HyperLOPIT for the Mouse Stem Cell Proteome

When researchers want to know more about pluripotency and the push towards stem cell differentiation, they need to know more than just which genes are firing up. Christoforou et al. (2016) suggest that they also need to know what proteins are on the go, and furthermore, where they are spatially located within the cell.¹  

The researchers propose an upgrade of a technique used to determine spatial proteomics: hyperLOPIT, which stands for a high-throughput mass spectrometric analysis of the products of localization of organelle proteins by isotope tagging (LOPIT). Using a combination of density ultracentrifugation, sample multiplexing and a liquid chromatography–mass spectrometry (LC-MS) proteomics approach, the new workflow can identify proteins and pinpoint location with organelles or sub-organelle fractions. 

The research team used the pluripotent mouse embryonic stem cell line E14TG2A for the experiments. First, they lysed the cells using a detergent-free system that would preserve organelle structure with minimal disruption to contents. They then subjected the lysate to extensive fractionation, using density gradient ultracentrifugation. Since each membrane and organelle group within a cell displays distinct enrichment patterns, this step separated them into fractions that could be prepared individually for subsequent MS evaluation. 

Following fractionation, the researchers chose 10 fractions for MS analysis. First, they digested the fractionated lysates using trypsin and then labeled the proteolytic digests using isobaric tandem mass tagging (TMT) techniques (Thermo Scientific). Using the most recent TMT technology available, this meant that the digests could be multiplexed to a 10-plex analytical preparation, thus maximizing data forthcoming from the cell studies. 

Christoforou et al. initially focused on optimizing the LC-MS analytical workflow, using a Proxeon Easy-nLC 1000 system in combination with an Orbitrap Fusion Tribrid mass spectrometer (both Thermo Scientific). In order to collect as many peptide identities as possible, the team looked to synchronous precursor selection (SPS) with MS3 instrument operation. They found that using an SPS approach balanced the gains in quantitation brought by MS3 operation with the sensitivity required for peptide identification. Furthermore, by increasing the frequency notches in the isolation waveforms, they could enhance signals from TMT reporter ions and thus improve organelle resolution in addition to boosting the accuracy of the quantitation data gathered. Although the number of peptide spectral matches decreased from 137,912 to 61,090 with SPS-MS3, the team considered that the approach brought benefits over MS2 data acquisition, since they achieved an acceptable number of quantifiable protein groups (7,114 for MS2 vs. 5,489 for SPS-MS3). 

Once the team optimized the workflow and ran the experimental replicates, they analyzed the data using Proteome Discover software v1.4 for analysis (Thermo Scientific). They explored manual localization of protein identities against a constructed software algorithm for determining spatial identity. From the spectral data they found that fewer than 5% of proteins were assigned to contradictory locations within the cell. Overall, the workflow and subsequent analysis located 2,855 proteins with 14 organelles and sub-organelles, giving new spatial data for around 350 proteins. This represented more than 50% of the proteins identified during the SPS-MS3 run. Of the remainder, Christoforou et al. found that these located to cytoskeletal elements or multiple compartments, or were in transit. 

In conclusion, the authors believe strongly that the workflow and methodology presented are valuable tools for investigating organelle structure, protein complexes and functional networks within cells. Furthermore, they suggest that hyperLOPIT is also useful for determining protein isoform localization and looking at other post-transcriptional events important as regulators of differentiation in pluripotent stem cells. 

 

The software behind hyperLOPIT data analysis is available as an Open Source resource for visualizing and annotating spatial proteomics data.  

 

Reference 

1. Christoforou, A., et al. (2016) “A draft map of the mouse pluripotent stem cell spatial proteome,” Nature Communications, 7(8992), doi: 10.1038/ncomms9992.