Development of a Bioanalytical Method for Quantification of Amyloid Beta Peptides in Cerebrospinal Fluid

Transcription

1 Development of a Bioanalytical Method for Quantification of Amyloid Beta Peptides in Cerebrospinal Fluid Joanne ( 乔安妮 ) Mather Senior Scientist Waters Corporation Data courtesy of Erin Chambers and Mary Lame 211 Waters Corporation 1

2 Background: Amyloid β Peptides Clinical significance of amyloid β peptides In normal healthy individuals, Aβ peptides are present only in small quantities as soluble monomers that circulate in cerebrospinal fluid and blood. Amyloid beta (Aβ) are peptides of amino acids that are processed from the Amyloid precursor protein Aβs are the main component of amyloid plaques - deposits found in the brains of patients with Alzheimer's disease Drug development strategies aimed at lowering production of these peptides or enhancing their clearance o Inhibition/modulation of gamma secretase enzyme 211 Waters Corporation 2

3 Amyloid β Peptides for Quantification in CSF Amyloid β 1-38 DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGG MW 4132, pi 5.2, HPLC index 96 Amyloid β 1-4 DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVV MW 433, pi 5.2, HPLC index 13 Amyloid β 1-42* DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA MW 4516, pi 5.2, HPLC index 117 *Aβ 42 occurs more frequently and forms fibrillar aggregates far more readily than the Ab4 peptide Internal standards are N15 labeled versions of 1-38, 1-4, and Waters Corporation 3

4 Outline Introduction Mass Spectrometry Liquid Chromatography Choice of Standard Curve Matrix Solid Phase Extraction Partial Validation 211 Waters Corporation 4

5 Why LCMSMS? Why an LCMSMS based assay? ELISA assays not practical for discovery, no antibodies available yet Challenges with ELISA assays o time consuming, expensive to develop o require separate assay for each peptide o limited linear dynamic range o Possible cross reactivity Benefits of LC/MS/MS for peptides LCMSMS provides single assay for multiple amyloid peptides Broad linear dynamic range Accurate, precise Universal Faster, cheaper method development 211 Waters Corporation 5

6 Specific Challenges in Developing an LCMSMS Assay for Amyloid β Peptides Extremely hydrophobic Very poor peptide solubility High level of aggregation in solution & plasma High level of non-specific binding (NSB) Low MS sensitivity Very large and hydrophobic Form many low abundance fragments Specificity in matrix 211 Waters Corporation 6

7 ESI- MSMS Spectra for Amyloid β M-4H 4- Daughters of 1127ES- 1.55e5 -H 2 O Non-specific water loss m/z Negative Ion mode ES Water loss most commonly reported MRM in literature to date Sensitive but not specific, Immuno-precipitation used prior to SPE 211 Waters Corporation 7

8 Positive Ion versus Negative Ion Detection: Specificity 1-42 in Human CSF 25pgmL_hCSF_3291_2x 2.1 2: MRM of 3 Channels ES > (a Beta 1-42) 7.5e Amyloid β 1-42 ESI pgmL_hCSF_3291_2x 1: MRM of 3 Channels ES > (a Beta 1-42) 1.2e4 Amyloid β ESI Time 211 Waters Corporation 8

9 Positive Ion versus Negative Ion Detection: Specificity 1-38 in Human Plasma Amyloid β : MRM of 3 Channels ES- 825 > (a Beta 1-38) 3.35e ESI : MRM of 3 Channels ES > 1.3 (a beta 1-38) 2.99e4 Amyloid β ESI Time 211 Waters Corporation 9

10 Positive Ion versus Negative Ion Detection: Specificity 1-4 in Human Plasma 5.7 2: MRM of 3 Channels ES > 86.9 (a Beta 1-4) 4.22e Amyloid β 1-4 ESI : MRM of 3 Channels ES > (a Beta 1-4) 1.99e4 Amyloid β ESI Time 211 Waters Corporation 1

11 Use of Higher m/z Precursors and Fragments Human Plasma Extract 2.48 MRM of 2 Channels ES > e4 4 th charge state > MRM of 2 Channels ES > e th charge state 825 -> Time 211 Waters Corporation 11

12 ESI+ MSMS Spectra for Amyloid β b b Daughters of 1129ES+ 1.62e4 M+H 4+ b38 b b32 b b33 b b b b35 b38 b b4 m/z 4+ fragments 3+ fragments ESI+ Sequence fragments - more specific High masses - more specific 211 Waters Corporation 12

13 Amyloid β Peptide 1-42: Identification of Fragments with BioLynx In Silico Fragmentation Model Fragment choice is 4+ b 4 ion 211 Waters Corporation 13

14 Xevo TQ-S MSMS Conditions Xevo TQ-S MS operated in ESI+ mode MRM Transitions: Peptide Name Precursor Ion 4+ Product Ion 4+ Product Ion ID Cone voltage (V) Collision energy (ev) Amyloid β b Amyloid β 1-38 N15 IS Amyloid β b Amyloid β 1-4 N15 IS Amyloid β b Amyloid β 1-42 N15 IS Waters Corporation 14

15 Outline Introduction Mass Spectrometry Liquid Chromatography Choice of Standard Curve Matrix Solid Phase Extraction Partial Validation 211 Waters Corporation 15

16 UPLC Conditions: Final Method Column: 2.1 X 15mm, ACQUITY BEH C18 3Å, 1.7µm Mobile phase A:.3 NH 4 OH by volume, or.1 absolute Mobile Phase B: 9/1 ACN/mobile phase A Temperature: 5 C SNW: 65/25/1 ACN/water/NH 4 OH SNW volume: 6 µl WNW: 9/1 water/acn +.3 NH 4 OH WNW volume: 6 µl Flow rate:.2 ml/min Injection mode: partial loop Injection Volume: 1 µl Injection Solvent for standards: SPE elution solvent diluted with water; SPE elution solvent = 75 ACN, 15 water, 1 NH 4 OH Gradient Table 211 Waters Corporation 16

17 Xevo TQ-S LOD: 12.5 pg/ml solvent standard in 5/5 elution solvent/water 12pt5pgmL_abetas_7261_1 Sm (SG, 3x5) MRM of 3 Channels ES > (Amyloid Beta 1-42) 5.58e3 Area Amyloid β 1-42 Peak Area pt5pgmL_abetas_7261_1 Sm (SG, 3x5) MRM of 3 Channels ES > 154 (Amyloid Beta 1-4) 5.58e3 Area Amyloid β 1-4 Peak Area pt5pgmL_abetas_7261_1 Sm (SG, 3x5) MRM of 3 Channels ES ; > 1.75 (Amyloid Beta 1-38) 5.58e3 Area Amyloid β 1-38 Peak Area Time 1µL injection 211 Waters Corporation 17

18 Representative Chromatogram: Extracted Spiked Artificial CSF Peak Widths 2-3 seconds Amyloid β e4 5.6 Amyloid β 1-42 Amyloid β Time 211 Waters Corporation 18

19 Representative Chromatography: Extracted Basal (Endogenous) Level Human CSF Sample hcsf_pool1_basal_421_ N15 IS MRM of 6 Channels ES > (a Beta 1-42 N15) 2.77e hcsf_pool1_basal_421_1 MRM of 6 Channels ES > (a Beta 1-4 N15) e hcsf_pool1_basal_421_1 MRM of 6 Channels ES > (a Beta 1-42) 8.54e ng/mL 1-4 N15 IS hcsf_pool1_basal_421_ hcsf_pool1_basal_421_ hcsf_pool1_basal_421_ ng/ml 1-38 N15 IS ng/mL MRM of 6 Channels ES+ 183 > (a Beta 1-4) 1.16e5 211 Waters Corporation MRM of 6 Channels ES+ 146 > (a Beta 1-38 N15) 4.5e MRM of 6 Channels ES > 1.3 (a beta 1-38) 2.31e Time

20 Outline Introduction Mass Spectrometry Liquid Chromatography Choice of Standard Curve Matrix Solid Phase Extraction Partial Validation 211 Waters Corporation 2

21 Choice of Standard Curve Matrix Artificial CSF chosen as surrogate matrix Lower cost Higher availability Does not influence Aβ quantification No spinal tap required No standard addition calculations Artificial CSF Issues with recovery caused by non-specific binding 5 Rat plasma added Plasma increased recovery & reproducibilty Plasma also helps to mimic real CSF (similar proteins) 211 Waters Corporation 21

22 Average Baseline Levels in Pooled Human CSF: Comparison of Results with Different Standard Curve Preparations Excellent correlation between artificial CSF & human CSF Standard curves prepared from either artificial CSF or by over-spiking human CSF Conc.(ng/mL) Source 1 Conc. (ng/ml) Source 2 Amyloid β 1-38 Conc from curve in art CSF Conc from curve in human CSF.679 n/a Conc.(ng/mL) Source 1 Conc. (ng/ml) Source 2 Amyloid β 1-4 Conc from curve in art CSF Conc from curve in human CSF n/a average.75 std deviation.1 RSD average 4.24 std deviation.44 RSD 1.93 Conc.(ng/mL) Source 1 Conc. (ng/ml) Source 2 Amyloid β 1-42 Conc from curve in art CSF Conc from curve in human CSF average std deviation RSD *Human CSF corrected for basal level 211 Waters Corporation 22

23 Outline Introduction Mass Spectrometry Liquid Chromatography Choice of Standard Curve Matrix Solid Phase Extraction Partial Validation 211 Waters Corporation 23

24 Sample Extraction: Pretreatment Sample Pretreatment Spike human or artificial CSF + 5 rat plasma samples, mix Equilibrate samples at room temperature for 3 minutes Remove 5 µl spiked artificial CSF (to which 5 rat plasma was added) for standard curves, or human/monkey CSF (basal level or over-spike QC s) Add 5µL 5M guanidine HCl (denatures & eliminates protein binding) Shake samples at room temperature for 45 minutes Add 5µL 4 H 3 PO 4 in water, mix 211 Waters Corporation 24

25 Sample Extraction: Final Method Oasis MCX µelution Plate Condition: 2 µl MeOH Equilibrate: 2 µl 4 H 3 PO 4 Amyloid β Peptide SPE Recovery Load: 15 µl diluted sample (pretreated sample: 5 µl human CSF, 5µL 5M guanidine HCl, 5µL 4 H 3 PO 4 in water) Wash 1: 2 µl 4 H 3 PO 4 Wash 2: 2 µl 1 ACN in water (by volume) Elute: 2 X 25 µl 75/1/15 ACN/conc. NH 4 OH/water (by volume) Dilute: 25 µl water Inject: 1 µl 211 Waters Corporation 25

26 Key Attributes for High Recovery Extraction Wash with no more than 1 ACN Removes polar interferences Does not impact recovery of 1-38 (earliest eluting/least hydrophobic of these 3 amyloid β peptides) Elute with no less than 75 ACN Provides required elutropic strength for 1-4 and 1-42 (the more hydrophobic of these 3 peptides) Elute with no less than 1 NH 4 OH Provides required solubility to fully elute 1-42 (least soluble/most hydrophobic of these 3 peptides) Addition of rat plasma to artificial CSF to eliminate nonspecific binding increases SPE recovery in this matrix from 6 on average to >9 Guanidine HCl denaturation Improves reproducibility of method; eliminates aggregation and protein binding 211 Waters Corporation 26

27 Outline Introduction Mass Spectrometry Liquid Chromatography Choice of Standard Curve Matrix Solid Phase Extraction Partial Validation 211 Waters Corporation 27

28 Summary of Samples Extracted Standard curves prepared in artificial CSF + 5 Rat plasma Calibration points are.25,.5,.1,.25,.35,.5,.75, 1, 5, 7.5, and 1 ng/ml Basal levels in human CSF 3 individual sources pooled human CSF samples, 1 source cynomalgous monkey 6 replicates from each source QC samples Prepared in each of the human CSF matrices 7 concentrations:.4,.75,.15,.2,.8, 2 and 6 ng/ml over-spike N=3 for each concentration, from each source of CSF 211 Waters Corporation 28

29 Representative Chromatogram: Basal Levels of Amyloid β 1-38 in Human and Monkey CSF hcsf_pool1_basal_421_ MRM of 6 Channels ES > 1.3 (a beta 1-38) 2.31e4 Human CSF, pooled sample hcsf_pool2_basal_421_1 MRM of 6 Channels ES > 1.3 (a beta 1-38) e4 Human CSF, pooled sample hcsf_pool3_basal_421_1 MRM of 6 Channels ES > 1.3 (a beta 1-38) 1.33e cyno_csf_pool_basal_421_1 MRM of 6 Channels ES > 1.3 (a beta 1-38) 2.43e4 Human CSF, pooled sample 3 Monkey CSF, pooled sample Time 211 Waters Corporation 29

30 Representative Chromatogram: Basal Levels of Amyloid β 1-4 in Human and Monkey CSF hcsf_pool1_basal_421_1 Human CSF, pooled sample MRM of 6 Channels ES+ 183 > (a Beta 1-4) 1.16e hcsf_pool2_basal_421_1 MRM of 6 Channels ES > (a Beta 1-4) 1.47e5 Human CSF, pooled sample 2 hcsf_pool3_basal_421_1 MRM of 6 Channels ES > (a Beta 1-4) 8.91e4 Human CSF, pooled sample 3 cyno_csf_pool_basal_421_1 MRM of 6 Channels ES > (a Beta 1-4) 1.37e5 Monkey CSF, pooled sample Time 211 Waters Corporation 3

31 Representative Chromatogram: Basal Levels of Amyloid β 1-42 in Human and Monkey CSF hcsf_pool1_basal_421_1 Human CSF, pooled sample MRM of 6 Channels ES > (a Beta 1-42) 6.61e hcsf_pool2_basal_421_1 MRM of 6 Channels ES > (a Beta 1-42) 8.77e3 Human CSF, pooled sample hcsf_pool3_basal_421_1 MRM of 6 Channels ES > (a Beta 1-42) 4.87e3 Human CSF, pooled sample cyno_csf_pool_basal_421_1 MRM of 6 Channels ES > (a Beta 1-42) 8.13e3 Monkey CSF, pooled sample Time 211 Waters Corporation 31

32 Baseline Levels of Amyloid β in Pooled Human and Monkey CSF Amyloid Beta 1-38 Human CSF Pool 1 ng/ml Human CSF Pool 3 ng/ml Human CSF Replicate # Pool 2 ng/ml Mean Std. Deviation Cyno CSF Pool 1 ng/ml CV Amyloid Beta 1-42 Human CSF Pool 1 ng/ml Human CSF Pool 3 ng/ml Human CSF Replicate # Pool 2 ng/ml Mean Std. Deviation Cyno CSF Pool 1 ng/ml CV Amyloid Beta 1-4 Human CSF Pool 1 ng/ml Human CSF Pool 3 ng/ml Human CSF Replicate # Pool 2 ng/ml Mean Std. Deviation CV Cyno CSF Pool 1 ng/ml 211 Waters Corporation 32

33 Representative Standard Curve: Amyloid β 1-42 Compound name: Amyloid Beta 1-42 Correlation coefficient: r = , r^2 =.9976 Calibration curve: * x Response type: Internal Std ( Ref 2 ), Area * ( IS Conc. / IS Area ) Curve type: Linear, Origin: Exclude, Weighting: 1/x, Axis trans: None Res sidual Conc Response 5.. Conc Waters Corporation 33

34 Representative Standard Curve: Amyloid β 1-42 Name Type Std. Conc RT Area IS Area Response Conc. Dev blank artificial CSF pg/ml artificial CSF Standard pg/ml artificial CSF Standard pg/ml artificial CSF Standard pg/ml artificial CSF Standard pg/ml artificial CSF Standard pg/ml artificial CSF Standard ng/ml artificial CSF Standard ng/ml artificial CSF Standard ng/ml artificial CSF Standard ng/ml artificial CSF Standard Waters Corporation 34

35 Representative Basal Levels and QC s in Human CSF: Amyloid β 1-42 Name Type Std. Conc RT Area IS Area Response Conc. Dev Human CSF 1 Basal Human CSF 1 Basal Human CSF 2 Basal Human CSF 2 Basal human CSF 1 QC 4 pg/ml QC human CSF 2 QC 4 pg/ml QC human CSF 1 QC 75 pg/ml QC human CSF 2 QC 75 pg/ml QC human CSF 1 QC 15 pg/ml QC human CSF 2 QC 15 pg/ml QC human CSF 1 QC 2 pg/ml QC human CSF 2 QC 2 pg/ml QC human CSF 1 QC 8 pg/ml QC human CSF 2 QC 8 pg/ml QC human CSF 1 QC 2 ng/ml QC human CSF 2 QC 2 ng/ml QC human CSF 1 QC 6 ng/ml QC human CSF 2 QC 6 ng/ml QC Waters Corporation 35

36 Average Deviation Values for all Overspike QC Samples QC.4 ng/ml QC.75 ng/ml QC.15 ng/ml QC.2 ng/ml QC.8ng/mL QC 2 ng/ml QC 6 ng/ml Amyloid β 1-38 Human CSF 1 and Amyloid β 1-4 Human CSF 1 and Amyloid β 1-42 Human CSF 1 and Waters Corporation 36

37 Conclusions Single flexible LC/MS/MS platform developed for simultaneous quantification of multiple amyloid peptides Highly selective sample preparation based on mixed-mode SPE Improved MS selectivity using positive ion mode and sequence ion fragments High sensitivity using new MS platform Highly reproducible, accurate, and precise Sample pretreatment and choice of SPE and LC solutions eliminate handling (NSB, losses, etc.) problems Quantitation of amyloid beta peptides Aβ(1-38), Aβ(1-4), and Aβ(1-42) in human cerebrospinal fluid by ultra-performance liquid chromatography-tandem mass spectrometry. Authors: Mary E Lame, Erin E Chambers, Matthew Blatnik Analytical biochemistry. 8/211; 419(2): Waters Corporation 37