Dipeptidyl peptidase-4

The enzyme dipeptidyl peptidase 4 ( DPP 4 short, DP IV or DPP IV) was first described in 1964/66 of Hopsu - Havu and Glenner. The authors described the newly found enzyme as dipeptide Naphthylamidase. Some time later, the enzyme was independent of Hopsu - Havu and Glenner by Schulz and Barth wiederentdeckt.Entscheidend was for enzyme purification ion exchange chromatography with DEAE Sephadex A- 50th As a result of various studies on the chemistry of the enzyme was called dipeptidyl peptidase IV [DP IV]. In further studies, it was first noted that it is a serine protease. In the course of time has been recognized that this enzyme exerts major functions in the human and animal organism. After eating different hormones are released in the stomach, the so-called incretin hormones, which enhance the release of insulin in the body. The DPP 4 splits these incretins, thus stopping their effect. It follows that inhibitors of DPP are 4 potential drugs for diabetes mellitus type II, there can help an increased release of insulin to lower chronically high blood sugar. As a result, the study of the dipeptidyl peptidase -4 inhibitors and their two drugs, sitagliptin and vildagliptin, by the Committee for Human Use (CHMP ) have already been the European Medicines Agency approved as therapeutics for the treatment of diabetes mellitus type 2.

Against type I diabetes mellitus, inhibition of this enzyme would result in nothing: Here do not know any more insulin to be produced, because the islet cells of the pancreas are completely destroyed.

The functional mechanism from chemical point of view ( chemistry )

The enzyme DPP 4 is ubiquitous in the human, but also in the animal organism. It is a proteinase which cleaves pairs of amino acids ( dipeptide ) from the N -terminal end of a peptide. Since there is the amino acid serine in the active site of the enzyme, it belongs to the serine proteinases. It can only depend on the amino terminus (N- terminus) of a peptide cleave dipeptides, but not from the carboxy terminus, the C-terminus (see peptide bond ).

An exemplary peptide as substrate for DPP 4 is as follows, where Xaa 2, Xaa1 and X'aa1 for any amino acids and are the vertical bar to the point at which the enzyme cleaves the peptide bond. With P2, P1 and P'1 the positions of the amino acids are described relative to the cleavage site:

N- Xaa 2 - Xaa1 - | - X'aa1 ... P2 P1 P'1 N- Xaa2 - Xaa1 - C N X'aa1 ... The DPP 4 preferentially cleaves peptides when a proline (Pro) is located in the P1 position.

The substrate specificity in the P1 position ( Xaa1 )

In addition to peptides with proline in the P1 position, splits the DPP 4 also peptides with other amino acids in this position.

First, it was found that in addition to proline - hydroxyproline and Dehydroprolinreste also be very well accepted. Substrates with Pipecolinsäureresten ( Pip ) in P1- position are similarly active. In Scheme 1, a " Prolinstammbaum " is presented which leads directly from proline to glycine. Tentatively, the individual di- peptide -para- nitroanilides (pNA ) were exchanged in P1- position against the listed aminoacyl residues of the " family tree ". It could be shown that all these derivatives are more or less active substrates. Were tested successfully in Ala -pNA Xaa1 particularly Xaa1 = Thz, oxa, pipecolic acid ( Pip ) and azetidine-2 -carboxylic acid ( Aze ).

Further, the tested Xaa1 residues of glycine (Gly ), 2- aminobutyric acid (Abu ), norvaline ( Nva ), sarcosine ( Sar ), N -ethyl- glycine, N-(n -propyl ) glycine, N-methyl -2- aminobutyric acid, N- methyl alanine, and N-ethyl- alanine. It is interesting that there are 4 substrates is a gradation in the following direction with respect to the effectiveness of the DPP:

With respect to the kcat / Km value of the Aze derivative is similar to the corresponding pro- substrate. In one case it is a five- in the other case by a four-membered ring. Also of interest was the examination of the compounds Ala- Xaa1 -pNA with Xaa1 = a- 5MeOxa, and s- 5MeOxa. Ala- (S- 5MeOxa ) -pNA served as the substrate, Ala- (a- 5MeOxa ) non -pNA [ S = SYN, a = anti- ]. The proline is not built planar but is in a configuration similar to scheme 2 (left) before.

Consequently, the positions in the ring system of proline are to be evaluated differently. This explains why Xaa 2 -Hyp -pNA, a substrate, Xaa 2 - (a- 5MeOxa ) -pNA but is not hydrolyzed by DPP 4 enzymatically.

It is observed that the rate of cleavage depends in substrates of the type Xaa 2 - per - X'aa1 of the nature of the radical X'aa1. Obviously shifts in this case the rate-determining step ( see below ) from the deacylation (k3 ) in previous steps. Conceivable an obstacle to the influx of these substrates in the active site of DPP 4 is the slowest step here should possibly be before the actual catalytic process. Here is the simplest case (assuming ka much smaller than k3 and k3 is much smaller than k2 and k1)

Ka = rate constant of the influx of the substrate in the active site, ka = rate constant for the efflux of the substrate from the active site, k1 = rate constant of the tetrahedral intermediate, k2 = rate constant for the acylation step (TI1 ® acyl-enzyme ), k3 = rate constant for the deacylation step ( TI2 ® dipeptide ).

The substrate specificity in P2 position ( Xaa 2 )

The importance of the N- terminal amino acid residue in the P2 position ( Xaa 2 ) was elicited. Were examined, inter alia, in Xaa2 -Pro -pNA, the residues of: Pro, Abu, Leu, Val, Ala, Ile, Glu, Phe, Tyr, Ser, Gln, Lys, Asp, Asn and N, N -dimethyl- glycine, [( N, N ) - DMG ], N, N, N -trimethyl- glycine [ (N, N, N) - TMG ] and S, S -dimethyl- sulfonium -acetic acid [(S, S)- DMS]. All substances acted as substrates of DPP 4 (see Table 1).

How far can the N-terminal amino group of the peptide bond between P1 and P2 be? These amino acid residues of Ala, β -Ala, γ - Abu and ε - Ahx were used in Xaa 2 -Pro -pNA. The derivatives of Ala, β -Ala and γ - Abu are substrates of DPP 4 with a downward trend. The ε - Ahx connection is not hydrolyzed

The role of aminoacyl residues in P'1 position ( X'aa1 )

To clarify the meaning of aminoacyl residues in P'1 position, kinetic studies were performed with the tripeptides Ala-Pro- X'aa1. Surprisingly, it appears that no enzymatic hydrolysis takes place when X'aa1 = Pro, Hyp, Dehydroprolyl, pipecolyl, Aciridyl radicals or generally the scissile peptide bond -CO -NH- by-CO- N ( R) - replaced (R not equal to H ).

The " recognition structure " of the substrates of DPP 4

On the basis of the kinetic data extensive material intensive studies of the computer simulation was performed. A result of this effort was the formulation of a structure which must have all of the substrates to be recognized in the active site of the DPP 4 as substrates. This " detection structure " is shown in Figure 1.

Of importance is the internal H-bond. Any obstruction of their training leads to inactivity as a substrate. The fact that any substitution at the NH group of the scissile peptide bond, this prevented is the reason why all peptides having the structural feature -CO -N ( R) - are not accessible with R not equal to H have an enzymatic hydrolysis. It is expected that the internal hydrogen bond is prevented by:

It is known that have a measurable Prolinpeptide cis- configuration at the scissile peptide bond. We have calculated the ratio of trans to cis bond at the example of the substrate Ala -Pro -pNA and tracks the enzymatic hydrolysis using the rapid kinetics ( "Stopped - Flow Method "). It shows a biphasic turnover after time. In the first phase, a rapid dependent on the enzyme concentration reduction up to a percentage, which was found for the presence of the trans - peptide bond is observed, thereafter, a slow response curve which is determined by the enzyme - independent cis-trans conversion speed. Thus, as expected, cleavage takes place only on the trans bond, the cis bond is inactive.

The stereospecificity

Furthermore, the substrates in the P1 position are absolutely stereospecific. Ala- L-Pro -pNA, for example, is a good substrate, Ala- D-Pro - pNA is not enzymatically hydrolyzed. In compounds of the type D- Xaa 2 -Pro -pNA also no hydrolysis takes place. D- Xaa2 -Ala- pNA, but are substrates of DPP 4 Table 3 demonstrates on the example of the Ala-Ala -pNA, and D-Ala- Ala.pNA Aib - Ala -pNA. The phenomenon that D-Phe- Pro-pNA and D -Tyr- Pro-pNA are uncompetitive inhibitors of hydrolysis of Ala-Pro -pNA ( Ki values ​​are 0.35 ± 0.04 mM and 0.52 ± 0.04 mm ), after which both are not competitive inhibitors or mixed inhibitors and the compounds show at Ala-Ala -pNA as substrate no inhibitory effect, can be seen in connection with the discussion in this work theoretical conclusions.

It is striking that the stereoselectivity in P2, as shown in Table 3, are not comfortable but serious aufwirkt on the KM value, the kcat values ​​. In this effect, which is surprising, will be discussed further below.

The sequence of enzyme catalysis

It is believed that in the course of the enzymatic catalysis, the reaction sequence shown in Schena 3, is run through:

A distinction between an acylation and an Deacylierungprozess. The substrate enters the active site (see recognition complex) and it forms ES. It follows the tetrahedral intermediate of the acylation process TI1. After cleavage of the first cleavage product formed by addition of a water molecule, the second tetrahedral intermediate TI2 ( Deacylierungsprozess ). After cleavage of these ultimately the free enzyme and the second cleavage product ( dipeptide ) is released. The first tetrahedral intermediate TI1 is asymmetric. Theoretically, can form a S- and / or R form here. TI2 other hand, is symmetric. Here there are no antipodal structures.

First, the question of where is located in the substrates of the rate-determining step, whether in the field of acylation or deacylation. Rises Therefore investigations were carried out with the substrates Ala-Ala- Ala-Pro- anilide and anilides having different aryl rings substituted in the pH optimum. The Hansch Approach ( QSWA = Quantitative structure -activity analysis) did not yield any correlation in the substituted derivatives of the series Ala-Pro- anilide, but a such in the series Ala -Ala- anilide.

Substrate Ala -Ala -NH -C6H4 -R QSWA: lgkcat = 0.807 ± 0.186 s R: -H, pf, p- Cl, p- Br, p- CH3, p- OCH3, p- OC2H5, p- NO2, m- Cl, m- CH3, m- CF3, m- NO2 Substrate Ala -Pro -NH -C6H4 -R QSWA: no correlation It can be seen that the rate determining step is the Alaninreihe in the acylation in a range that is, where the substituted anilines are not cleaved (k2). In the Prolinreihe however, the rate-determining step should be the deacylation (k3 ). The measurements were made at the pH optimum ( see below) performed.

The question now is: can the conformation of the tetrahedral intermediate TI1 determine? For this purpose, an analysis QCAR ( Quantitative Conformation Activity Relationships ) was carried out in the Alaninreihe. The results suggest that the effects occur when the-NH- Ar radical exiting the tetrahedral intermediate TI1 (k2). From QCAR analysis namely follows a possible conformation of a " disabled ", the hydrogen atoms a and b, corresponding to the structure in Scheme 4, occurs during the rotation of the aromatic ring.

The Scheme 5 demonstrates the two antipodes of TI1. Only the form on the left leads to the steric hindrance of the above-mentioned hydrogen atoms, but not the structure shown on the right. This could be a reference to the stereospecificity of the tetrahedral intermediate TI1 be ( in conjunction with the results of the D isotope measurements - see below).

The studies on pH dependence, demonstrated by the DPP 4 from pig kidney cortex showed a pH optimum of about 6.7. The optimum temperature is about 30 ° C. The studies further revealed that the activity of the substrates a positive charge at the N -terminus is necessary. The protonation of the N-terminal primary amino group is necessary. But it is not a requirement. For substrate recognition, the existence of a positive charge is sufficient, as shown in Table 1. Strictly speaking, it is in the DPP 4 is not an aminopeptidase, but a Oniumacylaminoacyl peptidase. The catalytic cleavage of dipeptides from an oligo - or polypeptide from the N -terminal end but, seen physiologically, an essential process.

Secondary D- isotope effects and Solventisotopieeffekte in D2O

In Scheme 4 and Scheme 5 (left) a possible model of TI1 is shown. His confirmation should be possible by D - isotope effects, if on the one hand, an H / D exchange takes place in the center of asymmetry of the P1 amino acid and on the other hand carried a H / D exchange in the aromatic ring in the ortho position. It turns out that in both cases secondary D- isotope effects are to be measured.

With Ala-Ala -pNA (L -Ala- L-Ala- d1- pNA) secondary D- isotope can be measured at the pH optimum at room temperature. They amount in kcat = 1.27 and KM = 1.24. If the substrate is completely deuterated in the aromatic ring (Ala -Ala -NH- C6D4 - pNO2 ), then secondary D- isotope effects arise in kcat and KM in 1.05 ± 0.03.

It is interesting that a secondary D- isotope effect in kcat / KM, according to the scheme 6 occurs as a function of pH. The curve is remarkable. With falling pH values ​​of about 7.5 to about 5.0 no isotope effect is measurable. From pH 5.0 is observed an increasing D secondary isotope effect with decreasing pH, associated with a transition from k3 k2 according to the rate-determining step.

This result underlines the high probability of the structure shown in Scheme 4.

Measurements of the solvent D isotope found in the case of the substrate Ala-Pro -pNA and Gly-Pro- pNA with the located in the deacylation rate-determining step of a one- proton transfer, and the Ala-Ala -pNA, the rate-limiting step in the acylation lying, a two- proton transfer ( table 4).

Inhibitors

A number of dipeptides were assayed for their effect as (competitive ) inhibitor of the DPP 4. Table 5 gives an overview.

Of these dipeptides, which are cleavage products DPP 4 catalysis, the Ile-Pro with a Ki value of about 7.0 x 10-6 M, and the ε -Z (4- NO2) -Pro Lys (K is about 10 -6) interesting. By about an order of magnitude more effectively, the respective Pyrrolidide (Table 6).

X-ray crystal structure of dipeptidyl peptidase 4

Currently, intensive studies on the field of X-ray crystal structure analysis of the DPP 4 is in progress. The enzyme is typically composed of two identical subunits. Each of said subunits having an N-terminal peptide sequence that anchors the protein to the surface of a cell. It also was found higher aggregates of the soluble DPP 4 consisting of four identical subunits.

For certain statements a not so extensive in the detail of the Conolli surface is more favorable. This shows that the DPP 4 has an unusual surface structure. The active site is not on the surface outside, but in the interior of the protein. The access to the active site is characterized by a " tube " is possible, which has a larger and a smaller opening. The question is how and especially where carried out, for example, the substrate entrance?

If the electrostatic potential at the surface you look at, it is found that in the vicinity and within the large aperture a very strong negative potential with a gradient exists in the direction of opening ( Figure 3). The smaller opening other hand, has no such strong negative potential. It is therefore conceivable that the substrates with the positively charged N-terminus of the larger areas of the detection potential of the structure substrates to be stamped into this opening " drawn ". ( Representation of a subunit, human DPP 4).

The structure of the tetrahedral intermediate TI1

From the experimental finding that the functioning as substrate D-Ala -Ala- pNA acts on kcat, where km is opposite that of the substrate Ala-Ala -pNA is not affected (see Table 3 ), the explanation arises that the N terminal aminoacyl residue (probably with the positive charge ) directly contact the tetrahedral intermediate in the P1 position in interaction. Therefore, a hypothetical conformation was constructed as follows ( Scheme 7).

There seem to be two types of hydrogen bonds to give, namely those who are responsible for direct binding to the active site, that is, introduce the substrate or the inhibitor of the interaction- active residues of the protein and those that are directly part of the function mechanism. If in the course of the function mechanism of serine proteases tetrahedral intermediates TI1 and TI2 (see scheme 3) are formed, TI1 is asymmetric. As you raise the Carbonylatoms of the scissile peptide bond, two antipodal conformations can form from which a peptide as a substrate ( 1), which should be seen by others as an inhibitor (2 ) ( Scheme 5).

The decision as to the direction in which the carbonyl group rearing, that is, which is preferably formed of antipodal tetrahedral intermediates will depend on the direction in which the interactions of the protein side chains in the active site (eg, H -bonds ) are most effective or spatial adjustment of the effector most effectively carried out. So it is quite possible that an inhibitor molecule may also be substrate. A key driver of the separation of a part of the molecule from the tetrahedral intermediate is whether the principle of stereoelectronic control after Delongchamps is valid. It states that the cleavage of esters or amides via a tetrahedral intermediate may take place only if both remaining at the central sp3 carbon atom hetero atoms of one of its lone pairs antiperiplanar to align each scissile bond.

The X-ray structure analysis of the DPP 4/Pro-Pro-boronsäurekomplexes confirmed in an impressive way the structure adopted in Scheme 7 ( Figure 4). Moreover, it is from the hypothetical Scheme 7 shows clearly why the DPP 4 can not remove N- terminal amino acids, but only N -terminal dipeptides cleaved. Is of interest that the boronic acid complex is in fact a tetrahedral structure is formed, that an interaction takes place on the one hand with the Tyr547 of the protein and an O- atom of the TI ( error = 2.04 A), and on the other hand between the Asn710 of the protein, and the oxygen atom, carbonyl group of the N-terminal peptide bond interaction occurs ( error = 1.93 a). The latter H- bridge is useful. It promotes the establishment of the carbonyl group of the N -terminal peptide bond in the sense shown above (see Scheme 7). These ratios are shown in the section of the X-ray crystal structure of DPP 4/Boronsäure-Komplexes ( Figure 4).

See similar images by diprotin A ( Figure 5 - right) and tertBuGly -Pro -Ile (Figure 5 - left) structures produced. This is where asymmetric tetrahedral conformations. They present themselves as antipodal opposite adopted in the scheme 7 Structure dar. If these determined by the X-ray real antipodal structures be characteristic of transition -state inhibitors? Due to this hypothetical assumption could be explained why the compounds

The principle possess a substrate structure, inhibitors. The term " hypothetical " expresses that the X-ray structures shown here (Figure 5 ) represent the conformations of inhibitors, the conformations of the TI (TI1 ) of substrates can not be represented, under the experimental conditions of the X-ray crystal structure analysis.

The question: Are diprotin A or B diprotin substrates or inhibitors of DPP 4? - Was founded in 1991 by Rahfeld et al. provided. Umezawa et al. report that both Diprotine act as inhibitors of DPP 4. Rahfeld note that both Ile-Pro -Ile and Val-Pro -Leu are expected to be substrates of DPP 4. The authors speak of the inhibitory effect as a " kinetic artifact " The X-ray crystal structure studies provide evidence for the conformation in TI1 that should be assigned to the inhibitor molecule, because the substrate structure can not be detected with certainty under the crystallization conditions. Is all the more surprising that the Diprotine in the aqueous milieu (kinetic measurements) are completely hydrolyzed after a long time (after about 20.5 hours under the conditions reported in the literature ). Obviously, a balance of antipodal TI conformations in dilute aqueous solution is gradually getting back on again. About such a mechanism, the dualism in the Diprotinen is clear. Its Ki and Km values ​​are shown in Table 7.

Natural substrates of dipeptidyl peptidase 4

(see dipeptidyl peptidase 4)

Concluding Remarks

Dipeptidyl peptidase 4 is physiologically and pharmacologically are of great interest. Some active and specific inhibitors of the enzyme have been identified as potential drugs for diabetes mellitus type II. Meanwhile, the two dipeptidyl peptidase -4 inhibitors sitagliptin and vildagliptin as drugs for the treatment of type 2 diabetes mellitus are allowed. There is evidence that DPP 4 or related enzymes play a role in processes of wound healing, possibly in a number of cancers ( dipeptidyl peptidase IV and aminopeptidase N seem to be involved according to English language publications also in the training of Thyreoidalcarcinomen ) and in the pathogenesis of AIDS play.

Several years ago, the close relationship of the enzymes DP II and PSE was (proline -specific endopeptidase ), also PPCE (post- Proline Cleaving Enzyme ) called, described for DPP 4. Recently it has been found that the DPP 4 is a member of a family of dipeptidyl peptidases. It can be assumed that in this area through comparative mechanistic and physiological studies further insights into theoretical and applied issues will result.