Eur. J. Biochem. 214, 829-835 (1993)
`© FEBS 1993
`
`Dipeptidyl-peptidase IV hydrolyses gastric inhibitory polypeptide,
`glucagon-like peptide-1(7—36)amide, peptide histidine methionine
`and is responsible for their degradation in human serum
`
`Rolf MENTLEIN‘. Baptist GALLWITZ’ and Wolfgang E. SCHMIDT’
`‘ Anatomisches lnstitut and
`
`2 Abteilung Allgemeine Innere Medizin der Universitat Kiel, Germany
`
`(Received February 9/April 16, 1993) — EJB 93 0215/3
`
`Peptides of the glucagon/vasoactive-intestinal-peptide (VIP) peptide family share a considerable
`sequence similarity at their N-terminus. They either start with Tyr-Ala, His-Ala or His-Ser which
`might be in part potential targets for dipeptidyl-peptidase IV. a highly specialized aminopeptidase
`removing dipeptides only from peptides with N-terminal penultimate proline or alanine. Growth-
`honnone-releasing factor(1—29)amide and gastric inhibitory peptide/glucose-dependent insulino-
`tropic peptide (GIP) with terminal Tyr-Ala as well as glucagon-like peptide-1(7-36)amideIinsulino-
`tropin [GLP-1(7—36)amide] and peptide histidine methionine (PHM) with terminal His-Ala were
`hydrolysed to their des-Xaa-Ala derivatives by dipeptidyl-peptidase IV purified from human pla-
`centa. VIP with terminal His-Ser was not significantly degraded by the peptidase. The kinetics of
`the hydrolysis of GIP, GLP-1(7-36)amide and PHM were analyzed in detail. For these peptides K,.,
`values of 4-34 u.M and V,,,_,, values of 0.6-3.8 umol - min“ - mg protein" were detennined for the
`purified peptidase which should allow their enzymic degradation also at physiological, nanomolar
`concentrations. When human serum was incubated with GIP or GLP-1(7—36)amide the same frag-
`ments as with the purified dipeptidyl-peptidase IV, namely the des-Xaa-Ala peptides and 'Iyr-Ala
`in the case of GIP or His-Ala in the case of GLP-1(7-36)amide, were identified as the main
`degradation products of these peptide hormones. Incorporation of inhibitors specific for dipeptidyl-
`peptidase IV, 1 mM Lys-pyrrolidide or 0.1 mM diprotin A (Ile-Pro-Ile), completely abolished the
`production of these fragments by serum. It is concluded that dipeptidyl-peptidase IV initiates the
`metabolism of GIP and GLP-1(7-36)amide in human serum. Since an intact N-terminus is obligate
`for the biological activity of the members of the glucagon/VIP peptide family [e. g. GIP(3—42) is
`known to be inactive to release insulin in the presence of glucose as does intact GIP], dipeptidyl-
`peptidase-IV action inactivates these peptide hormones. The relevance of this finding for their
`inactivation and their determination by immunoassays is discussed.
`
`Dipeptidyl-peptidase IV (DPP IV) is a highly specialized
`aminopeptidase removing dipeptides from bioactive peptides
`and synthetic peptide substrates provided that proline or ala-
`nine are the penultimate N-terminal residues (Mentlein,
`1988, for review). Small peptides or chromogenic substrates
`with proline in this position are far better hydrolysed than
`those with alanine (Heins et al., 1988). DPP IV occurs in
`human serum, as an ectoenzyme on the surface of capillary
`endothelial cells, at kidney brush-border membranes, on the
`
`Correspondence to R. Mentlein, Univcrsitat Kiel, Anatomisches
`Institut. Olshausenstrassc 40-60, D-24118 Kiel, Germany
`Fax: +49 431 8801557.
`Abbreviations. DPP IV, dipeptidyl-peptidase IV; GIR gastric in-
`hibitory polypeptide or glucose-dependent insulinotropic polypep-
`tide; GLP-1(7-36)amide, glucagon-like peptide-1(7-36)amide or
`insulinotropin or preproglucagon(78-107)amide; GLP-2, glucagon-
`like peptide-2 or preproglucagon(126—159); GRF, growth-hob
`mone-releasing factor/hormone; PHI, peptide histidine isoleucine;
`Pl-IM, peptide histidine methionine; VIP, vasoactive intestinal pep-
`tide; PACAP, pituitary adenylate-cyclase-activating polypeptide.
`Enzyme. Dipeptidyl peptidase IV (EC 3.4.14.5).
`
`surface of hepatocytes (here termed also GP110 or OX-61
`antigen), on the surface of a subset of T-lymphocytes and
`thymocytes (here termed CD 26. or thymocyte-activating
`molecule) and other sites (Loijda, 1979; Nausch and Hey-
`mann, 1985; Mentlein et al., 1984; McCaughan et al., 1990).
`The enzyme has been shown to be responsible for the degra-
`dation and inactivation of circulating peptides with penulti-
`mate proline,
`like substance P (Heymann and Mentlein,
`1978; Ahmad et al., 1992), but also for gmwth-hormone
`releasing factor (GRF) with penultimate alanine (Frohman et
`al.,
`1989; Kubiak.
`1989; Boulanger
`et
`al.,
`1992).
`[Ala"]GRF(1-29)amide with penultimate Ala is even a
`comparably good substrate as a synthetic Pro’-containing de-
`rivative for purified DPP IV (Bongers et al., 1992). This sug-
`gests that the conformation or chain length may greatly influ-
`ence the cleavage of peptides with penultimate proline/ala-
`nine-residues by DPP IV.
`We therefore evaluated whether or not other peptide hor-
`mones related to GRF might be substrates for DPP IV, and
`whether this probable proteolytic degradation might be of
`relevance in the circulation. GRF belongs to the glucagonl
`
`Page 1 of 7
`
`Astraleneca Exhibit 2096
`
`Mylan v. Astraleneca
`IPR2015-01340
`
`AstraZeneca Exhibit 2096
`Mylan v. AstraZeneca
`IPR2015-01340
`
`Page 1 of 7
`
`

`
`830
`
`5
`
`10
`
`11: Alt Asp Ala Ile
`-— — Glu Gly Thr
`
`Phe Thr Ann Sor Tyr
`-
`Ila Sat Mp -
`
`...-29
`...-42
`
`hGflF(t-29)amide
`IIGIP
`
`— Th! set Aepval
`flip — Glu Gly Thr
`—
`Set Asp Glu Met
`fin — - Gty Set
`...-27
`flla - - Gly Val — - so: AspPho
`flip — - Gly Val
`— - Set up — ' ...-27
`
`...-30
`
`hGlF-1l730lI'n|th
`h¢3LP-2
`hPI-IM-27
`r PHI-27
`
`Val
`uh set — -
`flu Set - Gly Th:
`flu Bur Gln Gly Thr
`11]; Su_— Gly —
`
`— — AspAan -
`- — Set Glu Leu
`- — Set Asp -
`— — Asp — —
`
`...-28
`...-27
`...-29
`
`_
`hVlP
`hsecretin
`helucooon
`IIPACAP-38
`
`Fig. l. N-terminal sequencs of peptides related to growth hor-
`mone-releaslng factor (GRF). Penultimate alanine and serine resi-
`dues are in bold; N—terrninal tyrosine and histidine residues are un-
`derlinedz (—) identity to GRF. h. Human sequences: r. rat sequence.
`
`family
`secretin/vasoactive-intestinal-peptide(VIP) peptide
`(Fig. 1) which share N-terminal sequences of considerable
`similarity. A number of them begin with Tyr-Ala, namely
`GRF and gastric inhibitory polypeptidelglucose-dependent
`insulinotropic peptide (GIP), or with His-Ala, namely gluca-
`gon-like peptide-1(7-36)amideIinsulinotropin [GLP-1(7-
`36)amide]. glucagon-like peptide-2 (GLP-2), peptide histi-
`dine methionine (PHM) and peptide histidinc isoleucine
`(PHI, the rat counterpart of human PHM), whereas others
`have temiinal His-Ser (VIP and others). For biological activ-
`ity the N-temiinal moiety is supposed to be the detenninant
`for transducing the ligand message and the C-terminal moi-
`ety for playing the major role in specific binding (Christophe
`et al., 1989, for review). Thus, proteolytic truncation of the
`N-terrninus of the members of the glucagon/VIP family by
`DPP IV should inactivate them.
`
`EXPERIMENTAL PROCEDURES
`
`Peptides, inhibitors and enzymes
`
`Synthetic peptide hormones (human sequences) were ob-
`tained from Saxon Biochemicals, dipeptides and diprotin A
`were purchased from Bachem. Purity of peptides was
`checked by HPLC; their amino acid compositions were ana-
`lyzed by the manufacturer. Lys-pyrrolidide was a gift from
`Dr. Mike Schutkowski, Martin-Luther-Universitiit Hallel
`Saale, Gennany. Dipeptidyl-peptidase IV was purified from
`human placenta and free of contaminating proteases (Piischel
`et al., 1982).
`
`Degradation assays with purified enzyme
`
`5 nmol of the peptides (5 ul of a 1 mM solution in water)
`were incubated at 37 °C with 0.1 ug peptidase in 50 mM tri-
`ethanolamine/HCI, pl-17.8, for 10-60min in 5001;! (final
`peptide concentration 10 uM) or less (other peptide concen-
`trations). Enzymic reactions were terminated by addition of
`5 pl 10% trifluoroacetic acid, and the mixtures applied onto
`a Vydac C,. widepore (30-nm pores, 5-p.M particles)
`250 mm X 4.6 mm HPLC column and eluted at a flow rate of
`lml/min with gradients of acetonitrile in 0.1% trifluoro-
`acetic acid. Either a linear gradient of 0—80% acetonitrile
`formed within 42 min (GIP degradation), or a stepwise linear
`gradient of 0—32% acetonitrile formed in 17 min followed
`by linear gradient of 32-48% acetonitrile formed in 30 min
`
`Page 2 of 7
`
`(other peptides) were used for separations. In some HPLC
`separations, trifluoroacetic acid was replaced by heptafluoro-
`butyric acid. Peptides and their degradation products were
`monitored by their absorbance at 220 nm (peptide bonds) or
`280 nm (aromatic amino acids). They were quantified by in-
`tegration of their peak areas related to those of standards
`(synthetic Tyr-Ala or tumcated peptides made by complete
`dipeptidyl-peptidase IV digestion). The concentrations of all
`peptide solutions were routinely calculated from their absor-
`bance at 280nm relative to their content of Trp and Tyr
`(using additively the known absorption coefficients).
`Activities were determined front estimations with less
`than 10% substrate turnover. Catalytic constants were calcu-
`lated according to the statistical method of Wilkinson (1961).
`
`Degradation of peptides in serum
`
`200 pl serum of healthy males were incubated with 10 pl
`1 mM peptide solution in water (final concentration 20 uM)
`for 60 min at 37°C. Inhibitors were added as 10mM or
`100 mM stock solutions in water. Enzymic reactions were
`terminated by addition of 20 pl 10% trifluoroacetic acid.
`Samples were centrifuged (5 min 13000 X g), and the super-
`natant liquids applied to a C.. reverse-phase Sep-Pak car-
`tridge (Millipore-Waters) that had been previously activated
`and washed with 10 ml each of methanol, 80% acetonitrile
`in 0.1% trifluoroacetic acid and finally 0.1% trifluoroacetic
`acid. After washing of the serum-loaded cartridges with
`20 ml 0.1% trifluoroacetic acid, peptides were eluted with
`2 ml 80% acetonitiile in 0.1% trifluoroacetic acid. Acetoni-
`
`trile eluates were lyophilized, dissolved in 100 pl 0.1% triflu-
`oroacetic acid and analyzed as described above. Non-bound
`supcmatants and washings were combined, lyophilized, re-
`acted with 4-dimethylaminoazobenzene-4-sulphonyl chloride
`and separated by reverse-phase HPLC as described by
`Stocchi et al. (1985) for amino acids.
`
`Peptide chemistry and other assays
`
`Fragments separated by HPLC were collected and lyo-
`philized for chemical determinations. Amino acid composi-
`tion was determined by acid hydrolysis (6 M HCl in vacuo
`at 100°C for 24 h) followed by lyophilisation, reaction with
`4-dimethylarninoazobenzene-sulphonyl-chloride and HPLC
`separation of derivatized amino acids (Stocchi et al.. 1985).
`N-terminal amino acids were determined by manual microse-
`quencing with 4-N,N-dimethylaminoazobenzene-4’-isothio-
`cyanate (Wittmarm-Liebold et al., 1986).
`Hydrolysis of 0.5 mM Gly-Pro-4-nitranilide at pH 8.6
`and at 37°C was monitored as described (Mentlein and
`Struckhoff, 1989).
`
`RESULTS
`
`Digestion of peptides by purified DPP IV
`
`DPP IV purified from human placenta liberated Tyr-Ala
`from GRF(1—29)amide and GIP, and His-Ala from GLP-
`1(7—36)amide and Pl-lM (Table 1, Fig. 2). No further proteo-
`lytic cleavage of these peptides was observed indicating the
`high specificity of the DPP IV for N-tertninal Xaa-Ala (and
`Xaa-Pro) and the absence of contaminating proteases in the
`enzyme preparation. Liberated 'Iyr-Ala (Fig. 2) was iden-
`titied by its retention time and co-chromatography with a
`synthetic standard. His-Ala was adsorbed to the C”, column
`
`Page 2 of 7
`
`

`
`‘liable 1. Cleavage rates for proteolysis of peptides from the GRFI
`VIP family by DPP IV purified from human placenta. Data are
`means of threedeterminations. variations were less than 10%.
`
`Peptide
`
`Concentration
`
`Cleavage rate
`
`GRF(l —29)amide
`
`GRF(1 —44)amide
`GIP
`
`GLP-1(7-36)amide
`
`PHM
`
`VIP
`
`uM
`20
`150
`
`150
`20
`100
`
`20
`100
`
`20
`100
`
`20
`100
`
`pmol - min" - mg"
`4.4
`5.0“
`
`4.5“
`1.4
`2.9
`
`0.79
`0.35
`
`0.47
`0.58
`
`<0.02
`<0.02
`
`‘Data taken from Bongers et al. (1992).
`
`only with heptafluorobutyric acid as ion-pairing reagent (Ta-
`ble 2) which, however, resulted in a relatively high back-
`ground. Therefore, liberated His-Ala was also identified as
`its 4-dimemylaminoazobenzene-sulphonyl-derivative
`(ob-
`tained also with a synthetic dipeptide standard). Moreover,
`the truncated peptides could be separated from the non-de-
`graded ones in reverse-phase HPLC (Fig. 2, Table 2).
`Highest
`initial velocities for DPP-IV degradation at
`micromolar peptide concentrations were found for GRF(1-
`29)amide. whereas those for other members of the VIP/glu-
`cagon-related peptides with penultimate Ala were lower (Ta-
`ble 1). No significant cleavage was observed with VIP tested
`as a representative member of this peptide family with N-
`terminal His-Ser. DPP IV hydrolysed GIP, GLP-1(7—36)am-
`ide and PHM with K, values in the range 4-34 uM (Table
`
`831
`
`3). These values are of the same order of magnitude as those
`detemiined earlier for the cleavage of other bioactive pep-
`tides with N-terrninal Xaa-Pro or Xaa-Ala by DPP IV. Km
`values in the micromolar range have been generally found
`for other peptide-degrading proteases. Therefore. degradation
`rates at physiological peptide concentrations in the nanome-
`lar ranges are given by the rate (specificity) constants kc.../Km.
`High rate constants indicate high cleavage rates at nanomolar
`concentrations (below K... value). kw/Km values of about
`105M“-s“ for GIP, GLP-1(7-36)amide and PHM (Table 3)
`are lower than those determined earlier for good DPP-IV
`substrates like substance P, but still high enough to ensure a
`physiological action.
`
`Degradation of GIP and GLP-1(7—36)amide
`by human serum
`
`When human serum was incubated with 20 uM GIP, two
`major degradation products were observed (Fig. 3): one elut-
`ing at the position of Tyr-Ala, the other at that of des-Tyr-
`Ala-GIP. Identity of these peaks was ensured by identical
`retention times with standards (prepared by digestion with
`pure DPP IV) as well as by amino-acid analysis of the Tyr-
`Ala peak and detennination of the N—terminal amino acid of
`the GIP (3-42)-peak, both collected after separation. More-
`over, addition of the DPP-IV inhibitors 1 mM Lys-pyrroli-
`dide or 0.1 mM diprotin A abolished the generation of both
`GIP fragments by human serum nearly completely (residual
`areas < 5%). Hydrolysis of 0.5 mM Gly-Pro-4-nitranilide (an
`established chromogenic substrate of DPP IV) in the same
`serum sample was reduced to 2% in the presence of 1 mM
`Lys-pyrrolidide and to 9% after addition of 0.1 mM diprotin
`A. Lys-pyrrolidide (Lys-tetrahydropyrrole), a substrate ana-
`log, and diprotin A (lle-Pro-lle). a bad, but high-affinity
`(K... = 4 uM) substrate (Rahfeld et al., 1991 a), are competi-
`tive inhibitors specific (as far tested) for DPP IV. Concluded
`from their influence and from the fragments generated, GIP
`
`A220
`o2
`
`Q1
`
`
`
` [Aeetonitrfle](96)
`
`Time (min)
`separation of an incubations of GIP with DPP IV purified from human placenta. The positions of
`Fig. 2. Reverse-phase HPLC
`liberated Tyr-Ala and of the truncated peptide hormone are indicated. The C,. HPLC column was eluted with a gradient of 0-80%
`acetonitrile in 0.1% trifluoroacetic acid as described in Experimental Procedures. Peptides were monitored in the eluate by their absorbance
`at 220 nm.
`
`Page 3 of 7
`
`Page 3 of 7
`
`

`
`832
`
`Table 2. Separation of DPP IV cleavage products of gastric inhibitory polypeptide (GIP), glucagon-like peptide-_l(7-36)amide_[GLP-
`1(7—36)amide] and peptide hhtldine methionine (Pl-IM) by reverse-phase HPLC on a C,. column. For conditions see Experimental
`Procedures, retention times varied :03 min. The first 20 min of gradients are identical.
`
`
`
` Peptide Retention time Gradient
`
`
`
`GIP
`GIP(3-42)
`Tyr-Ala
`His-Ala
`
`min
`
`27.4
`27.1
`14.3
`3.8
`18.2
`
`0-3 min 0% + 3-45 min 0—80% acetonitrile
`in 0.1% trifluoroacetic acid
`
`0-3 min 0% + 3-45 min 0—80% acetouitrile in 0.1% heptafluorobutyric acid
`
`40.7
`GLP-l(7—36)amide
`41.7
`GLP-l(9-36)amide
`44.1
`PHM
`44.8
`PHM(3-27)
`35.2
`VIP
`
`His-Ala 3.8
`
`0-3 min 0% + 3-20 min ()--32% + 20-50min 32-48% acetonitrile
`in 0.1% trifluoroacetic acid
`
`Table 5. Catalytic constants for the degradation of bioactive peptides by human DPP IV. Assays were performed in 50 mM triethano-
`laminell-ICI, pH 7.8, at 37°C. Values of kc. were calculated using a molecular mass of 120 kDa for one identical subunit of the human
`placental DPP IV dimer (Plischel et al., 1982). GLP-1(7-36)arnide shows substrate inhibition above 50 p.M. catalytic constants (.tSD)
`were calculated from the linear ranges of Lineweaver-Burk plots.
`
`Peptide
`
`N-ter-
`minus
`
`S.,
`
`uM
`
`No. of
`runs
`
`K.,,
`
`uM
`
`V...“
`
`umol ~ min"
`. Ing“
`
`GIP
`GLP—](7—36)-
`amide
`PHM
`[Ala"]GRF(1 -29)-
`amide
`fl-Casomorphin
`Substance P
`
`YA-E...
`
`1-100
`
`I-[A-E...
`HA-D...
`
`YA-D...
`YP-F...
`RP-K...
`
`5-100
`5-100
`
`2-350
`20-500
`25-200
`
`7
`
`7
`6
`
`12
`
`34
`
`1' 3
`
`3.8 i 0.2
`
`4.5 2 0.6
`6.5 I 0.5
`
`0.97 1 0.05
`0.62 I 0.03
`
`4.7 : 0.3
`59
`22
`
`4.7 t 0.1
`90
`10
`
`k,..
`
`s“
`
`7.6
`
`1.9
`1.2
`
`9.5
`180
`20
`
`kc,/K...
`
`Reference
`
`M" -s"
`
`0.22 - 10‘
`
`this study
`
`0.43 - 10‘
`0.19- 10‘
`
`this study
`this study
`
`2.0 - 10°
`3.1
`- 10°
`0.91 - 10‘
`
`Bongers et al., 1992
`Nausch et al., 1990
`Nausch et al., 1990
`
`A220
`
`
`
`50%
`
`3
`
`0.2
`
`0.1
`
`E.
`
`5.1:
`
`2 §
`
`Time (min)
`Fig.3. Reverse-phase HPLC analysis of an incubation assay of 20 ,uM gastric GIP with human serum (G1? -1» Serum) compared to
`a serum blank (Serum, inset). Positions of GI? and its degradation products Tyr-Ala and GIP(3-42) are indicated. Experimental conditions
`as in Fig. 2.
`
`the des-His-Ala-peptide after reverse-phase HPLC (not
`is metabolized by DPP IV activity of human serum mainly
`shown). This fragment was identified by identical retention
`to Tyr-Ala and G[P(3-42).
`Incubation of human serum with 20 ttM GLP-l(7— time with a standard (obtained with pure DPP IV, Table 2)
`36)amide yielded one degradation product at the position of
`and by determination of the N-terminal amino acid. His-Ala
`
`Page 4 of 7
`
`Page 4 of 7
`
`

`
`833
`
`penultimate proline or alanine residues (Fig.4). Almost no
`other naturally occurring amino acid is accepted in this posi-
`tion. Replacement of penultimate Ala in a GRF(1—29)amide
`derivative by hydrophilic Ser or Gly resulted in dipeptidyl-
`peptidase-IV substrates of far lower km and higher K... values
`(Bongers et al., 1992). In contrast, substrates with synthetic
`hydrophobic derivatives of the proline ring (oxa- or thia de-
`rivatives) or short. unbranched hydrophobic alkyl derivatives
`in the P, position are good substrates for DPP IV (Rahfeld et
`al., 1991 b; Schutkowski, 1991). This indicates a hydropho-
`bic substrate (P.) recognition site for DPP IV where Set is
`less well (or not) bound than Ala or Pro (Fig. 4). Moreover,
`a bulky N-temtinal amino acid with free amino group (P,
`position) as with Tyr or His in the peptides investigated here
`is optimal for high DPP-IV activity. This together with ef-
`fects of the C-tenninal part of the peptides might account for
`the relatively low K... and high k... values of DPP IV for the
`29-42 residue hormones GRF, GIP, GLP-1(7 -36)arnide and
`Pl-IM as compared to those found earlier for small chromo-
`genie substrates with penultimate Ala (Heins et al., 1988).
`GIP released postprandially into the blood from intestinal
`endocrine K cells inhibits the secretion of gastric acid and
`stimulates insulin release from pancreatic B-cells in the pres-
`ence of elevated glucose levels. Schmidt et al. (1986, 1987)
`have clearly shown that N-terminal Tyr-Ala is absolutely re-
`quired for the insulin-releasing activity (the main physiologi-
`cal eflect) of GIP. Pure des-'Iyr-Ala-GIP (3 -42) unlike in-
`tact GIP did not increase insulin secretion in the presence of
`16.7 mM glucose from rat pancreatic islets at physiological
`or higher concentrations even up to 250 nM. Therefore, trun-
`cation of GIP by DPP IV results in its inactivation with re-
`spect to its major physiological, the insulinotropic. action.
`Cleavage products and influence of specific inhibitors
`clearly show that dipeptidyl peptidase IV is the main degra-
`dation and. considering the above findings, inactivation en-
`zyme for GIP in human serum. The enzyme should be still
`more active on this peptide hormone at other sites, e. g. endo-
`thelial cells of blood vessels, hepatocytes, kidney brush-bor-
`der membranes (podocytes of the glomerular basement mem-
`brane and proximal tubule cells), lymphocytes, chief cells of
`gastric glands, or epithelial cells of the intestine, where it is
`found in high concentrations as an ectoenzyme of the plasma
`membranes (Loijda, 1979; Harte] et al., 1988; Gossrau,
`1979; McCaughan et al., 1990; Mentlein et al., 1984). Active
`hydrolysis by DPP IV might therefore explain why GIP(3-
`42) has been isolated as a second component (relative yield
`about 20-30%) beside intact GIP from porcine intestine and
`has been found as a contaminant of natural GIP preparations
`(Jiirnvall et al., 1981; Schmidt et al., 1987).
`GLP-1(7—36)a1nide is a product of the tissue-specific
`post-translational processing of the glucagon precursor. It is
`released postprandially from intestinal endocrine L cells and
`stimulates insulin secretion. Gallwitz et al.
`(1990) have
`shown that the C-terminal fragment of the peptide is impor-
`tant for receptor binding of the honnone, but is not sufficient
`to transduce a biological action as does the intact peptide
`(raise in cyclic AMP levels in rat insulinoma RINm5F cells).
`It appears that as in the case of glucagon (Unson et al., 1989),
`of GIP (Schmidt et al., 1986, 1987) and of other members of
`the VIP/glucagon peptide family (Christophe et al., 1989;
`Robberecht et al., 1992) also for GLP—1(7—36)atnide an in-
`tact N-terrninus is needed for signal transduction and biologi-
`cal action. Provided this, action of DPP IV inactivates GLP-
`1(7-36)amide.
`
`........ on
`
` 4
`
`4:-43
`
`Fig.4. Schematic representation of the substrate-binding and
`substrate-cleaving (arrow) sites of DPP IV. Proline and alanine
`fit in the hydrophobic P.-substrate-binding pocket. whereas serine
`appears to be too hydrophilic to yield appreciable binding. In the P,
`position bulky amino acids with an obligate free amino group are
`preferred. Peptides with Pro or Hyp in the P.’ position are not
`cleaved by DPP IV. Preferential amino acids for the P.’ position are
`not known.
`
`as further degradation product could be identified after deri-
`vatisation with 4-dimethylaminoazobenzenesulphonyl-chlo-
`ride (see Experimental Procedures) by identical retention
`time and co-chromatography with a derivatized, synthetic
`His-Ala standard. Again, in the presence of Lys-pyrrolidide
`(1 mM) and diprotin A (0.1 mM), the generation of the des-
`l-lis-Ala-fragment was abolished (<5%). Thus, as con-
`cluded from specific inhibition and generation of His-Ala
`and the des-His-Ala-peptide GLP-1(7 —36)amide is cleaved
`by human serum mainly by action of DPP IV.
`In sera of healthy males we measured a mean activity of
`551-12 pmol - min" - I" (n = 6) with the chromogenic
`substrate 0.5 mM Gly-Pro-4-nitranilide for DPP IV. No sig-
`nificant differences were found for the peptidase activities in
`preprandial an postprandial sera (n = 3). In a serum with an
`activity of 50 pmol - min" - I" for Gly-Pro-4-nitranilide, we
`estimated degradation rates of about 0.3 pmol - min" - 1“
`for Tyr-Ala liberation from 20 |.LM GIP and 0.4 umol-min“
`- 1" for His-Ala liberation from 20 uM GLP-1 (7 —36)amide.
`
`DISCUSSION
`
`Members of the VIP/glucagon peptide family with N-
`terminal penultimate alanine are good substrates for DPP IV.
`GRF(1-29)amide or GR.F(1 —44)amide as analyzed here and
`by Bongers et al. (1992), GIP, GLP-1(7-36)amide and PHM
`are cleaved to their des-Tyr-Ala or des-His- Ala derivatives
`by the highly purified human enzyme. In contrast, VIP with
`N-temrinal His-Ser was not significantly degraded. This fits
`well with the known. preferential specificity of DPP IV for
`
`Page 5 of 7
`
`Page 5 of 7
`
`

`
`834
`
`Based on the identification of cleavage products and in-
`fluence of specific inhibitors, DPP IV is the main degrada-
`tion enzyme for GLP-1(7-36)amide in human serum. Buck-
`ley and Lundquist (1992) have reported recently in an ab-
`stract the formation of GLP-1(9-37) by human plasma, but
`did not identify the peptidase responsible for its generation.
`As outlined for GIP above, plasma-membrane-bound DPP
`IV of endothelial and other cells might be still more irnpor-
`tant for the inactivation of GLP-1(7-36)amide than the
`plasma activity.
`PI-IM (rat counterpart PHI) and VIP are processing prod-
`ucts of a common precursor and are co-released from central
`and peripheral neurons. As far as is known, Pl-IM/PI-II have
`biological effects similar or identical
`to VIP. Since it
`is
`known that the biological actions of VIP critically depend on
`an intact N-terminus (Christophe et al., 1989; Robberecht et
`al., 1990), in analogy also PHM/PI-II might be inactivated by
`cleavage of the N-terminal dipeptide by DPP IV. Since serum
`concentrations of PHM like VIP are low and in contrast to
`GIP and GLP-1(7-36)amide do not rise postprandially (Bo-
`den and Shelmet, 1986), inactivation in serum is probably of
`minor importance and was not investigated. It can, however,
`be suspected that DPP IV cleaves the paracrine acting pep-
`tides PHM/Pl-II in other tissues where it is present on the
`surface of various epithelial and endothelial cells.
`In conclusion, members of the glucagon/VIP peptide
`family with N-tenninal 'l‘yr-Ala or I-Iis-Ala, namely GRF,
`GIP, GLP-1(7-36)amide and PHM, are inactivated by action
`of DPP IV in human serum. The truncated peptides could
`also be antagonists, because the binding specificity is di-
`rected by the C-terminal parts of these peptide hormones
`(Christophe et al., 1989; Gallwitz et al., 1990). Since the
`cleavage by this peptidase removes only 2 of 29-42 total
`residues of the hormones, antisera against these peptides not
`directed specially to the N-terrninus should cross-react also
`with the truncated peptides. Therefore,
`immunoassays for
`these hormones can be hampered by the measurement of bio-
`logically inactive, des-Xaa-Ala forms beside the active pep-
`tide, due a potential cross-reactivity of the antisera. Unless
`specific N-terminally directed antisera are available, semm
`samples should be stored for immunoassays at least in the
`presence of DPP-IV inhibitors (specific ones mentioned here
`or serine protease inhibitors like phenylmethanesulphonyl
`fluoride).
`DPP IV in human serum and at the surface of endothelial
`cells is known to be involved in the inactivation of other
`
`circulating bioactive peptides: removal of the N-terrninal
`tetrapeptide Arg-Pro-Lys-Pro of substance P (I-leymann and
`Mentlein, 1978) inactivates only some biological actions of
`this neuropeptide (e. g. histamine release from mast cells),
`but renders the peptide possible for the complete degradation
`by aminopeptidase M (Ahmad et al., 1992). Several other
`bioactive peptide with N-terminal Xaa-Pro including gastn'n-
`neleasing peptide, corticot:rophin-like intermediate lobe pep-
`tide and fl-casomorphin are excellent substrates for the puri-
`fied peptidase (Nausch et al.. 1990).
`
`We thank Martina von Kolszynski for her expert technical assis-
`tance. This work was supported by grants Me 758/2-3 and Ga 386/
`2-2 from the Deutsche Forschungsgemeinsehafi.
`
`REFERENCES
`
`Ahmad, S., Wang, L. & Ward, E. (1992) Dipeptidyl(amino)pept:idase
`IV and aminopeptidase M metabolize circulating substance P in
`vivo, J. Pharm. Exp. Ther. 260. 1257-1261.
`
`Page 6 of 7
`
`Boden, G. & Shelmet, J. J. (1986) Gastrointestinal hormones and
`carcinoid tumors and syndrome in Endocrinology and metabo-
`lism (Felig, P., Baxter, J. D., Broadus, A. E. & Frohman, L. A.,
`eds) pp. 1629-1662, McGraw-I-Iill, New York.
`Bongers, .l.. Lambros, T., Ahmad, M. & I-leimer, E. P. (1992) Kinet-
`ics of dipeptidyl peptidase IV proteolysis of growth honnone-
`releasing factor and analogs, Biochim. Biophys. Acta 1122, 147-
`153.
`
`Boulanger, L., Roughly, P. & Gaudreau, P. (1992) Catabolism of rat
`growth hormone-releasing factor(1—29)amide in rat serum and
`liver. Peptides (Elmsford) 13, 681-689. ,
`Bucldey, D. I. &. Lundquist, P. (1992) Analysis of the degradation
`of insulinotropin [GLP-1(7-37)] in human plasma and pro-
`duction of degradation resistant analogs, Regul. Peptides 40, 117.
`Christophe, J., Svoboda, M., Dehaye, J.-P., Winand, J., Vander-
`meets-Pire, M.-C., Vandermeers. A., Cauvin, A., Gourlet, P. &
`Robberecht, P. (1989) The VIP/PHI/secretin/helodermin/helos
`pectin./GRF family: structure-function relationship of the natural
`peptides, their precursors and synthetic analogues as tested in
`vitro on receptors and adenylate cyclase in a panel of tissue
`membranes, in Peptide hormones as prohormones: processing,
`biological activity, pharmacology (Martinez, 1., ed.) pp. 211-
`243, Ellis Horwood LTD. Chichester.
`Frohman. L. A., Downs, T. R., Heimer, E. P. & Felix, A. M. (1989)
`Dipeptidyl peptidase IV and trypsin-like enzymatic degradation
`of human growth hormone-releasing hormone in plasma, J. Clin.
`Invest. 83. 1533-1540.
`Gallwitz, B., Schmidt, W. E., Conlon, J. M. & Creutzfeldt, W. (1990)
`Glucagon-like peptide-1 (7—36)amide: characterization of the
`domain responsible for binding to its receptor on rat insulinoma
`RINm5F cells. J. Mol Endocrinol. 5, 33-39.
`Gossrau, R. (1979) Zur Lokallsation der Dipeptidylpeptidase IV
`(DPP IV). Histochemische und biochemisehe Untersuchungen,
`Histochemistry 60, 231-248.
`Hartel, S., Gossrau, R., I-Ianski, C. & Reutter. W. (1988) Dipeptidyl
`peptidase IV in rat organs. Comparison of immunohistochem-
`istry and activity histochemistry, Histochemistry 89, 151-161.
`I-leins, 1., Welker. P., Schonlein, C., Born, I., I-Iandtrodt, 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 Flavo-
`bacterium meningosepticum, Biochim. Biophys. Acta 954, 161 —
`169.
`Heymann, E. & Mentlein, R. (1978) Liver dipeptidyl peptidase IV
`hydrolyzes substance P, FEBS Lett. 9], 360-364.
`Jbrnvall, H., Carlquist, M., Kwauk, S., Otte, S. C., Mclntosch, C.
`H. S., Brown, I. C. & Mutt, V. (1981) Amino acid sequence and
`heterogeneity of gastric inhibitory polypeptide (GIP). FEBS Lett.
`123. 205-210.
`Kubiak, T. M., Kelly. C. R. & Krabill, L. F. (1989) In vitro metabolic
`degradation of a bovine growth hormone—releasing factor analog
`Leu27-bGRF(1-29)NI-I, in bovine and porcine plasma, Drug.
`Metab. Dispos. 17, 393-397.
`Loijda, Z. (1979) Studies on dipeptidyl(amino)peptidase IV. II.
`Blood vessels, Histochemistry 59, 153-166.
`McCaughan. G. W., Wickson, J. B., Cheswick, P. F. & Gorrell, M.
`D. (1990) Identification of the bile canicular cell surface mole-
`cule GP110 as the ectopeptidase dipeptidyl peptidase IV: an
`analysis by tissue distribution, purification and N-tenninal amino
`acid sequence, Heparology 11, 534-544.
`Mentlein, R. (1988) Proline residues in the maturation and degrada-
`tion of peptide hormones and neuropeptides, FEBS Lett. 234,
`251 - 256.
`Mentlein, R.. I-Ieymann, B., Scholz, W., Feller, A. C. & Flad, H.-D.
`(1984) Dipeptidyl peptidase IV as a new surface marker for a
`subpopulation of human 'l‘—lymphocytes, Cell. lmmunoL 89, 11-
`19.
`Mentlein, R. & Sttuckhoff. G. (1989) Purification of two dipeptidyl
`aminopeptidases II from rat brain and their action on proline-
`containing neuropeptides. J. Neurochem. 52, 1284-1293.
`Nausch, I. & Heymarm, E. (1985) Substance P in human plasma is
`degraded by dipeptidyl peptidase IV, not by cholinesterase, J.
`Neumchem. 44. 1354-1357.
`
`Page 6 of 7
`
`

`
`Nausch, I., Mentlein, R. & Heymann, E. (1990) The degradation of
`bioactive peptides and proteins by dipeptidyl peptidase IV from
`human placenta, Biol. Chem. Hoppe-Seyler 371, 1113-1118.
`Ptischel, G., Mentlcin. R. & Heymann, E. (1982) Isolation and char-
`acterization of dipeptidyl peptidase IV from human placenta,
`Eur. J. Biochem. I26, 359-365.
`Rahfeld, J., Schierhorn, M., Hartrodt, B., Neuben, K. & I-leins, J.
`(1991 a) Are dipmtin A (lle-Pro-Ile) and diprotin B (Val-Pro-
`Leu) inhibitors or substrates for dipeptidyl peptidase IV? Bia-
`chim. Biophys. Acta I076, 314-316.
`Rahfeld, J.. Schutkowski, M., Faust, J., Neubert, K., Barth, A. &
`Heins, J. (1991b) Extended investigation of the substrate speci-
`ficity of dipeptidyl peptidase IV from pig kidney, Biol. Chem.
`Hoppe-Seyler 372, 313-318.
`Robberecht, P., Cauvin, A., Gourlet, P. & Christophe, J. (1990)
`Heterogeneity of VIP receptors, Arch. Int. Pharmacodyn. 303,
`51 -66.
`Robberecht, P., Gourlet, P., De Neef, P., Woussen-Colle, M.—C.,
`Vandenneers-Piret, M.—C., Vandenneers, A. & Christophe, J.
`(1992) Structural requirements for the occupancy of pituitary ad-
`enylate-cyclase—activating peptide (PACAP) receptors and ade-
`nylate cyclase activation in human neuroblastoma NB-OK- 1 cell
`membranes, Em: J. Biochem. 207, 239-246.
`Schmidt, W. E., Siege]. E. G., Kiimm