Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Tresperimus (Cellimis), is a new immunosuppressive polyamine inducing a specific tolerance in rat cardiac allograft model after a short-term treatment. It is effective in the prevention of murine graft-versus-host disease (Dutartre et al., 1995; Andoins et al., 1996, 1997). The drug is mainly eliminated through an extensive nonhepatic metabolism that occurs mainly at the primary amino terminal group but also to a much lesser extent, at the middle (amide bond) of the molecule (Fig. 1) (Claud et al., 2001). The presence of oxidative deamination reactions in both biotransformation pathways suggests that the amine oxidases play a major role in tresperimus metabolism. Moreover, the significant increase of tresperimus plasma concentrations observed in one patient with graft-versus-host-disease, treated simultaneously with isoniazid and tresperimus, seemed to support this hypothesis. Indeed, the antitubercular drug isoniazid inhibits plasma and tissue-bound SSAO¹ with a relatively weak effect upon monoamine oxidase (MAO) (Blaschko, 1962; Lewinsohn et al., 1978). Besides, the ability of plasma SSAO to deaminate tresperimus has been shown in vitro (Claud et al., 2001).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Metabolic scheme of tresperimus in man (Claud et al., 2001). The chemical names have been previously published (Claud et al., 2001). M1, aldehyde derivative of tresperimus; M2, acid derivative of tresperimus; M3, desaminopropyl derivative of tresperimus; M4, aldehyde derivative of M3; M5, acid derivative of M4; M6, guanidinohexylamine; M7, 2-[[[[4-[(3-aminopropyl)amino]butyl]amino]carbonyl]oxy]-acetic acid; M8, acid derivative of M6.

The heterogeneous class of amine oxidases can be divided into two types according to the chemical structure of their cofactor (Strolin-Benedetti and Tipton, 1998). MAO-A and -B contain flavin adenine dinucleotide as a prosthetic group, whereas a second group is composed of diamine oxidase, lysyl oxidase, and SSAO with topaquinone as a cofactor (Janes et al., 1990; Buffoni and Ignesti, 2000).

Amine oxidases can be distinguished by their different sensitivity to inhibitors. Indeed, semicarbazide at concentrations around 1 mM has a relatively weak effect on MAO, whereas SSAO is not affected by acetylenic aromatic amines such as clorgyline or deprenyl, which cause selective inhibition of MAO-A and MAO-B at concentrations inferior to 1 µM (Johnston, 1968; Knoll and Magyar, 1972; Lyles, 1996).

As opposed to platelet-poor plasma previously used to study in vitro tresperimus metabolism, both amine oxidase classes are present within blood vessels. Indeed, smooth muscle cells within the vascular wall represent one of the main tissue location of SSAO, and MAO activity is also well characterized in this tissue (Precious and Lyles, 1988; Berry et al., 1994). The major proportion of tissue SSAO activity is membrane-bound with the greatest enrichment after subcellular fractionation being in microsomal fraction. On the other hand, MAO is primarily localized in the outer mitochondrial membrane although some MAO activity is also associated with microsomes (Gómez et al., 1988).

The umbilical artery could thus constitute a relevant in vitro model to assess the contribution of the different amine oxidase classes to tresperimus metabolism in man. The suitability of homogenates of human umbilical artery to study tresperimus metabolism was tested by determining the metabolic profiles of the unchanged drug and of two metabolites [desaminopropyl derivative of tresperimus (M3) and guanidinohexylamine (M6)] subject to oxidative deamination reactions (Fig. 1). Metabolites formed by vascular homogenates were identified by LC/MS analysis. The optimal pH for tresperimus and M3 metabolism was determined using an HPLC assay. Chemical inhibition experiments were then performed to identify the class of amine oxidase involved in tresperimus, M3, and M6 metabolism. The microsomal fraction of umbilical artery was used to characterize tresperimus and M3 deamination. The kinetic parameters of membrane-bound and plasma-soluble SSAO for the drug and the inhibitory constant of tresperimus toward benzylamine, the classical synthetic substrate of SSAO, metabolism were allowed to compare the kinetic behavior of both enzymes.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Chemicals. Tresperimus HCl was synthesized at Synkem (Chenôve, France). M3, M6, and LF 07-0109 as internal standard were synthesized by Laboratoires Fournier (Dijon, France). Benzylamine, R-()-deprenyl, clorgyline, hydralazine, and semicarbazide were purchased from Sigma Chemicals (Saint Quentin Fallavier, France), and potassium phosphate was obtained from Merck (Darmstadt, Germany).

Methods.

Preparation of Enzyme Sources Umbilical cords were supplied by the Maternity Unit, Chenôve Hospital (Chenôve, France). Arteries were dissected, washed in saline to eliminate blood as a potential source of contaminating plasma amine oxidases, and stored frozen at 80°C for up to 1 month before use in subsequent experiments. They were then homogenized in 10 mM potassium phosphate buffer, pH 7.4 (1 g of tissue/30 ml), and centrifuged at 600g for 10 min. The supernatants from this low-speed centrifugation were used as the enzyme source or centrifuged (12,000g for 20 min at 4°C) to eliminate mitochondria and big cellular debris. Subsequent centrifugation of the resulting supernatant (105,000g for 1 h) at 4°C produced the microsomal pellet, which was suspended in the same buffer (3 ml of buffer for 100 mg of membrane fractions). A pool of around eight arteries was constituted for each experiment. All the experiments were performed with an enzyme source freshly prepared.

Tresperimus, M3, and M6 Incubation. The incubation mixtures consisted of 1 ml of enzyme source (homogenates or microsomes of umbilical artery prepared in 10 mM potassium phosphate buffer and plasma). Sodium hydroxide was added to the vascular preparations before use for adjusting to the appropriate pH. The addition of sodium hydroxide did not significantly modify the ionic strength of the incubation medium. The reaction was initiated by the addition of 10 µl of substrate dissolved in potassium phosphate buffer at 10 mM, pH 7.4. At each incubation time, the reaction was stopped by transferring 0.4 ml of reaction mixture into tubes previously placed in ice. An acidic reagent containing citric acid (0.5 M)/NaH₂PO₄ (0.5 M) 70:30 (v/v) was added at a concentration of 5% (v/v), and the tubes were centrifuged at 2,400g for 5 min at 4°C and stored at 20°C until analysis.

Tresperimus, Desaminopropyl Tresperimus, and Guanidinohexylamine HPLC Assay. After a liquid-solid extraction on a C₁₈ cartridge of the sample, a reversed-phase ion-pairing HPLC method, similar to that developed previously, was used to quantify the unchanged tresperimus, M3, and M6 (Claud et al., 2001).

LC/MS Qualitative Analysis of Tresperimus Metabolism. LC/MS was used to identify the metabolites of tresperimus, M3, and M6 produced during incubations in homogenates and microsomes of umbilical artery. The treatment of samples and analytical conditions were previously published (Claud et al., 2001).

Effects of Tresperimus upon Benzylamine Metabolism in Microsomes of Umbilical Arteries and in Human Plasma. A typical incubation medium was composed of 300 µl of potassium phosphate (0.1 M, pH 7.8), 50 µl of 10 mM deprenyl in 0.01 M hydrochloric acid or hydrochloric acid alone, 50 µl of tresperimus in potassium phosphate buffer, pH 7.8, or buffer alone, and 50 µl of enzyme source. The reaction was started by the addition of 50 µl of benzylamine (in 0.01 M HCl). The reaction was terminated by the addition of 50 µl of 40% trichloroacetic acid, and the tubes were centrifuged for 5 min at 2,500g.

Protein Assay. Protein concentrations of homogenates and microsomes were measured by the method of Bradford with bovine serum albumin as the standard (Bradford, 1976).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Tresperimus, M3, and M6 Metabolic Profiles in Homogenates and Microsomes of Umbilical Artery. We checked that the tresperimus metabolic profile was not modified according to the pH. From the LC/MS spectra, the tresperimus metabolic scheme could be drawn in homogenates and microsomes of umbilical artery. This scheme appeared similar in both biological matrices. Only the metabolites formed from the spermidine moiety of the drug were detected, namely the aldehyde and acid derivative of tresperimus, M3, and the corresponding aldehyde and acid derivative of this metabolite. However, in microsomes the acid derivative of tresperimus (M2) was only detected in very low amounts, near the limit of detection. M6 formation was not observed in these cellular fractions no matter what the incubation conditions were, including pH value.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Selected ion monitoring. Mass chromatograms at m/z 330, 331, and 346 for M3 metabolic profile (A) and 158, 159, and 174 for M6 metabolic profile (B). M3 (3.0 µM) was incubated at 37°C for 45 min and M6 (12.7 µM) for 2 h at 37°C in homogenates of umbilical artery at pH 7.4 and pH 9.3, respectively. The metabolic profile of each compound was determined by LC/MS as described under Experimental Procedures.

Effect of pH on Tresperimus and M3 Metabolism. We first checked that chemical stability of tresperimus and M3 were not affected by the pH values we tested. The variations of initial velocity for tresperimus and M3 metabolism in homogenates as a function of pH values are presented in Fig. 3; they were very close for both substrates, with a similar optimum at pH 9.3 (Fig. 3B). On the other hand, the metabolism of both substrates was clearly reduced at physiological pH, especially for tresperimus metabolism. As shown by Fig. 3A, activity toward M3 metabolism appeared significantly higher than that for tresperimus no matter what the considered pH values were.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Effect of different pH values on tresperimus and M3 initial velocities in homogenates of umbilical artery. Tresperimus (; 2.6 µM) and M3 (; 3.0 µM) were incubated at 37°C in homogenates of umbilical artery brought to the appropriate pH by the addition of sodium hydroxide. The pH of the reaction mixtures was checked before and after the reaction and was found constant. The reactions for each compound occurred over a time period for which the linearity of the reaction velocity had been previously demonstrated. Each point is the mean (±S.D.) of incubations assayed in triplicate. The initial velocity is estimated by the compound disappearance assayed by HPLC. The activity is expressed in picomoles of product disappeared per minute per milligram of protein (A). The activity at the optimal pH was taken as 100% of activity (B).

Effects of Amine Oxidase Inhibitors on Tresperimus, M3, and M6 Metabolism. To discriminate between MAO and SSAO activities involved in tresperimus, M3, and M6 metabolism, the effect of different specific and irreversible inhibitors was tested; clorgyline and deprenyl were used toward MAO activity, semicarbazide for SSAO, and hydralazine for both enzymatic activities in the conditions used. We previously ascertained that these amine oxidase inhibitors could not interfere with the measure of unchanged tresperimus, M3, and M6 as well as in the liquid-solid extraction as in the HPLC assay. Therefore, we showed that the quantities of unchanged tresperimus, M3, and M6, measured in diluted human plasma used as a biological matrix for the calibration curves, were not modified in the presence of the amine oxidase inhibitors.

View larger version (51K): [in this window] [in a new window]   Fig. 4.   Effect of semicarbazide, deprenyl, clorgyline, and hydralazine upon tresperimus metabolism. Tresperimus was incubated at 2.6 µM for 2 h and 15 min at 37°C in the homogenate of umbilical artery at pH 8.2. Homogenate samples were preincubated for 1 h at 37°C with various inhibitor concentrations. Mean specific activity of control samples was 7.8 ± 1.0 pmol/min/mg of protein. Deaminating activities are expressed as a percentage of those in control samples preincubated without inhibitor. Each point is the mean (±S.D.) of incubations assayed in triplicate. The remaining activities are calculated from the drug disappearance measured by HPLC as described under Experimental Procedures.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Effects of semicarbazide, deprenyl, clorgyline, and hydralazine upon the metabolism of M3 and M6. M3 (3 µM) and M6 (12 µM) were incubated for 45 min and 2 h at 37°C, respectively. Homogenate samples were preincubated 1 h at 37°C with various inhibitor concentrations. Mean specific activities of control samples were 44.9 ± 7.9 and 48.2 ± 7.2 pmol/min/mg of protein for M3 and M6, respectively. Deaminating activities are expressed as a percentage of those in control samples preincubated without inhibitor. Each point is the mean (±S.D.) of incubations assayed in triplicate. The control activity is estimated by the compound disappearance assayed by HPLC.

K_m and V_max Determinations. Kinetic constants for tresperimus and M3 metabolism were determined by nonlinear regression according to the Michaelis-Menten rate equation. Kinetic constants for tresperimus metabolism were K_m (µM) 1.2 ± 0.3 at pH 9.3 in microsomes of umbilical artery and 1.9 ± 0.4 at pH 7.8 in plasma; the corresponding values for V_max (pmol/min/mg of protein) were 355.1 ± 18.3 in microsomes of umbilical artery and 12.6 ± 0.8 pmol/h/liter of plasma (equivalent to 0.2 ± 0.01 pmol/min/ml of plasma).

View larger version (7K): [in this window] [in a new window]   Fig. 6.   Tresperimus and M3 metabolism in microsomes of umbilical artery and human plasma. Left panel plot represents the Michaelis-Menten curves for tresperimus and M3 metabolism in microsomes of human umbilical artery at pH 9.3 and in plasma. The initial velocities are estimated by the compound disappearance assayed by HPLC as described under Experimental Procedures. Reactions were carried out over a time period for which the linearity of the reaction velocity was previously demonstrated. The parameters Km and Vmax, with standards errors, were estimated by fitting the Michaelis-Menten equation to the data using nonlinear regression analysis with the software GraphPad PRISM version 2.0 computer program. The substrate concentration velocity (from 0 to 30 µM) relationship for tresperimus metabolism was then analyzed by Hanes-Wolf plot (right panel) to reveal putative separate catalytic activities. , tresperimus metabolism in plasma; , tresperimus metabolism in microsomes; , M3 metabolism in microsomes.

Kinetic Parameters for Benzylamine Metabolism. Benzylamine is a substrate for MAO and SSAO. To evaluate the contribution of each enzymatic activity to benzylamine metabolism, kinetic data for benzylamine metabolism were determined in the presence or absence of MAO inhibitor (Table 1). In microsomes of umbilical artery, the K_m value for benzylamine metabolism obtained in the presence of deprenyl was greater (by 60%) than the value determined without this inhibitor. A similar phenomenon was observed with another preparation in which the K_m value without deprenyl (157 µM) was increased by 37% in deprenyl-treated samples (K_m, 215 µM). In contrast, V_max values were relatively unchanged. These results showing increases in apparent K_m values with deprenyl without any change in V_max values could be indicative of an apparent competitive action of deprenyl upon the total benzylamine-metabolizing activity. This does not support the notion that deprenyl is selectively inactivating a small MAO-B component contributing to the total benzylamine oxidase (BzAO) activity. Thus BzAO activity depends only on SSAO in our conditions.

Effects of Tresperimus upon Benzylamine Metabolism in Microsomes of Umbilical Arteries and in Human Plasma. The K_I value of tresperimus on benzylamine oxidase activity in microsomes of umbilical artery was determined in the presence or absence of deprenyl since this MAO inhibitor seemed to interact on BzAO activity. On the other hand, the K_I of tresperimus in plasma was determined only on total benzylamine oxidase activity since deprenyl did not seem to have an effect on plasma SSAO. Results were depicted as shown in the Lineweaver-Burk plot in Fig. 7. Tresperimus produced a mixed pattern of inhibition against benzylamine metabolism in both enzyme sources tested. The apparent K_m values for benzylamine metabolism were increased by tresperimus whereas the V_max values were decreased. The K_I value of tresperimus for total benzylamine metabolism in microsomes of umbilical artery (31.9 ± 3.9 µM) was close to that in plasma (39.5 ± 4.8 µM). As for the determination of K_m for benzylamine metabolism, this K_I value (89.0 ± 2.4 µM) was greater for deprenyl-treated microsome samples.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Lineweaver-Burk plot showing inhibition of benzylamine metabolism by treperimus in microsomes of umbilical artery and human plasma. Benzylamine was incubated from 10 µm to 1 mM in the presence of final Tresperimus concentrations ranged from 0 to 200 µM. Each point is the mean of triplicate determinations. The KI value for tresperimus was determined successively in microsomes of the umbilical artery after 30 min of preincubation with deprenyl at 1 mM (A); the comparison was made after a preincubation with the inhibitor vehicle alone (B) and in human plasma (C). The KI value for tresperimus was determined by linear regression from the slope replot (D). The benzylamine oxidase activity is measured by the appearance of benzaldehyde assayed by HPLC as described under Experimental Procedures. The KI values determined in deprenyl-treated microsomes (A), in microsomes (B), and in plasma (C) were 89 ± 2.4 µM, 31.9 ± 3.9 µM, 39.5 ± 4.8 µM, respectively.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

All the oxidative deamination reactions identified within the tresperimus metabolic scheme occurred in human umbilical artery. Indeed unchanged drug, M3, and M6 were deaminated in vascular homogenates as shown by the detection of the aldehyde and acid derivatives corresponding to these substrates. The deaminating activity toward tresperimus and M3 took place also in the microsomal fraction of this tissue. Besides, M6 metabolism by umbilical artery allowed the demonstration of the formation of the aldehyde derivative of M6, although this intermediary compound leading to the formation of M8 has never been detected in biological fluids in vivo. The oxidation of the aldehyde derivatives formed by these deamination reactions constituted clues to the presence of a functional NAD(P)-linked aldehyde dehydrogenase (ALDH; EC.1.2.1.3) despite the absence of supplementation in cofactors. On the other hand, the hydrolysis of the amide bond of tresperimus, leading to the release of M6, did not take place in these biological matrices no matter what the pH of the incubation medium was.

The use of umbilical artery confirmed the major involvement of SSAO in tresperimus deamination. Indeed, semicarbazide strongly inhibited tresperimus metabolism at 1 mM and 100 µM, whereas the oxidation of the drug was weakly sensitive to deprenyl or clorgyline at 100 µM and resistant to these inhibitors at 50 µM. The sensitivity of tresperimus metabolism to these inhibitors at 1 mM did not seem to reflect an MAO involvement in the drug-deaminating activity. Indeed, the A-form of the enzyme is sensitive to inhibition by low (in the nanomolar range) concentrations of clorgyline but is not inhibited by l-deprenyl until micromolar concentrations of this inhibitor are used. The reverse is true for the B-form with deprenyl (Fowler et al., 1982). Both forms of MAO activities should be inhibited at 100 µM and still highly decreased at 50 µM. Therefore the decrease of tresperimus metabolism in the presence of MAO inhibitors at high concentration could rather reflect an interaction between these inhibitors and tissue-bound SSAO, as shown by the apparent competitive action of 1 mM deprenyl on benzylamine metabolism in the presence or in the absence of tresperimus. Similar inhibition has already been reported with clorgyline in umbilical artery (Precious and Lyles, 1988). The contribution of MAO to tresperimus metabolism thus appeared negligible at physiologically relevant concentrations and could be reflected by the low percentage of remaining activity (5%) in the presence of 1 mM semicarbazide. In any event, tresperimus metabolism was quite blocked when MAO and SSAO activities were inhibited as shown by hydralazine up to 10 µM. Indeed, the antihypertensive drug hydralazine acted as an inhibitor of both forms of amine oxidases in the experimental conditions used (Lyles et al., 1983).

Furthermore the mixed inhibition caused by tresperimus on benzylamine metabolism in human plasma and in microsomes of umbilical artery, a substrate only for SSAO in the assay conditions, supported that both substrates are metabolized by the same catalytic entity. Indeed a mixed inhibition is common in multisubstrate systems (Frieden, 1964).

SSAO seemed to be also predominantly responsible for M3 and M6 metabolism. Indeed, MAO inhibitors at 50 µM did not affect significantly M3 and M6 metabolism although the metabolism of these two compounds was strongly decreased by semicarbazide at 1 mM and 100 µM with around 85 to 90% and 50% inhibition, respectively. The incomplete inhibition of M3 and M6 metabolism by high semicarbazide concentrations could reflect a limited contribution (around 10 to 20%) of MAO to their metabolism. Anyway M3 and M6 metabolism was completely suppressed by hydralazine, an inhibitor of both MAO and SSAO in our experimental conditions.

The microsomal fraction of umbilical artery, an enzyme source with a high specific SSAO activity and the contaminating MAO activity being partly eliminated, was used to determine the kinetic parameters of SSAO for tresperimus and M3 metabolism (Suzuki and Matsumoto, 1984). As shown by the Hanes-Wolf plots, tresperimus and M3 metabolism was only dependent on SSAO activity in the microsomal fraction. Membrane-bound enzyme showed a high affinity of SSAO for both compounds as demonstrated by the K_m values of 1.2 and 3.3 µM for tresperimus and M3 metabolism, respectively. These K_m values were much lower than that for aminoacetone metabolism, one of the lowest K_m values (92 µM) for a human physiological SSAO substrate determined under comparable assay conditions (Lyles and Chalmers, 1992). On the other hand, the maximal velocity of microsomal SSAO for tresperimus (0.36 nmol/min/mg of protein) was very low compared with those for M3 and benzylamine metabolism (2.77 nmol/min/mg of protein and 52.1 ± 2.0 nmol/min/mg of protein, respectively).

The kinetic behavior of the soluble and the membrane-bound SSAO toward tresperimus metabolism was compared since the relationships between both enzymes remains unclear (Lizcano and Unzeta, 2000). The membrane-bound enzyme and the soluble form of SSAO exhibited similar K_m values for the drug and K_I values of tresperimus toward benzylamine metabolism. Thus, this kinetic behavior for tresperimus metabolism of membrane-bound and soluble SSAO seemed to argue in favor of a same catalytic entity.

Nevertheless, both forms of the enzyme seemed to behave differently according to the pH. Indeed tresperimus metabolism showed a strong pH dependence; it was almost abolished in umbilical artery at pH values inferior to pH 7.8, although the plasma enzyme activity could be studied at this pH value. The pH dependence of the reaction mechanism of SSAO has been often reported and was also observed for M3 metabolism (Farnum et al., 1986; Lindström et al., 1974; Medda et al., 1995). Besides, SSAO activity is reduced or even suppressed for substrates that carry ionized groups close to the terminal amino group (McEwen, 1965; Buffoni et al., 1972). The presence of the secondary amine of tresperimus as a protonated form at physiological pH could explain the low catalytic efficiency of membrane-bound SSAO for tresperimus at physiological pH. The behavior of the membrane-bound enzyme according to the pH could result from modifications of the micro-environment of the enzyme during microsomal preparation. Indeed the breaking up of the membrane architecture notably the electrostatic bonding with phospholipids of a membrane enzyme can cause a shift of pH optimum (Silman and Karlin, 1967; Coleman, 1973; Houslay and Tipton, 1973).

Umbilical artery provided qualitative and quantitative information on tresperimus metabolism in man. Crude homogenates of this tissue allowed to demonstrate the predominant role of SSAO in tresperimus, M3, and M6 metabolism and to exclude significant contributions of MAO to the metabolism of these compounds at physiologically relevant concentrations. Furthermore, ALDH represents an important group of detoxification enzymes, and characterization of the ALDH form(s) present in homogenates and microsomes of umbilical artery would constitute an important step to further assess the interest of such metabolic models (Sladek et al., 1989; Lindahl, 1992). Since membrane-bound SSAO and plasma-soluble SSAO exhibited similar kinetic properties, the use of microsomal fraction, an enzyme source with a high specific SSAO activity, could be relevant for the assessment of quantitative data on drug metabolism or drug interaction potential.

Tresperimus metabolism highlights the important role that SSAO could play as a Phase I oxidative enzyme to scavenge certain exogenous amines entering the bloodstream. Such an important contribution has been rarely demonstrated in xenobiotic metabolism. Indeed, the contribution of amine oxidases has been often neglected, especially due to the lack of appropriate in vitro models to clearly demonstrate their involvement (Strolin-Benedetti, 2001).