Journal of Analytical Toxicology,Vol. 25, October2001
Hydrolysis of Conjugated Metabolites of
Buprenorphine. I. The Quantitative Enzymatic
Hydrolysis of Buprenorphine-3@D-Glucuronide in
Human Urine
Shixia Feng, Mahmoud A. ElSohly,and David T. Duckworth
EISohly Laboratories, Incorporated, 5 Industrial Park Drive, Oxford, Mississippi 38655
Abstract
Buprenorphine, which is a powerful analgesic, a substitution
drug for opioids widely used in Europe, and a promising new
drug currently undergoing clinical trials in the treatment of
opioid dependence in the U.S., is excreted in human urine
mainly as glucuronide conjugates. In gas chromatographic-mass
spectrometric analysis, the urine specimens must be first
hydrolyzed to release buprenorphine from its glucuronide
conjugates. Jn order to evaluate the existing hydrolysis methods
and to find the optimal hydrolysis conditions, buprenorphine3-13-D-glucuronide (B3G) was synthesized. Urine fortified
with synthetic B3G was hydrolyzed using acid, base, and
~-glucuronidases from different source species, including Helix
pomatia, Escherichia coli, and Patella vulgata. Glusulase| a
preparation containing both [3-glucuronidase (H. pomatia) and
sulfatase, was also tested. Whereas both acidic and basic
hydrolysis were ineffective, quantitative hydrolysis could be
achieved by using ~-glucuronidases under appropriate conditions.
However, we found that there was a marked difference in the
reactivity of these enzymes (E. coli > H. pomatia >> P. vulgata).
The optimal incubation conditions for enzymatic hydrolysis of
B3G were 2 h at 37~ for E. coli and 4 h at 60~ or 16 h at 37~
for H. pomatia. Using 1000 Fishman units of either of these two
enzymes, effective hydrolysis could be achieved even when the
B3G concentration was as high as 2000 ng/mL. Glusulase was
equally effective toward B3G if the fortified urine samples were
incubated with 25 pL of this enzyme for I h at 60~
Introduction
Buprenorphine is a synthetic thebaine derivative that has
both analgesic and opiate antagonist properties (1). Under the
trade names Buprex| in the U.S. and Temgesic| in European
countries, it has been widely prescribedas a pain-killersince the
early 1980s. Since 1996, under the trade name Subutex| high-
dose tablets (0.4-, 2-, and 8-mg tablets for sublingual use) have
been used in France as a maintenance drug for opioids as an
economic alternative to methadone (2). In the U.S., it is currently undergoing clinical trials for this specific indication,
and the results are very promising (3-5). Recently, fatal overdose and abuse of buprenorphine has been reported (6,7).
Buprenorphine is metabolized in humans to norbuprenorphine and to conjugated buprenorphine and norbuprenorphine. Cone et al. (8) reported that no free buprenorphine was
detected in urine after human subjects were administered
buprenorphine in various dosage forms and by different routes.
Another study (9) reported that concentrations of free
buprenorphine and norbuprenorphine in urine may be less
than 1 ng/mL after therapeutic administration, but can range
up to 20 ng/mL in abuse situations. Gas chromatographic-mass
spectrometric (GC-MS) analysis requires a hydrolysis step to
cleave buprenorphine and norbuprenorphine from their glucuronide conjugates prior to extraction. Recently,Vincent et al.
(10) reported the use of an enzymatic method commonly used
for morphine-3-13-D-glucuronide(M3G) to hydrolyze the urine
samples for buprenorphine analysis. In the method, urine samples were incubated with 13-glucuronidasefrom Helixpomatia
for 20 h at 37~ However,the extent of hydrolysisof buprenorphine conjugates was not known. It is well known that the
rates of enzymatic reactions are largely dependent on the substrates. Although buprenorphine and morphine share some
structural similarity in that they both have the morphinan
skeleton, buprenorphine is quite unique. It has 6-methoxy and
N-cyclopropylmethylgroups instead of 6-hydroxyand/V-methyl
groups in morphine, and it contains a 6,14-ethano bridge and
a 7-(2-hydroxy-3,3-dimethyl-2-butyl)group, both of which are
lacking in morphine. Therefore, it is questionable whether the
hydrolysis conditions (including both acidic and enzymatic hydrolysis) suitable for morphine glucuronide conjugates are also
suitable for buprenorphine glucuronide conjugates. The reported procedure (10) did not mention the use of any hydrolysis
Reproduction(photocopying)of editorialcontentof thisjournalis prohibitedwithoutpublisher'spermission.
589
Journal of Analytical Toxicology, Vol. 25, October 2001
control in the analysis. This is in contrast to the common practice of morphine analysis in which the procedures are validated by monitoring the extent of hydrolysis of a control
containing a known amount of M3G (11-13).
This study was designed to compare the effectivenessof commonly used hydrolysisconditions for opiates and to determine
the optimal hydrolysis conditions for B3G.
Urine samples
Negative urine (2 mL) was fortified with B3G equivalent to
100-2000 ng/mL of free buprenorphine. The calibrators at 50,
100, 150, and 2000 ng/mL were prepared by fortifyingnegative
urine with buprenorphine standard solutions. Buprenorphined4, the internal standard (40 IJL at 5 IJg/mL),was addedprior to
hydrolysis to give 100 ng/mL of buprenorphine-d4.
Hydrolysismethods
Experimental
Materials
[3-Glucuronidasesfrom Helb:pomatia (type H-l, catalog# G0751, Lot # 119H3380), Escherichia coli (type IX-A,catalog #
G-7396, Lot # 30K8612), and Patella vulgata (type L-l, catalog
# G-8132, Lot # 60K3780) were purchased from Sigma Chemical Company (St. Louis, MO) as dry powders, and enzyme solutions were prepared at 10,000 or 25,000 Fishman units/mL by
dissolvingeach in an appropriate buffer (TableI). Glusulasewas
obtained from DuPont Chemical Company as a solution that
contains approximately10,000 units/mL of sulfatase and 90,000
units/mL of [3-glucuronidase.All solvents were high-performance liquid chromatography (HPLC) grade and purchased
from Fisher Scientific (Pittsburgh, PA).Reference standard solutions of buprenorphine and buprenorphine-d4 were obtained
from Cerilliant (formerly Radian, Austin, TX) at 1001Jg/mL in
methanol and were diluted to working solutions of 5 1Jg/mL.
B3G, whose structure is shown in Figure 1, was synthesized
by EISohly Laboratories, Inc. (Oxford, MS) using buprenorphine and methyl(2,3,4-tri-O-acetyl-(z-D-glucopyranosylbromide)uronate as starting materials. The structure of the
synthetic material was characterized by spectral methods and
the chemical purity was greater than 99% by HPLC. Stock solutions of B3G at 137.5 tJg/mL (equivalent to 100 IJg/mL of
buprenorphine) and working solutions at 6.875 ]Jg/mL(equivalent to 5 tJg/mLof buprenorphine) were prepared in methanol.
Enzymatic hydrolysis. Each urine sample was aliquoted into
a 15-mL centrifuge tube, and an appropriate buffer was added.
Table I shows the type and volume of the buffer added for each
enzyme. Then, 1000 Fishman units of a [~-glucuronidase(100
1JL of 10,000 Fishman units/mL or 40 IJL of 25,000 Fishman
units/mL) were added. The samples were capped and incubated
in an oven for 1-20 h. The amount of enzyme,temperature, and
time varied in experiments to compare the effectivenessof the
hydrolysis conditions in order to determine the optimal conditions.
Acidic hydrolysis. To each sample was added 0.5 mL of concentrated HC1.The sample was heated. The temperature and
time varied in experiments. The sampleswere neutralized with
10N KOH followinghydrolysis.
Basic hydrolysis. To each sample was added 0.5 mL of 10N
KOH. The sample was incubated at 50~ for 0.5 h. The samples
were neutralized with concentrated HC1followinghydrolysis.
Extraction and derivatization
One milliliter of 40% phosphate buffer (pH 9) was added to
each hydrolyzedsample. The samples were gently shaken with
6 mL of CHClJisopropanol (9:1). The top aqueous layerwas discarded, and the organic layer was washed with 1 mL of deionized water. The organic layer was then transferred to a clean
10-mL tube and evaporated to - 1 mL at 50~ under nitrogen.
The samples were then transferred to GC vials, evaporated to
dryness, and derivatized with BSTFA containing 1% TMSC
Table I. Optimal pH and Buffers Used for Each
[3-Glucuronidase
EnzymeOrigin
pH
Buffer
Helix pomatia
Escherichia coli
Patella vulgata
Glusulase
5.0
6.8
3.8
5.2
1.0M Acetate buffer (1 mL)
0.1M Phosphate buffer (2 mL)
0.25M Phosphatebuffer (1 mL)
1.1M Acetate buffer (1 mL)
Ion 450.00 (449.70 to 450.70): P105-024.D
Ion 482.00 (481.70 to 482.70): P105-024.D
Ion 506.00 (505.70 to 506.70): P105-024.D
Ion 454.00 (453.70 to 454p0): P105-024.D
Ion 486.00 (485.70 to 486/}'0): P105-024.D
Ion 510.00 (509.70 to 510I~'0):P105-024.D
+oooo~
mmoo1
soooo~
4.sooo~
0
0
+I
39)00
25OOO
i
2oooo~
15OOO
H3)3
H
1
~ / / ~ " ~_,,)' ' 'CC(CH3)3
HO~C, . O ~ . . " - - - ' k
OH
HO.+..]--~-/.... O
u
OCH3
HO..../.~ ~-OH
2
Figure 1. Chemical structures of buprenorphine (1) and buprenorphine3@D-glucuronide (2).
590
1oo001J
5ooo!
O'
I
. . . . . .I . .
i
. . . .
r
.
.
.
.
.i .' .I .
.
i
. . . .
i . . . .
I
. . . .
~ . . . .
I . . . .
9.20 9.40 9.60 9.80 1@00 10.20 10.40 10.60 10.80 11.00 11,20
Time (min)
Figure 2. A typical GC-MS ion chromatogram of buprenorphine analysis.
I
Journal of Analytical Toxicology, Vol. 25, October 2001
(50 pL) at 70~ for 30 rain. After cooling to room temperature,
the samples were transferred to GC vial inserts, and the vials
were recapped. The TMS derivatives (1 ]aL) were analyzed by
GC-MS.
GC-MS analysis
Analysis was performed on a Hewlett-Packard (Palo Alto, CA)
5890 GC interfaced with a Hewlett-Packard 5970 mass selective
detector equipped with a 10-m x 0.18-ram i.d. DB-1 capillary
column (0.18-pro film thickness). The initial oven temperature
was held at 180~ for 0.5 min, and then increased to 265~ at
30~
with a final temperature hold of 9 rain. The ions
monitored were m/z 450, 482, and 506 for buprenorphine and
m/z 454, 486, and 510 for buprenorphine-d4. The retention
times for buprenorphine-d4 and buprenorphine were 10.27
rain and 10.23 min, respectively. A typical ion chromatogram is
shown in Figure 2. Concentrations of buprenorphine cleaved
from B3G by hydrolysis were calculated using calibration curves
of buprenorphine constructed by plotting drug/internal standard area ratios versus concentration ratios. The curves were
linear over the concentration range with r 2 > 0.999.
Results and Discussion
First, we tested the acid, base, and an enzymatic hydrolysis
procedure using [~-glucuronidase (H. pomatia) at 37~ for
20 h reported by Vincent et al. (10). The experiments were carried out in triplicate on the urine fortified with 100 ng/mL of
B3G. A control group that was not subjected to any hydrolysis
treatment was also included. As shown in Figure 3, the basic
treatment with 10N KOH at 50~ for 0.5 h and acidic treatment
with concentrated HC1 at room temperature for 16 h resulted
in no recovery of buprenorphine. At higher temperature (50~
acid is capable of cleaving the ether bond of B3G, but the rate
is very slow with only 2% of buprenorphine recovered after 2 h.
When the temperature was further increased to 70~ both internal standard (buprenorphine-d4) and buprenorphine suffered complete loss. This suggested the degradation of
buprenorphine had occurred with acid at this temperature.
One study (14) reported that buprenorphine and other 6,14endo-ethanotetrahydrooripavine compounds could undergo an
acid-catalyzed rearrangement, which involved the loss of a
molecule of methanol and formation of a tetrahydrofuran ring.
B3G is very likely to undergo such structural rearrangement
too. This is in contrast with morphine glucuronides which can
Iooi
.~ 95
i~
0
2000
80 9 A m -=25uon
t of enzyme (Fishman units)
Figure4. Effect of amount of I~-glucuronidase (H. pomatia) on the percent hydrolysis of B3G.
100
lOO
80
80
60
~ eo
r
40
~ 40
20
20
m
Base
Acid (1)
Acid (2)
m
i
NT
H.
pomatia
Treatment
Figure 3. Effectof different hydrolysis methods on B3G in human urine.
Acid (1) representssamplesthat were treatedwith 0.5 mL of concentrated
HCI at rt for 16 h. Acid (2) representssamplesthat were treatedwith 0.5
mL of concentrated HCI at 50~ for 2 h. NT representssamplesthat were
not subjected to any hydrolysis treatment.
4
6
8
10
T i m e (h)
12
lEo coil oH. pomatla AP.
14
16
18
vulgata[
Figure5. Effect of incubation time on the hydrolysis of B3G with three
different ~-glucuronidases from E. coli, H. pomatia, and P vulgata at
37~
591
Journal of Analytical Toxicology, Vol. 25, October 2001
be effectively hydrolyzed with concentrated HC1 at _> 100~
(13,15-17). As opposed to the acidic and basic hydrolysis, the
[3-glucuronidase treatment at 37~ for 20 h completely hydrolyzed B3G and gave quantitative recovery of buprenorphine.
These initial findings answered one of our original questions. That is, the acidic hydrolysis which is effective for morphine glucuronides is ineffective for B3G, and therefore, B3G
must be hydrolyzed enzymatically. In Vincent and co-workers'
procedure (10), the samples were hydrolyzed with 250 I~L of
I~-glucuronidase(H. pornatia) solution prepared from 15 mg of
enzyme powder dissolved in I mL of acetate buffer. The exact
10C
80
R
0
a.
9
20 -t
01
0
1
2
3
4
Time (h)
[oH. pomatia .P. vulgata]
Figure 6. Effect of incubation time on the hydrolysis of B3G with
~-glucuronidases from H. pomatiaand P. vulgataat 60~
100
g
8O
ffl
"~ 60
Re4o
g.
20
i
500
1000
2000
Concentration (ng/mL)
[mE. Coli IHelix pomatia E~Glusulase[
Figure 7. Effect of substrate concentration on the enzymatic hydrolysis.
The incubation time and temperature were 2 h and 37~ for E. coli, 4 h
and 60~ for H. pomatia,and 1 h and 60~ for Glusulase.
592
amount of enzyme in Fishman units was not clear. It is known
that enzymatic reactions sometimes can be problematic since
the enzyme activity varies because of the lot variation of the
preparation. Earlier studies (18-20) showed that enzyme concentration could influence the recovery of the desired analytes.
To determine roughly how much enzyme is needed to hydrolyze B3G, we added varying amounts off/. pomatia to 2 mL
of buffered urine fortifiedwith 100 ng/mL of B3G and incubated
samples for 20 h at 37~ The results (Figure 4) showed that the
average percent hydrolysis of B3G was higher for 1000-unit
(99.3%) and 2500-unit (99.5%) groups as compared with the
250-unit group (88.5%), and that 1000 Fishman units ofH. pomatia should be sufficient to effectively hydrolyze the urine
samples.
Besides/-/. pornatia, several other types of 13-glucuronidases
from source origins including E. coli and P. vulgata have been
studied for hydrolysisof morphine conjugates (11,13,16-18). All
three enzymes are reportedly able to quantitatively hydrolyze
M3G, although their rates of hydrolysis differ considerably. We
investigated the influence of source origins of l~-glucuronidases
on B3G hydrolysis.Figure 5 shows the percent hydrolysisof B3G
after a 16-h incubation at 37~ with [3-glucuronidases from
three different sources using 1000 Fishman units of each. The
experiments were carried out at the reported (21) optimal pH for
each enzyme (H. pomatia, 5.0; E. coli, 6.8;P. vulgata, 3.8). The
fastest rate was achieved by E. coli, which hydrolyzed ~ 95% of
B3G after incubating for only 2 h. For/-/. pomatia, hydrolysis
was complete after incubating for 16 h. ForP. vulgata, the rate
was much slower compared with E. coli and H. pomatia. Only
26% of B3G was hydrolyzed after a 16-h incubation.
One factor that influences the enzyme activityis temperature.
Combie et al. (18) reported that the incubation time for hydrolysis of morphine-3-glucuronide with P. vulgata could be reduced from 24 h to 3 h by increasing the temperature from
35~ to 65~ Romberg and Lee (17) reported the optimal temperature was 60~ for/-/, pomatia, and forE. coli, although the
enzyme showed highest activity at 50~ the difference in activity was very small between 37 and 50~ Therefore, we compared the influence on the percent hydrolysis of B3G by
H. pomatia and P. vulgata at 60~ Figure 6 shows that the time
required for complete hydrolysis with H. pomatia at 60~ was
only 4 h as compared to 16 h at 37~ With P. vulgata, although the rate at 60~ was five times the rate at 37~ the percent hydrolysis still only reached 35% after incubating for 4 h.
Therefore, the hydrolysis rate for B3G is much slower than
M3G with P. vulgata. These results further demonstrated the
substrate- and source-dependent nature of enzymatic hydrolysis
using ~-glucuronidases.
A recent study (7) reported that the concentrations of
buprenorphine in some postmortem urine samples could reach
over 1000 ng/mL. Because one of the objectives of this study
was to develop a practical hydrolysis procedure, we compared
the effectiveness of hydrolysis of B3G at three higher concentration levels (500, 1000, and 2000 ng/mL) with E. coli, H. pomatia, and Glusulase, a common alternative to H. pomatia. The
amount of each enzyme used was 1000 Fishman units for both
E. coli and H. pomatia and 25 lJL for Glusulase. Figure 7 shows
that the three enzymes were almost equally effective at all
Journal of Analytical Toxicology,Vol. 25, October 2001
levels. By using 25 I~L of Glusulase, quantitative hydrolysisof
B3G couldbe achieved after a 1-h incubation at 60~
In additionto buprenorphine glucuronide conjugates, there
are apparently significant levels of norbuprenorphine glucuronide conjugatespresent in human urine (7,8,10).The concentrations of the latter vary depending on the individual.
Studies are underway in this laboratory to evaluate the enzymatic hydrolysisof norbuprenorphine glucuronide conjugates
with [3-glucuronidases.
In conclusion,whereas acidicand basic hydrolysisof B3G are
ineffective,quantitative hydrolysis can be achieved enzymatically using[5-glucuronidasesfromE. coli or H. pomatia. Glusulase is equallyeffectiveunder appropriate conditions.
References
1. R.C. Heel, R.N. Brogden, T.M. Speight, and G.S. Avery. Buprenorphine: a review of its pharmacological properties and therapeutic
efficacy. Drugs 17:81-110 (1979).
2. P. Kopp, C. Rumeau-Pichon, and P.C. Le. The financial impact of
maintenance treatment in heroin addictive behavior: the case of
Subutex. Rev Epidemiol. Sante Publique 48:256-270 (2000}.
3. J.E. Henney. From the Food and Drug Administration: clarification
on buprenorphine. J. Am. Med. Assoc. 284:2178 (2000).
4. K.K. Downey, T.C. Helmus, and C.R. Schuster.Treatment of heroindependent poly-drug abusers with contingency management and
buprenorphine maintenance. Exp. Clin. Psychopharmacol. 8:
176-184 (2000).
5. R.S.Schottenfeld,J. Pakes, P. O'Connor, M. Chawarski, A. Oliveto,
and T.R. Kosten. Thrice-weekly versus daily buprenorphine maintenance. Biol. Psychiatry 47" 1072-1079 (2000).
6. G. Decocq, D. Fremaux, A. Small, M. Compagnon, and M. Andr~jak. Complications locales apr~s injection intraveineuse de
comprim~s dissous de bupr~norphine. Presse M~d. 26:1433
(1997).
7. A. Tracqui, P. Kintz, and 8. Ludes. Buprenorphine-related deaths
among drug addicts in France: a report on 20 fatalities. J. Anal
ToxicoL 23:270-279 (1998).
8. E.J.Cone, C.W. Gorodetzky, D. Yousefnejad, W.F. Buchwald, and
R.E.Johnson. The metabolism and excretion of buprenorphine in
humans. DrugMetab. Dispos. 12:577-581 (1984).
9. L. Debrabandere, M. Van Boven, and R Daenens. High-performance liquid chromatography with electrochemical detection of
buprenorphine and its major metabolite in urine. J. Chromatogr.
564:557-566 (1991).
10. F. Vincent, J. Bessard, J. Vacheron, M. Mallaret, and G. Bessard.
Determination of buprenorphine and norbuprenorphine in urine
and hair by gas chromatography-mass spectrometry. J. Anal. ToxicoL 23" 270-279 (1999).
11. M. Zezulak, J.J. Snyder, and S.B. Needleman. Simultaneous analysis of codeine, morphine, and heroin and ~-glucuronidase hydrolysis. J. Forensic 5cL 38:1275-1285 (1993).
12. W. Huang, W. Andollo, and W.L. Hearn. A solid phase extraction
technique for the isolation and identification of opiates in urine.
J. Anal ToxicoL 16" 307-310 (1992).
13. Z. Lin, P. Lafolie, and O. Beck. Evaluation of analytical procedures
for urinary codeine and morphine measurement. J. AnaL Toxicol.
18" 129-133 (1994).
14. E.J.Cone, C.W. Gorodetzky, W.D. Darwin, and W.E Buchward.
Stability of the 6,14-endo-ethanotetrahydrooripavine analgesics:
acid catalyzed rearrangement of buprenorphine. J. Pharm. ScL 73:
243-246 (1984).
15. F. Fish and T.S. Hayes. Hydrolysis of morphine glucuronide.
J. Forensic 5ci. 19:676-683 (1974).
16. T.A. Jennison, E. Wozniak, G. Nelson, and F.M. Urry. Quantitative
conversion of morphine-3-13-D-glucuronide to morphine using
~-glucuronididase obtained from Patella vulgata as compared to
acid hydrolysis. J. Anal. Toxicol. 17:208-210 (1993).
17. R.W. Romberg and L. Lee. Comparison of hydrolysis rates of morphine-3-glucuronide and morphine-6-glucuronide with acid and
~-glucuronidase. J. Anal. ToxicoL 19:157-162 (1995).
18. J. Combie, J.W. Blake, T.E. Nugent, and T. Tobin. Morphine glucuronide hydrolysis: superiority of 13-glucuronidasefrom Patella
vulgata. Clin. Chem. 28:83-86 (1982).
19. P.M. Kemp, I.K. Abukhalaf, J.E. Manno, B.R. Manno. D.D.
Alford, M.E. McWilliams, F.E.Nixon, M.J. Fitzgerald, R.R. Reeves,
and M.J. Wood. Cannabinoids in humans. II. The influence of
three methods of hydrolysis on the concentration of THC and
two metabolites in urine. ]. Anal ToxicoL 19:292-298 (1995).
20. M.A. ElSohly and S. Feng. Ag-THC metabolites in meconium:
Identification of 11-OH-Ag-THC, 8 ~, 11-di-OH-Ag-THC and 11-norAg-THC-9-COOH as major metabolites of delta-9-THC. J. Anal.
Toxicol. 22:329-335 (1998).
21. Biochemicals, Organic Compounds for Research and Diagnostic
Reagents. Sigma Chemical Co., St. Louis, MO, 1994, pp 485-487.
593