comprehensive metabolite profiling in a discovery...
TRANSCRIPT
Comprehensive Metabolite Profiling in a Discovery Environment, Coupled Workflows Enabled by HRMS
Jonathan L. Josephs1, C. Emily Luk1, Kate Comstock2,
Tim Stratton2, Yingying Huang2, and William G. Humphreys1
1Bristol-Myers Squibb, Hopewell, NJ 2Thermo Fisher Scientific, San Jose, CA
1 1
Thermo NE Users Meeting, Somerset, NJ October 12, 2011, 1:50pm
Where we look and why
2
The Success of SRM quantiation on Triple Quads has led to a mindset of targeted
quantitation methods
Establishment of departments
devoted to targeted quantitation
All quantitation done in a targeted manner
Where we look and why
3
A New Mechanism of Selectivity with Full Scan HRMS
Full Scan High Resolution Accurate Mass Complete spectrum acquired for every scan All data for all components acquired all the time
Selectivity achieved by isolating only the “exact mass” of the component of interest Like focusing a camera or microscope on the particular item
of interest – Difference is that “focusing” may be done post acquisition
such that you may focus on something that you had no knowledge of prior to the experiment.
“Method free”quantitation of multiple components High mass accuracy aids structural elucidation Faster structural assignments Improved certainty of assignments Greater suitability for automated structural assignment
approaches Integrated Metabolite ID/Quantitation datasets 5
286.5 287.0 287.5 288.0 288.5 289.0 289.5 290.0 290.5 291.0 291.5 292.0 m/z
288.04
289.12 290.06
287 288 289 290 291 292 m/z
288.0258
289.0292 290.0215
Where we look and why
6
Drug Discovery and Data Drug discovery and development is an integrated science
Expert scientists from a number of disciplines with a broad knowledge of drug discovery collaborate on integrated teams
HRMS allows the collection of more integrated datasets Streamlining of experiments and sample collection/distribution Interrelated data are obtained from the same system
– Metabolic rate/metabolite structures – PK/PD in the same animals – Metabolite profiling/estimation in FIH studies
Interpreted data streams normally integrated by a collaborative team Collection of integrated datasets increases the speed and
quality of interpreted data Facilitates the use of the combined knowledge in decision
making
7
Parent PK
TOX
Parent TK Biomarkers
Biotransfromation
Metabolite Profiles
Metabolite ID
Future ADMET Process?
DMPK Bioanalytical Analytical
BIOL CHEM
HRMS Data Collection Lead Profiling/Bioanalytical
/Analytical /Biotransformation Raw Data Acquisition
Discovery Working Group
Process/Interpret Data
Convert Interpreted Data into Knowledge
Drive Program Decisions
8
LC/MS is not a Universal Detector LC/MS is normally effective in detecting “drug like” compounds.
However it is not a universal detector Compound response is dependent on
– Compound structure – Mobile phase composition – Coeluting matrix
Quantitation by LC/MS Normally requires a calibration curve prepared from an authentic
standard. Quantitation of “unknowns” requires a “universal” detector
Radioactivity is considered to be universal UV detection at 220nm is more universal than MS response Preparation of “metabolite standards” from a high substrate
concentration in vitro incubation allows determination of UV response (or radioactivity response)
Sample dilution and analysis by LC/MS affords a UV:MS response factor (or RAD:MS response factor)
Samples run by LC/MS may then be corrected for response. 9
*Petia Shipkova, Jonathan Josephs, Mary Grubb, Robert Langish, Weiqi Chen, Mark Sanders, ASMS 2004
HRMS Integrated Qual/Quan Approach
In vivo quantitation Parent and metabolites
Metabolite “Standards” Buspirone Metabolite Time Course
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0Time (min)
UV
Co
rre
cte
d A
rea
0
1
10
100
10 20 30 40 50 60
0
1000
2000
3000
4000
5000
6000
7000
0 1 2 3 4 5 6 7 8 9
Time (hr)
Cal
cula
ted
UV
area
High substrate concentration in vitro
incubation
In vitro quantitation Parent and metabolites
Low substrate concentration in vitro incubation
In vitro metabolite ID
Accurate Mass LC/UV/MS/MS
10
* **Yanou Yang, Mary F. Grubb, Chiuwa E Luk, W. Griffith Humphreys and Jonathan L. Josephs, Rapid Commun. Mass Spectrom. 2011, 25, 3245
*
HRMS Integrated Qual/Quan Approach
t0 t45
+NADPH 30 µM
t0 t2.5 t5 t10 t15 t20 t30 t60
+NADPH 0.5 µM
t0 t2.5 t5 t10 t15 t20 t30 t60
+NADPH 0.5 µM
t0 t2.5 t5 t10 t15 t20 t30 t60
-NADPH 0.5 µM
= 2 Samples
= 24 Samples
1 compound = 2 + 1 + 24 = 27 Samples 6 compounds = 27 x 6 = 162 Samples Analysis time = 162 x 10 = 1620 mins = 27 hrs
t45/dil
Dilute = 1 Sample
Accurate Mass LC/UV/MS/MS RT-m/z MET ID
Accurate Mass LC/UV/MS/MS
RT-m/z MET ID
Accurate Mass LC/UV/MS/MS RT-m/z
MET ID
Accurate Mass LC/UV/MS/MS
RT-m/z MET ID
Accurate Mass LC/UV/MS/MS
RT-m/z MET ID
HRMS Integrated Qual/Quan Approach with Sample Pooling
Pooled Metabolite
“Standards”
High substrate concentration (30 µM) in vitro
incubations
Accurate Mass LC/UV/MS/MS
12
S
O
O
O
NHN
N
S
F
F F
O
O
HN N
N
FFF
O
ON
NH2
NO O
N
N
NN
F
O
HO OH
HN
O
N
N
F
t45 t45 t45 t45 t45 t45 t0 t0 t0 t0 t0 t0
RT-m/z MET ID
12 Qualitative samples generated
0.5 µM Parallel Incubations with Sample Pooling
13
t0
t2.5
t5
t10
t15
t20
t30
t60
Pooled Metabolite
“Standards”
t0
t2.5
t5
t10
t15
t20
t30
t60
8 + 1 = 9 Samples
Throughput improvements with pooled analysis Traditional Approach
6 compounds 25 Quantiative + 2 Qualitative samples/cmpd 6 x 27 = 161 Samples Analysis time = 162 x 10 = 1620 mins 27 hrs total
Pooling approach 6 compounds 25 Quantiative + 12 Qualitiative samples total 25 + 12 = 37 total samples Analysis time = 37 x 10 = 370 mins 6 hrs total 1 hr/compound
Full qualitative dataset 8 Time point t1/2 in duplicate with control 8 Timepoint metabolite formation quantitation in duplicate
Pooling makes sample analysis more challenging 6 x dilution factor Samples are now more complex
Instrument requirements For quantitation
High resolution (>30,000 FWHM) High mass accuracy/stability Fast scanning (> 3Hz at 30,000) Sensitive (detect metabolites at all time points in 6 x diluted 0.5 µM incubations)
For structural elucidation High resolution (>30,000 FWHM) in MS and MS/MS modes High mass accuracy in MS and MS/MS modes
Q Exactive: Max resolution: 140,000 (m/z 200) Scan speed: up to 12 Hz (17,500 @ m/z 200) <5ppm external <1 ppm internal +/- switching within 1 sec
Samples for pooled analysis
S
O
OO HN
NN
Esomeprazole
Chemical Formula: C17H19N3O3SExact Mass: 345.11471
S
F
F
F
O
OHN
NN
Lansoprazole
Chemical Formula: C16H14F3N3O2SExact Mass: 369.07588
F
O
N
NChemical Formula: C20H21FN2O
Exact Mass: 324.16379
Citalopram
F
FF
O
ON
NH2Chemical Formula: C15H21F3N2O2
Exact Mass: 318.15551
Fluvoxamine
F
O OH
OH
HN
Chemical Formula: C18H20FNO3Exact Mass: 317.14272
N
O
O
N NN
N
Chemical Formula: C21H31N5O2Exact Mass: 385.24778
BuspironeParoxetine
Esomeprazole
17
RT: 2.80 - 4.37
3.0 3.2 3.4 3.6 3.8 4.0 4.2Time (min)
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
200000
220000
240000
260000
280000
300000
320000
340000
uAU
Esomeprazole 0.5 µM Incubation 0 min
3.0 3.2 3.4 3.6 3.8 4.0 4.2
Time (min)
4.24
3.24
3.77
3.94
3.15
NL: 2.48E8 m/z= 346.1203-346.1237 F: FTMS + p ESI Full ms MS 042911_Esomeprazole_ RLM_HL_MQ_01
NL: 1.11E6 m/z= 362.1151-362.1187 F: FTMS + p ESI Full ms MS 042911_Esomeprazole_ RLM_HL_MQ_01
NL: 4.35E6 m/z= 332.1046-332.1080 F: FTMS + p ESI Full ms MS 042911_Esomeprazole_ RLM_HL_MQ_01
XICs +/- 5ppm
Esomeprazole
4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28 4.29 Time (min)
4.24
4.24 4.23
4.25
4.23
4.25
4.26 4.22
4.26
4.27 4.22 4.27 4.27
NL: 2.48E8 m/z= 346.1203-346.1237 F: FTMS + p ESI Full ms MS 042911_Esomeprazol e_RLM_HL_MQ_01
3 sec
12 scans/3s 4Hz
Theoretical 346.12199 Observed Mass
Error (ppm)
Variation from mean (ppm)
346.12103 -2.77 -0.08 346.12106 -2.69 0.01 346.12097 -2.95 -0.25 346.12097 -2.95 -0.25 346.12106 -2.69 0.01 346.12100 -2.86 -0.17 346.12097 -2.95 -0.25 346.12094 -3.03 -0.34 346.12088 -3.21 -0.51 346.12119 -2.31 0.38 346.12128 -2.05 0.64 346.12134 -1.88 0.82
Average 346.12106 -2.69 STD 0.00014 STD ppm 0.40787
0.5 uM Incubation 0 min
Esomeprazole
100 200 300 400 500 600 700 800 900 1000 m/z
346.1209
713.2178 198.0580
[M+H]+
[2M+H]+
0.5 µM Incubation 0 min
[M+Na]+
[M+H – H2O]+
346.0 346.5 347.0 347.5 348.0 348.5 m/z
346.1209 R=57513
347.1245 R=56547
348.06 348.08 348.10 348.12 348.14 348.16 348.18 348.20
348.1169
-2.4 ppm C17H20N3O3
34S = 348.1178
348.1275 12C16
13CH20N3O332S = 348.1332
-5.67 ppm
R = 59200
R = 56100
0
100
200
300
400
500
600
0 10 20 30 40 50 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Esomeprazole in RLM
2.96_332 3.07_362 3.60_362 3.79_362
Esomeprazole
21
0.5 µM Incubations
y = 78.478e-0.164x R² = 0.9627
1
10
100
0 10 20 30 40 50 60
%R
emai
ning
Time (min)
Half-Life of Esomeprazole in RLM %Remaining
%-NADPH
Expon. (%Remaining)
t1/2 4.2 min
y = 95.099e-0.151x R² = 0.9645
1
10
100
0 10 20 30 40 50 60
%R
emai
ning
Time (min)
Half-Life of Esomeprazole in RLM Pooled Analysis %Remaining
%-NADPH
Expon. (%Remaining)
t1/2 4.6 min
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Esomeprazole in RLM Pooled Analysis
2.96_332
3.07_362
3.60_362
3.79_362
0
100
200
300
400
500
600
0 10 20 30 40 50 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Esomeprazole in RLM
2.96_332 3.07_362 3.60_362 3.79_362
Fluvoxamine
22
y = 117.8e-0.104x R² = 0.9577
1
10
100
0 20 40 60 80
%R
emai
ning
Time (min)
Half-Life of Fluvoxamine in RLM
%Remaining
%-NADPH
Expon. (%Remaining)
t1/2=6.66 min
y = 100.76e-0.109x R² = 0.9869
1
10
100
0 20 40 60 80
%R
emai
ning
Time (min)
Half-Life of Fluvoxamine in RLM Pooled analysis %Remaining
%-NADPH
Expon. (%Remaining)
t1/2=6.36 min
0 20 40 60 80
100 120 140
0 10 20 30 40 50 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Fluvoxamine in RLM
5.14_305
5.20_319
0
20
40
60
80
100
120
140
0 10 20 30 40 50 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Fluvoxamine in RLM Pooled Analysis
5.14_305
5.20_319
0.5 µM Incubations
Buspirone
y = 85.318e-0.216x R² = 0.9629
0
1
10
100
0 20 40 60
%R
emai
ning
Time (min)
Half-Life of Buspirone in RLM
%Remaining
t1/2 3.2 min
y = 78.724e-0.211x R² = 0.9708
0
1
10
100
0 10 20 30 40 50 60
%R
emai
ning
Time (min)
Half-Life of Buspirone in RLM Pooled Analysis
%Remaining
%-NADPH
Expon. (%Remaining)
t1/2 3.3 min
0
500
1000
1500
2000
2500
0 20 40 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Buspirone in RLM Pooled analysis
3.03_402
3.22_402
3.67_402
3.77_418
4.01_402
4.27_418
4.44_402
4.93_402 0
200
400
600
800
1000
1200
1400
1600
0 20 40 60
Cor
rect
ed M
et A
rea
Time (min)
Metabolite Profile of Buspirone in RLM
3.03_402
3.22_402
3.67_402
3.77_418
4.01_402
4.27_418
4.44_402
4.93_402
0.5 µM Incubations
Conclusion Hybrid Quadrupole-Orbitrap platform (Q Exactive) is well suited to
integrated Qual/Quan workflows High resolution (>30,000 FWHM)
57,000 @ m/z 346 High mass accuracy/stability
< 5ppm/<1ppm @ 57,000 for mass m/z 346 Fast scanning (> 3Hz at 30,000)
4 Hz @ 57,000 for mass m/z 346
Sample pooling for analysis can greatly reduce the data acquisition time With a sufficiently selective and sensitive system comparable data may
be obtained compared to running samples individually Integrated Qual/Quan workflows may be feasible in a high
throughput environment Structural assignments are the limiting step Extensive Qual/Quan datasets may be more suited to computational
modeling than quantiative datasets alone
24
Beyond the horizon
Beyond the horizon
In Vitro Incubations and HRMS Data Collection Lead Profiling
In Vitro Incubations and HRMS Data Collection “Cloud Instrumentation”
Beyond the horizon
Automated Structural Assignment of Metabolites
Automated Calculation of Initial Rate of Formation of Primary Metabolites
Automated Determination of Relationship of Primary and Secondary Metabolites
Building of Metabolic Softspot SAR by Data Mining of Quan/Qual Datasets
Recommendation of Structural Solutions
In silico prediction of rate/site of metabolic stability
In Vivo Samples and HRMS Data Collection PCO
DWG
Scientists
Automated t1/2 Determination
ADME
Scientists
ADME
Scientists
In Vivo Samples and HRMS Data Collection “Cloud Instrumentation”
27
Parent PK
TOX
Parent TK Biomarkers
Biotransformation
Metabolite Profiles
Metabolite ID
2020 A Clearer Understanding of ADMET
MAP Bioanalytical Analytical
BIOL CHEM
Raw Data Acquisition
Discovery Working Group
Process/Interpret Data
Convert Interpreted Data into Knowledge
Drive Program Decisions
28
HRMS Data Collection Lead Profiling/Bioanalytical/ Analytical/Biotransfromation
Acknowledgements • Bristol-Myers Squibb
Bruce Car David Rodrigues Adrian Tymiak Harold Weller Mark Hillman Wilson Shu Ragu Ramanathan Silvi Chacko Yanou Yang Mary Grubb Hong Cai Kim Johnson Ben Johnson Yue-Zhong Shu Qin Ruan
• Thermo Fisher Scientific Bjoern Rose Markus Kellmann Kevin Cook Mark Sanders
29