NEW YORK(GenomeWeb) – A team led by researchers from the Swiss Federal Institute of Technology Zurich have completed a mass spec analysis of proteome-wide changes in Mycobacterium tuberculosis as the organism shifts between dormant and active states.
The work is part of a four-year, CHF3 million ($3.2 million) initiative funded by the Swiss government that aims to use various omics and systems biology approaches to characterize multiple strains of drug-resistant tuberculosis.
The mass spec analysis was detailed in a paper published this month in Cell Host & Microbe and led by the lab of ETH Zurich researcher Ruedi Aebersold, who is heading the initiative's proteomics efforts. In the study, the researchers used Swath mass spec to estimate absolute cellular protein concentrations for 2,023 TB proteins across organism growth stages including exponential growth, hypoxia-induced dormancy, and resuscitation.
Using this data the researchers were able to observe a number of shifts in proteome composition across the organism's different life stages, Aebersold said, noting that the hope was that some of the observations could provide insights on approaches for targeting TB when the organism is dormant or just beginning to reemerge from dormancy.
"One reason TB is so difficult to treat with antibacterial drugs is because it goes into this dormant state [during which] the [conventional drug treatment] does not help," he told GenomeWeb. "Then what happens is, at a certain point it comes out of dormancy and then the person gets sick. So if one knew how [TB] could be attacked in the dormant state or as it is coming out of the dormant state, one might be able to attack the disease in a promising way."
Proteomic analyses could help by, for instance, identifying proteins highly expressed during the dormant state that could serve as good antigens for vaccines targeting dormant TB. Likewise, a better understanding of proteins and pathways that are upregulated in the dormant form of the disease could lead to potential therapeutic targets.
The Cell Host & Microbe paper is the first publication from Aebersold's lab as part of the project and serves as a demonstration to establish the usefulness of the approach before applying it to actual clinical isolates, Aebersold said.
The study's method of estimating absolute per-cell protein quantities based on mass spec signal intensities is not new, he added, noting that it has been used previously by a number of groups, including his. The novelty of their approach, he said, is that they applied it to Swath data, whereas before the approach had been primarily used with standard shotgun mass spec methods.
In shotgun mass spec, the instrument performs an initial scan of precursor ions entering the instrument and selects a sampling of those ions for fragmentation and generation of MS/MS spectra. Because instruments can't scan quickly enough to acquire all the precursors entering at a given moment, many ions — particularly low-abundance ions — are never selected for MS/MS fragmentation and so are not detected.
In DIA methods like Swath, on the other hand, the mass spec selects broad m/z windows and fragments all precursors in that window, allowing the machine to collect MS/MS spectra on all ions in a sample. While DIA typically provides less complete proteome coverage than shotgun methods (though some recent studies have suggested this is no longer the case), because it fragments the same precursors each time, it provides more reproducible data, making it well-suited to quantitative experiments looking at large numbers of samples or samples under large numbers of conditions.
In the case of the TB work, using Swath allowed the ETH Zurichresearchers to obtain good quantitative data on roughly 2,500 proteins in M. tuberculosis H37Rv cultures grown in triplicate and analyzed at six time points covering exponential growth, hypoxia-induced dormancy, and resuscitation. Of those 2,500 proteins, the researchers were able to obtain per-cell protein copy estimates for 2,023, a figure covering roughly half of the 3,990 annotated TB proteins.
Obtaining absolute protein quantities as opposed to relative quantities allowed the researchers to better investigate how different proteins moved together over the course of the organism's growth cycle, Aebersold said.
"When you do relative quantitation, you can only make a statement about one protein and how this protein behaves in light of other samples," he said. "So if you measure 10 samples you can say, this protein goes up or down across these 10 samples, but you can make no statement about any other protein in there. In absolute [quantitation] you can make statements across the proteins — you can say protein A goes up five-fold from 20 to 100 and protein B up five-fold from 200 to 1,000 copies, and so you can start to calculate relationships across proteins."
Using genome-wide metabolic models, Aebersold and his colleagues worked to link the protein data to metabolic pathways to better understand the mechanisms involved in the shift to dormancy and then resuscitation.
The study is a proof of principle, he said, that "shows that we can quantify these proteins reasonably accurately on a large scale and identify differences and interpret these differences in terms of biochemical processes."
Now, Aebersold said, he and his colleagues are looking at actual clinical samples, working with epidemiologists and clinical microbiologists who are collecting strains around the world with different clinical phenotypes. They are currently working on an initial set of 20 different strains with more to be added.
Thus far they have analyzed a set of six strains looking to see if they are able to identify phenotypic differences by proteomic analysis. Specifically, they have looked at three older TB strains found primarily in Africa and compared them to three newer strains found primarily in Asia that, Aebersold said, tend to be more aggressive than the older varieties.
"We have done a relatively small cohort study with [these] strains and measured them in triplicate to see if we can reproducibly see these patterns from clinical isolates and if we can we classify them into groups [based on their proteomes]," he said. "And that actually works quite well. It is not a clinical relevant study — it is too small. But it shows that the method works."