Webinar Q&A transcription and follow-up: ‘Targeted protein quantitation by MS: strategies to achieve increased selectivity’
Thank you everyone who attended the live webinar: ‘Targeted protein quantitation by MS: strategies to achieve increased selectivity’. Below is a transcription of the Q&A session held during the webinar, as well as responses to the questions posed during the live event that we did not have time to answer. We hope this is a useful resource and thank our webinar attendees and our speaker, Mark Jairaj (UCB Celltech), for their time.
1. Were these techniques easy to set-up and can you see these techniques fitting into routine use?
These techniques are quite straight forward to set-up, as outlined in the presentation. You don’t need any specific reagents and also it takes about 2–3 days to create the methods. They are already in routine use.
2. To analyze the MRM3 data, does analyst provide the capability to sum several granddaughter ions?
It does, that’s shown also in the presentation. The final method that was used summed two of the secondary product ions. The advantage of this is you again decrease the noise and improve the signal to noise ratio.
3. Do you think the use of other techniques such as MRM3/SelexION could potentially allow the simplification of sample preparation or reduce the amount of LC development needed, as you can be more selective?
Yes, certainly in reality sample preparation for the analysis I have just presented is already as simple and straight-forward as it can be. All the samples shown in the presentation where just crude digests without any other sample preparation.
For the method development you have to be a little bit careful, as just because you can’t see something it doesn’t mean it can’t negatively impact your analysis. So you have to be aware there is a massive background that is basically invisible in the MRM, MRM3 or DMS traces and you have to do your routine checks for any matrix interference.
4. Have you tried SPE to concentrate the samples?
Not yet, you need to be careful with nonspecific binding and losses, but of course if you can’t run the advantages in terms of sensitivity gained and much cleaner samples. This makes this a promising thing to try but we haven’t investigated that yet.
5. What chromatography do you use for these analyses?
So typically it’s normal flow chromatography run on very short reversed phase columns. The analysis is run on C18 monoliths with pretty high flow rates 0.5ml per min to ~1.5ml per min.
6. What range of flow rates do you use for routine biosample analysis?
We currently run 0.5 to 1.5 ml per min, we are investigating lower flow rates but that’s not established routinely yet.
7. Do you need pure protein reference standard to optimize the mass spec parameters?
You don’t, that’s an ideal scenario to have either the pure protein or peptide as it makes tuning so much easier, but in some cases, for example for some of the biomarkers, you aren’t in that luxurious position so you have to optimize out of the digested biological matrix. It’s not straight forward to do that but you have software like MRMPilot or Skyline to help you with that task.
8. ELISA results are 2 fold higher compared with MRM, what may cause this result?
I have not shown ELISA data. If you find a discrepancy like that it may be due to antidrug antibodies. Although this ELISA data is going the other way so I am not quite certain what could cause that. You have to be a little careful as those two assays measure different things so there are multiple reasons the results could be different.
9. What are the selection criteria for the surrogate peptides?
You have to avoid peptides that are too short, anything shorter than 6-8 amino acids. That’s basically because if they’re shorter than that, you are unlikely to find peptides that are unique in your complex matrix.
You do want to avoid peptides with side chains that have known post-translation modifications such as phosphorylation. You want to select peptides based on their attributes, so peptides containing methionine for example because they can be oxidized. Cysteine’s can become carbamidomethylated or oxidated. Aspartic acid glycine pairs aren’t great as they can undergo deamidation, or N- terminal glutamines because you can form pyroglutamic acid residues during the incubation.
10. Will there be cases when MS3 decrease sensitivity as compared with MRM?
It depends on the analytes and on your overall method setup. The sensitivity is analyte specific, if you just go on an ion count basis, it is typically lower than MRM as you are fragmenting, so the ion counts go lower, but I have shown you can compensate for that by Q0 trapping. But of course absolute sensitivity doesn’t matter, it’s the signal to noise that matters and that really depends on your method.
11. What can we do to ensure the digestion to form peptides is complete?
You are never quite certain if the digest has gone to completion, you try to correct for that by including either internal standard proteins or elongated peptides, which contain a cleavage site, but you’re never 100% certain that the digest has gone to completion. However, you can ensure the overall digestion is reproducible, that’s what we have always done in the methods we have shown. Personally I think that digestion isn’t far from complete because you wouldn’t get the reproducibility that we have seen when not correcting using internal standard, but that data is not shown.
12. Is it long or difficult to optimize parameters for ion mobility?
No, once you’ve made the swap and the instrument is stable. It’s a fairly straightforward process; the difficulty with the MS is that you ideally have the peptide to infuse. It’s not impossible but it is quite complex to infuse; the digest, the neat digest and either the digested protein or the plasma. That’s because there are so many isobaric compounds that you need to separate out, which really make it difficult to do this for the MS. But overall the methods described are straightforward.
13. Have you tried HRMS/MS?
We’ve not really employed it. The high res example I’ve given was on the full scan analysis, and that’s something we need to try.
14. The digestion depends on PTMs, do we need protein standards with all the same modification?
I am not quite certain what the first part of the question means. The digestion doesn’t only depend on the post-translation modifications and, as I’ve outlined earlier, we do try to avoid peptides which contain PTM simply because it makes confirmation of the full protein more difficult.
15. Do you plan to try to use nanoLC?
We’re always chasing sensitivity, of course, and decreasing the flow rate is an obvious thing to do. We are investigating that, but I am not sure if the instruments currently available are robust enough to allow us to go down to a nanoflow range. The best sizes that we use are typically quite large as I have shown in the presentation and both for throughput reasons and robustness reasons I am not sure if nano is currently up there.
16. Using M3 and DMS do you think we can completely remove the depletion of immunoenrichment for plasma proteins quantitation?
As I have shown M3 and DMS certainly give you two ways to improve sensitivity and especially selectivity for these samples. Depletion is something I’ve shied away from. Immunoprecipitation has the added advantage of giving you a massive boost in sensitivity so if you are really after low abundant species I think that’s the approach I’d take if you have the reagents for that. If you’re after fast method development, then you don’t need the sensitivity, then yes MRM.
17. Have you tried quantitation of a QTOF?
We have not, we evaluated the instruments a couple of years ago for purchase but found that the dynamic range wasn’t quite as good as what we needed and hence we decided to invest in Qtraps.
18. Have you thought about microLC instead of nanoLC, which can be robust enough while delivering you a lower flow rate?
Yes that’s something that we are establishing currently or have established, you gain sensitivity of course when you shrink the column in use – squeeze the peaks if you like. The other thing you can do is look into multidimensional chromatography, so using the normal flow LC to trap and backflush and elute onto the microflow analytical column and that’s certainly something that has promise.
19. Any comments on the sensitivity difference using the TRAP function between 4000 vs 5500 or 6500?
I can’t answer the question for the 5000, we have both the 5500 and 6500 QTRAPS in the lab, the sensitivity increase is analyte dependent but it is a sum in the range of 5–10 fold from the 5500 to the 6500 QTRAP.
20. Do you recommend to always use MRM3 prior to MRM in development, or to use MRM3 only if selectivity is insufficient?
Yes it’s the latter, the first port of call is MRM – it is much easier to develop, but also if you have sufficient selectivity it is bound to be a little more sensitive. With MRM3 you need a MRM method anyway so you start with that as an initial method. So always start with MRM then go onto MRM3.
21. Have you done a direct comparison of MS and LBA for a given protein that demonstrates greater specificity for the MS method?
We have done a direct comparison of MS and LBA, it’s difficult to answer the specificity for the MS method because, as outlined in the presentation, they do measure different things.
Both assays were characterized to the same extent. LBA was slightly more sensitive, so you’d think that the background was lower, but the results were completely comparable for that particular example in the first instance. Later on in the in vivo time we did observe differences, which were possibly explained by the presence of antidrug antibodies.
22. Can DMS cell be installed on SCIEX 5500 triple quadrupole?
I am not sure, ask a SCIEX representative.
1. Do you see any need for improvement of LC separation to achieve better selectivity in bioanalysis?
Yes, improved LC separation is always an advantage and it is worth investigating the impact of different column chemistries on a given separation.
2. Is optimization of parameters (SV or CV) automated?
Yes.