the final step towards automated lc-ms screening ......automated lc-ms screening analysis. the aim...
TRANSCRIPT
• To our knowledge, the presented workflow is the first Toxtyper / TargetScreener approach featuring complete automation from sample preparation to evaluation of data. The extraction time of about 14 min fits into the runtime of the QTOF-Screening (20 min) and only slightly exceeds the Toxtyper runtime (11 min).
• LC-MSn screening of fortified urine using µSPE led to similar or better results than the routine sample preparation. This results could be confirmed in a batch of urine samples from real cases. All µSPE-LC-MSn screening results were in good agreement with the initial routine analysis.
• LODs of the QTOF are probably lower than the ones found for the Toxtyper and will be determined separately.
• Evaluation of ME showed a maximum ion suppression of 50% which was considered acceptable for a screening approach.
• Implementation of a cleavage step would overcome the extraction issues seen for some of the glucuronides and therefore might lower the LOD of the respective parent substance but would also prolong the overall extraction time.
• Direct injection of the µSPE eluate limits the choice of solvents that could be used for extraction. Optimization of the protocol, including incorporation of an evaporation unit, might increase the overall performance of the DAU cartridge.
o Routine LC-MSn screening (Toxtyper®, TT) o LC-QTOF-MS screening (TargetScreener, TS) o Elute UHPLC equipped with a PAL RTC o Three types of smartSPETM cartridges (ITSP Solutions):
UCT C18 endcapped cartridges 10 mg UCT DAU cartridges 10 mg UCT C18 endcapped cartridges 30 mg
The Final Step Towards Automated LC-MS Screening - Implementation of an Online µSPE for Urine Screening
Michaela Schmidt1,2, Marina Schumacher3, Birgit Schneider3, Laura M. Huppertz2, Jürgen Kempf2
1Faculty Medical and Life Sciences, Furtwangen University, Schwenningen, Germany 2Institute of Forensic Medicine, Forensic Toxicology, Medical Center - University of Freiburg, Germany 3Bruker Daltonik GmbH, Bremen
Institute of Forensic Medicine
Forensic Toxicology
Systematic toxicological analysis (STA) using comprehensive screening approaches is a major part of everyday work in forensic toxicology, usually involving different analytical techniques.
The great benefit of immunoassays (IA) in routine screening of body fluids is the high degree of automation regarding sample preparation and reporting of results. In recent years, LC-MS has become a key technique in STA, but in contrast to IA an appropriate sample preparation is crucial for screening of biological samples. Offline liquid-liquid extraction (LLE), solid phase extraction (SPE) or protein precipitation (PP) are often laborious but mandatory steps and their integration into the analytical workflow is the missing piece towards a fully automated LC-MS screening analysis.
The aim of this project is to implement an online µSPE into an existing LC-MS method to achieve a fully automated LC-MS screening of urine samples.
LC Eluents
TT A: H2O, 0.1% HCOOH, 2 mM NH4+COO-, 1% acetonitrile TT B: Acetonitrile, 2 mM NH4
+COO-, 0.1% HCOOH, 1% H2O
TS A: H2O, 0.2% buffer concentrate, 1% eluent B TS B: Methanol, 0.2% buffer concentrate
µSPE Steps
1. Wash: 0.2% NH4acetate in MeOH
2. Condition: 20% NH4 acetate in H2O
3. Sample load: 200 µl
4. Desalt: 20% NH4 acetate in H2O
5. Elution: 0.2% NH4 acetate (V = 75 µl)
TT LM B 989 mL Acetonitrile 10 mL NH4 formate (0.2 M) 1 mL HCOOH
TT LM A 979 mL H2O 10 mL Acetonitrile 10 mL NH4 formate (0.2 M) 1 mL HCOOH
TS LM B
1 L MeOH 2 mL Buffer
Bruker
Elu
te
UH
PLC
- - - -
LC Tool 1 979 mL H2O dest. 10 mL Acetonitrile 10 mL Ammonium formate (200 mM) 1 mL HCOOH
Fast Wash 2 300 mL Acetonitrile 300 mL Methanol 300 mL Isopropanol 100 mL HCOOH (0.1%)
Fast Wash 1 0.1 % HCOOH
LC Tool 2 300 mL Acetonitrile 300 mL Methanol 300 mL Isopropanol 100 mL HCOOH (0.1%)
Stop
TS LM A
1 L H2O 2 mL Buffer 10 mL DE
Waste
Back
PAL RTC
B R U K E R
B R U K E R
PAL RTC µSPE System and Elute UHPLC
ESI (alternating polarity) UltraScan: 70 - 800 Da Scan Speed: 32,500 Da/s AutoMSn mode (n = 3)
Bruker amaZon Speed Ion Trap
ESI positive mode Scan range: 30 - 1000 Da Full Scan TOF MS / bbCID
Bruker impact II QTOF
Automated evaluation of HR data using a database containing over 1700 drugs, drugs of abuse and metabolites
TargetScreener HR QTOF-Screening (TS)
Toxtyper Screening (TT)
Fully automated data evaluation and reporting of results
LC Columns
TT: Acclaim® RSLC C18 2.2 µm 120A 100x2.1 mm
TS: Intensity Solo 1.8 C18-2 100x2.1 mm
Starting Equipment
Method Development
General Proof of Concept Comparison of µSPE Cartridges Reproducibility Recovery, Matrix Effects, LOD Real Samples
C18-10 C18-30
DAU
3x 75/
01
SARSTEDT
uri
ne
Proba
n
d 3
75/
01
SARSTED
T
urin
e
Proba
nd
2
75/0
1
SARSTED
T
urin
e
Proba
nd
6
75/
01
SARSTE
DT
urin
e
Proba
nd
5
75/0
1
SARSTED
T
urin
e
Proba
nd
1
75/0
1
SARSTEDT
urin
e
Proba
nd
4
1x
75/0
1
SARSTEDT
urin
e
Proba
nd
3
75/0
1
SARSTEDT
urin
e
Proba
nd
2
75/0
1
SARSTEDT
urin
e
Proba
nd
6
75/0
1
SARSTEDT
urin
e
Proba
nd
5
75/0
1
SARSTEDT
urin
e
Proba
nd
1
75/0
1
SARSTEDT
urin
e
Proba
nd
4
n = 50
75/0
1
SARSTEDT
urin
e
Proba
nd
2
75/0
1
SARSTEDT
urin
e
Proba
nd
5
75/0
1
SARSTEDT
urin
e
Proba
nd
3 7
5/0
1
SARSTEDT
urin
e
Proba
nd
6
75/0
1
SARSTEDT
urin
e
Proba
nd
6
75/0
1
SARSTEDT
PO
OLED
UR
IN
E
N = 6
75/0
1
SARSTEDT
urin
e
Proba
nd
1
75/0
1
SARSTEDT
urin
e
Proba
nd
2
75/0
1
SARSTEDT
urin
e
Proba
nd
5
75/0
1
SARSTEDT
urin
e
Proba
nd
1 7
5/0
1
SARSTEDT
urin
e
Proba
nd
2
75/0
1
SARSTEDT
urin
e
Proba
nd
5
75/0
1
SARSTEDT
urin
e
Proba
nd
3 7
5/0
1
SARSTEDT
urin
e
GS 2
3-19
75/0
1
SARSTEDT
urin
e
Proba
nd
6
75/0
1
SARSTEDT
PO
OLED
UR
IN
E
N = 6
75/0
1
SARSTEDT
urin
e
Proba
nd
2
75/0
1
SARSTEDT
urin
e
Proba
nd
5
75/0
1
SARSTEDT
urin
e
GS 1
25-18 7
5/0
1
SARSTEDT
GS 4
2-19
UR
IN
E
n = 137
Cocain
e
Am
pheta
min
e
Low, medium, high concentration
Comparison of µSPE and PP using urine from routine cases (post / ante mortem)
75/0
1
SARSTEDT
PO
OLED
UR
IN
E
N = 6
n = 6
Low, med, high concentration
Cocain
e
Am
pheta
min
e
10x
75/
01
SARSTEDT
uri
ne
Proba
n
d 3
75/
01
SARSTED
T
urin
e
Proba
nd
2
75/0
1
SARSTED
T
urin
e
Proba
nd
6
75/
01
SARSTE
DT
urin
e
Proba
nd
5
75/0
1
SARSTED
T
urin
e
Proba
nd
1
75/0
1
SARSTEDT
urin
e
Proba
nd
4
75/0
1
SARSTEDT
PO
OLED
UR
IN
E
N = 6
n = 6
25 and 100 ng/ml 27.5, 275 and 500 ng/ml for ME and RE
Cocain
e
Am
pheta
min
e
75/0
1
SARSTEDT
urin
e
Proba
nd
2
75/0
1
SARSTEDT
urin
e
Proba
nd
6
75/0
1
SARSTEDT
urin
e
Proba
nd
5
75/0
1
SARSTEDT
urin
e
Proba
nd
1
75/0
1
SARSTEDT
urin
e
Proba
nd
4
75/0
1
SARSTEDT
urin
e
Proba
nd
3
n = 6
75/0
1
SARSTEDT
urin
e
GS 0
07-18
Comparison of µSPE and routine PP
2x (ME, RE) 3x (LOD)
Cocain
e
Am
pheta
min
e
Introduction Methods
General Proof of Concept
In comparison to routine PP with acetonitrile, the identification rate of the LC-MSn screening could be improved from 74% to 84% at low concentration levels and from 90% to 96% at high concentration levels, when using µSPE.
Due to the higher sensitivity of the QTOF-MS system all spiked compounds could be detected even at low concentrations.
Comparison of µSPE Cartridges
A set of 18 compounds of different compound classes covering the retention time and mass range of the method was chosen for further testing of different µSPE cartridges. For six analytes, higher S/N ratios could be observed using the DAU cartridges. For all other analytes, no preferable cartridge could be determined.
C18-30 cartridges led to higher S/N ratios for EME and norbuprenorphine but to low absolute peak areas, probably due to higher amounts of sorbent. Higher eluent volumes might enhance the elution but would also dilute the extract injected to the LC-MS system. Therefore, the C18-30 cartridge was excluded from further method development.
n = 18
Reproducibility First tests using two different PAL tools for sample and solvent handling - to circumvent carry over at all costs - led to poor results concerning reproducibility. So, a new 250 µl LC-MS Tool was tested for all liquid handling including the injection step. Injection reproducibility of different volumes (1, 2, 5 and 10 µl) ranged from 1.5 to 7%. Optimizing the cleaning procedures after the different extraction steps led to no detectable carryover caused by the µSPE system.
Reproducibility of the complete extraction process using a single tool was tested by tenfold preparation of pooled urine fortified with a chosen set of compounds. Morphine-glucuronide was the only outlier in this test with RSDLow of 23.3 and 19.8% and RSDHigh of 75.7 and 38.3%.
RSD of ten fold extraction
C18-10 DAU
Low 5.6 - 10.9% 10.0 - 14.9%
High 4.3 - 11.2% 1.8 - 6.8%
Matrix Effects (ME) and Recovery (RE) As ion source and ion transfer of both instruments are identical, ME and RE were only evaluated by QTOF-MS using a protocol adapted from Matuszewski et al.. For the C18-10 cartridge, maximum ion suppression in six tested urine matrices was around 50%. DAU cartridges showed comparable matrix effects in a pooled urine matrix. While ion suppression will have negative effects on the LOD, ion enhancement is not an issue in screening approaches.
For morphine-glucuronide and EME the overall recovery was very poor. The C18-10 material seems to have problems properly retaining these early eluting compounds. Average RE and ME of tested analytes are depicted below.
Limits of Detection (LOD)
LOD for the µSPE-LC-MSn screening approach were determined in pooled blank urine (n = 10) fortified with compounds most often found in routine during the last year, in decreasing concentrations down to 25 ng/ml. The lowest concentration automatically identified (n = 3) was set as LOD.
Met
ham
ph
.
MD
A
Am
ph
.
Su
pp
ress
ion
RE cLow RE cMed RE cHigh Loss
ME cLow ME cMed ME cHigh
MD
MA
BE
EME
THC
-CO
OH
Met
had
on
e
EDD
P
Mo
rph
ine
6-M
AM
Mo
rph
.-G
luc
No
rco
dei
ne
Fen
tan
yl
No
rfen
tan
yl
Alp
razo
lam
α-O
H-A
lpra
z.
Bro
maz
epam
3-O
H-B
rom
az.
No
rdaz
epam
Oxa
zep
am
Zolp
idem
85
1
05
1
14
56
7
0
79
En
han
cem
en
t
16
8
17
3
16
4
89
1
23
1
76
77
9
2
11
8
79
7
7
80
13
6
13
2
13
2
10
3
92
8
4
66
9
2
10
8
70
9
5
10
4
10
7
80
6
8
72
9
2
95
76
6
9
66
59
7
6
78
95
8
1
75
81
8
0
79
84
1
09
1
18
89
1
15
1
27
10
8
13
2
14
6
94
1
28
1
30
20
0
16
7
13
3
15
7 x
3
83
x3
3
6 x
3
Average ME and RE (n = 6) of 22 Selected Analytes at Three Concentrations
Re
cove
ry
Mat
rix
Effe
cts
0%
25%
50%
75%
100%
0%
50%
100%
150%
200%
250%
Real Samples
The µSPE-LC-MSn screening results of 28 urine and 22 post-mortem urine samples from real cases were in good agreement with the findings from routine analysis. Using µSPE, 90% (C18-10) and 88% (DAU) of the substances could be identified in accordance to routine analysis. Routine LC-MSn screening could identify 80% of the compounds. The sum of all different analytes identified by µSPE and PP and confirmed by further routine analysis corresponds to 100%.
Unfortunately, neither of the cartridges could extract ethyl glucuronide (EtG) and ethyl sulfate (EtS) which were therefore excluded from this evaluation.
In one post-mortem case, the antiparkinson drug pramipexole could be detected in both µSPE runs but not in the initial routine screening. The routine Toxtyper only identified pramipexole in the corresponding cardiac blood and vitreous humor.
In a second case, amphetamine could be found in both µSPE runs but not in the routine Toxtyper. The amphetamine finding was confirmed by the more sensitive LC-QTOF-MS.
Screening Results in Real Samples (n = 50)
Conclusion
≈
≥ ID Rate
ID Rate
-
-
-
-
µS
PE
To
ol
?
-
-
-
-
LC
P T
oo
l
100 µ
L
-
-
-
-
LC
P T
oo
l
250 µ
L
10 µL 5 µL 2 µL 1 µL
Analytical Results
6 6 4 5 6 8 5
23
5 4 10 9
5 8
11 15
9 8 5
8 10 15 13
7 11
5
13 17
5 6 5 4 8
4 4
12 11 10 6
9 8 7 13
10 10 6 5 5 5
22 7 7
4 6 6
9
4
27
4 4
10 10
7
10 10
17
11 11
6
12 12
16
13
8
12
6
13
19
6 10
6 6
9
8 5
13 11
9
5
9 11 8
15
10 9
5 7 5 6
28
6 6
4
6 4
6
4
24
5 3
12
7
5
10
11
19
11 12
6
13 12
16
11
10
12
6
15
21
6
8
5 5
9
9
4
12
11
9
5
9 8
9
16
10 10
5 6
5 6
25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
Total # of analytes in the sample
7 7 7 6
5
9
5 5 5
27
12
12
12 11
5
19
11 12
6
13 15
17
13 13
10
6
18
22
6
10
6 6 6 5
8
10 13
17
28
9 9
5
10
7 7
11 10 10
13 16
0
25
50
75
100
3-O
H-B
rom
aze
pam
4-F
orm
ylam
ino
anti
pyr
ine
4-M
eth
ylam
ino
anti
pyr
ine
7-A
min
ocl
on
aze
pam
7-A
min
ofl
un
itra
zep
am
9-H
ydro
xyri
sper
ido
ne
Ace
tam
ino
ph
en
Am
bro
xol
Am
isu
lpri
de
Am
itri
pty
line
Am
itri
pty
lino
xid
Am
lod
ipin
e
a-O
H-M
idaz
ola
m
Ari
pip
razo
le
Be
nzo
yle
cgo
nin
e
Bis
op
rolo
l
Bro
maz
epam
Bu
pre
no
rph
ine
Bu
pro
pio
n
Car
bam
azep
ine
Clo
naz
ep
am
Co
cain
e
Co
de
ine
Co
de
ine-
6-G
luc.
De
smet
hyl
mir
taza
pin
e
De
xtro
me
tho
rph
an
Dih
ydro
cod
ein
e
Dip
hen
hyd
ram
ine
Do
xep
in
Ecgo
nin
emet
hyl
este
r
Flu
nit
raze
pam
Flu
oxe
tin
e
Flu
pir
tin
e
Gab
apen
tin
Hyd
roco
do
ne
Hyd
roxy
zin
e
Ke
tam
ine
Lam
otr
igin
e
Lora
zep
am LSD
MD
A
MD
EA
MD
MA
Me
lper
on
e
Me
thad
on
e
Me
top
rolo
l
Mid
azo
lam
Mo
rph
ine
Nal
oxo
ne
Nic
oti
ne
No
rbu
pre
no
rph
ine
No
rdaz
ep
am
No
rdo
xep
in
No
rfe
nta
nyl
No
rmo
rph
ine
No
rtili
din
e
No
rtri
pty
line
Op
ipra
mo
l
Oxy
cod
on
e
Par
oxe
tin
e
Pip
amp
ero
ne
Pre
gab
alin
e
Pro
maz
ine
Rit
alin
ic a
cid
Ro
curo
niu
m
Sert
ralin
e
Tem
azep
am
THC
-CO
OH
-Glu
c
Traz
od
on
e
U-4
77
00
LODs of µSPE-LC-MSn and Routine LC-MSn Analysis (70 Selected Compounds)
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* n.d. at conc. ≤ 100 ng/ml with µSPE
* n.d. at conc. ≤ 100 ng/ml with PP
ng/ml
25, 50, 75 and 100 ng/ml for LOD
These LODs are suitable for STA in emergency and post mortem toxicology, but not sufficient for trace analysis as required for sobriety testing or analysis of DFC cases.
n = 12
n = 22