 · 10 )5 M to 3 · 10 )4 M ) on the hydrolysis of five fixed substrate concentrations Ala-Pro-pNA (s,1· 10 )5 M ; h,1.5· 10 )5 M ; n, 2 · 10 )5 M ; ,,4 · 10 )5 M ; e,8 · 10 )5 M ) represented as a Dixon plot (1/v vs. [I]). (B) Slope replot of Dixon plot (slopes vs. 1/[S]). The enzymatic assays contained 0.04 M Tris/HCl buffer (pH 7.6, I ¼ 0.125), 4.04 · 10 )8 M DP IV, different inhibitor concentrations and were incubated at 30 °C. The hydrolysis of Ala-Pro-pNA was measured by detecting the released pNA at 390 nm over 120 s. Ó FEBS 2003 DP IV inhibition by XXP containing oligopeptides (Eur. J. Biochem. 270) 2151 the prolyl oligopeptidase family, we constructed a 3D model of the C-terminal region of DP IV containing the active site [30]. In order to characterize the noncompetitive binding sites of Tat(1–9) and Trp2-Tat(1–9) on DP IV, docking studies with the ES complex on the basis of this 3D model were performed. Figure 5 illustrates the results of these docking studies showing the most stable interactions of Tat(1–9) with this model of DP IV including the substrate Ala-Pro-pNA bound to the active site with Ala at the S1 and Pro at the S2 binding sites. In the presence of the docked substrate at the active site some strong interactions of Tat(1–9) with DP IV could be detected. Salt bridges were formed between the positively charged N-terminus (Met1) of Tat(1–9) and Asp709 as well as Asp739 of DP IV, between the C-terminal Glu9 of Tat(1–9) and the side chains of Arg560 and Lys554 of DP IV. Furthermore, the side chain of Asp5 of Tat(1–9) was also able to interact with the side chain of Lys554 of DP IV. Additionally, hydrogen bonds were formed between the backbone carbonyl group of Val4 of the peptide and the backbone amide hydrogen of Ala743 of the enzyme, between the side chain carbonyl group of Asn7 of Tat(1–9) and the side chain of Lys554 of DP IV and finally, between the carbonyl group of Ile8 of the peptide interacting with the Tyr752 hydroxyl group of the enzyme. A hydrophobic interactionofthesidechainofMet1ofTat(1–9)withthe phenyl ring of the substrate Ala-Pro-pNA resulted in a fixation of the aromatic leaving group. Rather similar interactions were obtained for docking of Trp2-Tat(1–9) to DP IV in the presence of the substrate Ala-Pro-pNA. The salt bridges formed between the ligand and DP IV were identical to those described above for Tat(1–9). However, the side chain of Trp2 of this peptide forms additional hydrophobic interactions with Ile742 of DP IV. Furthermore, Trp2 interacts with the aromatic moiety of the substrate. Discussion DP IV cleaves oligopeptides at their N-termini by removing two amino acids, and has a preference for the penultimate amino acid residue to be proline [4]. Peptides containing amino acids other than proline in this position (Ala, Gly, Ser, Thr) are also cleaved but with strongly reduced efficiency [36,37]. DP IV is not able to catalyze the hydrolysis of peptides with proline as the third amino acid. In these cases the oligopeptides function as inhibitors of the enzyme [25,38]. From the biomedical point of view the importance of DP IV as a costimulatory molecule in T cell activation processes [8–11], as a hydrolytic enzyme of regulatory peptides [39] and as an adhesion molecule [12] is well characterized. Furthermore, it was shown that synthetic DP IV inhibitors induce immunosuppressive effects result- ing from the reduction of DNA synthesis and cytokine production (IL-2, IL-10, IL-12 and IFN-c)ofstimulated peripheral blood mononuclear cells [11]. Therefore, it was assumed that DP IV participates in signal transduction processes. The inhibition of DP IV by the HIV-1 Tat protein, a viral protein responsible for transactivation of viral genes, has been shown previously [21,22]. We demon- strated that the N-terminal amino acid sequence of Tat represents an important motif for DP IV inhibition [22]. Analogous to synthetic DP IV inhibitors, Tat(1–9) suppres- ses the DNA synthesis of stimulated peripheral blood mononuclear cells reflecting the possible role of Tat–DP IV interactions in AIDS [22]. The function of the viral protein Tat as an immunomodulatory oligopeptide implies the existence of soluble or cell surface-expressed endogenous DP IV-inhibitory molecules. One of them could be the thromboxane A2 receptor carrying the strong inhibitory sequence MWP at the N-terminus [24]. The compounds examined in this work are mainly Tat (1–9)-derived peptides as well as other oligopeptides with the N-terminal XXP motif. All peptides were characterized as inhibitors of DP IV. While in earlier studies the N-terminal XXP was described to be the essential sequence motif of DP IV inhibitory peptides [25], we found that oligopeptides with special proline substitutions in the third Fig. 4. Dixon plot and slope replot of a competitive inhibition. (A) Influence of eight fixed concentrations of TXA2-R(1–9) (1 · 10 )6 M to 2 · 10 )5 M ) on the hydrolysis of five fixed substrate concentrations Ala-Pro-pNA (s,1· 10 )5 M ; h,1.5· 10 )5 M ; n,2· 10 )5 M ; ,, 4 · 10 )5 M ; e,8· 10 )5 M ) represented as a Dixon plot (1/v vs. [I]). (B) Slope replot of Dixon plot (slopes vs. 1/[S]). The enzymatic assays contained 0.04 M Tris/HCl buffer (pH 7.6, I ¼ 0.125), 4.04 · 10 )8 M DP IV and different inhibitor concentrations and were incubated at 30 °C. The hydrolysis of Ala-Pro-pNA was measured by detecting the released pNA at 390 nm over 120 s. 2152 S. Lorey et al. (Eur. J. Biochem. 270) Ó FEBS 2003 position are also inhibitors of DP IV, though with lower potency than known product analogues as DP IV inhibitors [28,40,41]. Surprisingly, although the tested compounds are structurally highly related, they differed not only in their K i values but also in their inhibition type. Therefore, the present investigations were focused on the mechanistic analysis of the inhibition mode of DP IV in order to obtain a deeper insight into the possible enzyme–inhibitor interactions based on kinetic measurements and molecular modeling studies. For DP IV/CD26 several inhibition modes are known: competitive, noncompetitive, mixed-type, irreversible, etc. [28]. In all these cases, the enzyme inhibition takes place by binding to one site located in or out of the active site. Until now the inhibition of DP IV/CD26 via binding to two dif- ferent sites at the enzyme was unknown. This work showed for the first time that certain peptides may function as DP IV inhibitors according to a parabolic mixed-type mechanism that is characterized by the formation of an IEI complex consisting of one enzyme molecule and two inhibitor molecules one of them bound in the active site and one of them at an alternative site. This special, rare type of inhibi- tion is therefore worth examining although the kinetic con- stants characterize these inhibitors rather as weak inhibitors. The Tat(1–9)-related peptides inhibiting DP IV accord- ing to this in the literature as yet rarely described parabolic mixed-type mode were characterized by identical amino acids in position 2 and from positions 4–9 as well as by poor K i values in the range 10 )3 to 10 )4 M . In comparison to Tat(1–9), these compounds differ in only one amino acid position, either position 1 [Trp1-Tat(1–9)] or position 3 [Gly3-Tat(1–9) and Ile3-Tat(1–9)]. The negatively charged aspartic acid in position 2 seems to disturb binding of the corresponding peptide to the noncompetitive binding site of DP IV. In positions 1 and 3, a greater variability of the amino acids is allowed. Supporting this theory, the Tat(1– 9)-related peptides Lys2-Tat(1–9) and Trp2-Tat(1–9) derived by substitution of Asp2 inhibited DP IV with clearly lower K i values (10 -5 )10 -6 M ) and according to the linear mixed-type inhibition mode characterized by inhibitor binding only to the noncompetitive binding site. There- fore, the substitution of only one amino acid (Asp2) in the Tat(1–9) sequence resulted in a change of the inhibition mode in conjunction with a gain of the ability to bind at the noncompetitive site. The determination of the parabolic mixed-type inhibition mode raised questions according inactivation or precipita- tion of the enzyme and according enzyme and inhibitor stabilities under assay conditions. However, HPLC analysis demonstrated that the inhibitory peptides are not hydro- lyzed but are stable under test conditions (data not shown). For DP IV, dilution experiments showed that it retained its full biological activity at different inhibitor concentrations thereby excluding precipitation or inactivation. Moreover, for other structurally related peptidergic inhibitors under similar assay conditions, more common inhibition modes were observed suggesting that the measurements for the inhibitors following parabolic mixed-type mechanism did not have basic deficiencies, such as enzyme inactivation or precipitation. The kinetic data also provide evidence that the range of K i values (10 )3 M to 10 )6 M ) and the different modes of Fig. 5. Stereo-representation of the interaction of Tat(1–9) with the substrate Ala-Pro-pNA at the active site of a model of DP IV. Carbon atoms are colored orange [Tat(1–9)], magenta (Ala-Pro-pNA) or gray (DP IV). For clarity only amino acid residues of DP IV essential for the interaction with the ligands are depicted. Ó FEBS 2003 DP IV inhibition by XXP containing oligopeptides (Eur. J. Biochem. 270) 2153 inhibition of DP IV by the oligopeptides are not only affected by the XXP (or XXG, XXI, XXV) sequence motif but also by the subsequent amino acids. Trp2-Tat(1–9), TXA2-R(1–9), Met-Trp1-G-CSF(1–8), Met-Trp1-IL-2(1–8) and the tripeptide MWP contain the identical N-terminal sequence MWP. Nevertheless, Trp2-Tat(1–9) and Met- Trp1-G-CSF(1–8) represented linear mixed-type inhibitors, whereas TXA2-R(1–9), Met-Trp1-IL-2(1–8) and MWP inhibited DP IV competitively. Therefore, it seems to be most probable that the MWP motif alone is not responsible for the inhibition mode especially with regard to the different subsequent amino acid sequences of these peptides. On the other hand, the K i value seems to be strongly influenced bythe amino acid in the second position indicating compounds with tryptophan in this position as the most potent inhibitors compared to those without tryptophan in the second position as shown in the present study. Trp2- Tat(1–9) was the inhibitor with the lowest K i value (2.12 · 10 )6 M ) of all compounds with the N-terminal XXP sequence tested so far. This inhibition constant is in the same range as the K i values of inhibition of human recombinant DP IV by the product analogue amino acid pyrrolidides, e.g. Val-pyrrolidide (K i ¼ 1.08 · 10 )6 M )and Lys[Z(NO 2 )]-pyrrolidide (K i ¼ 0.42 · 10 )6 M )(A.Sto ¨ ckel- Maschek, unpublished results) as well as of the inhibitors TMC-2A and TSL-225 (K i values of 5.3 · 10 )6 M and 3.6 · 10 )6 M , respectively), which exert anti-inflammatory effects on experimentally induced arthritis in rat [42]. The K i values of all oligopeptides with the N-terminal MWP motif were determined to be in the range 10 )6 M to 10 )5 M indicating that compounds with tryptophan in position 2 were the most potent inhibitors we examined. Comparing Tat(1–9) and Trp2-Tat(1–9) by conformational analysis, we have shown that the backbone conformations of these two oligopeptides are not significantly altered [24]. Therefore, the side chain of Trp2 is clearly responsible for the enhanced inhibitory potency. Conformational alterations of the peptide backbones have to be taken into consideration especially in the case of different amino acid sequences from positions 4–9 in some peptides. The flexibility of the peptide backbone of Tat(1–9) is restricted by two proline residues at positions 3 and 6 resulting in a relatively rigid conformation. This is likely to contribute to the nature of enzyme inhibitor interactions. Concordantly, all nonapeptides inhibiting DP IV according to the linear mixed-type mechanism contained both of these proline residues whereas peptides inhibiting DP IV com- petitively contained only one proline residue in positions 3 or 6 (Table 1). Comparing inhibition of DP IV by Trp2- Tat(1–9) with that of Trp2,Ile3-Tat(1–9) it was shown that the substitution of proline in the third position resulted in a change of the inhibition mode from a linear mixed-type to a competitive mechanism. Using chromogenic substrates such as Ala-Pro-pNA allowing online measurement of enzymatic hydrolysis, here we identified Tat(1–9) as a parabolic mixed-type inhibitor with a K i of 2.67 · 10 )4 M (Table 1). In previous studies, in a DP IV assay using capillary electrophoresis-based analysis of the hydrolysis of a more physiological substrate, the N-terminal peptide IL-2(1–12), Tat(1–9) was found to be a competitive inhibitor with a K i of (1.11 ± 0.12) · 10 )4 M [23]. One possible explanation for these, on the first view contradictory results, could be the usage of different substrates. Therefore, the influence of steric requirements of different substrates on DP IV inhibition was examined using the substrate Gly-Pro-R110-CO-(CH 2 ) 4 Cl containing a chain length roughly corresponding to a pentapeptide. In comparison to the small substrate Ala-Pro-pNA, the larger substrate Gly-Pro-R110-CO-(CH 2 ) 4 Cl did not affect the type of inhibition and the K i value of the oligopeptides Tat(1–9) and Trp2-Tat(1–9). On the other hand, corres- ponding to the fourfold difference of the K m values of the hydrolysis of both substrates, the factors a and d were reduced using Gly-Pro-R110-CO-(CH 2 ) 4 Cl. Therefore, in the presence of the latter substrate, a decreased substrate affinity resulted in an increased affinity of the inhibitor to the noncompetitive binding site of the enzyme implying possible interactions between the ligand and the substrate. Presumably, because of the definitely shorter chain length of Gly-Pro-R110-CO-(CH 2 ) 4 Cl in comparison to the dodeca- peptide IL-2(1–12), the different results for DP IV inhibi- tion by Tat(1–9) obtained with IL-2(1–12) and the chromogenic substrates could not be explained with Gly- Pro-R110-CO-(CH 2 ) 4 Cl. In order to examine the binding of inhibitory peptides to the noncompetitive binding site of substrate-loaded DP IV, docking studies of Tat(1–9) and Trp2-Tat(1–9) with DP IV in the presence of the substrate Ala-Pro-pNA, located at the active site, were carried out on the basis of our 3D model of the DP IV active site. From this, the preference for interac- tion of the acidic C-terminus (Glu9) of both peptides with the basic amino acid residues of DP IV Arg560 and Lys554, as we postulated earlier [30], was demonstrated. These inter- actions might be mainly responsible for the binding of Tat(1–9)-related peptides. Furthermore, it could be demon- strated that the protonated, positivelycharged N-terminus of the peptides is able to interact with both Asp709 and Asp739 of DP IV resulting in a considerable stabilization of the complex. The interactions of the C-terminus as well as the N-terminus of Tat(1–9)-related peptides permitted the dock- ing of the inhibitor close to the active site but not directly inside, thereby allowing the substrate Ala-Pro-pNA to bind to the active site. However, it can be assumed that the binding of larger substrates directly interferes with binding of Tat(1–9). This could be a possible explanation for the competitive character of DP IV inhibition by Tat(1–9) we observed in previous studies using the long IL-2(1–12) substrate [23]. Additionally, multiple interactions of Tat(1–9) with DP IV contributed to the attractive interaction between the inhibitor and the enzyme. In the case of Tat(1–9) containing Asp2, however, this negatively charged residue was only able to interact with the N-terminus of the peptide itself but not with the enzyme. Interestingly, a hydrophobic interactionofthesidechainofMet1ofTat(1–9)withthe phenyl ring of Ala-Pro-pNA may hinder the DP IV- catalyzed cleavage of the substrate. In comparison, the salt bridges between Trp2-Tat(1–9) and DP IV were identical to those determined for Tat(1–9). However, the side chain of Trp2 could interact with Ile742. Furthermore, the indole ring of Trp2 can form strong hydrophobic interactions with the phenyl ring of Ala-Pro- pNA resulting in a fixation of the substrate These additional attractive hydrophobic interactions seem to be responsible for the improved inhibitory capacity of Trp2-Tat(1–9). 2154 S. Lorey et al. (Eur. J. Biochem. 270) Ó FEBS 2003 Together with former studies, describing competitive binding of Tat(1–9) directly to the empty active site of DP IV [23], the docking studies demonstrate the possible binding of Tat(1–9) to two different binding sites, one binding site at the active site and one at the noncompetitive site close to the Ala-Pro-pNA-loaded active site, thus confirming the results of the inhibition studies with Tat(1–9). Moreover, the docking studies give a suggestion for the different inhibition modes observed for Tat(1–9) with both the chromogenic substrates and the longer substrate IL-2(1–12). Trp2-Tat(1–9) exhibiting increased inhibitory capacity interacts with DP IV at the noncompetitive binding site close to the active site, and its lower K i can be explained by additional attractive interactions formed between the Trp side chain and DP IV. Furthermore, the inhibitors stabilize the leaving group of the bound Ala-Pro–pNA by inter- actions either with the side chain of Met1 in Tat-(1–9) or with Trp2 in Trp2-Tat(1–9), thereby hindering the DP IV- catalyzed cleavage of the substrate. Very recently, the crystal structure of human DP IV in complex with the competitive inhibitor valine-pyrrolidide (Val-Pyr) has been reported [43]. As it was outlined by both Rasmussen et al. [43] and Gorrell [44], the structure of the active site of the DP IV model developed by us is in good agreement with that of the reported crystal structure of DP IV, particularly concerning the oxyanion hole formed by Tyr547 together with the backbone NH of Tyr631. In conclusion, the kinetic investigations presented here revealed different modes of DP IV inhibition by peptides with the N-terminal XXP motif. We detected peptides inhibiting DP IV according to the until now rarely described parabolic mixed-type mechanism indicating binding of two inhibitor molecules to two different binding sites at the enzyme; furthermore, we could show that single amino acid substitutions at certain positions of the parent structure alter the mode of inhibition indicating binding of the peptide to another binding site. In addition to differences in binding behavior, the compounds varied in inhibitor potency over three orders of magnitude. The inhibition of DP IV by the peptides Trp2,Ile3-Tat(1–9), Gly3-Tat(1–9), Ile3-Tat(1–9) and MWV demonstrated that the N-terminal XXP sequence is not the essential structural motif. Tat(1–9)-related peptides with the sub- stition of Pro3 by other amino acids (Gly, Ile, Val) also inhibited DP IV, though with lower inhibitory capacity indicated by higher K i values. Furthermore, it was shown that enzyme–inhibitor interactions depend on multiple factors such as the amino acid sequence and the conformation of the peptide backbone of the inhibitor or specific interactions between the inhibitor and the bound substrate. On the basis of the active site-containing 3D model of the C-terminal region of DP IV developed by us [30], evidence for possible molecular interactions of the inhibitory molecules with DP IV was presented. The recently reported crystal structure of DP IV [43] provides a framework for future work and the basis for the investigation of the protein-bound, pharmacophore con- formation of the ligands. The stronger inhibitory potency of MWP-containing peptides [Trp2-Tat(1–9) and TXA2- R(1–9)] towards DP IV activity and DNA synthesis among those peptides studied in the present work underlines the importance of interactions between endog- enous peptidergic ligands and DP IV, especially with regard to the role of DP IV in activation and proliferation of lymphocytes [24]. Our investigations demonstrate for the first time the existence of different inhibitor binding sites of DP IV indicating the complex manner of DP IV– inhibitory peptide interactions and therefore contribute to the understanding of physiological effects mediated by Tat(1–9) and its analogs. Additional knowledge of the molecular mechanisms of inhibitor–DP IV interactions is important for the development of more potent and more selective DP IV inhibitors as therapeutics in diseases including diabetes and multiple sclerosis. Acknowledgements Financial support was obtained from the Deutsche Forschungs- gemeinschaft, SFB 387 and NE 501/2-1, and is gratefully acknowledged. References 1. Hopsu-Havu, V.K. & Glenner, G.G. (1966) A new dipeptide naphthylamidase hydrolyzing glycyl-prolyl-beta-naphthylamide. Histochemie 7, 197–201. 2. Ikehara, Y., Ogata, S. & Misumi, Y. (1994) Dipeptidyl-peptidase IV from rat liver. Methods Enzymol. 244, 215–227. 3. Darmoul, D., Voisin, T., Couvineau, A., Rouyer-Fessard, C., Salomon,R.,Wang,Y.,Swallow,D.M.&Laburthe,M.(1994) Regional expression of epithelial dipeptidyl peptidase IV in the human intestines. Biochem. Biophys. Res. Commun. 203, 1224– 1229. 4. Heins, J., Welker, P., Scho ¨ nlein, C., Born, I., Hartrodt, B., Neubert, K., Tsuru, D. & Barth, A. (1988) Mechanism of proline- specific proteinases: (I) Substrate specificity of dipeptidyl peptidase IV from pig kidney and proline-specific endopeptidase from Flavobacterium meningosepticum. Biochim. Biophys. Acta 954, 161–169. 5. Scho ¨ n, E. & Ansorge, S. (1990) Dipeptidyl peptidase IV in the immune system. Cytofluorometric evidence for induction of the enzyme on activated T lymphocytes. Biol. Chem. Hoppe-Seyler 371, 699–705. 6. Bu ¨ hling, F., Junker, U., Reinhold, D., Neubert, K., Ja ¨ ger, L. & Ansorge, S. (1995) Functional role of CD26 on human B lym- phocytes. Immunol. Lett. 45, 47–51. 7. Bu ¨ hling, F., Kunz, D., Reinhold, D., Ulmer, A.J., Ernst, M., Flad, H D. & Ansorge, S. (1994) Expression and functional role of dipeptidyl peptidase IV (CD26) on human natural killer cells. Nat. Immun. 13, 270–279. 8. Morimoto,C.,Torimoto,Y.,Levinson,G.,Rudd,C.E.,Schrie- ber, M., Dang, N.H., Letvin, N.L. & Schlossman, S.F. (1989) 1F7, a novel cell surface molecule, involved in helper function of CD4 cells. J. Immunol. 143, 3430–3439. 9. Dang, N.H., Torimoto, Y., Deusch, K., Schlossman, S.F. & Morimoto, C. (1990) Comitogenic effect of solid-phase immo- bilized anti-1F7 on human CD4 T cell activation via CD3 and CD2 pathways. J. Immunol. 144, 4092–4100. 10. Dang, N.H., Torimoto, Y., Shimamura, K., Tanaka, T., Daley, J.F., Schlossman, S.F. & Morimoto, C. (1991) 1F7 (CD26): a marker of thymic maturation involved in the differential regula- tion of the CD3 and CD2 pathways of human thymocyte acti- vation. J. Immunol. 147, 2825–2832. 11. Ka ¨ hne, T., Lendeckel, U., Wrenger, S., Neubert, K., Ansorge, S. & Reinhold, D. (1999) Dipeptidyl peptidase IV: a cell surface peptidase involved in regulating T cell growth. Int. J. Mol. Med. 4, 3–15. Ó FEBS 2003 DP IV inhibition by XXP containing oligopeptides (Eur. J. Biochem. 270) 2155 12. Kameoka, J., Tanaka, T., Nojima, J., Schlossman, S.F. & Mori- moto, C. (1993) Direct association of adenosine deaminase with a T cell activation antigen, CD26. Science 261, 466–469. 13. Oravecz,T.,Pall,M.,Roderiquez,G.,Gorrell,M.D.,Ditto,M., Nguyen,N.Y.,Boykins,R.,Unsworth,E.&Norcross,M.A. (1997) Regulation of the receptor specificity and function of the chemokineRANTES(regulatedonactivation,normalTcell expressed and secreted) by dipeptidyl peptidase IV (CD26)- mediated cleavage. J. Exp. Med. 186, 1865–1872. 14. Proost, P., Struyf, S., Schols, D., Durinx, C., Wuyts, A., Lenaerts, J P., De Clercq, E., De Meester, I. & Van Damme, J. (1998) Processing by CD26/dipeptidyl-peptidase IV reduces the chemo- tactic and anti-HIV-1 activity of stromal-cell-derived factor-1al- pha. FEBS Lett. 432, 73–76. 15. Holst, J.J. & Deacon, C.F. (1998) Inhibition of the activity of dipeptidyl-peptidase IV as a treatment for type 2 diabetes. Dia- betes 47, 1663–1670. 16. Reinhold,D.,Hemmer,B.,Gran,B., Born,I.,Faust,J.,Neubert,K., McFarland, H.F., Martin, R. & Ansorge, S. (1998) Inhibitors of dipeptidyl peptidase IV/CD26 suppress activation of human MBP-specific CD4+ T cell clones. J. Neuroimmunol. 87, 203–209. 17. Sodroski, J.G., Rosen, C.R., Wong-Staal, F., Salahuddin, S.Z., Popovic, M., Arya, S., Gallo, R.C. & Haseltine, W.A. (1985) Trans-acting transcriptional regulation of human T-cell leukemia virus type III long terminal repeat. Science 227, 171–173. 18. Ensoli, B., Barillari, G., Salahuddin, S.Z., Gallo, R.C. & Wong- Staal, F. (1990) Tat protein of HIV-1 stimulates growth of cells derived from Kaposi’s sarcoma lesions of AIDS patients. Nature 345, 84–86. 19. Viscidi, R.P., Mayur, K., Lederman, H.M. & Frankel, A.D. (1989) Inhibition of antigen-induced lymphocyte proliferation by Tat protein from HIV-1. Science 246, 1606–1608. 20. Subramanyam, M., Gutheil, W.G., Bachovchin, W.W. & Huber, B.T. (1993) Mechanism of HIV-1 Tat induced inhibition of anti- gen-specific T cell responsiveness. J. Immunol. 150, 2544–2553. 21. Gutheil, W.G., Subramanyam, M., Flentke, G.R., Sanford, D.G., Munoz, E., Huber, B.T. & Bachovchin, W.W. (1994) Human immunodeficiency virus 1 Tat binds to dipeptidyl aminopeptidase IV (CD26): a possible mechanism for Tat’s immunosuppressive activity. Proc. Natl Acad. Sci. USA 91, 6594–6598. 22. Wrenger, S., Reinhold, D., Hoffmann, T., Kraft, M., Frank, R., Faust, J., Neubert, K. & Ansorge, S. (1996) The N-terminal X–X-Pro sequence of the HIV-1 Tat protein is important for the inhibition of dipeptidyl peptidase IV (DP IV/CD26) and the suppression of mitogen-induced proliferation of human T cells. FEBS Lett. 383, 145–149. 23. Wrenger, S., Hoffmann, T., Faust, J., Mrestani-Klaus, C., Brandt, W.,Neubert,K.,Kraft,M.,Olek,S.,Frank,R.,Ansorge,S.& Reinhold, D. (1997) The N-terminal structure of HIV-1 Tat is required for suppression of CD26-dependent T cell growth. J. Biol. Chem. 272, 30283–30288. 24. Wrenger, S., Faust, J., Mrestani-Klaus, C., Fengler, A., Sto ¨ ckel- Maschek, A., Lorey, S., Ka ¨ hne,T.,Brandt,W.,Neubert,K., Ansorge, S. & Reinhold, D. (2000) Down-regulation of T cell activation following inhibition of dipeptidyl peptidase IV/CD26 by the N-terminal part of the thromboxane A2 receptor. J. Biol. Chem. 275, 22180–22186. 25. Hoffmann, T., Reinhold, D., Ka ¨ hne, T., Faust, J., Neubert, K., Frank, R. & Ansorge, S. (1995) Inhibition of dipeptidyl peptidase IV (DP IV) by anti-DP IV antibodies and non-substrate X–X-Pro- oligopeptides ascertained by capillary electrophoresis. J. Chro- matogr. A. 716, 355–362. 26. Mrestani-Klaus, C., Lorey, S., Faust, J., Bu ¨ hling, F. & Neubert, K. (2002) Detection of the activity of the ectopeptidases DPIV and APN using sensitive fluorogenic substrates. In Ectopeptidases: CD13/Aminopeptidase N and CD26/Dipeptidylpeptidase IV in Medicine and Biology (Langner, J. & Ansorge, S., eds), pp. 1–24. Kluwer Academic/Plenum Publishers, NY, USA. 27. Lorey, S. (1999) Fluorogenic substrates and inhibitors for the detectionofDPIVactivityonimmunecells.PhDThesis,Martin- Luther-University Halle-Wittenberg, Germany. 28. Sto ¨ ckel-Maschek, A., Stiebitz, B., Born, I., Faust, J., Mo ¨ gelin, W. & Neubert, K. (2000) Potent inhibitors of dipeptidyl peptidase IV and their mechanisms of inhibition. In Cellular Peptidases in Immune Functions and Diseases (2) (Langner, J. & Ansorge, S., eds), pp. 117–123. Kluwer Academic/Plenum Publishers, NY, USA. 29. Segel, I.H. (1993) Enzyme Kinetics. Behavior and Analysis of Rapid Equilibrium and Steady-State Enzyme Systems. John Wiley & Sons, Inc., NY, USA. 30. Brandt, W. (2000) Development of a tertiary-structure model of the C-terminal domain of DPP IV. In Cellular Peptidases in Immune Functions and Diseases (2) (Langner, J. & Ansorge, S., eds), pp. 97–101. Kluwer Academic/Plenum Publishers, NY, USA. 31. Clark, M., Cramer, R.D. III & Van Opdenbosch, N.J. (1989) Validation of the general purpose Tripos 5.2 force field. Comp. Chem. 10, 982–1012. 32. Gasteiger, J. & Marsili, M. (1980) Iterative partial equalization of orbital electronegativity – a rapid access to atomic charges. Tetrahedron 36, 3219–3228. 33. Tabatabai, L.B. & Graves, D.J. (1978) Kinetic mechanism and specificity of the phosphorylase kinase reaction. J. Biol. Chem. 253, 2196–2202. 34. Stephens, D.T., Whaley, K.J., Klimkow, N.M., Goh, P. & Hoskins, D.D. (1986) Kinetic characterization of the inhibition of purified cynomolgus monkey lactate dehydrogenase isozymes by gossypol. J. Androl. 7, 367–377. 35. Fu ¨ lo ¨ p, V., Bo ¨ cskei, Z. & Polga ´ r, L. (1998) Prolyl oligopeptidase: an unusual b-propeller domain regulates proteolysis. Cell 94, 161–170. 36. Bongers, J., Lambros, T., Ahmad, M. & Heimer, E.P. (1992) Kinetics of dipeptidyl peptidase IV proteolysis of growth hor- mone-releasing factor and analogs. Biochim. Biophys. Acta 1122, 147–153. 37. Martin, R.A., Cleary, D.L., Guido, D.M., Zurcher-Neely, H.A. & Kubiak, T.M. (1993) Dipeptidyl peptidase IV (DPP-IV) from pig kidney cleaves analogs of bovine growth hormone-releasing factor (bGRF) modified at position 2 with Ser, Thr or Val. Extended DPP-IV substrate specificity? Biochim. Biophys. Acta 1164, 252–260. 38. Harada, M., Fukasawa, K.M., Fukasawa, K. & Nagatsu, T. (1982) Inhibitory action of proline-containing peptides on Xaa-Pro- dipeptidylaminopeptidase. Biochim. Biophys. Acta 705, 288–290. 39. Mentlein, R. (1999) Dipeptidyl-peptidase IV (CD26) – role in the inactivation of regulatory peptides. Regul. Pept. 85, 9–24. 40. Scho ¨ n, E., Demuth, H U., Eichmann, E., Horst, H J., Ko ¨ rner, I J., Kopp, J., Mattern, T., Neubert, K., Noll, F., Ulmer, A.J., Barth, A. & Ansorge, S. (1989) Dipeptidyl peptidase IV in human T lymphocytes. Scand. J. Immunol. 29, 127–132. 41. Augustyns, K., Bal, G., Thonus, G., Belyaev, A., Zhang, X.M., Bollaert, W., Lambeir, A.M., Durinx, C., Goossens, F. & Hae- mers, A. (1999) The unique properties of dipeptidyl-peptidase IV (DPP IV/CD26) and the therapeutic potential of DPP IV inhibi- tors. Curr. Med. Chem. 6, 311–327. 42. Tanaka, S., Murakami, T., Nonaka, N., Ohnuki, T., Yamada, M. & Sugita, T. (1998) Anti-arthritic effects of the novel dipeptidyl peptidase IV inhibitors TMC-2A and TSL-225. Immuno- pharmacology 40, 21–26. 43. Rasmussen, H.B., Branner, S., Wiberg, F.C. & Wagtmann, N. (2003) Crystal structure of human dipeptidyl peptidase IV/CD26 in complex with a substrate analog. Nat. Struct. Biol. 10, 19–25. 44. Gorrell, M.D. (2003) First bite. Nat. Struct. Biol. 10,3–5. 2156 S. Lorey et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . Different modes of dipeptidyl peptidase IV (CD26) inhibition by oligopeptides derived from the N-terminus of HIV-1 Tat indicate at least two inhibitor. competitive binding of Tat( 1–9) directly to the empty active site of DP IV [23], the docking studies demonstrate the possible binding of Tat( 1–9) to two different