Hamilton LC-MS User Manual

LC-MS Sample Preparation Proteomics

High-Throughput Automation of the PreOmics iST
Technology for Proteomics LC-MS Sample Preparation

Authors: 1 Russell Golson, 1 Nils A. Kulak, 2 Piotr Soczynski, 2 Thomas Howe, 2 Andreas Essig

1

PreOmics GmbH, Am Klopferspitz 19, D-82152 Planegg/Martinsried, Germany, [e-mail]

2

Hamilton Bonaduz AG, Via Crusch 8, 7402 Bonaduz, Switzerland, [e-mail]

Introduction

Proteomics workflows are becoming increasingly important in the clinical diagnostics and biotech industries, such as in thera­peutic drug monitoring or biomarker detections. Liquid Chromatography Mass Spectrometry (LC-MS)-based assays in particu­lar offer the great advantage of measuring multiple analytes at once (and is quantitative).

Proteomics workflows have been traditionally limited by LC-MS measurement time and sample preparation throughput. While
advances in LC-MS instrument technologies and workflows have now significantly increased the number of samples that can
be processed on a weekly basis, the bottleneck has shifted to efficient, robust, and reliable high-throughput sample preparation.

In this Application Note we demonstrate for the first time a completely automated high-throughput LC-MS sample preparation
workflow, combining the Hamilton liquid handling technology with the PreOmics iST workflow.

n Save time and costs with a maximized walk away time for 96-Well
LC-MS sample preparation

n Full flexibility to process 1 to 96 samples with minimal tip usage

n Standardized workflow with high data reproducibility and process safety

Workflow Description

Automated iST workflow

Microlab® VANTAGE Liquid Handling System

Figure 1: The Hamilton

®

1.3 m

Figure 2: Automated iST workflow – In this figure, the reagents, labware and modules are described for processing 96 samples on a Hamilton liquid handling platform. The tem­perature and run time on the robot are displayed for each step. The conditions for lysis may var y, according to the sample material. After resolubilization of the digested peptides,
the 96-well plate can be directly loaded onto a LC-MS autosampler.

Deck Layout Description

Results

2

7

8

3

6

13

4

9

11

1

5

10

12

1. [ M P E ]2 (Monitored Multi-Flow Positive Pressure Evaporative
Extraction module)

2. ODTC (On-Deck Thermo Cycler)

3. Tube module (reagents)

4. DWP (Deep-Well Plate) module (collection plate)

5. MTP (Micro-Titer Plate) module (sample plate)

6. Tip Waste (2 Tracks)

7. HHS (Hamilton Heater Shaker)

8. [MPE]2 Evaporator module parking position

9. QCG (Quad CO-RE Gripper) & 60 ml trough module (buffers)

10. MTP module (PCR Comfort lid)

11. 60 ml trough module (buffers)

12. DWP module (frame for filter plate)

13. Tip carrier (50 µl and 300 µl Tips)

Figure 3: De ck layout

Application Software

Via a Graphical User Interface, an operator can define the parameters of each run, such as the digest time or the volume for
resolubilization, prior to LC-MS.

Kit Description

The PreOmics iST kit contains proprietary reagents to denature, reduce and alkylate proteins in one step, as well as the en­zymes to perform a tryptic digestion. The final peptide clean-up includes two positive-pressure 96-well plates (iST-REG-PSI
96HT (192 samples): P.O. 00108; iST-REG-PSI 96HT (384 samples): P.O. 00112).

Technology

LC-MS sample preparation assays, such as the PreOmics iST kit, often use organic and volatile liquids in their workflows. A
distinct pressure system in the Hamilton channels allows for the reliable monitoring and control of such liquids with, for example,
Anti-Droplet Control (ADC).

Four experiments were performed to assess the reproducibility and the robustness of the automated workflow (Fig. 2):

(I) A cross-contamination test was performed with

Saccharomyces cerevisiae

protein extracts, demonstrating that there is

no cross-contamination occurring during the [MPE]2 procedure or any other step in the workflow (Fig. 4).

(II) Proteins from
a single experiment. A 100% inter-day overlap of identified proteins was achieved for both

Pichia pastoris

and commercial human plasma were digested on two different days and the data was acquired in

P. pastoris

and human plasma

(Fig. 5).

(III) Aliquots of

P. pastoris

samples were also digested, following the standard manual protocol. 98% of the identified proteins

were detected in both the manual and inter-day runs (Fig. 5A).

(IV) To test a full 96-well plate automation run, aliquots from 48

P. pastoris

and 48 commercial human plasma protein extracts

were processed. The normalized protein intensities obtained demonstrated a mean correlation of 0.97 and 0.98 for

ris

and human plasma, respectively (Fig. 6).

1 5 93 7 112 6 104 8 12

A

B

C

D

E

F

G

H

A)

P. pastoris P. pastoris

C)

570

540

553 579

568547

B)

D)

Human PlasmaHuman Plasma

577

580

S. cerevisiae

Blank

Figure 4: Cross-contamination – 8 aliquots of
tracts (manually prepared, 5 µg) and 88 blank samples (only LYSE
buffer) were arranged as shown and processed according to a custom­ized procedure: 100 µl of protein extract or buf fer, respectively, and 10 µl
of DIGEST were incubated for 1.5 h in the ODTC.

The dried samples were dissolved in 20 µl LC-LOAD and 2 µl was in­jected on a UPLC, coupled with an Orbitrap Fusion Lumos (Thermo
Scientific). A 20 min gradient of 5-35% acetonitrile was applied to sepa­rate the peptides. Protein identification and visualization were performed
using Mascot (Matrix Science) and Scaffold (Proteome Software).

After filtering for 0.1% peptide/1% protein False Discovery Rate (FDR) and
3 peptides/protein, the result showed 540 to 580 proteins identified in the

S. cerevisiae

Figure 5: Inter-day reproducibility and comparison to manual processing
– (A and B) 8 aliquots of manually prepared
(50 µg each) we re processed on t wo different days, and 4 additional
samples were digested with the manual procedure. The automation was
customized as follows: 50 µl of proteins, 50 µl of DIGEST, and 10 µl of
additional RESUSPEND were incubated for 1.5 h in the ODTC.

The LC-MS measurement was performed as described for the cross-con­tamination assay with 1 µl injected on the LC-MS system. Protein identi­fication and quantification was per formed using MaxQuant 1 (5% protein
FDR; 2 peptides/protein). 1060 proteins we re identified for the automated
runs and 1043 for the manually processed samples (A) with a mean value
of 0.12 for day 1, 0.10 for day 2 and 0.07 for manual when considering the
protein intensitie s (median-normalized scaled) (B). Day 1 and 2 revealed
an intersection of 100% and 98%, as compared to the manual run (A).

(C and D) 8 samples of commercially available human plasma (each
~70 µg; Sigma-Aldrich, P9523) were processed on day 1 and 2. The
LC-MS measurement and analysis were executed as described for the

P. pastoris

for days 1 and 2, with an intersection of 100% (C) and a mean value of

0.22 for day 1 and of 0.25 for day 2 with regard to the protein intensities
(median-normalized scaled) (D). All plots were generated in R 2.

samples and no proteins detected in the blanks.

proteins, with a gradient of 35 min. 172 proteins were identified

S. cerevisiae

P. pastoris

protein extracts

P. pasto-

protein ex-

The CO-RE (Compressed O-Ring Expansion) technology integrated in the channels permits the transport to-and-from each
module on-deck without the need for an additional transport tool.

The ODTC, in combination with the proprietary Hamilton PCR Comfort lid, controls the temperature during protein digestion with
high-precision and uniformly, without the loss of liquid, due to evaporation. The [MPE]2 provides the positive-pressure functional­ity to process filter plates and the evaporator module to rapidly dry down samples, eliminating the need for a centrifuge.

Figure 6: 96 sample run – (A and B) 48 aliquots of manually extracted
teins were processed together with 48 samples of commercially available human plas­ma. For the automation run and LC-MS measurement, the same parameters were used
as for the inter-day experiments, except for a 1 to 10 dilution of the plasma samples
prior to the LC-MS measurement. The normalized protein intensities exhibited Pearson
correlation values between 0.94 and 0.98 for the

0.93 and 0.99 for the plasma samples (B). All plots were generated in R.

P. pastoris

samples (A) and between

P. pastoris

pro-

A)

P. pastoris

41

31

21

11

B) Human Plasma

41

1.0 01.0 0

31

0.980.98

0.960.96

21

0.940.94

11

Summary

2

1

42322212

2

1

42322212

The qualification experiments and results demonstrate that the PreOmics iST technology effectively runs on a Hamilton liquid
handling platform. The user is supported with a fully automated standardized and reproducible workflow, ultimately resulting in
LC-MS grade peptides in less than 4 hours of total sample processing time. Successful processing of yeast, human plasma and
cerebrospinal fluid (CSF) samples (not shown), demonstrates the applicability in clinical, biotech and research settings.

System Requirements Part Number

Microlab VANTAGE 1.3m, INSTINCT V Sof tware v1.9

(the power cord must be ordered for the specific country)

818 0 5 0 A13

Arm Channel/IPG 81800 9

8x Standard Pipetting Channels 196 0 05

2T waste block 818047

Ejector plate for 2T waste 10088368

System Dimensions

Width: 1448 mm (including left extension [MPE]2)

Height: 1360 mm (door open)

De pt h: 1010 m m

Labware Requirements Part Number

[MPE]2 with evaporator module
(mounted on base plate with ODTC)

ODTC 96 kit left VANTAGE
(mounted on base plate with [MPE]2)

96160 - 04

100 67561

Integration Kit [MPE]2 and ODTC left VANTAGE (This kit is
based on 95952-01 and 10066706. Panel for cosmetic is

10 113 023

included)

2x MultiFlex Carrier Base Plate 188039

MultiFlex Tube / Cup Module 188048

2x MultiFlex DWP Module 188042

2x MultiFlex MTP Module 188228

Shaker Carrier Base 1870 01

HHS 3.0 mm orbit flat bottom 10068482

Consumables Part Number/ Provider

iST-REG-PSI 96HT (192 samples) 00108 / PreOmics

iST-REG-PSI 96HT (384 samples) 00112 / PreOmics

50 µl Conductive Tips without Filter 235966

300 µl Conductive Tips without Filter 235902

96 Well PCR FramePlate 814302

60 ml Reagent Container 1940 51

PCR Comfor t Lid 814300

1.5 ml Eppendorf Tubes 0030123328 / Eppendorf

DWP (Collection plate) 186005837 / Waters

To find a subsidiary or distributor in your area,
please visit, www.hamiltoncompany.com/support.

Web:

www.hamiltoncompany.com/robotics

Email: [e-mail]

MFX Trough Module QCG Pos 1 0113 3 03

QCG on MFX position 96006-01

MultiFlex Module Bracket 7T 18 813 3

MultiFlex Reagent Trough Module 188404

Frame for filter plate 182712

Tip Carrier 18208 5

Citations:

1) Cox, J. and Man n, M. MaxQu ant enable s high pepti de identifi cation rate s, individu alized p.p.b.-rang e
mass ac curaci es and proteome-w ide prote in quant ificati on. Nat Bio techno l, 2008, 26, pp 1367-72.

2) R Core Team (2020). R: A la nguag e and environme nt for statistical comput ing. R Foun dation fo r
Statistical Computing, Vienna, Austria.

Acknowledgments:
Mass spectrometry measurements and data analysis were performed at the Functional Genomics
Center Zurich (ETH Zurich and University of Zurich, Switzerland). Special thanks go to Antje Dittmann,
Claudia Fortes and Paolo Nanni.

United States

[phone]

United Kingdom, Ireland
[phone]

Brazil

[phone]

China

[phone]

All trademarks are owned and/or registered by Hamilton Company in the U.S. and/or other countries.

Denmark, Norway,
Sweden, Finland

[phone]

France

[phone]

Germany, Switzerland,
Austria, Benelux

[phone]

© 2021 Hamilton Company. All rights reserved.

Lit. No. AN-2102-03 — 02/2021

Italy

[phone]

Japan

[phone]

Spain, Portugal

[phone]