2007 Cpsa - Poster - Automated Bioanalytical Sample Preparation

An Automated Dynamic BioAnalytical Sample Preparation Routine for Integration with Liquid Handlers 1

2

1

1

Charles E. Taylor , Qiner Yang , John P. Walsh and Robyn A. Rourick 1 2 Kalypsys, Inc. Pharmaceutical Sciences Department and Informatics 10420 Wateridge Circle, San Diego, CA 92121

•

500

250

20

250

375

625

375

10

250

250

500

250

4

250

375

625

625

Table 1: Serial dilution used in the construction of the standard curves in each plasma type.

Curves are to exceed R values of 0.99

LLOQ for every curve is not to exceed CS20 after data exclusion •

Legend Liquid Handler Prepared Data Compilation

Replicates need to be consistent and inter trial tolerability CV for a given concentration should no exceed 15%

25uL of blank plasma was combined with 100uL of IS and placed in the same precipitation plate as the standard concentration samples.

•

Placed in the same precipitation plate as the standard concentration samples.

0.15

0.10

0.05

0

20

40

60

80

100µL aliquot of the internal standard was added to each of the wells in the plate to track the extraction, analysis and to act as a precipitating agent.

Precipitation Plate Preparation

3.0

0.02 0.00 -0.02

0

20

40

60

80

Mettler Toledo AX26

Liquid Handler

Tecan Freedom EVO 100 Tip Configuration

Tip Sizes Used

8 Disposable Tip Mounts (DiTi) 200 µL DiTi Tips 1000 µL DiTi Tips

Pump

Agilent HP1100 Binary Pump

Degasser Autosampler

Agilent HP1100 CTC PAL

Table 2: Equipment used in the preparation and analysis.

HPLC Conditions Flow Rate

1000µL/min

Mobile Phase

A: 0.1% Formic Acid in Water B: 0.1% Formic Acid in Acetonitrile

Gradient

0.0 – 1.0 min:

10%B to 95%B

1.0 – 1.9 min:

95%B

2.0 – 2.7 min:

10%B

Injection Volume 5 µL Column

R2

Slope

3.5

Curve

R2

Slope

1st

.9978

.000247

3.0

1st

.9971

.000285

2nd

.9987

.0005

2.5

2nd

.9969

.000304

3rd

.9992

.00046

3rd

.9994

.000264

2.0

Phenomenex Polar-RP 2.0x30mm, 4µm

Table 3: HPLC parameters used in the analysis of the compounds.

0.06 0.04 0.02

4000.0 6000.0 Analyte Conc. / IS Conc.

5.0

1.0

8000.0

0.0 0.0

1.0e4

2000.0

4000.0 6000.0 Analyte Conc. / IS Conc.

8000.0

•

50µL of each supernatant from precipitation plate transferred to a deep-well microtitre plate. 50µL Milli-Q water was added

Source Type Ionization Source Temperature

Turbospray ESI+ (positive)

4.0 3.0

5.5 kV Medium

Curtain Gas Ion Gas 1 Ion Gas 2 Entrance Potential

10 50 60 22

Scan Type

Multiple Reaction Monitoring (MRM)

Table 4: Mass spectrometer parameters used in the MS/MS detection of the analyte and internal standard.

60

80

A naly te A rea / IS A rea

0.15 0.10

0

20

40

60

80

Slope

1st

.9977

.00034

2nd

.9962

.000504

3rd

.9951

.000522

0.020 0.000 -0.020

20

40

60

80

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 6a: Expanded view of Rabbit Plasma LLOQ

.9974

.0122

2nd

.9952

.00366

3rd

.9973

.00374

4000.0 6000.0 Analyte Conc. / IS Conc.

3.5 2.0 3.0

8000.0

1.0e4

2.5 1.5 2.0

0.020

4000.0 6000.0 Analyte Conc. / IS Conc.

4.0 2.0

8000.0

40

60

80

1.0e4

Figure 9: Automated Liquid Handler Generated Calibration Curve for Cyno Monkey Plasma

3.0 1.5 2.0

R2

Curve

0.20 0.15 0.10 0.05

1st

.9983

.000363

2nd

.9982

.000298

3rd

.9980

.000287

0.0 0.5 0.0

5.0 4.0

2000.0

4000.0 6000.0 Analyte Conc. / IS Conc.

8000.0

1.0e4

Figure 10: Automated Liquid Handler Generated 0.0 2000.0 4000.0 6000.0 8000.0 1.0e4 Calibration Curve Cyno Monkey Plasma Analyte Conc.for / IS Conc.

80

.000276

2nd

.9949

.000277

3rd

.9972

.000281

0.5 0.0 0.5 0.0

2000.0

4000.0 6000.0 Analyte Conc. / IS Conc.

8000.0

8: Automated Liquid Handler Generated 2000.0 4000.0 6000.0 8000.0 1.0e4 Calibration Curve forConc. Mouse Analyte / IS Conc.Plasma

0.040

3.0

R2

Curve

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 9a: Expanded view of Cyno Monkey Plasma LLOQ

Figure 14: The analyte peaks were consistent in retention time, with the higher concentrations maintaining good peak shape (14a). The IS trends for the studies are shown in Figure 14b, with the colors corresponding to the replicate as shown in the calibration graph section. The attenuation in intensity (Figure 14b) is due to the presence of two IS batches. The crosses in Figure 14b correspond to manually prepared trials; the circles represent automated trials.

0.020

Conclusion

0.000 -0.020

20

40

60

80

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Automated Preparation 100807_Trial21_FullAuto_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000292 x + 0.00119 (r = 0.9979)�� 101507_Trial41_FullAuto_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.00031 x + 0.000469 (r = 0.9997)�� 101507_Trial47_FullAuto_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000276 x + 0.00183 (r = 0.9994)��

Curve

R2

Slope

1st

.9979

.000292

1st

.9964

.000544

3.0 2.0

2nd

.9953

.000233

2.5

2nd

.9997

.000310

3rd

.9983

.000465

2.0 1.5

3rd

.9994

.000276

1.0 0.0

2000.0

4000.0 6000.0 Analyte Conc. / IS Conc.

0.02 0.00 -0.02

8000.0

1.0e4

Figure 11: Automated Liquid Handler Generated Calibration Curve for Human Plasma

0.20 0.15 0.10 0.05

0

20

40

60

80

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 10a: Expanded view of Cyno Monkey Plasma LLOQ

0.00

•

Data points within these curves passed tolerability for inclusion using the 85% to 115% rule

•

•

1.0 1.0 0.5 0.0 0.5 0.0

2000.0

4000.0 6000.0 Analyte Conc. / IS Conc.

8000.0

1.0e4

Figure 12: Automated Liquid Handler Generated 0.0 2000.0 4000.0 6000.0 8000.0 1.0e4 Calibration Curve for /Human Plasma Analyte Conc. IS Conc.

0.25

0.04

Standard curves had greater than 6 points with weighted regression values exceeding 99%

•

1.5

0.100 0.080

Automated construction of standard curves by the liquid handler for different animal model plasmas was successful.

•

100807_Trial15_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000243 x + 0.00118 (r = 0.9910) 3.5

2.0

The lower limits of quantification (LLOQ) for these automated prep samples were very linear, proving the ability for the method to carry out a precise serial dilution in viscous medium Server based liquid handler file generator is able to create production capable files This ability for an automated platform to produce injection ready samples from raw materials positions this method as a front running technology that is ready to be utilized in the Kalypsys BioAnalytical workflow.

Acknowledgements The authors would like to thank the following individuals for their input and contributions: Nahid Yazdani, Michael Herman and Tiffany Chea. Document Reference Subject AC-MethDev-2007-015v1 BioA Tecan Sample Prep

0.060 0.040 0.020

0

20

40

60

80

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 11a: Expanded view of Human Plasma LLOQ

ies

Figure 14b

Figure 14a

Figure 8a: Expanded view of Mouse Plasma LLOQ

Slope

Stud

Time (min.)

1.0e4

0.120

-0.04 60

.9995

10000 9500 9000 8500 8000 7500 7000 6500 6000 5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0

1.0 1.0

0.06

40

1st

0

101407_Trial35_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000544 x + 0.00965 (r = 0.9964)�� 100807_Trial15_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000233 x + 0.00187 (r = 0.9953)�� 101407_Trial29_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000465 x + 0.00446 (r = 0.9983)��

Slope

1.0 1.0

0.25

20

Slope

Manual Preparation

0.09 0.08

0

R2

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 7a: Expanded view of Beagle Plasma LLOQ

101407_Trial23_FullAuto_Cyno.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000363 x + 0.000573 (r = 0.9983)�� 101507_Trial40_FullAuto_Cyno.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000298 x + 0.00228 (r = 0.9982)�� 101507_Trial46_FullAuto_Cyno.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000287 x + 0.000583 (r = 0.9980)��

2.0

2000.0

Curve

•

20

4 10 20 50 100 250 500 1000 2500 5000 7500 10000

1.5

Figure 0.0

0.040

0

Conc.

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

100807_Trial19_FullAuto_Beagle.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000276 x + 0.00138 (r = 0.9995)�� 101507_Trial39_FullAuto_Beagle.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000277 x + 0.000817 (r = 0.9949)�� 101507_Trial45_FullAuto_Beagle.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000281 x + 0.00205 (r = 0.9972)��

0.060

Automated Preparation

4.0

0.0

80

Human Plasma

Slope

1st

60

0.060

0.000

100807_Trial15_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000243 x + 0.00118 (r = 0.9910)

R2

40

-0.040 0

101407_Trial34_Manual_Cyno.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.00103 x + 0.0122 (r = 0.9974)�� 100807_Trial13_Manual_Cyno.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000231 x + 0.00366 (r = 0.9952)�� 101407_Trial28_Manual_Cyno.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000556 x + 0.00374 (r = 0.9973)��

Curve

20

Automated Preparation

1.0

2000.0

R3 95.6% 113% 91.3% 110% 103% 101% 96.1% 94.1% 97.4% 94.1% 92.8% 112%

Figure 4a: Expanded view of Rat Plasma LLOQ

0.080

Manual Preparation

0.00

-0.02

0.076

Male Cynomolgus Monkey Plasma

6.0

0.00

0

Figure 7: Automated Liquid Handler Generated Calibration Curve for Beagle Plasma

0.040

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 5a: Expanded view of Rabbit Plasma LLOQ

8.0

0.02

100807_Trial15_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000243 x + 0.00118 (r = 0.9910)

R2

R2 502% 98.5% 107% 89.1% 97.4% 108% 98.1% 98.1% 104% 103% 96.5% 121%

Table 5: The back-calculated concentrations of the standard curve plasma samples were made by entering the analyte peak area into the calibration curve, which those samples were used to construct. The tolerability of the standard curve point to be included in the curve is measured by it’s statistical difference from the theoretical concentration. Exclusion from the standard curve occurs when the point is outside of the 85% to 115% tolerability range; the LLOQ is allowed 80% to 120% tolerability. The red cells denote excluded data points. The rat model is the most pertinent at Kalypsys, which is why it is displayed here. The results confirm that the automated method of preparing rat plasma samples is consistent in producing tight standard curves with minimal point exclusion due to intolerability.

0.04

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

2.0

0.070 0.060

500

Spray Voltage Collision Gas (CAD)

40

-0.040

15µL injected according to the listed analysis method. MS Parameters

20

Curve

0.0 0.0

1.0e4

Figure 6: Automated Liquid Handler Generated Calibration Curve for Rabbit Plasma

0.20

0.00

1.0e4

-0.04

101207_Trial11b_Manual_Beagle.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.00034 x + 0.00237 (r = 0.9977)�� 101407_Trial27_Manual_Beagle.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000504 x + 0.00544 (r = 0.9962)�� 101407_Trial33_Manual_Beagle.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000522 x + 0.00987 (r = 0.9951)��

0.5 2000.0

8000.0

4: Automated Liquid Handler Generated 4000.0 6000.0 8000.0 2000.0 1.0e4 Calibration Curve forConc. Rat / ISPlasma Conc. Analyte

Manual Preparation

1.5

0.25 A naly te A rea / IS A rea

•

•

HPLC System

0.08

Figure 3a: Expanded view of Rat Plasma LLOQ

100807_Trial18_FullAuto_Rabbit.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000285 x + 0.00107 (r = 0.9971)�� 101507_Trial38_FullAuto_Rabbit.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000304 x + 0.000751 (r = 0.9969)�� 101507_Trial44_FullAuto_Rabbit.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000264 x + 0.00164 (r = 0.9994)��

1.0

Injection Plate Preparation

Analytical Balance

0.10

0

0.05

Equipment List

Figure 0.0

4000.0 6000.0 Analyte Conc. / IS Conc.

0.06

Automated Preparation

Figure 5: Automated Liquid Handler Generated Calibration Curve for Rabbit Plasma

Centrifuged at 35000rpm for 20min and

2000.0

Male Beagle Plasma

Curve

0.0 0.0

Mixed thoroughly on a multitube vortexer.

0.0 0.5 0.0

0.08

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 2a: Expanded view of Mouse Plasma LLOQ

2.0

CS0 Controls and the Double Blank Controls added to the plate

1.0e4

Figure 3: Automated Liquid Handler Generated Calibration Curve for Rat Plasma

0.04

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Analyte Area / IS Area

25µL of each spiked standard concentration was transferred to a deep well plate.

Analyte Area / IS Area

Each level was thoroughly mixed via aspiration/dispensation repetitions

4.0

8000.0

0.12

Manual Preparation 5.0

4000.0 6000.0 Analyte Conc. / IS Conc.

0.14

Male Rabbit Plasma

Serial dilution of a concentrated, spiked solution was performed in blank plasma

Plasma Preparation

•

Automated Liquid Handler Generated 4000.0 8000.0 2000.0 1.0e4 Calibration Curve for 6000.0 Mouse Plasma Analyte Conc. / IS Conc.

2000.0

0.00

100807_Trial9_Manual_Rabbit.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000247 x + 0.0011 (r = 0.9978)�� 101407_Trial26_Manual_Rabbit.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.0005 x + 0.00472 (r = 0.9987)�� 101407_Trial32_Manual_Rabbit.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.00046 x + 0.0133 (r = 0.9992)��

Plasma Standard Curve Generation

•

0.0 0.0

1.0e4

0.08

Figure 1a: Expanded view of Mouse Plasma LLOQ

25uL of blank plasma was combined with 100uL of acetonitrile

•

8000.0

-0.04

•

•

2:

0.0

4000.0 6000.0 Analyte Conc. / IS Conc.

0.06

Double Blank Controls

•

Figure

0.20

0.00

Tolerability of data points included in the curve are to be between 85% and 115% Minimum of 6 points in a calibration curve

•

Automated Liquid Handler Generated Calibration Curve for Mouse Plasma

2000.0

A naly te A rea / IS A rea

•

250

1:

0.0 0.5 0.0

Analyte Area / IS Area

•

250

Figure

1.0e4

A naly te A rea / IS A rea

•

50

CS0 Controls

8000.0

R1 97.9% 104% 101% 104% 101% 101% 101% 95.2% 94.1% 91.8% 96.3% 112%

1.502

375

4000.0 6000.0 Analyte Conc. / IS Conc.

1.485

625

2000.0

1.468

375

0.0 0.0

Rat Automatic CV AVG 100.92% 231.83% 6.96% 105.17% 7.94% 99.77% 10.65% 101.03% 2.82% 100.47% 3.91% 103.33% 2.5% 98.4% 2.16% 95.8% 5.12% 98.5% 6.14% 96.3% 2.19% 95.2% 4.52% 115%

1.452

250

Vortexed and sonicated

1.435

100

•

0.5

1.418

250

1.0 1.0

1.402

500

1.5

1.385

250

.000284

1.368

250

0.5

.9965

1.352

250

1.0

3rd

1.335

250

1.0 1.0

2.0

1.5 2.0

1.318

500

1.0

1.5

.000511

.000307

1.302

250

2.0

.9990

.9980

1.285

250

500ng/mL deuterated Analyte (Analyte-D6) in acetonitrile

3rd

2nd

1.268

500

•

3.0

2.5

1.252

375

.000276

.000498

1.235

625

0.9971

.9969

1.218

375

3rd

2nd

.000232

1.202

250

.000295

.9983

1.185

1000

.000511

0.9975

1st

1.168

250

.9986

2nd

.000471

1.152

500

3rd

2.5 1.5 2.0

.9958

Slope

R3 97.8% 104% 100% 108% 103% 101% 98.6% 97% 105% 95.3% 94.6% 96.2%

1.135

250

.000501

1st

R2

1.118

250

.9989

4.0

Curve

1.102

2500

3.0

2nd

.000299

3.5 2.0 3.0

1.085

Internal Standard (IS)

0.9977

Slope

1.068

290

1st

R2

1.052

540

.000244

100807_Trial15_Manual_Human.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000243 x + 0.00118 (r = 0.9910)

Curve

1.035

180

.9957

5.0

R2 102% 86.9% 112% 103% 97.4% 113% 103% 96.6% 96.3% 97.1% 95.7% 95.7%

1.018

360

1st

Slope

R1 93.1% 115% 120% 110% 125% 101% 107% 103% 97.1% 91.2% 92% 90.2%

1.002

5000

4.0

R2

Peak Area

240

3.5 2.0 3.0

Curve

A naly te A rea / IS A rea

600

(K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000243 x + 0.00118 (r = 0.9910) 100807_Trial15_Manual_Human.rdb 3.9

A naly te A rea / IS A rea

150

Slope

A naly te A rea / IS A rea

450

R2

Analyte Analyte Area / ISArea Area/ IS Area

7500

Liquid handler is to use file generated from the server based script generator 2

Vortexed and sonicated

•

Curve

Analyte Analyte Area / ISArea Area/ IS Area

•

Quantification is to be performed through the generation of a standard curve

5.0

4 10 20 50 100 250 500 1000 2500 5000 7500 10000

101207_Trial22_FullAuto_Rat.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000232 x + 0.00116 (r = 0.9983)�� 101507_Trial37_FullAuto_Rat.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000307 x + 0.00293 (r = 0.9980)�� 101507_Trial43_FullAuto_Rat.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000284 x + 0.00217 (r = 0.9965)��

Analyte Analyte Area / ISArea Area/ IS Area

250

Conc.

Automated Preparation

100507_Trial3_Manual_Rat.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000471 x + 0.00225 (r = 0.9958)�� 101407_Trial25_Manual_Rat.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000498 x + 0.00391 (r = 0.9969)�� 101407_Trial31_Manual_Rat.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000511 x + 0.0126 (r = 0.9990)��

A naly te A rea / IS A rea

700

100807_Trial17_FullAuto_Mouse.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000299 x + 0.00112 (r = 0.9977)�� 101507_Trial36_FullAuto_Mouse.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000295 x + 0.00198 (r = 0.9975)�� 101507_Trial42_FullAuto_Mouse.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000276 x + 0.00187 (r = 0.9971)��

A naly te A rea / IS A rea

665

100807_Trial7_Manual_Mouse.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000244 x + 0.00143 (r = 0.9957)�� 101407_Trial24_Manual_Mouse.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000501 x + 0.004 (r = 0.9989)�� 101407_Trial30_Manual_Mouse.rdb (K932961): "Linear" Regression ("1 / (x * x)" weighting): y = 0.000511 x + 0.0162 (r = 0.9986)��

Analyte Area / IS Area

35 of stock

50:50

Manual Preparation

Analyte Area / IS Area

10000

in

A naly te A rea / IS A rea

(uL)

Analyte

Automated Preparation

Analyte Analyte Area / ISArea Area/ IS Area

(uL)

200ug/mL of acetonitrile:water

A naly te A rea / IS A rea

(uL)

•

Rat Manual AVG CV 97.63% 4.56% 101.97% 13.89% 110.67% 9.1% 107% 3.37% 108.47% 13.45% 105% 6.6% 102.87% 4.08% 98.87% 3.63% 99.47% 4.83% 94.53% 3.2% 94.1% 2.02% 94.03% 3.54%

Sprague-Dawley Male Rat Plasma

Manual Preparation

Analyte Area / IS Area

(uL)

C57 Male Mouse Plasma

Stock Solution

A naly te A rea / IS A rea

•

Plasma samples for this study are to be prepared by protein precipitation and monitored with an internal standard

Vol of Prev Vol of Total Remaining Vol Blank Vol

(ng/mL)

•

•

Decision Point

Conc.

•

Objectives

Manually Prepared

Plasma

Analyte Area / IS Area

The Analytical Chemistry department at Kalypsys has developed and implemented an automated sample preparation method for its routine BioAnalytical work. The purpose of this system was to assume the work effort of protein precipitation sample preparation, reducing the FTE time required per study. Designed to process sample plates with any combination of matrices and layouts into methodical injection plates, the system is able to achieve much of the sample arrangement that a human analyst would provide. The system depends on proper plate map registration; a tool was developed to allow users to populate the database with a platemap containing animal number, compound number, dose type, matrix, model, time point and concentration for each well of a plate. A server based program crafts a liquid handler script that constructs the corresponding injection plate(s) from the sample plate(s), once the desired calibration curve is chosen on the web based user interface. The resulting script is a standalone file, not a work list, which is recognized and executed by Tecan’s Gemini software. The script generator optimizes the liquid handling by mimicking the logic a human analyst would use in injection plate construction. One representation is the ordering of samples into the densest arrangement. This allows, in some cases, multiple sample plates to be condensed into one injection plate. Another is the detection of multiple compounds in the plate, which triggers the generator to program the construction of corresponding calibration curves and CS0 samples. With the capabilities in place, the ruggedness testing and performance trials were initiated. The system met specification for standard curve construction and consistency of internal standard concentrations in multiple matrices. Tractability of preparation error is achieved by monitoring the liquid handling error log files, and making database tags of any samples that initiated a liquid handling error. Expansion efforts are targeted for the system and include: automated SPE, hemolysis tracking, liquid-liquid extraction, matrix expansion and high density plate use. These efforts look to capture a larger portion of the workflow and position this system as a multipurpose BioAnalytical platform.

Results

Analyte Analyte Area / ISArea Area/ IS Area

Experimental

Abstract

0.000

0

20

40

60

80

100 120 140 160 180 200 220 240 260 Analyte Conc. / IS Conc.

Figure 12a: Expanded view of Human Plasma LLOQ

Poster Contact

Email

Charles Taylor

[email protected]

Product Reference

Manufacturer

Freedom EVO 100

Tecan AG

Costar Deep Well Plate

Corning

Costar Shallow V-Well Plate

Corning