metformina y glp-1 efecto hinke 2002

7/27/2019 Metformina y GLP-1 Efecto Hinke 2002

1/7

Metformin Effects on Dipeptidylpeptidase IV Degradationof Glucagon-like Peptide-1

Simon A. H inke,* Kerst in K u hn-Wa che, Torst en H offma nn, Ra ymond A. P ederson,*Christopher H. S. McIntosh,* and Hans-Ulrich Demuth ,1

Pr obiodr ug R esear ch, Bi ocent er , Wein ber gweg 22, D -06120 H all e (Saal e), Ger m any; an d Department of Physiology, Universi ty of Bri t ish Columbia, Vancouver, Canada V6T 1Z3

Received February 14, 2002

There is current interest in the use of inhibitors of ipeptidyl peptidase IV (DP IV) as therapeutic agents

o normalize glycemic excursions in type 2 diabeticatients.Data indicating thatmetformin increases theirculating amount of active glucagon-like peptide-1GLP-1) in obese nondiabetic subjects have recentlyeen presented, and it was proposed that metformin

might act as a DP IV inhibitor. This possibility haseen investigated directly using a number of i n vi t r o

methods. Studies were performed on DP IV enzymerom three sources: 20% human serum, puried por-ine kidney DP IV, and recombinant human DP IV.nhibition of DP IV hydrolysis of the substrate Gly-

Pro-pNA by metformin was examined spectrophoto-metrically. Effects of metformin on GLP-1 [7-36NH2] deg-adation were assessed by mass spectrometry. In ad-ition, surface plasmon resonance was used to estab-ish whether or not metformin had any effect on GLP-

[7-36NH2] or GLP-1 [9-36NH2] interaction with immobilizedorcine or human DP IV. Metformin failed to alter the

kinetics of Gly-Pro-pNA hydrolysis or GLP-1 degrada-ion tested according to established methods. Surfacelasmon resonance recordings indicated that both

GLP-1 [7-36NH2] and GLP-1 [9-36NH2] show micromolar afnityK D) for DP IV, but neither interaction was inuencedy metformin. The results conclusively indicate that

metformin does not act directly on DP IV, thereforelternative explanations for the purported effect of

metformin on circulating active GLP-1concentrationsmust beconsidered. 2002 Elsevier Science (USA)

K ey W or ds: incretin; entero-insular axis; CD 26; DPPV; MALDITOF mass spectrometry; BIAcore; surfacelasmon resonance.

Dera ngement of glucose homeosta sis a ffects a pprox-mately s ix percent of the inhabi tants of the Uni ted

States of America, and s imilar pro ject ions are madew orldwide (1). Of t hose diagnosed w ith diabetes m elli-tus, i t is thought that type 2 diabetes (T2D), dened

prima rily by peripheral insulin resista nce with concur-rent hyperglycemia , a ccounts for ninety to n inety- vepercent of diagnosed diabetic patients (1). Therapiesfor T2D include insulin injection and various oral phar-ma ceutica ls (sulfonylureas, metformin, aca rbose, a ndcertain glitazones), however, resistance to monothera-pies as the disease progresses usually results in thenecessity of combinatorial treatment in order to im-prove blood glucose levels (2). As such, there is addedpressure on the pharmaceut ical industry to developmore potent forms of existing therapies and new oralagents w i th novel cellular ta rgets t ha t can be used as

monotherapies or in combination with other antidia-betic dr ugs (3).

One such novel molecular target with potential an-tihyperglycemic effects is the ubiquitous proteolyticenzyme, dipeptidyl peptidase IV (DP IV, or known a sCD 26 to immu nologist s; E C3.4.14.5). The un ique prop-erty of DP IV with respect to diabetes melli tus is tha tit is t he prima ry enzym e responsible for degrada tion ofthe incretins i n v i v o (4). Incretins are the hormonala rm of the enteroinsular axis, the l ink betw een t he guta nd th e endocrine pa ncreas (5). G lucose-dependent in-sulinotropic polypeptide (G IP ) a nd a mino-terminallytr uncat ed gluca gon-like peptide-1 (G LP -1 [7-36NH 2] ) a r eth e only hormones wh ich ha ve been proven to fulll th erequ i remen ts to be dened as an incre t in : they a rereleased into t he blood stream in response t o luminalnutrients, and act to augment nutrient-induced insulinrelease in a glucose-dependent fashion (6). Mentlein et a l . (7) r st s how e d t h a t G IP a n d G LP -1 w ere s u b-s t ra tes fo r DP IV i n v i t r o , and short ly thereafter, i n v ivo degrada t ion was a l so demons t ra ted (4 ) . I t wasPauly and colleagues who rst postulated the l ink be-tw een the poss ib le bene ts of DP IV inhibi t ion an dglycemic control due to enhancement of the incretineffect (8). The hy poth esis tha t DP IV inhibition w ould

1

To whom correspondence should be addressed. Fax: 49-345-559901. E-mail: [email protected].

Biochemical and Biophysical Research Communications 291, 13021308 (2002)

oi:10.1006/bbrc.2002.6607, a va ilable online at ht tp://ww w .idealibr a ry.com on

1302006-291X/02 $35.00 2002 Elsevier Science (USA)ll r ight s reserved.


2/7

mprove glucose tolerance was later shown to be cor-ect in both Wistar rats and diabetic fatty Zucker rats9, 10). These ndings ha ve been corrobora ted by sim-lar studies in mouse, rat and pig (11 13).

Metformin is a derivative of the antidiabetic bigua-nide alkaloids found in French lilac ( Galeg ofcinalis ),

medieval t reatment for d iabetes (14). I t has beenommercially available since the 1950s, and is com-

monly used wor ldwide as an in i t ia l monotherapy fornewly d iagn osed T2D pat ients, a s it is equa lly effective

s sulfonylur ea t rea tm ent (2, 14). How ever, metforminwa s not a vai lab le for cl in ical use in the U nited S ta tesuntil 1995 (15). The speci c molecular target o f i t ist i ll unknown, a l though b iguanides general ly act toensitize periphera l tissues to insulin a ction (in pa rt ic-

ular skeletal muscle) and inhibit hepatic gluconeogen-sis and glycogenolysis (2, 3, 14). Notably, unlike thencretins, metformin does not improve glucose toler-nce via an increase in circulating insulin levels, im-licating other modes of action.Recent ly, da ta were presented demonstra t ing t he

ffect of metformin on plasma a ctive (a mino-term ina llyn tact) G LP -1 concentra t ions in obese non-d iabet ic

male pat ien ts (16) ( r s t appear ing in abs t r ac t fo rm17)). In th is st udy, a dminist ra tion of metformin (2550

mg/da y) over a tw o w eek period a ppeared to signi -antly increase active GLP-1 levels relative to the con-rol group a fter a n ora l glucose loa d w ith a euglycemic

hyperinsulinemic clamp protocol, but did not affectasal act ive G LP-1 concentra t ion . Fur thermore, dur-ng incubat ion of G LP-1 [7-36NH2] in h u ma n s er u m oruffer conta in ing porcine DP IV i n v i t r o , metforminoncentra tions tha t w ould be expected in v ivo appearedo dose-dependently preserve inta ct G LP -1 (a s mea-ured using an N-terminally speci c ELISA) (16). Theurpose of the current st udy w as to reinvestigate th ese

ndings us ing a l ternat ive b iochemical methods . Ex-eriments were designed such that the effect of met-ormin on DP IV ac t iv ity in human serum, pu r i edorcine DP IV, and pur i ed recombinant human DPV could be determined. Gly-Pro-para-nitroaniline was

used as a DP IV substra te for spectrophotometric stud-es , and synthet ic GLP-1 [7-36NH 2] wa s employed dur ing

kinetic studies with ma trix-assisted laser-desorptiononiza tion-time of ight mass spectrometry (MALDI-

TOF MS). Surface plasmon resonance is able to detecteal-time intera ctions betw een proteins, a nd thus thisechnique was appl ied to es tab l ish i f metformin af -ected GLP-1 [7-36NH2] or GLP-1 [9-36NH2] binding to DP IV.

MATERIALS AND METHODS

Enzyme pu ri cati on. Pu r ied pork kidney dipeptidyl peptidaseV was prepared by the method of Wolf et a l . (18). Recombinantol ub le h u ma n D P I V w a s k in d ly p rov id ed b y J . B ar (P robio-rug, German y). Using the chromogenic substra t e , H-Gly-Pro-4-

itroaniline (GP -pNA; P robiodrug, G erman y), the speci c activity ofDP IV used in the current st udy w as m easured t o be 31.2 units/mg

(puri ed porcine) and 32.4 units/mg (recombina nt h uma n; rhu ma n).One un i t o f DP IV ac t iv i ty i s dened as the re lease of 1.0 mol/lni troaniline (yellow product) per minute measured spectrophoto-metr ica lly a t 390 nm under sta ndard condit ions (de ned below).Human serum was obtained from healthy donors, pooled and storeda t 20 C until use, described previously (19).

Eff ect of metform in on DP I V hydr olysis of GP-pNA. Experimentswere ca r r ied ou t under s tanda rd condi t ions : 30 C i n p H 7. 6 40mm ol/l H EP ES (Sigma -Aldrich) buffer conta ining 0.4 mmol/l H -G ly-

Pro-4-nitroaniline, and 2.5 mU of DP IV (porcine or rhuman) or 20%huma n serum. Metformin (1,1-dimethylbiguan ide; S igma -Aldrich)was added over the concentra t ion range of 0 to 100 mol/l. Nit ro-anil ine production was monitored using a HTS 7000 microplatereader (PerkinElmer, U berlingen, Germany).

E ffect of met formi n on DP IV h y d rol y si s o f GL P -1 [ 7 - 36NH2] using M A L D I -T O F M S . Similar to spectrophotometric studies, matrix-assisted laser-desorption ionization time of ight mass spectrometry(MALDI-TOF MS) experiments were carried out at 30 C at pH 7.6,but in 0.1 mol/l Tris/HC l (Sigma -Aldrich) buffer with 12 mol/lGLP-1 [7-36NH2] . The degrada t ion f a te of GLP-1 was measured bymonitoring the signal intensity of the pseudomolecular ion peaks ofGLP-1 [7-36NH2] ([M H] 3299.7) and GL P -1 [9-36NH2] ([M H] 3090.4)versus time when incubat ed with 2.5 mU DP IV (porcine or rhuman) or

20%human serum, with or without metformin (0 1 mm ol/l). The ma ssspectrometer employed was a Hewlett-Packard G2025 model with alinear time of ight analyzer; samples (4 l) were mixed 1:1 v/v w ithmatrix (44 mg diammonium-hydrogen-citrate and 30 mg 2 ,6 -dihy-droxyacetophenone in 1 ml aqueous solution containing 50%acetoni-trile and 0.05% tri uoroacetic acid; Sigma-Aldrich), transferred to aprobe tip and immediately evaporat ed using the H ewlett-P ackard G2024Asample preparation vacuum chamber. 250 single laser-shot spectrawere accumulated. This method of monitoring degradation has beenvalidated in several prior publications (8, 19, 20), and allows the generalcomparison of half-degrada tion t imes (t 1/2 ) under various conditions.

Effect o f met formin on subst rate b ind ing to DP IV using surface pl asmon resonance. Surface plasmon resonance is a highly sensi-tive technique which measures biomolecular interactions by detect-ing th e change in refra ctive properties a t t he surface of a sensor chip.Pu r ied pork DP IV and recombinant human DP IV were immobi-lized on the surface of a CM5 chip (BIAcore AB, Uppsala, Sweden)using amine coupling chemistry, precisely as previously described(19). Baseline values for porcine DP IV and recombinant human DPIV w ere 5000 a nd 3500 resonan ce units (RU), respectively. B aselinevalues affect the maximal possible change in RU upon analyte bind-ing (proportional to the ra t io of molecular masses of the a nalyte tothe imm obilized liga nd m ultiplied by the ba seline value), however itdoes not in theory a lter th e outcome of kinetic ana lyses. Experimentswere carried out using a ow ra te of 20 l/min, in HB S-EP buffer (10mm ol/l H EP E S, 150 mmol/l Na Cl, 3 mm ol/l E DTA, 0.005%v/v S ur-factant P-20, pH 7.4; BIAcore AB). Apparent K D values were mea-sured a t 4 C and 25C for both G LP -1[7-36NH2] and GLP-1 [9-36NH2] , over a

concentration range of 1.56 mol /l t o 100 mol/l, a nd mea sured fromnon-linea r regr ession curves on plots of R Eq (steady -sta te equilibriumdifference in resonance units) versus peptide concentration. To es-tablish if metformin had an effect on GLP-1 [7-36NH2] or GLP-1 [9-36NH2]interaction with DP IV, 20 mol/l of peptide w as co-injected w ithmetformin in the concentration range 0 0.3 mm ol/l.

Pepti de synt hesis. GLP-1 [7-36NH2] and GLP-1 [9-36NH2] were synt hesizedin house, using the automated Symphony peptide synthesizer (RaininInstrum ent Co., Woburn, MA). P eptides w ere puri ed to 95%purityby HPLC (Merck-Hitachi, Darmstadt, Germany) and MALDI-TOF MSwas used to conrm identity and purity of the products.

Data analysis. Data points represent compiled data from at leastthree independent measurements, given as the mean s tanda rd

error of the mean (SEM). Data were analysed using the Pr ism 3.0(GraphPad, San Diego), BIAevaluation 3.0.1 (BIAcore AB) or Excel

Vol. 291, No. 5, 2002 B IOCH E MICAL AND B IOP HYS ICAL RE SE ARCH COMMU NICATIONS

1303


3/7

7 (Microsoft) softw ar e packages for P C. MALDI-TOF MS degrad a-ion curves were t t ed to rst-order exponential decay equations,

whereas BIAcore binding curves were t t ed to rst-order exponen-ia l association equat ions, both using a ppropriate n on-linear r egres-ion software. Signicance of difference was ascertained using anal-sis of variance (ANOVA) or a student s t test , w here a ppropr ia te ,

wi th P 0.05 considered signi cant.

RES ULTS

H ydrolysis of G ly-Pro-pNA is not a l tered by met- o rmin . A broad concentra t ion range of metformin

wa s used to monitor any in uence i t ha d on the s ta n-da rd colorimetric determina tion of D P IV a ctivity. Ta -

le 1 shows the effect of metformin on Gly-Pro-pNAhydrolysis by puri ed pig kidney DP IV, recombinanth u m a n D P I V a n d b y h u m a n s e r u m . N o s ig n i cant

ffects were observed at any concentration tested. Theoncentration range of metformin used includes clini-ally relevant concentrations, as well as those higherha n found in v ivo (normally less than 18 mol/l; (21)).

With either competitive or non-competitive enzyme in-hibition, one w ould expect d ose dependent effects.

M easur ement of GL P-1 [7 -36NH2] degradation kin etics y M AL DI -TOF ma ss spectr ometr y shows no effect of

metformin. Mat rix-a ssisted laser-desorption time ofigh t mass spectrometry wa s used to monitor the hy -drolysis of inta ct GL P -1 [7-36NH2] to GLP -1 [9-36NH2] by serum

n d p u r i ed DP IV homologs f rom p ig and human .MALDI-TOF MS has been used to measure classical

nzyme kinetic constants (8, 20), however, more rou-inely performed is the comparison of half degradationim e (t 1/2 ) in t he presence or a bsence of a n inh ibitor (19,0). Figur e 1 depicts represent a tive spectr a obta ined inh e presence a nd a bsence of 10 mmol/l metformin a t 0

min a nd 60 min a fter incubat ion w ith porcine, rhuma nDP IV or 20%huma n serum at 30 C, pH 7.6. Qualita-ively, metformin appears not to prevent GLP-1 [7-36NH2]

hydr olysis by D P IV or serum . Compar ison of exponen-ia l decay curves quant i ta t ively ver i es this conclu-ion; t 1/2 va lues w i th or w i thou t met formin d id notignicantly differ (Table 2).

M etform in fa i l s to a l ter in teract ion between GLP- 1 [7 -36NH 2] or GLP-1 [9 -36NH2] and DP IV, as measured wi th sur face plasmon resonance. I n it ia l s t ud ies u sin gsurface p lasmon resonance examined the apparentb inding afn i ty o f GLP-1[7-36NH2] a n d G L P - 1 [9-36NH2] for

FIG. 1. Representa tive MALDI-TOF m ass spectra of GLP -1 deg-ra da tion by (A) puri ed pork DP IV, (B) recombinant human DP IV,or (C) 20% huma n serum, w ith or w ithout 1 mol/l m etformin. Theabscissa is relative peak intensity , and th e ordina te is mass to char ge

ra tio (m/z). S ee text for deta iled methods. Quan tita tive kinetic pa-rameters can be found in Table 2.

TABLE 1

Spectrophotometric Studies Using Gly-Pro-pNAas a Subs t ra te

Metformin( mol/l)

DP IV act ivity (%)

pDP IVa rhDP IVb 20%huma n plasma

0 100 2 100 2 100 7

1.25 98 1 92 2 110 710 99 2 89 2 101 9

100 95 2 97 1 102 10

Note. Hydrolysis w as monitored under standa rd condit ions a s de-cribed under Materials and Methods.

a Pu r ied porcine DP IV.b Pur i ed recombinant human DP IV.


1304


4/7

dextra n immobilized D P IV (puri ed porcine and re-ombinan t huma n), a s previously described for gluca -

gon analogs (19). Apparent K D values w ere obta ined bylo t t ing change in R E q from baseline versus peptideoncentr a tion (Fig. 2). As immobilized DP IV r etainsnzymatic activity, experiments rst measured at 25 C

w ere also performed a t 4 C, to obta in more accura te K Dva lues. This w a s hypothesized to be more importa nt forhe measurement of GLP-1 [7-36NH2] binding afn i ty a t5 C, w here the measur ed K D w ould a lso be inuencedy the hydro lys is to GLP-1 [9-36NH2] at the sensor ch ipurface. In fact , the measured K D for GLP-1 [7-36NH2] a p-eared t o be only modera tely reduced (i.e. higher a f n-ty) a t 4C relative to 25 C (Ta ble 3). F urt hermore, t he

net effect of reduction of t empera ture wa s t o decreaseB ma x : G LP -1[7-36NH2] , r educed by 35.6%a nd 58.2%, G LP -

[9-36NH2] , r educed by 48.9%a nd 61.9%, for porcine a ndhuman DP IV isoforms, respectively. Metformin (0 .3 m mol/l), ha d no effect on either G LP -1 [7-36NH2] or

GLP-1 [9-36NH2] (20 mol/l) bindin g t o immobilized DP IVFig. 3). Metformin concentr at ions a bove t his ra ngentera cted non-speci ca l ly w i th t h e d ext r a n m a t r ixth e reference cha mber) in t he a bsence of peptide, pre-

ven t ing the test ing of h igher doses (a l though 0.3mmol/l metformin is a lrea dy a suprapha rma cological

oncentration). Results also indicated that constant 30mol/l met formin did n ot produ ce consisten t effects onther concentr a tions of G LP -1 [7-36NH2] or GLP-1 [9-36NH2]ver the concentra t ion range 0 100 mol/l ( n 2,

data not shown).

DISCUSSION

The curren t manuscr ip t addresses whether or no tmetformin a cts directly on dipeptidyl peptida se IV (DPV) in order t o retar d degra da tion of G LP -1 [7-36NH2] to itsnact ive N-terminal ly t runcated form, G LP-1 [9-36NH2] .

The r ecent ma nuscript by Man nucci et a l . (16) sug-gested tha t because G LP -1 [7-36NH2] levels were elevatedn metformin-tr eat ed non-dia betic obese males relat iveo non-trea ted subjects, metformin ma y a ct as a DP IV

inhibitor, thus explaining the anorectic effect of met-formin a nd t he concurrent improvement in glucose tol-erance. By their own admiss ion , in the Mannucci re-por t on ly prel iminary n d in gs a r e in clu ded , a n d

TABLE 2

Degradation Kinetics of GLP-1 [7-36NH2] Measuredby MALDI-TOF Mass Spectrometry

Metformin( mol/l)

Half-time of degradation (t 1/2 ; min)

pDP IVa rhDP IVb 20%huma n plasma

0 37.1 3.5 37.6 2.6 49.8 3.3

10 44.5 3.7 30.3 1.7 NDc

1000 37.0 1.3 34.7 2.6 53.5 3.3

Note. See text for detailed methods.a Pu r ied porcine DP IV.b Pur i ed recombinant human DP IV.c ND, not determined.

FIG.2. B inding kinetics of G LP -1 for dextra n-immobilised DP IVmeasured by surface plasmon resonance. (A) A representative sen-sorgram showing binding of GLP -1 [7-36NH2] to DP IV immobilised onthe surface of the sensor chip versus t ime ( ow ra te 20 l/min ,25 C, pH 7.4) . Baseline measurements were taken a t 60 s and R Eq wa s mea sured a t 360 s, at the end of the 5-min peptide injection. Atleast 10 min of wash out was allowed in between peptide injections

to a l low re turn to baseline . (B) Satura tion binding curves of GLP-1 [7-36NH2] or GLP -1 [9-36NH2] and DP IV at 25C using equilibrium surfaceplasmon resonance de ections plotted versus peptide concentr at ion.See text and Table 3 for complete quantitative comparisons.

TABLE 3

Binding Constants (K D) of G LP -1[7-36NH2] or GLP -1 [9-36NH2] forDextran-Immobilised Porcine and Recombinant Human DPIV, Measured Using Surface Plasmon Resonance

pDP IVa r h D P I Vb

4C 25C 4C 25C

GLP-1 [7-36NH2]( mol/l)

14.3 0.9 33.4 2.2 47.3 5.7 153 17

GLP-1 [9-36NH2]( mol/l)

22.8 1.6 32.0 2.8 57.7 7.4 105 20

a

Pur i ed porcine DP IV.b Pu r ied recombinant human DP IV.


1305


5/7

imilar experiments have not yet been carr ied out inhea lthy or diabetic subjects, a nima l models, or in v i t r o.We h a v e a d d r es sed t h e la t t e r d e ciency, and per-ormed severa l direct i n v i t r o enzymologica l experi-

ments t o determine i f metformin inh ib its DP IV orlters t he substra te-enzyme intera ction. In contr a st t o

ndings by Mannucci and co-workers , we have beenuna ble to show t ha t metformin ha s an y effect on DP IV,

nd thus we offer a l ternat ive explanat ions for theirndings.

Mannucci et al. (16) continued the work of Lugari et l . (22), with respect to the effect of metformin on

G LP -1 levels in obese or T2D pat ien ts . The la t termanuscript examined the effect of metformin (1 week,

00 mg three times per day) on plasma glucagon andotal GLP-1 (GLP-1 [7-36NH2] GLP-1 [9-36NH 2] ) i n t y p e 2

diabet ic subjects af ter a tes t meal (550 kcal). Met-ormin signi cantly increased both glucagon and total

GLP-1 levels after one week; glucagon release was notl tered by the tes t meal , bu t w as s ign i cant ly greaterhan pa i r ed da ta ob ta ined p r io r to met fo rmin t r ea t -

ment (22). While plasma GLP-1 increased postprandi-lly in both contr ol a nd metformin t reat ed subjects, inhe m etformin trea ted group G LP -1 levels w ere signif-cant ly greater tha n the contro l g roup a t several t imeoints (22). P erha ps th e most simplistic int erpreta tionf these ndings is tha t m etformin either increases the

glucose sensit iv ity of the pancreat ic a lpha cell an denteroendocrine L-cell , or the secretory rate of thesecells , resulting in greater hormone release with met-formin t reatment . The work of Lugar i et a l . is sup-ported by studies published previously (23), wh ichfound t ha t metformin signi cantly increased release ofpancreat ic an d gut g lucagon (g licen t in an d oxynto-modulin intestinal products of proglucagon processingrelea sed in equa l a mounts t o G LP -1 from enteroendo-crine L-cells in response to lumina l nut rients (24)).

Mann ucci a nd colleagues tested obese non-diabeticsubjects using a euglycemic h yperinsulinemic clamptes t p rotocol , a s opposed to a tes t meal , in order toavoid glycemia induced alterations in GLP-1 release,ra ther tha n d irect effects of metformin (16). U nderthese condit ions , i t was found that GLP-1 [7-36NH2] w a ssigni cant ly greater in the metformin t reated group ,us ing a commercia lly ava i lab le as say speci c f orN-terminally intact GLP-1 (16). Unfortunately, totalGLP -1 levels were not measu red . An increase inN-terminally intact GLP-1 was interpreted as indicat-ing p rotect ion f rom deg rada t ion by DP IV, and thepossibility of a n increas e in tota l G LP -1 levels, yieldinga proportional rise in intact GLP-1 concentrations, wasnot considered. This possibility is consistent with priors tud ies examining g lucagon and G LP-1 levels af termetformin treatment (22, 23). Experiments were con-

FIG. 3. Effect of graded metformin concentra tions on interactions of GLP -1 [7-36NH2] or GLP -1 [9-36NH2] with porcine or recombinant hu ma n DPV. See t ext for specic methods.


1306


6/7

inued i n v i t r o using huma n p lasma f rom heal thy do-nor subjects and pur i ed p ig k idney DP IV, w i th o rwithout graded metformin concentrations (16). After a

0 min incubat ion , detect ion of G LP -1 [7-36NH2] b y t h eN-terminally speci c E LISA wa s reduced by 24% inhuman serum and 84% in pur i ed D P I V in t h e a b -ence of metformin, while addition of 0.5 g/ml met -ormin (approx. 3 mol/l) appear ed t o modera tely re-

verse t he loss in detection to 12%a nd 55%respectively.These ndings compelled us to perform i n v i t r o exper-ments using a lternat ive enzymologica l methods t o re-ssess th e role of metformin on D P IV.In contra s t to the i n v i t r o ndings of Mannucci et al.

16), w e w ere una ble to detect a ny signi cant effect ofmetformin on G ly-P ro-pNA hydrolysis, the prototy pica lDP IV substra te, in hea lthy huma n serum, puri ed pigkidney DP IV or recombinant human DP IV. Fur ther

xperiments using MALDI-TOF ma ss spectrometrywhich can concurrently detect disappearance and ap-

ea ra nce of t he molecular species corresponding toGLP-1 [7-36NH2] and GLP-1 [9-36NH2] , incubated in the same

nzyme prepara t ions , y ielded a t 1/2 para meter f romxponentia l deca y curves wh ich can be compared un der

d ifferen t exper imenta l condit ions . Consisten t withG ly-P ro-pNA enzym ological experiments, metformindid not a l ter the degrada t ion k inet ics of G LP -1 [7-36NH2]

ver a wide ra nge of concentra t ions in a ny of t he en-yme preparations.

Sur face plasmon resona nce (SP R) w as used to exam-ne the interaction between puri ed DP IV homologsnd GLP -1, i rr espect ive of ca ta lyt ic ac t iv ity. Themine-coupling reaction does not affect enzyme activ-ty (19), and SPR allows measurement of intermolecu-ar interactions not necessarily con n ed t o t h e c a t a -ytic site. From these studies, apparent K D values were

measu red for GLP -1 [7-36NH2] a n d G L P -1 [9-36NH2] . Met -ormin failed to alter the binding interaction between

GLP-1 [7-36NH2] or GLP-1 [9-36NH 2] and DP IV. Remarkably,he apparen t a f nity of GLP-1[9-36NH2] is very similar toh a t of G L P -1[7-36NH2] , even a t r educed tempera tu re .

Binding constants for N-terminally truncated glucagonragmen ts and pur i ed DP IV w ere not p rev ious lyested, however, studies indica ted t ha t substitution or

modi cation of the penultimate amino acid of glucagoneduced the apparent K D by approximately 10-fold, andhat a l ter ing the ch ira l i ty of Gln 3 produced more pro-

nounced effects on glucagon/DP IV intera ctions (19).St udies compar ing DP IV binding consta nts for G IP 1 42K D 1.7 mol/l ) a nd DP IV hydrolysis product ,

G IP 3 42 (K D 3.2 mol/l), r esulted in simila r ndingso those presented here (K. Ku hn-Wache, unpublished

dat a) . K D values determined for GLP-1 [7-36NH2] a n d G L P -[9-36NH2] shown a re one order of ma gnitude grea ter tha n

hose for GIP, measured under ident ical condit ions .However, s imilar to G IP , N-terminal t runcat ion doesnot d ramat ica l ly a f f ect b ind ing a fn ity f or D P I V.Hence, i t can be concluded that the substrate binding

with in the cat a ly t ic s ite of DP IV contr ibu tes l i t t le tothe overal l a fnity of the interaction between enzymeand substra te .

I n s um ma r y, w e h a ve a t t em pt ed t o d et er min ew h eth er m et f or m in h a s d ir ect ef fect s on D P I V-mediated GLP-1 degradation in v i t r o, and addi t ional lyha ve enhan ced our understa nding of G LP -1/DP IV in-tera ctions. We ha ve been una ble to support the claim

that metformin inhibits DP IV activity by a number ofdifferent experimenta l a pproa ches. The most likely ex-planation for the ndings of Ma nnucci et al. (16) w ithrespect to preservation of N-terminally intact GLP-1with metformin treatment is that metformin increasesthe secretion of total GLP-1, and thus a proportionalincrease in inta ct G LP -1 would be expected. I t is dif -cult to explain the disparate ndings i n v i t r o betweenthe current report a nd t ha t published previously, how-ever, the primary experimental omission of measuringonly in tact GLP-1 [7-36NH2] , and no t to ta l GLP-1 under-mines t he ear l ier dat a , as sample recovery cann ot beassessed. In contrast, MALDI-TOF mass spectrometryal lowed d irect detect ion of both in tact and inact iveG LP -1, and hence is more convincing. Su rfa ce plasmonresonance allowed measurement of af nity of interac-tion between enzyme and substrate. This was not sig-n icantly altered by metformin or the presence of aninta ct N-terminus.

In conclusion , i t appear s tha t metformin ma y in-crease hormone secretion from both the pancreatic a l-pha cell an d int estinal L-cell, resulting in great er glu-cagon and tota l GLP -1 levels in met formin t r ea tedindividua ls. The la tt er effect m ay be one of t he mech-a nisms by wh ich metformin improves glucose t oler-ance. With the emergence of poten t speci c D P IVinhibitors for th e trea tm ent of type 2 dia betes mellitus,an even g rea te r poten t ia l may l ie in combina to r ia ltrea tment w ith both metformin and DP IV inhibitors tomaximize the incretin effect.

ACKNOWLEDG MENTS

This work w as funded in pa r t by D epa r tment of Sc ience a ndTechnology of Sa chsen Anha lt (HU D G ra nt 9704/00116) a nd by t heM ed ica l Res ea r ch Cou n ci l of Ca n a d a (CH SM a n d RAP G r a n t

590007) and the Canadian Diabetes Associa t ion. Simon Hinke isgrateful for the support of the Killam Trusts, the Medical ResearchCouncil of Canada, and the Deutscher Akademischer Austausch-dienst (DAAD). The authors tha nk Madeleine Speck, J oachimB ar, Michae l Wermann, and Dr. Susanne Manhar t for technica lassistance .

REFERENCES

1. (1999) Diabetes Statistics. NI H Publ ica t ion No. 99-3892.2. DeFronzo, R. A. (1999) Pha rma cologic thera py for ty pe 2 diabe-

tes mellitus. Ann. I n tern . M ed. 131, 281 303.3. Zhang, B. B. , and Moller, D. E. (2000) New approaches in the

trea tment of type 2 diabetes.Cur r. Op i n . Chem . B i o l . 4,

461 467.


1307


7/7

4. K ieffer, T. J . , McIntosh, C. H. S. , and Pederson, R. A. (1995)Degra dat ion of glucose-dependent insulinotropic polypeptideand truncated glucagon-like peptide 1 in vitro and in vivo bydipeptidyl peptida se IV. Endocrinology 136, 3585 3596.

5. Unger, R. H., and Eisentraut, A. M. (1969) The entero-insularaxis . Arch. In te rn . M ed. 123, 261 266.

6. DAlessio, D . (1997) P eptide hormone regulat ion of islet cell.H orm. M etab. Res. 29, 297 300.

7. Mentlein, R., Ga llwitz, B ., and Schmidt, W. E. (1993) Dipeptidyl-

peptida se IV hydr olyses ga stric inhibitory polypeptide, glucagon-like peptide-1 (7-36)amide, peptide histidine methionine and isresponsible for their degradation in human serum. Eur. J . B i o- chem. 214, 829 835.

8 . Pau ly, R. P. , Rosche , F. , Wermann, M. , McIn tosh , C. H . S . ,Pederson, R. A. , a nd Demuth, H.-U. (1996) Investigation ofG I P 1 42 and GLP -1 7-36 degradation in vitro by dipeptidylpeptidase IV (DP IV) using Matr ix-Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry (MALDI -TOFMS): A novel kinetic approach. J. Biol . Chem. 271, 23222 23229.

9. Pederson, R. A., White, H. A., Schlenzig, D., P auly, R . P . , McIn-tosh, C. H. S., and Demuth, H.-U. (1998) Improved glucose tol-erance in zucker fa t ty ra ts by ora l a dministra t ion of the dipep-tidyl peptida se IV inhibitor isoleucine th iazolidide. Diabetes 47,

1253 1258.0. Pa uly, R. P. , Demuth, H.-U. , Rosche, F. , Schmidt, J . , White ,

H. A., Lynn, F., McIntosh, C. H. S., and Pederson, R. A. (1999)Improved glucose tolerance in ra t s t rea ted with the dipeptidylpeptidase IV (CD26) inhibitor ile-thiazolidide. M etaboli sm 48,385 389.

1. Ahren, B . , Holst , J . J ., Ma rtensson, H . , a nd Ba lkan, B . (2000)Improved glucose tolerance and insulin secretion by inhibition ofdipeptidyl peptidase IV in mice. Eur. J . Pharmacol . 404, 239 245.

2. Dea con, C . F ., H ughes, T. E ., a nd Holst, J . J . (1998) Dipeptidylpeptidase IV inhibition potentiates the insulinotropic effect ofglucagon-like peptides 1 in the anesthetized pig. Diabetes 47,764 769.

3. B alka n, B ., Kwa snik, L., Miserendino, R., Holst, J . J . , and Li, X.(1999) Inhibition of dipeptidyl peptidase IV with NVP-DPP728increases plasma GLP -1 (7-36 a mide) concentra t ions and im-proves oral glucose tolerance in obese Zucker ra ts. Diabetologia 42, 1324 1331.

14. Ba iley, C. J . , a nd Turner, R. C. (1996) Metformin. N . E n g l .J . Med. 334, 574 579.

15. DeFronzo, R. A., and Goodman, A. M. (1995) Ef cacy of met-formin in pat ients w ith n on-insulin-dependent diabetes mellitus.N. Eng l . J . M ed . 333, 541 549.

16. Mannucci, E., Ognibene, A., Cremasco, F., Bardini, G., Mencucci,A. , Pierazzuoli , E. , Ciani , S. , Messer i , G. , and Rote lla , C. M.(2001) E ffect of met formin on glucagon-like peptide 1 (G LP -1)and leptin levels in obese nondiabetic subjects. Di abetes Care 24,

489 494.17. Mannucci, E., Ognibene, A., Cremasco, F., Bardini, G., Mencucci,

A. , Pierazzuoli , E. , Ciani , S. , Messer i , G. , and Rote lla , C. M.(2000) Effects of metformin on bas eline an d ora l glucose-inducedglucagon-like peptide-1 (GLP-1) in obese non-diabetic subjects,Abstract #472. Diabetes 49 (Suppl. 1).

18. Wolf, B ., Fischer, G., and B a rt h, A. (1978) [Kinet ics of dipeptidyl-peptida se I V] (Germa n). Acta Bi ol . M ed. Ger. 37, 409 420.

19. Hinke, S. A. , P ospisi l ik, J . A. , D emuth, H.-U. , Mannha r t , S. ,K u hn-Wache, K ., Hoffman n, T., Nishimura , E ., P ederson, R. A.,and McIntosh, C. H. S. (2000) Dipeptidyl peptidase IV degrada-tion of glucagon: Chara cterization of glucagon degrada tion prod-ucts and DP IV resistant a nalogs. J. Bi ol . Chem. 275, 3827 3834.

20. Rosche, F. , Schmidt, J . , H offmann, T. , P auly, R. P. , McIntosh,C. H. S. , Pederson, R. A. , and Demuth, H. U. (2000) Kineticanalysis of enzymatic and nonenzymatic degradation of peptidesby MALDI-TOF MS. M ethods Mol . Bi ol . 146, 251 272.

21. Scheen, A. J . (1996) C linical pharm acokinetics of m etformin.Clin. Pharmacokinet . 30, 359 371.

22. Lugar i , R. , DellAnna, C. , Sar t i , L. , Coppi, S. , Ver la to, C. A. ,Sbordone , P . , Bianco, M., Gnudi , A. , and Zandomeneghi , R.(1998) Effects of metformin on intestinal and pancreatic endo-crine secretion in type 2 (non-insulin-dependent) diabetes. I n Molecular and Cell Biology of Type 2 Diabetes and Its Compli-cations (Belore, F. , Lorenzi, M., Molinat ti, G . M., and P orta, M.,Eds.), pp. 161 163, Karger, Basel.

23. Molloy, A. M., Ardill, J ., an d Tomkin, G . H . (1980) The effect ofmetformin t rea tment on gastr ic a c id secretion and gastrointes-tina l hormone levels in n ormal subjects. Diabetologia 19, 93 96.

24. Kieffer, T. J ., and Habener, J . F. (1999) The glucagon-like pep-tides. Endocr. Rev. 20, 876 913.


1308

metformina y glp-1 efecto hinke 2002

Documents