Accepted Manuscript Title: PRESSURIZED LIQUID EXTRACTION FOR THE DETERMINATION OF CANNABINOIDS AND METABOLITES IN HAIR: DETECTION OF CUT-OFF VALUES BY HIGH PERFORMANCE LIQUID CHROMATOGRAPHY-HIGH RESOLUTION TANDEM MASS SPECTROMETRY Author: Camilla Montesano Maria Chiara Simeoni Gabriele Vannutelli Adolfo Gregori Luigi Ripani Manuel Sergi Dario Compagnone Roberta Curini PII: DOI: Reference:
S0021-9673(15)00868-7 http://dx.doi.org/doi:10.1016/j.chroma.2015.06.021 CHROMA 356579
To appear in:
Journal of Chromatography A
Received date: Revised date: Accepted date:
4-2-2015 20-5-2015 10-6-2015
Please cite this article as: C. Montesano, M.C. Simeoni, G. Vannutelli, A. Gregori, L. Ripani, M. Sergi, D. Compagnone, R. Curini, PRESSURIZED LIQUID EXTRACTION FOR THE DETERMINATION OF CANNABINOIDS AND METABOLITES IN HAIR: DETECTION OF CUT-OFF VALUES BY HIGH PERFORMANCE LIQUID CHROMATOGRAPHY-HIGH RESOLUTION TANDEM MASS SPECTROMETRY, Journal of Chromatography A (2015), http://dx.doi.org/10.1016/j.chroma.2015.06.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Highlights
2
1) Fast method for the determination of cannabinoids and metabolites in hair.
4
2) Extraction from hair using pressurized liquid extraction with water modified with SDS.
5
3) Detection of 11-nor-9-carboxy- THC in hair below cut-off value by HPLC-HRMS/MS.
6
4) Validation following SWGTOX guidelines for confirmatory analysis.
Ac ce p
te
d
M
an
us
cr
7
ip t
3
Page 1 of 33
7 8 9 10
PRESSURIZED LIQUID EXTRACTION FOR THE DETERMINATION OF CANNABINOIDS AND METABOLITES IN HAIR: DETECTION OF CUT-OFF VALUES BY HIGH PERFORMANCE LIQUID CHROMATOGRAPHY-HIGH RESOLUTION TANDEM MASS SPECTROMETRY
11 Camilla Montesanoa, Maria Chiara Simeonib, Gabriele Vannutellia, Adolfo Gregoric, Luigi Ripanic, Manuel
13
Sergib*, Dario Compagnoneb, Roberta Curinia
14
a
Sapienza University of Rome, Department of Chemistry, 00185, Rome, Italy
15
b
Carabinieri, Department of Scientific Investigation (RIS), 00191 Rome, Italy
16
c
17
Mosciano Sant'Angelo (TE), Italy
us
an
University of Teramo, Faculty of Bioscience and Technology for Food, Agriculture and Environment, 64023
18
Corresponding author’s contacts: [email protected] Tel.: +39 0861266912; fax: +39 0861266915
M
19
cr
ip t
12
d
20
Abstract
22
Hair analysis has become a routine procedure in most forensic laboratories since this alternative
23
matrix presents clear advantages over classical matrices; particularly wider time window, non-
24
invasive sampling and good stability of the analytes over time. There are, however, some major
25
challenges for the analysis of cannabinoids in hair, mainly related to the low concentrations of 11-
26
nor-9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH), that is the major metabolite. In this study a
27
fast, accurate and sensitive method for the determination of cannabinol, cannabidiol, THC and
28
THC-COOH in hair has been developed. The extraction of analytes from hair (50 mg) is based on
29
an automated pressurized liquid extraction (PLE) using water modified with the surfactant sodium
30
dodecyl sulphate as eluent phase. PLE extract is then cleaned up by SPE using polymeric reversed
31
phase cartridges Strata XL before the injection in the HPLC-HRMS/MS system. Chromatographic
32
conditions obtained with a fused-core column allowed a good separation of the analytes in less than
33
4 min. The whole procedure has been validated according to SWGTOX guidelines. The LLOQs
34
obtained for THC-COOH and the other analytes were respectively 0.1 and 2 pg/mg. To the best of
35
our knowledge, this is the first LC-MS/MS based method that allows the detection of THC-COOH
36
in hair at values lower than the cut-off.
Ac ce p
te
21
Page 2 of 33
Keywords: Cannabinoids; 11-nor-9-carboxy-THC; Hair; Pressurized liquid extraction; Validation;
38
LC-HRMS/MS
Ac ce p
te
d
M
an
us
cr
ip t
37
Page 3 of 33
39 40
1. Introduction Hair analysis has become a routine procedure in most forensic laboratories since this alternative
42
matrix presents clear advantages over classical biological matrices; particularly wider time window,
43
non-invasive sampling and
44
straightforward. Major challenges arise from i) low concentration of Δ9-tetrahydrocannabinol
45
(THC) and even lower amounts of the main metabolite 11-nor-9-carboxy-THC (THC-COOH;
46
expected concentration in the fg/mg range) [2-4] ; ii) the necessity to develop procedures taking
47
into account only real use excluding environmental contamination as cannabis smoke can
48
condensate on the hair surface and can be easily adsorbed, [5] ; iii) the time consuming steps needed
49
for extraction and derivatization: in fact, gas chromatography coupled with tandem mass
50
spectrometry GC-MS(/MS) is still the technique of choice for the analysis of cannabinoids in hair.
51
The determination of THC-COOH has been shown to be crucial to distinguish between passive drug
52
exposure and active consumption since this molecule is an exclusive product of metabolism and can
53
be considered as marker of drug abuse.
54
Other cannabinoids, i.e. cannabinol (CBN) and cannabidiol (CBD) are often quantified as well,
55
since it has been reported that the sum of the major cannabinoids may provide a better indication of
56
drug use than THC alone [6] ; however, their presence alone cannot fully exclude passive exposure.
57
Recently, in order to identify external contamination, the detection of a specific marker, i.e. THCA-
58
A has been proposed; this non-psychoactive precursor of THC is the main cannabinoid in fresh
59
plant material and it is not significantly incorporated into hair after cannabis intake [7] .
60
To date, the techniques generally used to reach the cut-off of 0.2 pg/mg recommended by the
61
Society of Hair Testing (SoHT) [8] for THC-COOH are GC-MS/MS using either electron impact
62
(EI) ionization mode [9, 10] or negative ion chemical ionization (GC-NCI-MS-MS) that allows a
63
further increase of the sensitivity [2-4] and GC/GC-MS [11] . Liquid chromatography coupled with
ip t
41
Ac ce p
te
d
M
an
us
cr
stability [1] . However, cannabinoids analysis in hair is still not
Page 4 of 33
64
mass spectrometry (LC-MS/MS) in cannabinoids hair analysis has only been recently reported [12-
65
14] , and THC-COOH was included in
66
significantly higher than the cut-off value and thus the method can be applied only for chronic use
67
studies [15] ; in the other paper the LOQ reported, obtained by using a MS3 experiment, was exactly
68
at the cut-off [16] .
69
The most used approach for sample preparation is based on digestion in alkaline medium using
70
NaOH followed by liquid/liquid extraction (LLE) [2-4, 9, 13, 15-19] , solid phase extraction (SPE)
71
[10, 11, 20] , solid phase micro extraction (SPME) [21-25] or solid phase dynamic extraction
72
(SPDE) [26, 27] . Enzymatic digestion [28] and direct methanol extraction [5, 7, 29] have been also
73
reported but require longer times, up to 5 hours. A drawback of NaOH digestion is that the stability
74
of the analytes might be affected during the digestion procedure, for example CBD is not stable
75
under the severe conditions of alkaline digestion [19] . Large volume of organic solvents are,
76
instead, used for LLE procedure and the reproducibility is poor.
77
Pressurized liquid extraction (PLE) has reached an increasing attention in the last years since it
78
showed significant advantages over competing techniques as time saving, solvent use, automation
79
and efficiency [30] . Initially the technique has been mainly focused on environmental samples [31]
80
. Other applications included food samples [32] and biological matrices such as tissues [33] ; we
81
have recently demonstrated the potential of PLE for the extraction of illicit drugs from hair,
82
allowing automation of the extraction and a significant reduction of total analysis time [34] .
83
The aim of the present work was to develop a fast and accurate method for the determination of
84
CBN, CBD, THC and THC-COOH in hair. The extraction of analytes from hair is based on an
85
automated PLE using water modified with the surfactant sodium dodecyl sulphate as eluent phase.
86
The PLE eluent is then cleaned-up by SPE that allows both the reduction of matrix effect and the
87
enrichment of the analytes which is particularly useful for the detection of THC-COOH. The
88
chromatographic conditions obtained with a fused-core column allowed a good separation of the
Ac ce p
te
d
M
an
us
cr
ip t
few studies. The LOQ obtained in one work was
Page 5 of 33
89
analytes in less than 5 min. The whole procedure has been validated according to SWGTOX
90
guidelines. To the best of our knowledge, this is the first LC-MS/MS based method that allows the
91
detection of THC-COOH in hair at lower values than the cut-off.
Ac ce p
te
d
M
an
us
cr
ip t
92
Page 6 of 33
92 93
2. Experimental
94
2.1.
ip t
Standard and reagents
THC, THC-COOH, CBD, CBN drugs standards and THC-d3 and THC-COOH-d3 internal standards
96
were purchased from LGC standard (Sesto San Giovanni, Milan, Italy). The purity of the reference
97
compounds was ≥99%. All standards were provided at a concentration of 1 mg mL-1 with the
98
exception of ISs that were at a concentration of 0.1 mg mL-1. Individual stock solutions were
99
prepared in methanol at 0.1 mg mL-1, except for the ISs which were diluted at 1 μg mL-1. Working
100
standard mixtures were prepared by appropriate dilution of the standards in methanol. All solutions
101
were stored at -20 °C in the dark. Ultrapure water, formic acid, acetonitrile and methanol used to
102
prepare the mobile phases were of HPLC grade and were purchased from Fisher Scientifc (Fair
103
Lawn, NJ, USA). Sodium dodecyl sulphate (SDS) and sodium hydroxide were from Sigma-Aldrich.
104
Ultrapure water used for sample preparation was produced by a Milli-Q Plus apparatus from
105
Millipore (Bedford, MA, USA).
us
an
M
d
te
Ac ce p
106
cr
95
2.2.
External decontamination procedure
107
Crudely cut hair (≈1 cm) was placed in a 50 ml Falcon cone tube in 5 mL of phosphate buffer (0.1
108
M, pH 6). The mixture was vortexed for one minute and, after removal of aqueous buffer, it was
109
sequentially washed with isopropanol (5 mL) and dichloromethane (5 mL). The last wash was
110
collected in a vial and evaporated under a gentle stream of N2; the residue was reconstituted with 1
111
mL of 10 mM formic acid in methanol and was stored for further analysis in order to assess the
112
performance of the procedure.
113
2.3.
Pressurized liquid extraction
Page 7 of 33
Fifty milligrams of hair sample, cut into 1-2 mm segments, were homogenized with diatomaceous
115
earth (Sigma Aldrich,Milan, Italy) by means of a mortar. The sorbent was previously powdered and
116
washed with the same PLE extraction conditions used for hair sample. The mixture was then placed
117
in a 1 mL pressure resistant stainless steel cell that was sealed at both ends with glass-fiber filters.
118
Void volumes in the cell were filled up with diatomaceous earth and 25 µL of methanol containing
119
the ISs at a concentration of 2 ng mL-1 for THC-COOH-d3 and 20 ng mL-1 for THC-d3 were added.
120
PLE was carried out performing a single extraction cycle using as extraction solvent a 90:10 (v/v)
121
water-methanol mixture containing SDS 25 mM. The extraction conditions were: pressure, 100 bar;
122
temperature, 150 °C; preheat time, 1 min; heat time, 7 min; static time, 5 min; flush volume, 0%;
123
purge time, 60 s. The PLE extract (5-6 mL) was automatically collected in glass vial with caps
124
solvent resistant (PTFE) septa.
cr
an
us
2.4.
Solid phase extraction
M
125
ip t
114
PLE extract was collected into a graduated tube and centrifuged at 6000 g for 5 min at 25°C. Five
127
mL were then cleaned up by SPE using polymeric reversed phase cartridges Strata XL 30 mg/1 mL
128
from Phenomenex (Torrance, CA, USA). The cartridge, installed on vacuum system (Visiprep), was
129
previously conditioned with 1 mL of methanol and 1 mL of water-methanol (90:10 v/v). After
130
loading, the cartridge was washed with 5 mL of water-methanol (90:10 v/v) and 3 mL of water-
131
methanol (50:50 v/v); the analytes were then eluted using 1 mL of methanol. The eluate was
132
directly injected in the HPLC-HRMS/MS system (6 μL).
te
Ac ce p
133
d
126
2.5.
NaOH digestion
134
For comparative purposes three authentic positive hair samples were pretreated with alkaline
135
digestion as well. The digestion was carried out according with an existing method described in
136
literature with slight modifications[15] . Briefly, 50 mg of decontaminated sample were transferred
137
into an amber glass vial and 25 μL of ISs solution was added. The samples were then subjected to
Page 8 of 33
digestion in 1 mL of 2.5 M NaOH at 60 °C for 25 min; after cooling to room temperature, the
139
solution was neutralised with formic acid. The mixture obtained was extracted by vortex mixing
140
using 3 mL ethyl acetate. After centrifugation the organic supernatant was separated, evaporated to
141
dryness under a gentle flow of nitrogen and the residue was finally reconstituted in 200 μL of
142
methanol. 10 μL were then injected in the HPLC-HRMS/MS system. 2.6.
HPLC-HRMS/MS analysis
cr
143
ip t
138
HPLC-HRMS/MS analysis was performed on a Dionex UltiMate® 3000 Rapid Separation LC
145
system from Thermo Fisher Scientific (San Jose, CA, USA) coupled to a Thermo Scientific Q
146
Exactive Mass spectrometer (Thermo Fisher Scientific, Bremen, Germany).
147
The Q Exactive mass spectrometer was equipped with a heated electrospray ionization source (H-
148
ESI) and operated in the positive ionization mode. The spray voltage was 4 kV, capillary
149
temperature 375°C, heater temperature 450°C, S-lens RF level 60, sheath gas flow rate 60, auxiliary
150
gas flow rate 20 (arbitrary units). Nitrogen was used for spray stabilization, for collision–induced
151
dissociation experiments in the higher energy collision dissociation (HCD) cell and as the damping
152
gas in the C-trap. The instrument was calibrated in the positive and negative mode at the beginning
153
of each working day. The mass spectrometer operated in Targeted-MS/MS mode at a resolution of
154
35,000 (full width at half maximum at m/z 400). Automatic Gain Control (AGC) was set at 2x105
155
and maximum injection time at 100 ms. Using this scan mode, precursor ions are selected in the
156
quadrupole with a 0.4 m/z window and subsequently fragmented in the HCD cell. A full scan of all
157
fragmented ions originating from the precursor ion is performed, and two specific product ions used
158
for data analysis with a mass tolerance of 5 ppm (Table 1). Targeted-MS/MS experiments were
159
segmented in three windows (one for each compound, excluding THC and CBD which were
160
detected in the same window) in order to have the maximum sensitivity from the detector.
Ac ce p
te
d
M
an
us
144
Page 9 of 33
The analytes were separated using a Kinetex C18-XB column (10 cm x 2.1 mm ID) from
162
Phenomenex packed with core-shell particles of 2.6 μm. A Phenomenex security Guard Ultra
163
Cartridge (packed with C18 particles) was also used to protect the column from damaging
164
contaminants and microparticulate. The mobile phases were: (A) acetonitrile and (B) water both
165
containing 0.1% formic acid. The flow rate was 0.5 mL min-1 entirely driven into the ion source.
166
The following gradient elution scheme was used: phase A was increased from the initial 60% to
167
75% in 2.4 min, then up to 90% in 1.2 min and in the following 0.4 min brought to 100%. The latter
168
was maintained for 1.2 min and then switched back to the initial 60% in 2.3 min. The complete
169
separation of all substances occurred in 4 min. All source and instrument parameters were tuned by
170
injecting a standard solution of THC-COOH at a concentration of 0.1 ng µL-1 at 10 µL min-1 by a
171
syringe pump in flow injection analysis with the same chromatographic conditions (flow and
172
solvent composition). Peak areas for the selected ions were determined using Thermo Xcalibur
173
QUAN Browser and quantitation was performed by the internal standard method. The selected
174
transitions, together with the main HPLC-HRMS/MS parameters, are reported in Table 1.
cr
us
an
M
d
te
2.7.
Validation
Ac ce p
175
ip t
161
176
The method was validated according to SoHT
177
modifications. Linearity, recovery, matrix effect, precision, accuracy, limits of detection (LODs)
178
and lower limits of quantification (LLOQs) were evaluated.
179
[8] and SWGTOX guidelines [35] with few
2.7.1. Quantification and identification
180
Calibration standards were prepared in methanol. Quality control samples (QC) were prepared by
181
spiking the standard solution on the hair in the PLE cell; in addition four positive hair samples were
182
used for precision and recovery studies. Two product ions were selected for each analyte, one for
183
quantitation and one for qualitative analyte confirmation. Identification criteria included retention
184
time (tr) within ±0.2 min of average calibration standard (tr), presence of two product ions, and ion
Page 10 of 33
ratio between the quantifying and qualifier ion within ±20 % of that established by the calibration
186
standards.
187
Linearity was investigated in the concentration range from LLOQ to 50 pg/mg for THC-COOH and
188
to 750 pg/mg for the other analytes, using six calibration standards for each group. Replicates (n=5)
189
at each concentration level were analyzed during three days. The calibration curves were derived by
190
plotting the peak area of analytes to ISs versus the hair concentration using a least squares
191
regression model.
192
cr
ip t
185
us
2.7.2. Extraction recovery, absolute recovery and matrix effect
The extraction recoveries (RE) and absolute recoveries were determined at two concentration levels
194
(0.1 and 50 pg/mg for THC-COOH; 2 and 500 pg/mg for the other analytes); 5 hair samples of
195
different types (blond, black, straight and curled) were used. For each concentration and hair type,
196
RE were obtained by the ratio of the absolute peak area for each analyte spiked on the hair before
197
extraction (A) and the absolute peak area of the analyte for blank samples spiked after the extraction
198
step (B). Accordingly, RE(%) = A/B*100. Absolute recoveries were calculated by comparing A
199
with the absolute peak area for each analyte in a pure standard solution at the same concentration
200
(C).
201
Matrix effects (ME) were estimated as ME(%) = 100 × ((B/C)-1) in five different hair matrices at
202
three concentration levels (0.1, 10, 50 pg/mg for THC-COOH and 2, 250, 750 pg/mg for the other
203
analytes). A ME value < 0 indicated ionization suppression, and a value > 0 ionization
204
enhancement.
M
d
te
Ac ce p
205
an
193
2.7.3. Selectivity and carry-over
206
Interferences from endogenous hair components were evaluated by the analysis of 10 different
207
cannabinoids-free hair samples collected among employees and some family members . If analytes
208
were not detected (
S0021-9673(15)00868-7 http://dx.doi.org/doi:10.1016/j.chroma.2015.06.021 CHROMA 356579
To appear in:
Journal of Chromatography A
Received date: Revised date: Accepted date:
4-2-2015 20-5-2015 10-6-2015
Please cite this article as: C. Montesano, M.C. Simeoni, G. Vannutelli, A. Gregori, L. Ripani, M. Sergi, D. Compagnone, R. Curini, PRESSURIZED LIQUID EXTRACTION FOR THE DETERMINATION OF CANNABINOIDS AND METABOLITES IN HAIR: DETECTION OF CUT-OFF VALUES BY HIGH PERFORMANCE LIQUID CHROMATOGRAPHY-HIGH RESOLUTION TANDEM MASS SPECTROMETRY, Journal of Chromatography A (2015), http://dx.doi.org/10.1016/j.chroma.2015.06.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Highlights
2
1) Fast method for the determination of cannabinoids and metabolites in hair.
4
2) Extraction from hair using pressurized liquid extraction with water modified with SDS.
5
3) Detection of 11-nor-9-carboxy- THC in hair below cut-off value by HPLC-HRMS/MS.
6
4) Validation following SWGTOX guidelines for confirmatory analysis.
Ac ce p
te
d
M
an
us
cr
7
ip t
3
Page 1 of 33
7 8 9 10
PRESSURIZED LIQUID EXTRACTION FOR THE DETERMINATION OF CANNABINOIDS AND METABOLITES IN HAIR: DETECTION OF CUT-OFF VALUES BY HIGH PERFORMANCE LIQUID CHROMATOGRAPHY-HIGH RESOLUTION TANDEM MASS SPECTROMETRY
11 Camilla Montesanoa, Maria Chiara Simeonib, Gabriele Vannutellia, Adolfo Gregoric, Luigi Ripanic, Manuel
13
Sergib*, Dario Compagnoneb, Roberta Curinia
14
a
Sapienza University of Rome, Department of Chemistry, 00185, Rome, Italy
15
b
Carabinieri, Department of Scientific Investigation (RIS), 00191 Rome, Italy
16
c
17
Mosciano Sant'Angelo (TE), Italy
us
an
University of Teramo, Faculty of Bioscience and Technology for Food, Agriculture and Environment, 64023
18
Corresponding author’s contacts: [email protected] Tel.: +39 0861266912; fax: +39 0861266915
M
19
cr
ip t
12
d
20
Abstract
22
Hair analysis has become a routine procedure in most forensic laboratories since this alternative
23
matrix presents clear advantages over classical matrices; particularly wider time window, non-
24
invasive sampling and good stability of the analytes over time. There are, however, some major
25
challenges for the analysis of cannabinoids in hair, mainly related to the low concentrations of 11-
26
nor-9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH), that is the major metabolite. In this study a
27
fast, accurate and sensitive method for the determination of cannabinol, cannabidiol, THC and
28
THC-COOH in hair has been developed. The extraction of analytes from hair (50 mg) is based on
29
an automated pressurized liquid extraction (PLE) using water modified with the surfactant sodium
30
dodecyl sulphate as eluent phase. PLE extract is then cleaned up by SPE using polymeric reversed
31
phase cartridges Strata XL before the injection in the HPLC-HRMS/MS system. Chromatographic
32
conditions obtained with a fused-core column allowed a good separation of the analytes in less than
33
4 min. The whole procedure has been validated according to SWGTOX guidelines. The LLOQs
34
obtained for THC-COOH and the other analytes were respectively 0.1 and 2 pg/mg. To the best of
35
our knowledge, this is the first LC-MS/MS based method that allows the detection of THC-COOH
36
in hair at values lower than the cut-off.
Ac ce p
te
21
Page 2 of 33
Keywords: Cannabinoids; 11-nor-9-carboxy-THC; Hair; Pressurized liquid extraction; Validation;
38
LC-HRMS/MS
Ac ce p
te
d
M
an
us
cr
ip t
37
Page 3 of 33
39 40
1. Introduction Hair analysis has become a routine procedure in most forensic laboratories since this alternative
42
matrix presents clear advantages over classical biological matrices; particularly wider time window,
43
non-invasive sampling and
44
straightforward. Major challenges arise from i) low concentration of Δ9-tetrahydrocannabinol
45
(THC) and even lower amounts of the main metabolite 11-nor-9-carboxy-THC (THC-COOH;
46
expected concentration in the fg/mg range) [2-4] ; ii) the necessity to develop procedures taking
47
into account only real use excluding environmental contamination as cannabis smoke can
48
condensate on the hair surface and can be easily adsorbed, [5] ; iii) the time consuming steps needed
49
for extraction and derivatization: in fact, gas chromatography coupled with tandem mass
50
spectrometry GC-MS(/MS) is still the technique of choice for the analysis of cannabinoids in hair.
51
The determination of THC-COOH has been shown to be crucial to distinguish between passive drug
52
exposure and active consumption since this molecule is an exclusive product of metabolism and can
53
be considered as marker of drug abuse.
54
Other cannabinoids, i.e. cannabinol (CBN) and cannabidiol (CBD) are often quantified as well,
55
since it has been reported that the sum of the major cannabinoids may provide a better indication of
56
drug use than THC alone [6] ; however, their presence alone cannot fully exclude passive exposure.
57
Recently, in order to identify external contamination, the detection of a specific marker, i.e. THCA-
58
A has been proposed; this non-psychoactive precursor of THC is the main cannabinoid in fresh
59
plant material and it is not significantly incorporated into hair after cannabis intake [7] .
60
To date, the techniques generally used to reach the cut-off of 0.2 pg/mg recommended by the
61
Society of Hair Testing (SoHT) [8] for THC-COOH are GC-MS/MS using either electron impact
62
(EI) ionization mode [9, 10] or negative ion chemical ionization (GC-NCI-MS-MS) that allows a
63
further increase of the sensitivity [2-4] and GC/GC-MS [11] . Liquid chromatography coupled with
ip t
41
Ac ce p
te
d
M
an
us
cr
stability [1] . However, cannabinoids analysis in hair is still not
Page 4 of 33
64
mass spectrometry (LC-MS/MS) in cannabinoids hair analysis has only been recently reported [12-
65
14] , and THC-COOH was included in
66
significantly higher than the cut-off value and thus the method can be applied only for chronic use
67
studies [15] ; in the other paper the LOQ reported, obtained by using a MS3 experiment, was exactly
68
at the cut-off [16] .
69
The most used approach for sample preparation is based on digestion in alkaline medium using
70
NaOH followed by liquid/liquid extraction (LLE) [2-4, 9, 13, 15-19] , solid phase extraction (SPE)
71
[10, 11, 20] , solid phase micro extraction (SPME) [21-25] or solid phase dynamic extraction
72
(SPDE) [26, 27] . Enzymatic digestion [28] and direct methanol extraction [5, 7, 29] have been also
73
reported but require longer times, up to 5 hours. A drawback of NaOH digestion is that the stability
74
of the analytes might be affected during the digestion procedure, for example CBD is not stable
75
under the severe conditions of alkaline digestion [19] . Large volume of organic solvents are,
76
instead, used for LLE procedure and the reproducibility is poor.
77
Pressurized liquid extraction (PLE) has reached an increasing attention in the last years since it
78
showed significant advantages over competing techniques as time saving, solvent use, automation
79
and efficiency [30] . Initially the technique has been mainly focused on environmental samples [31]
80
. Other applications included food samples [32] and biological matrices such as tissues [33] ; we
81
have recently demonstrated the potential of PLE for the extraction of illicit drugs from hair,
82
allowing automation of the extraction and a significant reduction of total analysis time [34] .
83
The aim of the present work was to develop a fast and accurate method for the determination of
84
CBN, CBD, THC and THC-COOH in hair. The extraction of analytes from hair is based on an
85
automated PLE using water modified with the surfactant sodium dodecyl sulphate as eluent phase.
86
The PLE eluent is then cleaned-up by SPE that allows both the reduction of matrix effect and the
87
enrichment of the analytes which is particularly useful for the detection of THC-COOH. The
88
chromatographic conditions obtained with a fused-core column allowed a good separation of the
Ac ce p
te
d
M
an
us
cr
ip t
few studies. The LOQ obtained in one work was
Page 5 of 33
89
analytes in less than 5 min. The whole procedure has been validated according to SWGTOX
90
guidelines. To the best of our knowledge, this is the first LC-MS/MS based method that allows the
91
detection of THC-COOH in hair at lower values than the cut-off.
Ac ce p
te
d
M
an
us
cr
ip t
92
Page 6 of 33
92 93
2. Experimental
94
2.1.
ip t
Standard and reagents
THC, THC-COOH, CBD, CBN drugs standards and THC-d3 and THC-COOH-d3 internal standards
96
were purchased from LGC standard (Sesto San Giovanni, Milan, Italy). The purity of the reference
97
compounds was ≥99%. All standards were provided at a concentration of 1 mg mL-1 with the
98
exception of ISs that were at a concentration of 0.1 mg mL-1. Individual stock solutions were
99
prepared in methanol at 0.1 mg mL-1, except for the ISs which were diluted at 1 μg mL-1. Working
100
standard mixtures were prepared by appropriate dilution of the standards in methanol. All solutions
101
were stored at -20 °C in the dark. Ultrapure water, formic acid, acetonitrile and methanol used to
102
prepare the mobile phases were of HPLC grade and were purchased from Fisher Scientifc (Fair
103
Lawn, NJ, USA). Sodium dodecyl sulphate (SDS) and sodium hydroxide were from Sigma-Aldrich.
104
Ultrapure water used for sample preparation was produced by a Milli-Q Plus apparatus from
105
Millipore (Bedford, MA, USA).
us
an
M
d
te
Ac ce p
106
cr
95
2.2.
External decontamination procedure
107
Crudely cut hair (≈1 cm) was placed in a 50 ml Falcon cone tube in 5 mL of phosphate buffer (0.1
108
M, pH 6). The mixture was vortexed for one minute and, after removal of aqueous buffer, it was
109
sequentially washed with isopropanol (5 mL) and dichloromethane (5 mL). The last wash was
110
collected in a vial and evaporated under a gentle stream of N2; the residue was reconstituted with 1
111
mL of 10 mM formic acid in methanol and was stored for further analysis in order to assess the
112
performance of the procedure.
113
2.3.
Pressurized liquid extraction
Page 7 of 33
Fifty milligrams of hair sample, cut into 1-2 mm segments, were homogenized with diatomaceous
115
earth (Sigma Aldrich,Milan, Italy) by means of a mortar. The sorbent was previously powdered and
116
washed with the same PLE extraction conditions used for hair sample. The mixture was then placed
117
in a 1 mL pressure resistant stainless steel cell that was sealed at both ends with glass-fiber filters.
118
Void volumes in the cell were filled up with diatomaceous earth and 25 µL of methanol containing
119
the ISs at a concentration of 2 ng mL-1 for THC-COOH-d3 and 20 ng mL-1 for THC-d3 were added.
120
PLE was carried out performing a single extraction cycle using as extraction solvent a 90:10 (v/v)
121
water-methanol mixture containing SDS 25 mM. The extraction conditions were: pressure, 100 bar;
122
temperature, 150 °C; preheat time, 1 min; heat time, 7 min; static time, 5 min; flush volume, 0%;
123
purge time, 60 s. The PLE extract (5-6 mL) was automatically collected in glass vial with caps
124
solvent resistant (PTFE) septa.
cr
an
us
2.4.
Solid phase extraction
M
125
ip t
114
PLE extract was collected into a graduated tube and centrifuged at 6000 g for 5 min at 25°C. Five
127
mL were then cleaned up by SPE using polymeric reversed phase cartridges Strata XL 30 mg/1 mL
128
from Phenomenex (Torrance, CA, USA). The cartridge, installed on vacuum system (Visiprep), was
129
previously conditioned with 1 mL of methanol and 1 mL of water-methanol (90:10 v/v). After
130
loading, the cartridge was washed with 5 mL of water-methanol (90:10 v/v) and 3 mL of water-
131
methanol (50:50 v/v); the analytes were then eluted using 1 mL of methanol. The eluate was
132
directly injected in the HPLC-HRMS/MS system (6 μL).
te
Ac ce p
133
d
126
2.5.
NaOH digestion
134
For comparative purposes three authentic positive hair samples were pretreated with alkaline
135
digestion as well. The digestion was carried out according with an existing method described in
136
literature with slight modifications[15] . Briefly, 50 mg of decontaminated sample were transferred
137
into an amber glass vial and 25 μL of ISs solution was added. The samples were then subjected to
Page 8 of 33
digestion in 1 mL of 2.5 M NaOH at 60 °C for 25 min; after cooling to room temperature, the
139
solution was neutralised with formic acid. The mixture obtained was extracted by vortex mixing
140
using 3 mL ethyl acetate. After centrifugation the organic supernatant was separated, evaporated to
141
dryness under a gentle flow of nitrogen and the residue was finally reconstituted in 200 μL of
142
methanol. 10 μL were then injected in the HPLC-HRMS/MS system. 2.6.
HPLC-HRMS/MS analysis
cr
143
ip t
138
HPLC-HRMS/MS analysis was performed on a Dionex UltiMate® 3000 Rapid Separation LC
145
system from Thermo Fisher Scientific (San Jose, CA, USA) coupled to a Thermo Scientific Q
146
Exactive Mass spectrometer (Thermo Fisher Scientific, Bremen, Germany).
147
The Q Exactive mass spectrometer was equipped with a heated electrospray ionization source (H-
148
ESI) and operated in the positive ionization mode. The spray voltage was 4 kV, capillary
149
temperature 375°C, heater temperature 450°C, S-lens RF level 60, sheath gas flow rate 60, auxiliary
150
gas flow rate 20 (arbitrary units). Nitrogen was used for spray stabilization, for collision–induced
151
dissociation experiments in the higher energy collision dissociation (HCD) cell and as the damping
152
gas in the C-trap. The instrument was calibrated in the positive and negative mode at the beginning
153
of each working day. The mass spectrometer operated in Targeted-MS/MS mode at a resolution of
154
35,000 (full width at half maximum at m/z 400). Automatic Gain Control (AGC) was set at 2x105
155
and maximum injection time at 100 ms. Using this scan mode, precursor ions are selected in the
156
quadrupole with a 0.4 m/z window and subsequently fragmented in the HCD cell. A full scan of all
157
fragmented ions originating from the precursor ion is performed, and two specific product ions used
158
for data analysis with a mass tolerance of 5 ppm (Table 1). Targeted-MS/MS experiments were
159
segmented in three windows (one for each compound, excluding THC and CBD which were
160
detected in the same window) in order to have the maximum sensitivity from the detector.
Ac ce p
te
d
M
an
us
144
Page 9 of 33
The analytes were separated using a Kinetex C18-XB column (10 cm x 2.1 mm ID) from
162
Phenomenex packed with core-shell particles of 2.6 μm. A Phenomenex security Guard Ultra
163
Cartridge (packed with C18 particles) was also used to protect the column from damaging
164
contaminants and microparticulate. The mobile phases were: (A) acetonitrile and (B) water both
165
containing 0.1% formic acid. The flow rate was 0.5 mL min-1 entirely driven into the ion source.
166
The following gradient elution scheme was used: phase A was increased from the initial 60% to
167
75% in 2.4 min, then up to 90% in 1.2 min and in the following 0.4 min brought to 100%. The latter
168
was maintained for 1.2 min and then switched back to the initial 60% in 2.3 min. The complete
169
separation of all substances occurred in 4 min. All source and instrument parameters were tuned by
170
injecting a standard solution of THC-COOH at a concentration of 0.1 ng µL-1 at 10 µL min-1 by a
171
syringe pump in flow injection analysis with the same chromatographic conditions (flow and
172
solvent composition). Peak areas for the selected ions were determined using Thermo Xcalibur
173
QUAN Browser and quantitation was performed by the internal standard method. The selected
174
transitions, together with the main HPLC-HRMS/MS parameters, are reported in Table 1.
cr
us
an
M
d
te
2.7.
Validation
Ac ce p
175
ip t
161
176
The method was validated according to SoHT
177
modifications. Linearity, recovery, matrix effect, precision, accuracy, limits of detection (LODs)
178
and lower limits of quantification (LLOQs) were evaluated.
179
[8] and SWGTOX guidelines [35] with few
2.7.1. Quantification and identification
180
Calibration standards were prepared in methanol. Quality control samples (QC) were prepared by
181
spiking the standard solution on the hair in the PLE cell; in addition four positive hair samples were
182
used for precision and recovery studies. Two product ions were selected for each analyte, one for
183
quantitation and one for qualitative analyte confirmation. Identification criteria included retention
184
time (tr) within ±0.2 min of average calibration standard (tr), presence of two product ions, and ion
Page 10 of 33
ratio between the quantifying and qualifier ion within ±20 % of that established by the calibration
186
standards.
187
Linearity was investigated in the concentration range from LLOQ to 50 pg/mg for THC-COOH and
188
to 750 pg/mg for the other analytes, using six calibration standards for each group. Replicates (n=5)
189
at each concentration level were analyzed during three days. The calibration curves were derived by
190
plotting the peak area of analytes to ISs versus the hair concentration using a least squares
191
regression model.
192
cr
ip t
185
us
2.7.2. Extraction recovery, absolute recovery and matrix effect
The extraction recoveries (RE) and absolute recoveries were determined at two concentration levels
194
(0.1 and 50 pg/mg for THC-COOH; 2 and 500 pg/mg for the other analytes); 5 hair samples of
195
different types (blond, black, straight and curled) were used. For each concentration and hair type,
196
RE were obtained by the ratio of the absolute peak area for each analyte spiked on the hair before
197
extraction (A) and the absolute peak area of the analyte for blank samples spiked after the extraction
198
step (B). Accordingly, RE(%) = A/B*100. Absolute recoveries were calculated by comparing A
199
with the absolute peak area for each analyte in a pure standard solution at the same concentration
200
(C).
201
Matrix effects (ME) were estimated as ME(%) = 100 × ((B/C)-1) in five different hair matrices at
202
three concentration levels (0.1, 10, 50 pg/mg for THC-COOH and 2, 250, 750 pg/mg for the other
203
analytes). A ME value < 0 indicated ionization suppression, and a value > 0 ionization
204
enhancement.
M
d
te
Ac ce p
205
an
193
2.7.3. Selectivity and carry-over
206
Interferences from endogenous hair components were evaluated by the analysis of 10 different
207
cannabinoids-free hair samples collected among employees and some family members . If analytes
208
were not detected (
