Journal of Analytical Toxicology,Vol.
20, July/August1996
An Enzymic Digestion and Solid-PhaseExtraction
Procedure for the Screeningfor Acidic, Neutral,
and Basic Drugs in Liver Using Gas Chromatography
for Analysis
Zeng-Ping Huang, Xiao-Hua Chen, JaapWijsbeek, Jan-PietFranke*, and RokusA. de Zeeuw
Department of Analytical Chemistry and Toxicology, Groningen Institute for Drug Studies, University Centre for Pharmacy,
A. Deusinglaan 1, 9713 AV Groningen, The Netherlands
Abstract
Analysisof liver specimensis an important issue in forensic
toxicology, but suitable workup and extraction methods for
general screening purposes have been lacking until now. A work-
up and extraction scheme based on a recently developed
procedure for the screeningof biological fluids was developed that
can be used for the screeningof acidic, neutral, and basic drugs in
liver. This method usesa single solid-phaseextraction (SPE)
column and gas chromatography-flame ionization detection
(GC-FID) for the final analysis. First, the homogenized liver
sample is sonicated and centrifuged; the resulting supernatant is
applied to the SPE column. Elution of acidic, neutral, and some
weakly basic drugs is then performed with acetone-chloroform
and analyzed by GC-FID. Next, the pellet of tissue material
obtained from the centrifugation is enzymically digested by
subtilisin Carlsberg. This frees the drugs bound to the liver tissue.
The resulting clear liquid is brought to the reconditioned SPE
column. A wash step is introduced to remove acidic and neutral
interferences and the basic drugs can then be eluted with
ammoniated ethyl acetate. Using 100-rag wet liver samples spiked
with 2 pg of amounts of various drugs, recoveries were 70-102%
with relative standard deviations less than 9%. The resulting
GC-FID chromatograms were virtually free of endogenous
interferences. GC-nitrogen-phosphorous detection detected
smaller amounts of nitrogen-containing drugs, again without
endogenous interferences. With the SPE columns currently used,
which contain a bed mass of 130 rag, the liver samples should be
smaller than 200 mg because the endogenous compounds obtained
after the digestion of the tissuewill overload the column, which
results in a lower recovery of the drugs of interest. Drugs that
decompose under the digestion conditions (pH 10.5 at 60~ for
1 h) may be lost in the present procedure. This phenomenon is
being investigated further.
Introduction
In forensic toxicology, blood and urine may be available only
in small amounts or not at all. Moreover, interpretation may be
~ 1o whom correspondence should be addressed
24'8 Reproduction(photocopying)of editorialcontentof thisjournalis prohibitedwithoutpublisher'spermission.
Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019
difficult because blood levels are frequently site-dependent,
with heart blood giving higher levels than femoral blood (1).
Therefore, analysis of tissue material is of interest. Before car-
rying out the analysis, the relevant drugs must be extracted
from tissue material. Extraction techniques must be rapid,
efficient, and give good recoveries for most of the relevant
drugs. In addition, they must provide clean extracts. Meeting
these requirements has proven to be a very difficult task for
forensic scientists.
Until now, the commonly used methods for extracting drugs
from tissue samples were direct solvent extraction or
liquid-liquid extraction after pretreatment of the tissue with a
protein precipitant or acid hydrolysis of the tissue. However, di-
rect extraction suffered from solvent emulsion and resulted in
dirty extracts; protein precipitation gave low yields of drugs; and
acid hydrolysis was not suitable for some acid or heat labile
drugs (2). In 1977, Osselton et al. (3-5) first reported on enzyme
digestion for releasing some acidic, neutral, and basic drugs
from liver tissue. Compared with other extraction methods,
the enzymic digestion yielded far higher recoveries, which in-
dicated that enzymic digestion may be the best pretreatrnent
method for general screening. Since then. some other enzymic
digestion methods have been published (6-8). Generally, the en-
zymic digestion is followed by liquid-liquid extraction.
Over the past decade, solid-phase extraction (SPE) has
emerged as a powerful tool for the extraction of biological
fluids for toxicological drug analysis (9-11). SPE of tissue sam-
ples has been reported as well (12-14). However, these methods
are less suitable for drug screening because they were estab-
lished for a single drug or a group of relevant drugs. The only
SPE method for drug screening of tissue has been reported by
Cordonnier et al. (15).
Recently, a single column SPE procedure was developed in
our laboratory for the screening for acidic, neutral, and basic
drugs in plasma, whole blood, and urine. It has been demon-
strated to be useful for drug screening (16-20). In this study,
we investigated whether the screening procedure for biological
fluids could be adapted for drug screening in liver tissue.
Journal of Analytical Toxicology, Vol. 20, July/August 1996

Materials and Methods

A B ~
Materials
All reagents were analytical-grade (Merck,
Darmstadt, Germany), with the exceptions 6
of ethyl acetate, acetone (Lab-Scan, Dublin,
Ireland), and chloroform (Westburg,
Leusden, The Netherlands), which were :. 4
HPLC-grade. The test drugs were from
commercial suppliers and were of pharma-
t "
l
ceutical quality. Stock solutions of the in-
dividual drugs (1 mg/mL) were prepared by

Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019

dissolving the appropriate amount of drug
in methanol-ethyl acetate (1:1, v/v). For v

spiking the samples, the stock solutions 5 10 15 20 0

were diluted 10 times with methanol-ethyl Time (min) Time (min)
acetate. The chromatographic standard so-
lution was prepared by diluting the stock Figure 1. GC-NPD chromatograms of Fraction A and Fraction B after SPEof liver tissue digested by
solution of prazepam with ethyl acetate to subtilisin and using the same procedure as Chen et al. (17). Peak 1 is from prazepam (chromatographic
standard).
20 IJg/mL.
Subtilisin Carlsberg (11 units/rag) was
purchased from Sigma Chemical (St. Louis,
MO), and papain (0.57 units/rag) was pur-
chased from Fluka (Buchs, Switzerland). Sample SPE Analysis
Bond-Elut Certify columns with a capacity
of 10 mL (130 mg sorbent) and the Vac Elut I I
vacuum manifold were supplied by Varian
Sample Preparation Products (Harbor City,
0.4 mL homogenizedliver I
spiked with 2 ~tganalyte~
centrifuge
I 1 I Columnconditioning
2 mL methanol
2 mL bufferpH 6.0
CA). Phosphate buffer (0.1 tool/L, pH 6.0) I I
was prepared by dissolving6.81 g potassium
dihydrogen phosphate in 450 mL deionized
[ Supernatant ~ "-[ Sample application

water, adjusting the pH with I mol/L potas- I

sium hydroxide, and making the total I Colunmwash: I mL water
0.5 mLbufferpH 4.0
volume up to 500 mL with deionized water.
Acetic acid (1 mol/mL,pH 2.4) was prepared Enzymicdigestion I I
by diluting 5.75 mL glacial acetic acid to of pelletat pH 10.5 [ Columndrying
50 I~Lmethanol,air
100 mL with deionizedwater. Acetate buffer
(0.1 tool/L, pH 4.0) was prepared by mixing I |

0.575 rnL glacial acetic acid with 1.6 mL ElutionFraction A [ EvaporationeluateA

4 mL ~.,etone-chloroform GC analysis
1 mol/L potassium hydroxide and 80 mL
deionized water. If necessary, the pH was
I
I
adjusted to 4.0, and the volume was made Columnconditioning I
up to 100 mL with deionized water. Tris- 2 mL bufferpH 6.0
base solution 1 mol/L was prepared by dis- I
Phosphoricacid to pH 6-7; I I
solving 12.1 g tris(hydroxymethyl)amino- centrifuge ]------1 Sampleapplication [
methane in 100 mL deionized water.
Subtilisin tris-base solution was prepared I
by dissolving 2 mg subtilisin Carlsberg in Columnwash: 1 mL water
1 mL IM HAc pH 2.4
4 mL 1 mol/L tris-base solution before use. 2 mL acetone-chloroform
Papain solution was prepared by dissolving
250 mg papain in 3 mL phosphate buffer I
(pH 7.4) and then adding 0.25 mL cysteine Columndrying ]
50 pL methanol,air
(0.1 tool/L) and 0.1 mL EDTA (0.1 rnol/L).
Ammoniated ethyl acetate (2%) was pre- I I

pared dailyby adding I mL ammonia (25%) Ehtion FractionB I I EvaporationeluateB

2 mL ethylacetate,NFI3 GC analysis
to 49 mL ethyl acetate in an Erlenmeyer I
flask with a glass stopper and sonicating for
Figure 2. Outline of the SPEprocedure for screening liver samples using GC for analysis.

249
 Journal of Analytical Toxicology, Vol. 20, July/August 1996

5 min. Acetone-chloroform was prepared by mixing equal vol-
umes of each. Phosphoric acid (10%) was prepared by mixing
7.35 mL phosphoric acid (85 wt%) with 92.65 mL deionized
water9
Instrumentation
A Sonicor sonic bath was purchased from 8onicor Instru-
ment Corporation (Farmingdale, N'Y). A Megafuge 1.0 cen-
trifuge, which was purchased from Heraeus Sepatech
(Osterode/Harz, Germany), was used at 6000 x g (4000 rpm).
An Ultra Turrax homogenizer was purchased from Wilten
(Utrecht, The Netherlands). Analyses were performed with a
HP-5890 series II GC (Hewlett-Packard,Avondale,PA) equipped
(FID), and a nitrogen-phosporous detector (NPD). The column
was a 30-m HP-1 fused-silica capillarycolumn with a 0.53-ram
internal diameter and a film thickness of 0.88 I~m.The oven
temperature program for FID was 2 rain at 80~ increased
20~ to 215~ increased 5~ to 285~ and held 5
min at the final temperature. For NPD, the program was 2 rain
at 150~ increased 20~ to 215~ increased 5~ to
285~ and held 5 min at the final temperature. The injector
and detector temperatures were 275 and 310~ respectively.
Spiking liver tissue
Calf liver (1 g) was finely minced and mixedwith 3 mL water
in an Ultra Turrax homogenizer. Of the homogenized liver,

Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019

with a HP-7673 automatic sampler, a flame-ionizationdetector 0.4 mL (equivalentto 100 mg of wet liver)was spikedwith 2 I~g
of each test drug (20 I~Lof 0.1 mg/mL test
A B drug solution). The spiked samples were
vortex mixed for 30 s and equilibrated
on a waterbath at 37~ for 1 h. Livers were
also spiked at I m g and 0.1 IJg/100 mg (see
Figure 1).

Sample pretreatment
One milliliter of deionized water was
added to 0.4 mL of spiked liver homogenate.
The mixture was sonicated for 5 min and
then centrifuged at 6000 x g for 10 rain.
C The supernatant was applied to the SPE
column at step 2 of the procedure below.
Subtilisin tris-base solution (1 mL) was
added to the pellet. The mixture was vortex
10 15 20 25 mixed for 10 s and digested at 60~ for 1 h.
'n' ~ Time (mJn) After cooling down to room temperature,
the pH was adjusted to 6-7 with 10%
Figure 3. GC-NPD chromatograms of Fraction A and Fraction B after SPE of blank tissue digested by
sublilisin Carlberg an(] using the same procedure as Chen et al. (17). Peak 1 is from prazepam (chro-
H3PO4. The suspension was centrifuged at
6000 x g for 10 rain. The supernatant was
applied to the SPE column at step 7 of the
procedure below.
A B 5~
SPE procedure for screening liver tissue
The extraction procedure is outlined in
Figure 2. The procedure consisted of 13
steps: 1. The column was preconditioned
I 6 with 2 mL of methanol and then with 2 mL
2
1 of phosphate buffer (pH 6.0). 2. The super-
9
":17 8 natant was applied onto the column after
sonication and passed through completely
at a flow rate of 1-1.5 mL/min. 3. The
column was washed sequentially with 1 mL
deionized water and 0.5 mL acetate buffer
(pH 4.0). 4. The column was dried under
full vacuum (15 in. Hg) for 4 rain. After
J rinsing the column with 50 IlL methanol, it
0 5 10 15 20 25 0 5 tO 15 20 25
was dried again for 1 rain. 5. A labeled evap-
Time (min) Time (min)
oration tube was placed into the manifold
Figure 4. GC-FID chromatograms of Fraction A and Fraction B after SPEof liver tissue according to basin after wiping the column outlet and
the scheme of Figure 2. The liver was spiked with 2 pg of various drugs. Peak identification: 1, ben- manifold basin with tissue. Then 4 mL of
zocaine; 2, secobarbital; 3, diazepam; 4, flunitrazepam; 5, methamphetamine; 6, methadone; 7, nor- acetone-chloroform (1:1) was added to the
triptytine; 8, codeine. Peak 9, prazepam, is the chromatographic standard. column reservoirand etuted completelyat a

250
Journal of Analytical Toxicology, Vol. 20, July/August 1996

flow rate of 0.8 mL/min. This eluate is Fraction A. 6. The
column was reconditioned with 2 mL of phosphate buffer (pH
6.0). 7. After digestion, the supernatant was applied to the
column and passed through at a flow rate of 1-1.5 mL/min. 8.
The column was washed sequentially with I mL of deionized
water, 1 mL 1 mol/L HAc (pH 2.4), and 2 mL acetone--chloro-
form (1:1). 9. The column was dried as in step 4. 10. After
wiping the column outlet and manifold basin with tissue, an-
other labeled evaporation tube was placed into the manifold
basin. Ammoniatedethyl acetate (2%, 2 mL) was added to the
column reservoir; the flow rate was 0.5 mL/min. This eluate is
1 calculated by summing up the percentage in both Fractions.
Fraction B. 11. To each tube, 100 IJL of chromatographic stan-
dard solution (prazepam, 20 IJg/mL) was added. The eluate
was then evaporated in a water bath at 37~ under a nitrogen
stream until about 100 I~Lof solvent remained in each tube. 12.
One microliter of the remaining solvent was injected onto the
GC. The ratio of the peak heights of the analyte to that of
prazepam was used for quantitation. 13. Absolute extraction
yields were determined by comparing the response ratios of
the extract with the calibration graph from the pure substance
in the concentration range of 4-24 IJg/mLwith a constant con-
centration of prazepam (20 pg/mL). For some weakly basic
drugs that appeared in both Fractions, the extraction yield was

Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019

m

,n

Results and Discussion

Homogenized liver tissue cannot be applied directly onto the

2 SPE columns because this will result in the clogging of the
column. Different pretreatment procedures of liver samples
t. 6 were carried out. Criteria for judging quality of the extracts
r
were cleanliness of the extracts and the extraction yields of the
drugs in the spiked samples. It must be emphasized that high
extraction yields of the drugs in the spiked samples does not
automatically mean that these yields will also be obtained in
m
authentic specimens; however,high extraction yields in spiked
sample are at least a prerequisite. For the calculation of abso-
lute extraction yields, a chromatographic standard (prazepam)
was added at the end of extraction procedure, just prior to the
injection step in order to correct for difference in the final
volume of the extracts and for difference in the chromato-
,i i
graphic process.
0 5 10 15 20 25
Time (rain)
Sonication and centrifugation
Sonication followed by centrifugation as the only pretreat-
Figure 5. GC-FID chromatograms of Fraction B with phosphate buffer in- ment of the homogenized liver sample gave low recoveries,
stead of tris-buffer used for the enzymic hydrolysis procedure of liver
especiallyfor basic drugs. Apparently,some basic drugs remain
tissue spiked with 2 pg amounts of various drugs. Peak identification:
bound to the particulate cell material and are subsequently
1, methamphetamine; 2, methadone; 3, nortriptyline; 4, codeine; 5,
diazepam. Peak 6, prazepam, is the chromatographic standard used for spun down in the pellet. However,acidic drugs such as barbi-
recovery calculations. turates showed good recoveries,usually between 80 and 100%.

Enzymicdigestion
120 - Allobarbital ~ Trlmlpramine Different types of enzymes were used to
Glutethimide --x-- l[etazolam digest the liver tissue before sample appli-
lO0- Meplvacaine cation. Shankar et al. (7,8) found that pa-
.d
pain gave the best result. However, diges-
'~ 80
tion with papain resulted in very dirty
~ 60- extracts, which caused substantial interfer-
o
ences in the subsequent GC analyses.
~ 40- Shankar et al. (7,8) used high-performance
~ 20- liquid chromatography (HPLC) with ultra-
violet (UV) detection at 254 nm for basic
drugs and 240 nm for barbiturates. The im-
100 200 300 400 500
purities detected by GC may not be detected
A m o u n t of l i v e r , r n g
under these conditions. In agreement with
Figure 6, Influence of the amount of liver in the digestion-extraction procedure on the extraction yield the literature (3-5), the protease subtilisin
of five drugs. The amount of drug per liver sample was kept constant at 2 lag per analyte. Carlsberg at a pH of 10.5 and a temperature

251
 Journal of Analytical Toxicology, Vol. 20, July/August 1996

of 55-60~ digested the tissue very efficiently.It must be noted
that during digestion the pH drops to 8-9. However,when ap-
plying the digested material to the SPE procedure used for
blood and urine (17), the first fraction, Fraction A, which con-
tained the acidic and neutral drugs, gave chromatograms with
many interfering peaks when analyzed with GC-FID (Figure
3A). These interferences appear to be a result of ]ipophilic
compounds such as fatty acids. When using GC-NPD, cleaner
chromatograms were obtained (Figure 1A),but they were still
not considered to be useful analyses. In contrast, Fraction B
was much cleaner and contained only a small number of im-
purity peaks (Figures 1B and 3B). Because the lipoid interfer-
ences are of acidic to neutral character, they were washed away
Therefore, pH values between 7.5 and 10.5 were investigated.
However, at pH values less than 10.5, the digestion did not
result in clear solutions. The remaining particulate material
clogged the columns, and centrifugation to remove the
particulate material gave low recoveries of some drugs in the
supernatants. Thus, a digestion pH of 10.5 appeared to be
necessary.
Influence of the type of buffer in the enzymic digestion step
The use of a tris buffer pH ]0.5, which is usually recom-
mended for subtilisin digestion, resulted in a few impurities in
Fraction B (Figure 4B, retention time, 5-7 min). Replacingthe
tris buffer with a 1 mol/L [(2HPO4solution with the pH ad-

Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019

by the acetone-chloroform eluent used to obtain Fraction A.
The interferences seen in Fraction B may then be due to com-
pounds with a basic character. It must also be noted that
GC-MS analysis of Fraction A is very complicated, and more-
over, the many interferences present may easily contaminate
the instrument.
Combination of sonication/centrifugation and enzymic
digestion with subtilisin Carlsberg
The two tissue pretreatment procedures outlined previously
were combined, which resulted in an extraction procedure as
described in the Method section and schematicallypresented in
Figure 2. Typical chromatograms of spiked samples that were
analysed by GC-FID are depicted in Figure 4. In order to get
optimal results, the following criteria were investigated in
more detail.
The pH of the enzymic digestion
According to the manufacturer, Novo, the optimal pH for
subtilisin digestion is 9-11 (22). Other authors stated that
lower pH values can be used (6-8). At pH 10.5, one can expect
that some drugs, such as esters like atropine, may decompose.
justed to 10.5 with 1 mol/L KOH, resulted in a much cleaner
chromatogram for Fraction B (Figure 5) without affecting the
efficiencyof the digestion itself.
Influence of the amount of liver tissue
Extractions were carried out according to the procedure
outlined in Figure 2 with liver portions of 100-500 mg and a
fixed analyte concentration. Figure 6 shows that with 100 mg
liver tissue, recoveries of approximately 70-100% were found
for all drugs investigated.With higher amounts of tissue the ex-
traction yields dropped substantially, to as low as 20%, except
for mepivacaine, which remained rather constant (between
70-80%). Moreover,with higher amounts of liver matrix, more
interfering peaks were seen in the chromatograms. Another
phenomenon was observed: ketazolam usually eluted in Frac-
tion B when 100 mg liver was applied, but it appeared in Frac-
tion A when 200 mg of tissue or more was used. This indicates
Table II. Extraction Yields of Twenty Drugs Extracted
from Liver
Yield
RSD*
Drug Fraction A FractionB Total (%)

Table I. Postmortem Concentrations of Drugs in Liver AIIobarbital* 80.3 -* 80.3 1.8

Benzocaine 45.5 - 45.5 3.6
Postmortemliver Codeine - 87.3 87.3 6.2
concentrations(pg/g) Number of Diazepam 49.4 35.3 84.7 3.9
range mean fatalities Doxepin - 94.7 94.7 2.5
Flunitrazepam 68,8 14.6 83.4 7.4
Codeine 0.1-45 5 22 Glutehimide* 84.2 - 84.2 2.5
Diazepam 11-13 12 2 Heptabarbital 101.8 - 101.0 7.6
Doxepin 22-38 32 4 Ketazolam t 51.5 18.5 70.0 7.5
Clutethimide 63 141 89 11 Lidocaine - 86.6 86.6 1.4
Lidocaine 20-96 62 3 Mepivacaine* - 91.0 91.0 1.4
Mepivacaine 75 1 Meprobamate 100.2 - 100.2 2.6
Meprobamate 58-360 148 16 Methamphetamine - 74.6 74.6 4.8
Methamphetamine 0.4-175 44 4 Methadone - 86.5 86.5 5.7
Methadone 0.05-50 6 23 Methaqualone 75.1 8.5 83.6 4.3
Methaqualone 26-89 55 6 Nortriptyline - 87.5 87.5 8.4
Nortriptyline 8-220 90 9 Pentobarbita[ 79,7 - 79.7 3.5
Pentobarbital 23-550 130 55 Promethazine - 52.1 52.1 5.0
Promethazine 23-31 27 2 Seconbarbital 86.4 - 86.4 5.1
Secobarbital 25-605 271 4 Trimipramine t - 81.4 81.4 3.3
Trimipramine 224-544 384 2
* Relative standard deviation; n = 4 unless otherwise indicated.
* Data from Moffat (21). These concentrations are associated with suspected drug fn=3
overdose. * Not detected.

252
Journal of Analytical Toxicology, Vol, 20, July/August 1996
that higher amounts of tissue material on the SPE column
cover the binding sites of the functional groups in the Bond-
Elut Certifysorbent, thus preventing the latter from adequately
retaining the drugs of interest. This hypothesis would also ex-
plain the constant recovery of mepivacaine in Figure 6 if one
assumes that the affinity of mepivacaine to the binding sites of
the SPE column is larger than that of the matrix components.
Therefore, only 100 mg of liver tissue was used in subsequent
experiments. In general, with this amount of liver, detection of
1 IJg of drug in 100 mg of wet tissue is possible with GC-FID,
which should be quite sufficient in forensic toxicology because
drug concentrations in liver often exceed 10 IJg/g in fatal cases
(Table I). If desired, lower concentrations can be detected with
GC-NPD or GC-MS.
Recoverieswith the optimized sonication-centrifugation-
digestionmethod
Utilizing the procedure of Figure 2 and optimized as de-
scribed above, recoveries for 20 drugs were determined using
GC-FID for the final analysis at a concentration level of 20 IJg/g
liver (i.e., 100 mg liver with 2 IJg drug). The results are pre-
sented in Table II. With the exception of benzocaine and
promethazine, extraction yields of 70-100% were obtained.
The precision expressed as relative standard deviation (RSD)
was below 8.4%. Care should be taken for some weakly basic
drugs (e.g., benzodiazepines and methaqualone) with PKa
values around 3 that appear in both Fractions. Part of these
drugs may be washed away by acetone-chloroform before
eluting Fraction B. Benzocaine is a weakly basic substance
(PKa, 2.8). Moreover,benzocaine has an ester functional group.
We would expect benzocaine to appear in both Fractions (Frac-
tions A and B). The absence of benzocaine in Fraction B may be
a result of decomposition after enzymic digestion or of being
washed away during the washing step with acetone-chloroform
prior to the elution of Fraction B.
Promethazine is a phenothiazine drug that may decompose
under alkaline conditions, especially under the influence of
light. However, the recoveries still seem to be sufficient to de-
tect the presence of these types of drugs.
Conclusion
The presented work-up and extraction procedure for the
systematic analysis of liver tissue fulfilled the two major criteria
of providing sufficiently high recoveries of a wide selection of
toxicologically relevant drugs and giving GC-FID chro-
matograms with minimal interference from endogenous tissue
compounds. This is substantiated by the following observa-
tions: 1. The binding affinity of acidic and neutral drugs to liver
tissue is relatively weak. Consequently, these drugs can be lib-
erated and extracted by SPE after sonication centrifugation of
the liver homogenate. Coextraction of potentially interfering
endogenous compounds is minimal under these conditions. 2.
The binding affinity of basic drugs to liver tissue is much
stronger, which makes the digestion of the tissue material
necessary to liberate these drugs and available for SPE.
Although digestion also results in the freeing of a great many
endogenous compounds, the acidic or neutral character of
these compounds makes it possible to remove them by
applying an acidic wash to the SPE column. The basic drugs
can then be eluted in the next step by using ammoniated ethyl
acetate, and this basic fraction is virtually free of endogenous
interferences. 3. The combination of sonication and enzymic
digestion was found to be an effective technique for pretreat-
ment of liver. The optimal pH of subtilisin digestion is 10.5 at
a temperature of 55-60~
A phosphate buffer gave cleaner chromatograms for Fraction
B than the tris buffer. BondElut Certify columns can be used
for the extraction of a wide range of drugs in liver. For columns
Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019
with 130 mg sorbent, the amount of liver should be smaller
than 200 mg in order to obtain acceptable and reproducible re-
coveries.
It should be noted that the present work was carried out with
spiked liver samples that were left to equilibriate for at ]east
1 h at 37~ prior to preparation. Further work on liver samples
from case work is in progress. We are also investigating the fate
of ester drugs that may decompose during the enzymic diges-
tion.
Acknowledgment
The authors acknowledge Varian Sample Preparation Prod-
ucts (Harbor City, CA) for the extraction columns.
References
1. R.W. Prouty and W.H. Anderson. The forensic science implication
of site and temporal influences on postmortem drug concentra-
tions. J. Forensic Sci. 35:243-70 (1990).
2. L.J. Dusci and L.P. Hachett. A comparison of the borate-celite
column screening technique with other extraction methods in
forensic toxicology. J. Forensic Sci. 22:545-49 (1977).
3. M.D. Osselton, M.D. Hammond, and P.J.Twitchett. The extrac-
tion and analysis of benzodiazepines in tissues by enzymic di-
gestion and high-performance liquid chromatography. J. Pharrn.
Pharmacol. 29:460-62 (1977).
4. M.D. Osselton. The release of basic drugs by the enzymic diges-
tion of tissues in cases of poisoning. J. Forensic Sci. Soc. 17:
189-94 (1977).
5. M.D. Osselton, I.C. Shaw, and H.M. Stevens. Enzymic digestion
of liver tissue to release barbiturates, salicylic acid and other
acidic compounds in cases of human poisoning. Analyst 103:
1160-64 (1978).
6. M. Holzbecher and H.A. Ellenberger. Simultaneous determination
of diphenhydramine, methaqualone, diazepam and chlorpro-
mazine in liver by use of enzyme digestion: a comparison of di-
gestion procedures. J. Anal. Toxicol. 5:62-64 (1981 ).
7. V. Shankar, C. Damodaran, and P.C. Sekharan. Comparative eval-
uation of some enzymic digestion procedures in the release of
basic drugs from tissue. J. Anal. Toxicol. 11" 164-67 (1987).
8. V. Shankar, C. Damodaran, and P.C. Sekharan. Barbiturate an-
alysis in tissue by enzymic digestion and high-performance liquid
chromatography, Forensic Sci. Int. 40:45-55 (1989)
9. M.R. Harkey and M.L. Stolowitz. In Advances in Analytical Tox-
icology, Vol. 1., R.C. Baselt, Ed. Biomedical Publications, Foster
253

City, California, 1984, pp. 255-70.
10. B.K. Logan, D.T. Stafford, I.R. Tebbett, and C.M. Moore. Rapid
screening for 100 basic drugs and metabolites in urine using
cation exchange solid-phase extraction and high-performance
liquid chromatography with diode array detection. J. Anal. Tox-
icol. 14:154-59 (1990).
11. O. Suzuki, H. Seno, and T. Kumazawa. Rapid isolation of ben-
zodiazepines with Sep-Pak C18 cartridges. J. Forensic Sci. 33"
1249-53 (1988).
12. J. Scheurer and C.M. Moore. Solid-phase extraction of drugs from
biological tissues--a review. J. Anal. Toxicol. 16:264-69 (1992).
13. J. Balkon, B. Donnelly, and D. Prendes. A rapid isolation tech-
nique for drugs from tissues and fluids: use of Du Pont Prep 1
system. J. Forensic Sci. 27:23-31 (1982).
14. G. de Groot, B.C.A. Tepas, and G. Storm. High-performance
liquid chromatographic determination of doxorubicin in tissues
after solid phase extraction. J. Pharm. Biomed. Anal. 6:927-32
(1988).
15. J. Cordonnier, M. Van den Heede, and A. Heyndickx. A rapid iso-
lation of acidic and basic drugs from human tissues by coupled
Chem ElutTM extraction columns. Int. Anal. 3" 28-34 (1987).
16. X.-H. Chen, J. Wijsbeek, J.-P. Franke, and R.A. de Zeeuw. A single
column procedure on Bond-Elut Certify for systematic toxico-
254
Journal of Analytical Toxicology,Vol. 20, July/August1996
logical analysis of drugs in plasma and urine. J. Forensic 5ci. 37:
61-71 (1992).
17. X.-H. Chen, J.-P. Franke, J. Wijsbeek, and R.A. de Zeeuw. Isola-
tion of acidic, neutral, and basic drugs from whole blood using a
single mixed-mode solid-phase extraction column. J. Anal. Tox-
icol. 16:351-55 (1992).
18. X.-H. Chen. Mixed-mode solid-phase extraction for the screening
of drugs in systematic toxicological analysis. Ph.D. Thesis. Uni-
versity of Groningen, Groningen, The Netherlands. 1993.
19. P.G.M. Zweipfenning, A.H.C.M. Wilderink, P. Horsthuis,
J.-ff Franke, and R.A. de Zeeuw. Toxicological analysis of whole
blood samples by means of Bond-Elut Certify columns and gas
chromatography with nitrogen-phosphorus detection. J. Chro-
matogr. 674:87-95 (1994).
20. X.-H Chen, J.-P. Franke, and R.A. de Zeeuw. Solid-phase extrac-
Downloaded from https://academic.oup.com/jat/article-abstract/20/4/248/838531 by guest on 24 September 2019
tion for systematic toxicological analysis---an overview. Forensic
$ci. Rev. 4:147-59 (1992).
21. A.C. Moffat, Ed. Clarke's Isolation and Identification of Drugs. 2nd
ed., The Pharmaceutical Press, London, U.K., (1986)
22. Novo Industry AIS, Enzymes Division, Application Note, (1979)
Manuscript received June 9, 1995;
revision received October 25, 1995.