DMSO enhances electrospray response boosting sensitivity of proteomic experiments?

[MSENC]

A recent dataset (PXD000254) released on ProteomeCentral by Dr. Hannes Hahne contains an interesting claim that "low percentages of dimethylsulfoxide (DMSO) in liquid chromatography solvents lead to a strong enhancement of electrospray ionization of peptides, improving the sensitivity of protein identification in bottom up proteomics by up to tenfold". Couldn't wait for the publication, I decided to dive into the data and check out the results for my own curiocity.

My first look of the MaxQuant reports from the authors can be found at the http://rpubs.com/wyu/7590. Generally, the data does show significant improvements in LCMS components, peptides and proteins in the presence of DMSO.

Your thought?
http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD000254

[tsbatth]

Yeah I heard about this as well. I wonder if it enhances the ms/ms as well, also does it dirty up the system faster etc...

[MSENC]

Yes, there are improvements in MS/MS and peptide IDs due to DMSO presence. The tables below are from our own pipeline. The "assignments, sequence, totals" are the annotated MS/MS, distinct peptides and total MS/MS of the run, respectively. The benefits are especially large at small sample loading. We'd have to wait for the publication to find out the rational for the improvements.

8-13-2013 4-17-42 PM.png

8-13-2013 4-16-41 PM.png

[isobaric]

Hi MSENC,

your comparison looks interesting! Can you tell us more about how you acquired the data?( what type of Orbitrap(XL, Velos or Elite), MS2 was acquired in LTQ or Orbitrap?, how long was the measurement to acquire each raw file(I assumed the last three raw files in your tables were acquired in 4hrs, how about the others)?

[MSENC]

Sorry this is not my data. It was released by Dr. Hannes Hahne with limited information as described in the link. I'm anxiously waiting for more experimental details myself, too.

[Craig]

Here is my analysis of this data:

[TABLE="class: grid, width: 875, align: left"]
[TR]
[TD]Dataset[/TD]
[TD]MS/MS Spectra[/TD]
[TD]Target PSMs[/TD]
[TD]Distinct Target Peptides[/TD]
[TD]Target Protein Groups[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_100pg_R1.raw[/TD]
[TD]116[/TD]
[TD]1[/TD]
[TD]1[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_100pg_R2.raw[/TD]
[TD]146[/TD]
[TD]0[/TD]
[TD]0[/TD]
[TD]0[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_100pg_R3.raw[/TD]
[TD]154[/TD]
[TD]1[/TD]
[TD]1[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_1ng_R1.raw[/TD]
[TD]322[/TD]
[TD]0[/TD]
[TD]0[/TD]
[TD]0[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_1ng_R2.raw[/TD]
[TD]453[/TD]
[TD]0[/TD]
[TD]0[/TD]
[TD]0[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_1ng_R3.raw[/TD]
[TD]272[/TD]
[TD]1[/TD]
[TD]1[/TD]
[TD]1[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_10ng_R1.raw[/TD]
[TD]2430[/TD]
[TD]36[/TD]
[TD]33[/TD]
[TD]29[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_10ng_R2.raw[/TD]
[TD]2058[/TD]
[TD]34[/TD]
[TD]32[/TD]
[TD]26[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_10ng_R3.raw[/TD]
[TD]1929[/TD]
[TD]44[/TD]
[TD]40[/TD]
[TD]28[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_100ng_R1.raw[/TD]
[TD]10194[/TD]
[TD]3470[/TD]
[TD]2928[/TD]
[TD]771[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_100ng_R2.raw[/TD]
[TD]10260[/TD]
[TD]3502[/TD]
[TD]2973[/TD]
[TD]833[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_100ng_R3.raw[/TD]
[TD]10199[/TD]
[TD]3542[/TD]
[TD]2982[/TD]
[TD]830[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_1ug_R1.raw[/TD]
[TD]16324[/TD]
[TD]9640[/TD]
[TD]7544[/TD]
[TD]1603[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_1ug_R2.raw[/TD]
[TD]16315[/TD]
[TD]9345[/TD]
[TD]7451[/TD]
[TD]1538[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_1ug_R3.raw[/TD]
[TD]16215[/TD]
[TD]9245[/TD]
[TD]7317[/TD]
[TD]1508[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_4h_1ug_R1.raw[/TD]
[TD]54421[/TD]
[TD]16825[/TD]
[TD]9294[/TD]
[TD]1768[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_4h_1ug_R2.raw[/TD]
[TD]54129[/TD]
[TD]17017[/TD]
[TD]9282[/TD]
[TD]1786[/TD]
[/TR]
[TR]
[TD]130218_noDMSO_Hela_4h_1ug_R3.raw[/TD]
[TD]55483[/TD]
[TD]18126[/TD]
[TD]9851[/TD]
[TD]1857[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_100pg_R1.raw[/TD]
[TD]1482[/TD]
[TD]6[/TD]
[TD]6[/TD]
[TD]5[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_100pg_R2.raw[/TD]
[TD]1354[/TD]
[TD]10[/TD]
[TD]10[/TD]
[TD]9[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_100pg_R3.raw[/TD]
[TD]5344[/TD]
[TD]61[/TD]
[TD]48[/TD]
[TD]26[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_1ng_R1.raw[/TD]
[TD]1399[/TD]
[TD]9[/TD]
[TD]9[/TD]
[TD]7[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_1ng_R2.raw[/TD]
[TD]1192[/TD]
[TD]11[/TD]
[TD]11[/TD]
[TD]10[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_1ng_R3.raw[/TD]
[TD]1168[/TD]
[TD]13[/TD]
[TD]13[/TD]
[TD]10[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_10ng_R1.raw[/TD]
[TD]6920[/TD]
[TD]910[/TD]
[TD]734[/TD]
[TD]282[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_10ng_R2.raw[/TD]
[TD]7394[/TD]
[TD]747[/TD]
[TD]549[/TD]
[TD]218[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_10ng_R3.raw[/TD]
[TD]7963[/TD]
[TD]1135[/TD]
[TD]974[/TD]
[TD]317[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_100ng_R1.raw[/TD]
[TD]13762[/TD]
[TD]7532[/TD]
[TD]6075[/TD]
[TD]1381[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_100ng_R2.raw[/TD]
[TD]13391[/TD]
[TD]7199[/TD]
[TD]5583[/TD]
[TD]1276[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_100ng_R3.raw[/TD]
[TD]13823[/TD]
[TD]7122[/TD]
[TD]5745[/TD]
[TD]1360[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_1ug_R1.raw[/TD]
[TD]19371[/TD]
[TD]11377[/TD]
[TD]8763[/TD]
[TD]1919[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_1ug_R2.raw[/TD]
[TD]19674[/TD]
[TD]11749[/TD]
[TD]9136[/TD]
[TD]1933[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_1ug_R3.raw[/TD]
[TD]19759[/TD]
[TD]12167[/TD]
[TD]9386[/TD]
[TD]1948[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_4h_1ug_R1.raw[/TD]
[TD]61844[/TD]
[TD]32098[/TD]
[TD]15869[/TD]
[TD]2761[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_4h_1ug_R2.raw[/TD]
[TD]59320[/TD]
[TD]27721[/TD]
[TD]13201[/TD]
[TD]2383[/TD]
[/TR]
[TR]
[TD]130221_DMSO_Hela_4h_1ug_R3.raw[/TD]
[TD]59670[/TD]
[TD]29030[/TD]
[TD]13646[/TD]
[TD]2502[/TD]
[/TR]
[TR]
[TD]20130219_Hela_1ug_before_DMSO_R1.raw[/TD]
[TD]34669[/TD]
[TD]20329[/TD]
[TD]11049[/TD]
[TD]1840[/TD]
[/TR]
[TR]
[TD]20130219_Hela_1ug_before_DMSO_R2.raw[/TD]
[TD]34094[/TD]
[TD]20223[/TD]
[TD]11053[/TD]
[TD]1876[/TD]
[/TR]
[TR]
[TD]20130219_Hela_1ug_before_DMSO_R3.raw[/TD]
[TD]34149[/TD]
[TD]20113[/TD]
[TD]10834[/TD]
[TD]1833[/TD]
[/TR]
[TR]
[TD]20130221_Hela_1ug_after_DMSO_R1.raw[/TD]
[TD]38123[/TD]
[TD]22380[/TD]
[TD]13775[/TD]
[TD]2213[/TD]
[/TR]
[TR]
[TD]20130221_Hela_1ug_after_DMSO_R2.raw[/TD]
[TD]37418[/TD]
[TD]21602[/TD]
[TD]13240[/TD]
[TD]2143[/TD]
[/TR]
[TR]
[TD]20130221_Hela_1ug_after_DMSO_R3.raw[/TD]
[TD]36427[/TD]
[TD]20753[/TD]
[TD]13082[/TD]
[TD]2149[/TD]
[/TR]
[/TABLE]


The 130221*.raw data was acquired on an Orbitrap Elite, and the 201302*.raw data was acquired on an LTQ Orbitrap Velos. The 130221*.raw data without 4h in the name is all 60-minute runs, the rest are 225 minutes.

[isobaric]

Thanks Craig. The results look exciting. Cannot wait to see the paper and test it on our systems.

[botte]

Hannes works in Bernhard Küster's group in Munich, and part of this was presented as an ASMS poster in June. The ID improvements in signal were quite significant especially for lower sample loads (<250 ng), less so for higher sample loads - more or less as you would expect I guess. Gains observed also seem to depend on the instrument (interface?) used.

AB SCIEX/Eksigent also had a poster presentation at ASMS where they split in DMSO post-column using the cHiPLC, more or less to the same end.

[Gajanan]

http://www.ncbi.nlm.nih.gov/pubmed/22610994
This article also talking on the same line which was publish in J. Am. Soc. Mass Spectrom. (2012)

[Craig]

The paper has been published online in Nature Methods: http://www.nature.com/nmeth/journal/vaop/ncurrent/full/nmeth.2610.html.

[isobaric]

the improvement of the ion chromatogram looks phenomenal after adding DMSO as shown in the Figure 1a but I wonder if this just because the column was under-loaded. I don't know how much peptide was loaded in the figure 1a but if more peptides were loaded into the column, the red ion chromatogram should look similar to the blue one. If there is enough material loaded, the ion injection time may no longer be the bottleneck for deep identification. As botte mentioned, the gain might not be that much for the higher sample load.
I guess for most human proteome measurement, getting enough protein might not be a big problem. the real challenge is to deal with the dynamic range issue. The current shotgun proteomics can easily identify thousands of proteins within a few hours of the measurement, but if you want to go deeper and identify those low-abundance proteins, you will need to spend much more time on separation to pull those low-abundance peptides apart from those high-abundance ones.

[rtool]

Hello, just came across this post on Google. Thanks very much all for posting the paper along with the comparison.

I'm trying to replicate this on our Orbi Elite, but have come across a question about the lockmass.
The Nature Methods paper mentioned a lockmass of m/z = 401.9228 (it shows 401.92272 in the shared raw files).
However, the suggested composition, C6H10O14S3, corresponds to a mass of 401.92327.

Even though it's just a 1.6ppm difference, is 401.92272 the proper lockmass we should use?
Thanks in advance for your advice.

[Craig]

Hello, just came across this post on Google. Thanks very much all for posting the paper along with the comparison.

I'm trying to replicate this on our Orbi Elite, but have come across a question about the lockmass.
The Nature Methods paper mentioned a lockmass of m/z = 401.9228 (it shows 401.92272 in the shared raw files).
However, the suggested composition, C6H10O14S3, corresponds to a mass of 401.92327.

Even though it's just a 1.6ppm difference, is 401.92272 the proper lockmass we should use?
Thanks in advance for your advice.

401.92327 is the neutral mass. Since this is positive electrospray, you need to subtract off an electron (0.00055) because one of the hydrogens is actually just a proton. That gives 401.92272, which agrees with the paper and the raw files.

[rtool]

Thanks very much, Craig!
You've really just saved me at least a day or two trying to figure this out on my own.

What a great forum!

[Chipsa]

Hi!

Does anyone have an idea regarding the use of DMSO in mobile phases with the Ultimate RSLCnano system? I'm referring to the mention regarding PEEK swelling when in contact with DMSO. It does not, however, mention any critical concentrations. Is this something to worry about? Did anyone try to contact the manufacturer regarding this issue?

cheers!

[Sergo]

Is there any data on whether having DMSO in mass specs will not affect the performance of the instruments in long term. I'm going to stay very cautious. For dirt-sensitive instruments like the Q Ex this could be especially problematic.

[jessegmeyer]

I wrote the JASMS paper mentioned above. I've run samples with 5% DMSO using a PEEK tubing split-flow setup for at least a few months without replacing the PEEK tubing or PEEK unions. With 5% DMSO, the instrument (LTQ-XL w/ ETD) seemed fine with respect to sensitivity as compared to the traditional mobile phase, requiring only cleaning of the ion-transfer tube and re-calibration about once a month. Of course, it is very important to used only the highest purity DMSO available. I think I used Sigma's "Puriss." It is also important to use a high ion-transfer tube temperature (I used 275 deg. C) to ensure effective desolvation of the high-boiling DMSO. I think that DMSO did slightly increase the column back pressure but I can't remember.