JOURNAL OF MASS SPECTROMETRY
`J. Mass Spectrom. 2002; 37: 581–590
`Published online 7 May 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.316
`
`Identification of phenylbutyrylglutamine, a new
`metabolite of phenylbutyrate metabolism in humans
`
`Blandine Comte,1 Takhar Kasumov,1 Bradley A. Pierce,1 Michelle A. Puchowicz,1
`Mary-Ellen Scott,1 Williams Dahms,2 Douglas Kerr,2 Itzhak Nissim3 and
`Henri Brunengraber1∗
`
`1 Department of Nutrition, Case Western Reserve University, Cleveland, Ohio 44106, USA
`2 Department of Pediatrics, Case Western Reserve University and University Hospitals of Cleveland, Cleveland, Ohio 44106, USA
`3 Department of Pediatrics, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA
`
`Received 26 December 2001; Accepted 1 March 2002
`
`Phenylbutyrate is used in humans for treating inborn errors of ureagenesis, certain forms of cancer, cystic
`fibrosis and thalassemia. The known metabolism of phenylbutyrate leads to phenylacetylglutamine,
`which is excreted in urine. We have identified phenylbutyrylglutamine as a new metabolite of
`phenylbutyrate in human plasma and urine. We describe the synthesis of phenylbutyrylglutamine
`and its assay by gas chromatography/mass spectrometry as a tert-butyldimethylsilyl or methyl derivative,
`using standards of [2H5]phenylbutyrylglutamine and phenylpropionylglutamine. After administration of
`phenylbutyrate to normal humans, the cumulative urinary excretion of phenylacetate, phenylbutyrate,
`phenylacetylglutamine and phenylbutyrylglutamine amounts to about half of the dose of phenylbutyrate.
`Thus, additional metabolites of phenylbutyrate are yet to be identified. Copyright © 2002 John Wiley &
`Sons, Ltd.
`
`KEYWORDS: phenylacetate; phenylbutyrate; phenylacetylglutamine; phenylbutyrylglutamine; liver conjugation
`
`INTRODUCTION
`
`Phenylbutyrate (PB) has been used as a pro-drug of
`phenylacetate (PA) in the treatment of hyperammonemia
`related to inborn errors of urea synthesis1 and as a drug
`in the treatment of a number of malignancies,2 – 4 cystic
`fibrosis5 and sickle cell anemia.6 The main fate of PA in
`primates and humans is its conjugation as phenylacetyl-CoA
`with glutamine in the liver to form phenylacetylglutamine
`(PAGN), which is excreted in urine.7 – 9 The labeling pattern
`of PAGN has been used for the non-invasive probing of the
`labeling pattern of citric acid cycle intermediates in human
`and primate liver.10 – 14 Compared with PA, PB has a more
`acceptable taste and smell and is less toxic.2 However, after
`administration of PB, the combined urinary excretion of PB,
`PA and PAGN is less than half of the ingested amount
`of PB.15 We hypothesized that by analogy with PA, PB
`is activated in the liver to phenylbutyryl-CoA (PB-CoA),
`
`Ł
`Correspondence to: Henri Brunengraber, Department of
`Nutrition, Case Western Reserve University, Cleveland, Ohio
`44106-7139, USA. E-mail: hxb8@po.cwru.edu
`Contract/grant sponsor: NIH.
`Contract/grant sponsor: Cleveland Mt Sinai Health Care
`Foundation.
`Abbreviations: EI, electron ionization; GC/MS, gas
`chromatography/mass spectrometry; PA, phenylacetate; PAGN,
`phenylacetylglutamine; PAG, phenylacetylglutamate; PB,
`4-phenylbutyrate; PBGN, 4-phenylbutyrylglutamine; PBG,
`4-phenylbutyrylglutamate; PCI, positive chemical ionization; PP,
`3-phenylpropionate; PPG, 3-phenylpropionylglutamate; PPGN,
`3-phenylpropionylglutamine; TBDMS, tert-butyldimethylsilyl.
`
`581
`
`which could either undergo ˇ-oxidation to PA-CoA (the
`precursor of PAGN), or be conjugated with glutamine to
`form a new compound, phenybutyrylglutamine (PBGN).
`The latter would be excreted in urine with PAGN.
`We synthesized unlabeled and 2H-labeled PBGN, and
`present sensitive methods for its determination in biological
`fluids by GC/MS. Further, we assayed PB and its metabolites
`in plasma and urine from seven humans after an oral bolus
`of Na-PB. We demonstrated the formation and urinary
`excretion of PBGN.
`
`EXPERIMENTAL
`Materials
`Chemicals and solvents were obtained from Sigma-
`Aldrich Chemicals (Milwaukee, WI, USA). [2H7]Phenylacetic
`acid (99%) and [2H6]benzene were purchased from Isotec
`(Miamisburg, OH, USA). The derivatization agents N-
`methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide (MtB-
`STFA) and Methyl-8 (dimethylformamide dimethyl acetal)
`were supplied by Regis Chemical (Morton Grove, IL, USA)
`and Pierce (Rockford, IL, USA), respectively. All aqueous
`solutions were made with water purified with a Milli-Q
`system (Millipore).
`
`Preparation of unlabeled and deuterated standards
`(Table 1)
`PAGN and [2H7]PAGN were synthesized16 by reacting
`unlabeled or [2H7]phenylacetyl chloride with glutamine, as
`
`P LUPIN
`Copyright © 2002 John Wiley & Sons, Ltd.
`EX. 1025
`
`

`
`(1C,Cg3)
`(2C,Cb2,Cb4),31.54(1C,Cg4),27.11(1C,Cb3),26.98
`141.57(1C,Cipso(cid:2)ar),51.72(1C,Cg2),34.80,34.75
`174.5,173.65,178.46(3C,3C(O),Cg1,Cg5,Cb1),
`
`CH2b4),2.30(m,4H,2CH2),1.92(m,4H,2CH2)
`7.50(d,1H,NH),6.30(bs,2H,NH2),2.65(m,2H,
`
`127–128
`
`OH
`
`CHC
`
`HN
`
`O
`
`O
`
`H2
`
`C N
`
`CH2
`
`CH2
`
`O
`
`D
`
`D
`
`D
`
`D
`
`D
`
`582
`
`B. Comte et al.
`
`27.02(1C,Cg3)
`34.99(2C,Cb2,Cb4),31.72(1C,Cg4),27.46(1C,Cb3),
`Cmeta(cid:2)ar),125.66(1C,Cpara(cid:2)ar),51.78(1C,Cg2),35.24,
`141.57(1C,Cipso(cid:2)ar),128.33,128.15(4C,2Cortho(cid:2)ar,
`175.26,173.73,173.00(3C,3C(O),Cg1,Cg5,Cb1)
`
`(m,4H,CH2b2,CH2g4),1.90(m,4H,CH2b3,CH2g3)
`NH2),4.4(m,1H,CHg2),2.62(m,2H,CH2b4),2.28
`7.48(d,1H,NH),7.23(m,5H,Ph),6.28(s,2H,
`
`13CNMR(50MHz,υppm,CDCl3)b
`
`1HNMR(200MHz,υppm,CDCl3)a
`
` (PBGN)
`
`4-Phenylbutyrylglutamine
`
`NH2
`
`O
`
`C
`
`CH2
`
`CH2
`
`O
`
`126–127
`
`M.p.(°C)
`
`OH
`
`CHC
`
`HN
`
`O
`
`Compound
`
`1
`
`Row
`
`Table1.Characteristicsofsynthesizedcompounds
`
`2
`
`3
`
`4
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`Ccycle),20.84(1C,Cb3)
`Cb2),35.12(1C,Ccycle),31.97(1C,Ccycle),25.70(1C,
`125.96(1C,Cpara(cid:2)ar),57.70(1C,Ccycle),36.07(1C,
`Cipso(cid:2)ar),128.55,128.37(4C,2Cortho(cid:2)ar,2Cmeta(cid:2)ar),
`176.19,174.19,174.15(3C,3C(O),),141.88(1C,
`
`1.96(m,2H,CH2b3)
`CH2cycle),2.7–2.18(m,6H,CH2cycle,CH2b2,CH2b4),
`7.25(m,5H,Ph),4.76(m,1H,CHcycle),2.98(m,2H,
`
`57–58
`
`21.24(1C,Cb3)
`36.08(2C,Cb2,Cb4),35.1(1C,Cg4),25.71(1C,Cg3),
`2Cmeta(cid:2)ar),125.80(1C,Cpara(cid:2)ar),51.74(1C,Cg2),
`141.60(1C,Cipso(cid:2)ar),128.50,128.30(4C,2Cortho(cid:2)ar,
`175.03,173.70,172.92(3C,3C(O),Cg1,Cg5,Cb1),
`
`2.38–1.90(m,8H,2CH2b,2CH2g)
`NH),4.58(m,1H,CHg2),2.61(m,2H,CH2b4),
`8.5(bs,2H,2COOH),7.22(m,5H,Ph),6.92(d,1H,
`
`acid
`
`N-(4-Phenylbutyryl)pyroglutamic
`
`O
`
`O
`
`COH
`
`O
`
`N
`
`(PBG)
`
`4-Phenylbutyrylglutamate
`
`O
`
`H
`C O
`
`CH2
`
`CH2
`
`O
`
`OH
`
`CHC
`
`HN
`
`69–70
`
`O
`
`4-[2H5]Phenylbutyrylglutamine
`
`([2H5]PBGN)
`
`

`
`Identification of phenylbutyrylglutamine
`
`583
`
`b13CNMRspectrumofPPGNwasmonitoredinDMSO-d6.
`moietiesoftheappropriatecompound.
`aChemicalshiftsareassignedtocarbonsoftheglutamineandglutamate(g1–g5),pyroglutamate(cycle),butyryl(b1–b4),propionyl(p1–p3)orphenyl(ortho-ar,meta-ar,para-arandipso-ar)
`
`26.98(1C,Cg3)
`Cg2),36.76(C,Cp3),31.39(1C,Cg4),31.06(1C,Cp2),
`2Cortho(cid:2)ar,Cmeta(cid:2)ar),125.90(1C,Cpara(cid:2)ar),51.59(1C,
`141.35(1C,Cipso(cid:2)ar),128.32,128.23(4C,
`173.59,173.45,171.55(3C,3C(O),Cg1,Cg5,Cp1),
`
`2H,CH2g3)
`(t,2H,CH2p2),1.92(m,2H,CH2g4),1.84–1.65(m,
`NH2),4.16(m,1H,CHg2),2.62(t,2H,CH2p3),2.26
`7.26(d,1H,NH),6.95(m,5H,Ph),6.68(bs,2H,
`
`Cb2),30.12(1C,Cg4),27.29(1C,Cb3),27.02(1C,Cb3)
`Cg2),51.86,51.55(2C,2C3),35.57,35.14(2C,Cb4,
`2Cortho(cid:2)ar,2Cmeta(cid:2)ar),126.00(1C,Cpara(cid:2)ar),52.53(1C,
`141.46(1C,Cipso(cid:2)ar),128.53,128.42(4C,
`173.34,172.83,172.49(3C,C(O),Cg1,Cg5,Cb1),
`
`2H,CH2b4),2.35–1.9(m,8H,2C2g3,2,2C2b3,2).
`HCg2),3.70(s,3H,CH3),3.60(s,3H,CH3),2.62(t,
`7.21(m,5H,Ph),6.32(d,1H,NH),4.60(m,1H,
`
`(PPGN)
`
`3-Phenylpropionylglutamine
`
`O
`
`H2
`
`C N
`
`CH2
`
`CH2
`
`O
`
`OH
`
`CHC
`
`HN
`
`129–130
`
`O
`
`N-(4-phenylbutyryl)glutamate
`
`Dimethyl
`
`O
`
`Me
`
`C O
`
`CH2
`
`CH2
`
`O
`
`OMe
`
`CHC
`
`HN
`
`38–39
`
`O
`
` N-(4-phenylbutyryl)pyroglutamate
`
`Methyl
`
`O
`
`O
`
`37.75,25.81,22.80,20.84(5C,3Cb,2Ccycle)
`Cpara(cid:2)ar),57.84(1C,Ccycle),52.36(1C,CH3),38.08,
`Cipso(cid:2)ar),128.36(4C,2Cortho(cid:2)ar,2Cmeta(cid:2)ar),125.95(1C,
`174.08,174.02,171.85(3C,3C(O)),141.07(1C,
`
`CH3),2.98–1.98(m,10H,2C2cycle,3C2b)
`7.23(m,5H,Ph),4.75(m,1H,CHcycle),3.78(s,3H,
`
`45–46
`
`COMe
`
`O
`
`N
`
`5
`
`6
`
`7
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`584
`
`B. Comte et al.
`
`described previously. Other compounds, listed in Table 1,
`were prepared as follows. [2H5]PB was prepared by alu-
`minum chloride-catalyzed condensation of ((cid:3)-butyrolactone
`with a twofold excess of [2H6]benzene.17 The yield was
`94% based on 0.125 mol of ((cid:3)-butyrolactone. The mass iso-
`topomer distribution of [2H5]PB was 67% M5, 27% M4, and
`5.4% M3 (the mass isotopomer distribution of [2H6]benzene
`used for the synthesis was 88%M6, 11% M5, and 0.8%
`M4). PBGN, [2H5]PBGN and 3-phenylpropionylglutamine
`(PPGN) were synthesized by conjugation of glutamine with
`4-phenylbutyryl chloride, 4-[2H5]phenylbutyryl chloride and
`3-phenylpropionyl chloride, respectively. Unlabeled stan-
`dards of glutamate conjugates (formed during the analytical
`processing of the glutamine conjugates), i.e. phenylacetyl-
`glutamate (PAG), 3-phenylpropionylglutamate (PPG) and
`4-phenylbutyrylglutamate (PBG), were prepared by reacting
`the corresponding acid chlorides with glutamate. Note that
`other procedures have been described for the syntheses of
`PBGN, PBG, and PPG.18 – 20 Except for commercial pheny-
`lacetyl chloride, the unlabeled and labeled acid chlorides
`used in the above syntheses were prepared by reacting the
`acids with freshly distilled dichloromethyl methyl ether and
`distilled under vacuum. The yields of [2H5]phenylbutyryl
`chloride and [2H5]phenylbutyrylglutamine synthesis were
`98% and 45%, respectively.
`For the synthesis of N-(4-phenylbutyryl)pyroglutamic
`acid and its methyl ester, procedures described for the
`phenylacetyl analogs were adapted.21 Pyroglutamic acid
`was converted to its ditrimethylsilyl derivative, which
`was
`reacted with 4-phenylbutyryl
`chloride. Removal
`of
`the
`carboxyl
`trimethylsilyl group yielded N-(4-
`phenylbutyryl)pyroglutamic acid. For the synthesis of
`the methyl ester of N-(4-phenylbutyryl)pyroglutamic acid,
`pyroglutamic acid was converted to its methyl ester, which
`was activated by NaH before reaction with 4-phenylbutyryl
`chloride.22 Dimethyl N-(4-phenylbutyryl)glutamate was
`prepared by reacting 4-phenylbutyryl
`chloride with
`commercial L-glutamate dimethyl ester hydrochloride.
`
`Sample preparation
`For the determination of the concentrations of the free
`acids (PA and PB)
`in blood, plasma samples (0.5 ml)
`were spiked with 0.17 μmol of [2H7]PA, [2H5]PB and 3-
`phenylpropionate (PP) before deproteinization with 25 μl
`of saturated sulfosalicylic acid. The slurries were saturated
`with NaCl, acidified with one drop of 6 M HCl and extracted
`three times with 5 ml of diethyl ether. For the assays in urine,
`0.2 ml samples were spiked with 0.68 μmol of [2H7]PA and
`0.16 μmol [2H5]PB, acidified to pH 3 with HCl, saturated with
`NaCl and extracted three times with 3 ml of diethyl ether. The
`combined extracts were dried with Na2SO4 and evaporated
`before reacting the residues with 70 μl of MtBSTFA or
`Methyl-8 at 60 °C for 20 min. PA and PB were analyzed
`as their TBDMS or methyl derivatives.
`For the assay of
`the glutamine conjugates (PAGN
`and PBGN) in plasma, samples (0.5 ml) were spiked with
`0.8 μmol of PPGN, alkalinized to pH 12 and incubated at
`75 °C for 4 h to convert the glutamine conjugates to the
`glutamate derivatives. After the addition of 25 μl of saturated
`
`sulfosalicylic acid and 50 μl of 6 M HCl, the slurries were
`saturated with NaCl and extracted three times with 5 ml of
`ethyl acetate. The extracts were pooled, dried with Na2SO4,
`evaporated and the residues were reacted with 70 μl of
`Methyl-8 and incubated at 60 °C for 20 min.
`For the assay of PAGN and PBGN in urine, we tested two
`different methods. In the first assay, samples (0.2 ml) were
`adjusted at pH 12 with 1 M NaOH, spiked with 0.5 μmol
`of [2H7]PAGN and 0.25 μmol of [2H5]PBGN and incubated
`at 75 °C for 4 h to convert PAGN and PBGN to their respective
`glutamate conjugates (PAG and PBG).12 Then, the samples
`were acidified with 50 μl of 6 M HCl, saturated with NaCl
`and extracted three times with 5 ml of ethyl acetate. As
`previously described,23 the extracts were dried with Na2SO4
`and reacted with 70 μl of MtBSTFA at 60 °C for 20 min and
`analyzed as their TBDMS derivatives. In the second assay,
`samples (0.2 ml) were spiked with 1 μmol of PPGN and
`treated as for the first assay, except that PAG and PBG were
`derivatized with Methyl-8.
`
`GC/MS methods
`All the metabolites were analyzed as their TBDMS or methyl
`derivatives on a Hewlett-Packard Model 5890 gas chro-
`matograph equipped with an HP-5 capillary column (30 mð
`0.2 mm i.d., 0.5 mm film thickness; Hewlett-Packard) and
`coupled to a Model 5989A mass-selective detector. Sam-
`ples (0.2–1 μl) were injected with a splitting ratio of 20–50 : 1.
`(cid:2)1) and the column
`The carrier gas was helium (1 ml min
`head pressure was 32 kPa. The injector port temperature
`was at 250 °C, transfer line at 305 °C, source temperature
`at 230 °C and quadrupole at 150 °C. For the analysis of
`the TBDMS derivatives, the column temperature program
`(cid:2)1 to
`was initial temperature 150 °C, increased at 10 °C min
`(cid:2)1 to
`240 °C, held for 1 min at 240 °C, increased at 30 °C min
`305 °C and held for 18 min at 305 °C. After automatic cali-
`bration, the mass spectrometer was operated in the electron
`ionization (EI) mode (70 eV). Appropriate ion sets were
`monitored with a dwell time of 25–45 ms per ion, at
`m/z 193/200 (PA/[2H7]PA), 221/226 (PB/[2H5]PB), 304/311
`(PAG/[2H7]PAG cyclic form), 436/443 (PAG/[2H7]PAG lin-
`ear form), 332/337 (PBG/[2H5]PBG cyclic form) and 464/469
`(PBG/[2H5]PBG linear form).
`For the analysis of the methylated derivatives, the column
`program was slightly modified to initial temperature 90 °C
`(cid:2)1 to 190 °C, then at
`held for 1 min, increased at 15 °C min
`(cid:2)1 to 225 °C, then at 25 °C min
`(cid:2)1 to 305 °C and held
`5 °C min
`for 6 min at 305 °C. The injector port was at 280 °C, transfer
`line at 305 °C, source temperature at 200 °C and quadrupole at
`100 °C. The mass spectrometer was operated in the positive
`chemical ionization mode (CI, ammonia, 133 eV) with the
`appropriate selected ions at m/z 168/175 (PA/[2H7]PA), 182
`(PP), 196/201 (PB/[2H5]PB), 279/311 (PAG cyclic and linear
`form, respectively), 293/325 (PPG cyclic and linear) and
`307/339 (PBG cyclic and linear forms). All samples were
`injected two or three times.
`
`Clinical investigation
`The protocol was reviewed and approved by the Institutional
`Review Board of Case Western Reserve University and Uni-
`versity Hospitals of Cleveland. All subjects were free from
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`Identification of phenylbutyrylglutamine
`
`585
`
`any chronic or acute illness. Women had a negative preg-
`nancy test and were not breastfeeding. Seven subjects (three
`men, four women; age 31.7 š 5.0 years; height 171.3 š 3.4 cm;
`weight 79.5 š 5.9 kg) received detailed information on the
`purpose of the investigation and signed an informed consent
`form. After an overnight fast, the subjects were admitted
`to the Clinical Research Center at 7.30 a.m. They remained
`fasting until completion of the study. An intravenous line
`(cid:2)1)
`was installed in the forearm with a saline infusion (20 ml h
`and a short blood sampling catheter was inserted in a super-
`ficial vein of the contra-lateral hand. The hand was placed
`in a heating box at 60 °C for sampling of arterialized venous
`blood. At 8.00 a.m, after baseline blood and urine sam-
`(cid:2)1 (5 g per 75 kg)
`pling, each subject ingested 0.36 mmol kg
`of Na-PB. This dose corresponds to 15–25% of the dose com-
`monly used in the treatment of patients with inborn errors
`(cid:2)1 per day). Water intake was
`of urea synthesis (0.2–0.4 g kg
`adjusted to induce a diuresis of at least 100 ml in 30 min.
`Heparinized blood (10 ml) and urine samples were collected
`at 30 min intervals for the first 3 h and then every hour
`until 8 h. Plasma and urine samples were quickly frozen and
`stored at (cid:2)80 °C until analysis.
`
`RESULTS
`
`Synthesis and assay of PBGN
`To test our hypothesis that PBGN is formed in human
`subjects who have ingested PB, we synthesized unlabeled
`and [2H5]PBGN (see above). PBGN is a white, crystalline
`solid, with a melting-point of 126–127 °C (103–104 °C for
`PAGN). The structure of PBGN and [2H5]PBGN were
`confirmed by NMR spectroscopy (1H and 13C, Table 1, rows 1
`and 2).
`When planning our analytical strategy, we took into
`account that PBGN would probably have characteristics
`similar to those of the previously known PAGN. The lat-
`ter can be isolated from biological fluids either (i) as
`such by ion-exchange chromatography or (ii) more con-
`veniently by solvent acid extraction following mild alka-
`line hydrolysis to PAG. This is why we also synthesized
`PBG (Table 1, row 3). In addition, since methyl deriva-
`tization of PAG yields the methyl derivative of cyclic
`N-(phenylacetyl)pyroglutamic acid, and also the open-
`chain dimethyl N-(phenylacetyl)glutamate, we synthesized
`(Table 1, rows 4–6) N-(4-phenylbutyryl)pyroglutamic acid,
`methyl N-(4-phenylbutyryl)pyroglutamate and dimethyl N-
`(4-phenylbutyryl)glutamate. Lastly, we synthesized PPGN
`(Table 1, row 7) to serve as an analytical standard in addition
`to [2H7]PAGN and [2H5]PBGN.
`When PBGN was reacted with MtBSTFA or Methyl-
`8, each reagent yielded one derivative which, by analogy
`with PAGN, was tentatively identified as the TBDMS and
`methyl derivative, respectively, of the cyclic compound N-
`(4-phenylbutyryl)pyroglutamic acid (Table 1, row 4; Table 2,
`row 1). The formation of the cyclic compound was confirmed
`by reacting PBG with the same reagents, leading to the
`formation of the same cyclic derivatives (Table 2, compare
`row 1 with the first line of row 3). However, reaction of
`PBG with MtBSTFA or Methyl-8 also yielded non-cyclic
`
`derivatives, i.e. a di-TBDMS, a tri-TBDMS and a dimethyl
`derivative (Table 2, row 3, lines 2 and 3). Similar cyclic and
`open derivatives were obtained starting from [2H5]PBGN
`and [2H5]PBG (the latter derived from mild hydrolysis of the
`former; see Table 2, rows 2 and 4).
`cyclic
`the
`To confirm further
`the
`identity of
`derivatives of PBGN and PBG, we reacted the N-(4-
`phenylbutyryl)pyroglutamic acid that we had synthesized
`(Table 1, row 4) with MtBSTFA and with Methyl-8 and
`obtained the same cyclic derivative (Table 2, compare row 5
`with rows 1 and 3). In addition, we prepared the methyl ester
`of N-(4-phenylbutyryl)pyroglutamic acid (Table 1, row 5),
`which after injection without further processing into the
`GC/MS system yielded the same spectrum as PBGN, PBG
`and N-(4-phenylbutyryl)pyroglutamic acid derivatized with
`Methyl-8 (Table 2, compare row 6 with rows 1, 3 and
`5). Lastly, to confirm the identity of the linear dimethyl
`derivative obtained by reacting PBG with Methyl-8, we
`prepared dimethyl N-(4-phenylbutyryl)glutamate (Table 1,
`row 6), which after injection without further processing into
`the GC/MS system yielded the same spectrum as PBG
`derivatized with Methyl-8 (Table 2, compare row 7 with
`line 2 of row 3 in the methyl derivative column).
`The combination of
`information obtained from the
`derivatives listed in Table 2, rows 1–11, confirms the identity
`of the ions used to assay the concentration of PBGN in
`biological fluids.
`Table 2 (rows 10 and 11) also presents the derivatives
`used to assay the concentration of phenylacetylglutamine
`using [2H7]PAGN as internal standard. Mild hydrolysis of
`the analyte and internal standard yielded the corresponding
`unlabeled and deuterated PAG.
`In addition to the use of deuterated internal standards,
`we used unlabeled PPGN to compute the concentrations of
`PAGN and PBGN. Mild hydrolysis of PPGN yielded PPG
`which, after reaction with MtBSTFA and with Methyl-8,
`yielded the derivatives listed in Table 2, row 13. Figure 1(A)
`shows the ion chromatogram of a mixture of cyclical
`methylated derivatives of PAG, PPG and PBG resulting
`from the hydrolysis of the corresponding standards of
`glutamine conjugates before derivatization with Methyl-8.
`Figure 1(B) and (C) show similar ion chromatograms of
`identical derivatives formed by treatment of the urine from a
`subject who had ingested phenylbutyrate. The urine sample
`was spiked with a standard of phenylpropionylglutamine.
`The calibration graphs were linear from 10 to 700 nmol for
`both derivatives (r2 D 0.99).
`PA and PB were assayed using deuterated analogs and
`PP as standards. The derivatives are listed in Table 2, rows
`14–18. The calibration graphs of PA and PB concentrations
`were linear in the range 0.4–500 nmol. Assays of methylated
`PA and PB in 84 samples of plasma yielded identical data
`when computed using the deuterated internal standards
`versus the PP standard. The correlation slopes were 1.06 and
`1.05 with r2 D 0.99 and 0.97 for PB and PA, respectively.
`
`Clinical investigation
`In seven normal adults, the baseline plasma concentrations
`of PA (0.09 š 0.07 (SE) mM) and PAGN (0.6 š 0.6 μM) were
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`586
`
`B. Comte et al.
`
`269/286
`
`262/279
`322/339
`290/307
`
`290/307
`
`327/344
`
`295/312
`
`322/339
`
`290/307
`
`295/312
`
`290/307
`
`m/z
`
`AmmoniaPCI
`
`268
`
`261
`321
`289
`
`289
`
`326
`
`294
`
`321
`
`289
`
`294
`
`289
`
`m/z
`
`12.9
`
`12.9
`18.7
`16.0
`
`16.0
`
`18.7
`
`16.0
`
`18.7
`
`16.0
`
`16.0
`
`16.0
`
`RT(min)
`
`EI
`
`Methylderivatives
`
`Cyclicmethyl
`
`Cyclicmethyl
`
`Cyclicmethyl
`
`LinearDimethyl
`
`Cyclicmethyl
`
`LinearDimethyl
`
`Cyclicmethyl
`
`Cyclicmethyl
`
`Cyclicmethyl
`
`311
`
`304
`
`332
`583
`469
`
`337
`578
`464
`
`332
`
`337
`
`332
`
`m/z
`
`12.1
`
`12.1
`
`13.8
`17.5
`17.0
`
`13.8
`17.5
`17.0
`
`13.8
`
`13.8
`
`13.8
`
`TBDMS
`Cyclic
`TBDMS
`Cyclic
`
`TBDMS
`Cyclic
`Tri-TBDMS
`Di-TBDMS
`TBDMS
`Cyclic
`Tri-TBDMS
`Di-TBDMS
`TBDMS
`Cyclic
`TBDMS
`Cyclic
`TBDMS
`Cyclic
`
`[2H7]Phenylacetylglutamine
`
`Phenylacetylglutamine
`DimethylN-(4-phenylbutyryl)glutamate
`MethylN-(4-phenylbutyryl)pyroglutamicacid
`
`N-(4-Phenylbutyryl)pyroglutamicacid
`
`3-[2H5]Phenylbutyrylglutamate
`
`4-Phenylbutyrylglutamate
`
`4-[2H5]Phenylbutyrylglutamine
`
`Phenylbutyrylglutamine
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`RT(min)
`
`EI
`
`TBDMSderivatives
`
`Compound
`
`Row
`
`Table2.GC/MSanalysesa
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`Identification of phenylbutyrylglutamine
`
`587
`
`initaliccharactersinrows6and7aremethylesterswhichdonotrequirederivatization.
`toderivatizationofthetwocarboxylsandoftheamidenitrogenlinkingglutamatewitheitherthephenylacetyl,phenylpropionylorphenylbutyrylgroup.Thesynthesizedcompoundslisted
`showstheretentiontimes(RT)andtheionsmonitored(m/z).InthecaseofammoniaPCI,them/zvaluesreportedcorrespondtotheMC1andMC18ions.Thetri-TBDMSderivativesrefer
`aAlltheTBDMSandmethylderivativeswererunon30mHP-5columns(0.2mmi.d.,0.5μmfilmthickness).ThemassspectrometerwasoperatedinEIorammoniaPCImode.Thetable
`
`165/182
`158/175
`151/168
`184/201
`179/196
`
`308/325
`
`276/293
`
`276/293
`
`301/318
`
`269/286
`
`294/311
`
`262/279
`
`164
`157
`150
`183
`178
`
`307
`
`275
`
`275
`
`300
`
`268
`
`293
`
`261
`
`5.41
`4.45
`4.51
`6.33
`6.37
`
`17.0
`
`14.5
`
`14.5
`
`15.0
`
`12.9
`
`15.0
`
`12.9
`
`LinearDimethyl
`
`Cyclicmethyl
`
`Cyclicmethyl
`
`LinearDimethyl
`
`Cyclicmethyl
`
`LinearDimethyl
`
`Cyclicmethyl
`
`207
`200
`193
`226
`221
`564
`450
`
`318
`
`318
`557
`443
`
`311
`550
`436
`
`304
`
`5.48
`4.25
`4.31
`6.45
`6.49
`16.8
`16.2
`
`13.1
`
`13.1
`16.0
`15.3
`
`12.1
`16.0
`15.3
`
`12.1
`
`Tri-TBDMS
`Di-TBDMS
`TBDMS
`Cyclic
`TBDMS
`Cyclic
`Tri-TBDMS
`Di-TBDMS
`TBDMS
`Cyclic
`Tri-TBDMS
`Di-TBDMS
`TBDMS
`Cyclic
`
`3-Phenylpropionate
`[2H7]Phenylacetate
`Phenylacetate
`4-[2H5]Phenylbutyrate
`4-Phenylbutyrate
`
`3-Phenylpropionylglutamate
`
`3-Phenylpropionylglutamine
`
`[2H7]Phenylacetylglutamate
`
`Phenylacetylglutamate
`
`18
`17
`16
`15
`14
`
`13
`
`12
`
`11
`
`10
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`Abundance × 104
`
`5.5
`5.0
`4.5
`4.0
`3.5
`3.0
`2.5
`2.0
`1.5
`1.0
`.5
`
`00
`
`A
`
`C
`
`PAG, m/z 279
`
`PBG, m/z 307
`
`PPG, m/z 293
`
`13.0
`
`13.5
`
`14.0
`
`14.5
`
`15.0
`
`15.5
`
`16.0
`
`16.5
`
`17.0
`
`17.5
`
`B
`
`012345678
`
`012345678
`
`Abundance × 106
`
`Abundance × 106
`
`588
`
`B. Comte et al.
`
`13.0
`
`13.5
`
`14.0
`
`14.5
`
`15.5
`15.0
`Time (min)
`
`16.0
`
`16.5
`
`17.0
`
`17.5
`
`Figure 1.
`(A) Ion chromatogram of a mixture of phenylacetylglutamate (PAG), phenylpropionylglutamate (PPG) and
`phenylbutyrylglutamate (PBG) resulting from the hydrolysis of the corresponding standards of glutamine conjugates before
`derivatization with Methyl-8. (B) and (C) similar ion chromatograms of derivatives formed by treatment of the urine from a subject
`who had ingested phenylbutyrate. The urine sample was spiked with a standard of phenylpropionylglutamine.
`
`PA
`PB
`PAGN
`PBGN
`
`100
`
`300
`200
`Time (min)
`
`400
`
`500
`
`1.8
`
`1.5
`
`1.2
`
`0.9
`
`mM
`
`0.6
`
`0.3
`
`0.0
`
`0
`
`Figure 2. Profile over 8 h of phenylacetate (PA), phenylbutyrate (PB), phenylacetylglutamine (PAGN) and phenylbutyrylglutamine
`(PBGN) concentrations in human plasma (mean š SE, n D 7) after an oral ingestion of 0.36 mmol kg(cid:2)1 of Na-PB.
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`Identification of phenylbutyrylglutamine
`
`589
`
`0
`
`30
`
`60
`
`240
`90 120 150 180
`Time (min)
`
`300
`
`360
`
`420
`
`480
`
`A
`
`B
`
`20
`10
`0
`
`80
`
`60
`
`40
`
`20
`
`mmol PA
`
`mmol PB
`
`0
`-60
`
`Figure 3. Profile over 8 h of the urinary excretion of phenylacetate (PA) and phenylbutyrate (PB) in the same patients as in Fig. 2.
`
`0
`
`30 60 90 120 150 180
`240
`Time (min)
`
`300
`
`360
`
`420
`
`480
`
`A
`
`B
`
`2.5
`
`2
`
`1.5
`
`1
`
`0.5
`
`0
`
`2.5
`
`2
`
`1.5
`
`1
`
`0.5
`
`mmole PAGN
`
`mmole PBGN
`
`0
`-60
`
`Figure 4. Profile over 8 h of the urinary excretion of phenylacetylglutamine (PAGN) and phenylbutyrylglutamine (PBGN) in the same
`patients as in Fig. 2.
`
`similar to previously published data.23 PB and PBGN were
`not detected in basal plasma. Figure 2 shows the profiles
`of PA, PB, PAGN and PBGN concentrations in plasma
`after an oral bolus of Na-PB. The concentrations of PB and
`PBGN peaked at 50–90 min and those of PA and PAGN at
`150–180 min. The ratios of the areas under the curve were
`2.3 š 0.4 for PB/PA and 2.3 š 0.7 for PBGN/PAGN.
`Figure 3 shows the profiles of the urinary excretion of
`PB and PA. Eight hours following the ingestion of PB,
`the cumulative excretions of PB (278 š 61 μmol) and PA
`(79 š 12 μmol) amounted to 0.97 š 0.23 and 0.26 š 0.06% of
`the dose, respectively.
`Figure 4 shows the profile of urinary excretion of PBGN
`and PAGN. The cumulative excretions of PB metabolites over
`8 h are presented in Table 3. The excretions of PBGN and
`PAGN amounted to 21.5 and 32.6% of the dose, respectively.
`
`By then, excretion of PBGN was terminated, whereas
`excretion of PAGN was still ongoing, albeit at a low rate.
`The total recovery of the ingested dose of phenylbutyrate
`as identified urinary compounds (PACPBCPAGNCPBGN)
`was 53.4 š 4.5% after 8 h.
`
`DISCUSSION
`Methodological considerations
`Isotope dilution mass spectrometric assays are often made
`difficult by the non-availability and/or the high cost of
`internal standards labeled with stable isotopes (2H or 13C). An
`alternate strategy is to use a chemical analog of the analyte,
`preferably commercially available or easy to synthesize at a
`low cost. We were concerned that few laboratories would
`consider synthesizing [2H7]PAGN and [2H5]PBGN as part of
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`J. Mass Spectrom. 2002; 37: 581–590
`
`

`
`590
`
`B. Comte et al.
`
`Table 3. Recovery of PB and its metabolites in
`human urine (n D 7) after the oral ingestion of
`0.36 mmol kg(cid:2)1 of Na-PB
`
`Metabolite
`
`PA
`PB
`PAGN
`PBGN
`Total
`
`Amount (mmol)
`0.079 š 0.012
`0.278 š 0.061
`9.63 š 0.74
`6.39 š 0.50
`16.33 š 1.32
`
`Percentage
`0.26 š 0.06
`0.97 š 0.23
`32.6 š 1.9
`21.5 š 2.4
`53.4 š 4.5
`
`a study of the metabolism of PB, which is why we conducted
`our assays using the deuterated species and the chemical
`analog PPGN as standards. Similarly, we assayed PA and
`PB with [2H7]PA and [2H5]PB and also with the analog PP.
`Since the two types of standardization procedures yielded
`identical concentrations, investigators can select either type
`of standardization of the assays.
`We also showed that the analytes studied can be assayed
`as either methyl or TBDMS derivatives. Although TBDMS
`derivatives are more stable than methyl esters, the latter
`are more sensitive. Hence a decision between derivatives
`should be made depending on the expected concentrations
`and instrument availability.
`
`Physiological considerations
`PA has been used for many years for the treatment of inborn
`errors of the urea cycle.1,24 – 26 Patients treated with PA excrete
`waste N as PAGN instead of urea. This conjugation occurs
`only in humans and primates. The conjugation of PA with
`glutamine represents by far the main fate of exogenous PA.
`Indeed, in rhesus monkeys infused intravenously with PA
`for 5 h, the cumulative excretion of PAGN amounted to
`about 95% of the dose of PA infused.12
`The formation of PAGN has been used in clinical
`investigations using radioactive and stable isotopic tracers
`to probe non-invasively the labeling pattern of citric acid
`cycle intermediates in the livers of normal and diabetic
`humans.11,13,14,27 When 14C- or 13C-labeled lactate or pyruvate
`are administered, the labeling pattern of urinary PAGN is
`identical to that of ˛-ketoglutarate in the liver12.
`In recent years, the foul-smelling PA was replaced by
`PB for the treatment of inborn errors of the urea cycle.25,26
`It was assumed that PB undergoes one cycle of ˇ-oxidation
`forming PA-CoA and PA, the former being the substrate
`that actually conjugates with glutamine. In addition, PB
`was used extensively as (i) a cytostatic which potentiate the
`effect of cytotoxic agents such as fluorouracil on malignant
`tumors, probably via inhibition of histone deacetylases,2,3
`and (ii) an experimental drug for the treatment of cystic
`fibrosis,5 peroxisomal biogenesis disorders28 and sickle-cell
`anemia.6
`Investigators at the US National Cancer Institute pointed
`out to us that, following ingestion of PB, the cumulative
`excretions of PB C PA C PAGN amounted to less that
`half of the ingested dose of PB (L. Grochow, personal
`
`communication, 1999). This is why we postulated the
`formation of PBGN, an analog of PAGN. The data from
`our clinical investigation clearly demonstrate that PBGN is
`formed and excreted by normal humans who have ingested
`PB. A substantial amount (21.5 š 2.4%) of the Na-PB load
`was converted to PBGN. However, the cumulative excretions
`of PA, PB, PAGN and PBGN still account for only about half
`of the amount of PB ingested. Therefore, part of the ingested
`PB is converted to metabolite(s) which have not yet been
`identified.
`
`Acknowledgements
`This work was supported by grants from the NIH and the Cleveland
`Mt Sinai Health Care Foundation.
`
`REFERENCES
`1. Brusilow SW. Pediatr. Res. 1991; 29: 147.
`2. Gore SD, Carducci MA. Expert Opin. Investig. Drugs 2000; 9: 2923.
`3. Gilbert J, Baker SD, Bowling MK, Grochow L,
`Figg WD,
`Zabelina Y, Donehower RC, Carducci MA. Clin. Cancer Res. 2001;
`7: 2292–2300.
`4. Bar-Ner M, Thibault A, Tsokos M, Magrath IT, Samid D. Clin.
`Cancer Res. 1999; 5: 1509.
`5. Rubenstein RC, Zeitlin PL. Am. J. Physiol. Cell Physiol. 2000; 278:
`C259.
`6. Dover GJ, Brusilow S, Charache S. Blood 1994; 84: 339.
`7. Ambrose AM, Sherwin CP. J. Biol. Chem. 1933; 101: 669.
`8. Moldave K, Meister A. J. Biol. Chem. 1957; 229: 463.
`9. Webster LL Jr, Lucas SV, Strong JM, Mieyal JJ. J. Biol. Chem. 1976;
`251: 3352.
`Ekberg K,
`Schumann WC,
`10. Landau BR, Chandramouli V,
`Kumaran K, Kalhan SC, Wahren J. Diabetologia 1995; 38: 831–838.
`11. Magnusson I, Schumann WC, Bartsch GE, Chandramouli V,
`Kumaran K, Wahren J, Landau BR. J. Biol. Chem. 1991; 266:
`6975–6984.
`12. Yang D, Previs SF, Fernandez CA, Dugelay S, Soloviev MV,
`Hazey JW, Agarwal KC, Levine WC, David F, Rinaldo P,
`Beylot M, Brunengraber H. Am. J. Physiol. 1996; 270: E882–E889.
`13. Cline GW, Rothman DL, Magnusson I, Katz LD, Shulman GI. J.
`Clin. Invest. 1994; 94: 2369.
`14. Jones JG, Solomon MA, Cole SM, Sherry AD, Malloy CR. Am. J.
`Physiol. Endocrinol. Metab. 2001; 281: E848.
`15. Piscitelli SC, Thibault A, Figg WD, Tompkins A, Headlee D,
`Lieberman R, Samid D, Myers CE. J. Clin. Pharmacol. 1995; 35:
`368–373.
`16. Williams CM, Porter AH, Greer M, Scott KN, Sweeley CC.
`Biochem. Med. 1969; 3: 164.
`17. Truce WE, Olson CE. J. Am. Chem. Soc. 1952; 74: 4721.
`18. Garyaev AP, Kononov OV, Dumpis MA, Piotrovskii LB. Khim.
`Farm. Zh. 1990; 24: 20.
`19. Gunther R, Bordusa F. Chem. Eur. J. 2000; 6: 463.
`20. Goldman VS, Laney JW, Slife CW. US Patent 5, 932, 198, 1999.
`21. Rigo B, Lespagnol C, Pauly M. J. Heterocycl. Chem. 1988; 25: 49.
`22. Rigo B, Lespagnol C, Pauly M. J. Heterocycl. Chem. 1988; 25: 63.
`23. Yang D, Beylot M, Agarwal KC, Soloviev MV, Brunengraber H.
`Anal. Biochem. 1993; 212: 277.
`24. Batshaw ML, Thomas GH, Brusilow SW. Pediatrics 1981; 68: 290.
`25. Berry GT, Steiner RD. J. Pediatr. 2001; 138: S56.
`26. Burlina AB, Ogier H, Korall H, Trefz FK. Mol. Genet. Metab. 2001;
`72: 351.
`27. Diraison F, Large V, Bru