Ultimate 3000 (Dionex) coupled to an HCT-Ultra Ion Trap MS (Bruker) in use by Andrew Porter at Northumbria University
It’s now a year and a half since I introduced Andrew Porter and his PhD project that we are sponsoring at Northumbria University. Looking back, one of the biggest challenges we faced at the start was optimising the LC-MS system for label-free, quantitative analysis of complex mixtures.
The results of our efforts have now been turned into an application note. In it, we discuss how to get the best detection and quantification of parent peptide ions for a quantify-then-identify approach, and why this is important.
I hope that by sharing it we can help others struggling with the same challenges, so they can learn from our mistakes!
Until the project started, the LC-MS system was used exclusively to identify simple mixtures. Often, these were single, over-expressed recombinant proteins that were relatively easy to identify. Consequently, the first thing we chose to do was benchmark performance for what we needed. As I revealed last year, our first efforts did not go well!
Optimising quantification using a low resolution ion trap instrument
From this point, Andrew took on a steep learning curve, perfecting his LC-MS skills and using Progenesis LC-MS to optimise the data generated for analysis. The Bruker HCT-Ultra MS can be operated in different modes and scan rates optimised for quantification and/or identification of peptides and proteins. We compared analysis of data from two different scan modes — Ultra versus Standard (Enhanced) — and the results are shown here:
Comparing detection of features in MS spectra between Ultra scan mode vs. Standard (Enhanced) scan mode. (A) Ion intensity map of peptide elution over a whole LC-MS run in each mode with relative ion intensity represented by the grey scale. (B) Zoomed in view of the same high abundance peptide ion quantified in each mode. (C) Zoomed in view of the same low abundance peptide ion quantified in each mode.
In this case, the slower scan rate of Standard (Enhanced) mode resolved more features in MS spectra and allowed a greater number of fragmentation ion data to be correctly associated with these features. This improved the number of peptides we could use to confidently quantify and identify proteins of interest.
Optimising your LC separation
As well as identifying the optimal MS and MS/MS settings for our particular set-up, Andrew also used the ion intensity maps to help optimise the LC gradient. A variety of “Buffer B” (90% acetonitrile) gradients were run with standard peptide mix, specific column and flow rates. Observing the gradient profiles on an ion intensity map allowed us to:
- Select the best peptide elution profile over 120mins
- Refine the time needed for washing and column equilibration between runs
- Measure reproducibility of chromatography between runs
Maintaining your LC-MS performance
One thing that wasn’t covered by the application note, but that’s of interest for anyone doing LC-MS for proteomics, was how ion intensity maps were used to reveal carry-over between runs and spray instability within runs. By observing the frequency of blockages, Andrew has developed a routine maintenance protocol to eliminate the inconvenient blockages and avoid carry-over between experiments.
Get the application note
If you’re working with a low-resolution ion trap, I hope you find the application note useful.
And, of course, I look forward to any comments from your own experience of optimising LC-MS for label-free quantification.