Poster Presentation 28th Annual Lorne Proteomics Symposium 2023

Single cell in multiple cell types using a CellenOne and dia-PASEF combined with PaSER identifies thousands of proteins at scale (#164)

Chris Adams 1 , Christoph Krisp 2 , Anjali Seth 3 , David Hartlmayr 3 , Jonathan Krieger 4 , Tharan Srikumar 4 , Adam Rainczuk 5 , Pierre-Olivier Schmit 2 , Guilhem Tourniaire 3 , Markus Lubeck 2
  1. Bruker Daltonics, San Jose, CA, USA
  2. Bruker Daltonik GmbH, Bremen, Germany
  3. Cellenion, Lyon, France
  4. Bruker Ltd., Milton, ON, Canada
  5. Bruker Pty Ltd, Preston, VIC, Australia

Enhancements in trapped ion mobility spectrometry (TIMS) has the potential to make important contributions to the Single Cell Proteomics (SCP) field.  Paired with automated single cell sorting and sample preparation using the cellenONE platform, the timsTOF SCP allows for fast and sensitive proteome analyses at the single cell level. Data independent acquisition (DIA) mode data files using deep learning with neuronal networks (e.g., DIA-NN), can lead to further improvements in detectability and quantifiability of proteins from minimal input samples such as single cells.

Methods 

HEK 293 and HeLa cells were each sorted into groups of 20, 10, 5, or 1 cell using the cellenONE platform (Cellenion).  Cell lysis, and protein digestion was also performed using the cellenONE. Tryptic peptide injection from low protein binding autosampler vials, and the label-free proteoCHIP.  Injections onto a 25 cm x 75 µm Aurora C18 column (IonOpticks) using a nanoElute 2 (Bruker) and 30 minute gradient, were eluted into a timsTOF SCP instrument (Bruker).  Data was acquired in dia-PASEF mode (DIA - parallel accumulation serial fragmentation), with data analysis using TIMS-DIA-NN.  A spectral library generated by DDA from a deeply fractioned human cell line containing 573,610 precursors from 13,679 proteins without match-between-runs was used.

Preliminary Data 

An automated and seamless sample workflow where single cells can be transferred to a proteoCHIP using the CellenOne was tested. Sample processing is performed on the proteoCHIP, allowing transfer directly to the LC autosampler (nanoElute 2). We tested HEK293 and HeLa cells using the CellenOne/proteoCHIP and timsTOF SCP/dia-PASEF workflow. Triplicate samples, ranging down to a single cell of both cell types, were sorted, processed and analyzed. Using a 30 minute gradient, data was acquired by dia-PASEF and analyzed by PaSER (Bruker). For HEK293 cells the single cell sorted channel corresponded to 837 protein groups form 2877 precursors identified. Not surprisingly, for both cell types more cells corresponded to a linear increase in the number of proteins identified where 20 HEK93 cells identified 5,001 protein groups from 26,000 precursors. PCA analysis of HEK293 and HeLa cell types at the single cell level showed clustering and cell type discrimination. Finally, comparing known protein expression profiles of CDK1, CDK2, EEF1A1 and EEF1A2 in the two cell types examined showed preferential expression that is expected for HeLa cells.

Novel Aspect 

A start to finish single cell proteomics workflow, including cell sorting, processing, acquisition and analysis.