Having done this test on E. coli long time ago with 3 different vendors (nano 75um vs micro flow on .5 mm ID columns), large scale discovery type proteomics studies which require global coverage and quantitation, microflow will not do. It's just not sensitive enough even if you load more, the coverage was still limited to what you can achieve with nano flow. When taking into account PTM's it's even more difficult where large starting material is required to enrich for precious low amounts of PTM with low stoichiometry (phospho, acetyl). For these purposes nano still reigns, however for targeted studies, where sample is not a limiting factor there are many benefits of using larger micro scale columns. You can make up for the lack of sensitivity with higher loading amounts. In our case we found a sweet spot with 2.1mm columns using a high flow source and high temperatures/sheath gas (depending on the source/vendor ). There is a very good paper on this you can read:

http://www.ncbi.nlm.nih.gov/pubmed/22547352

Also a shameless plug for one of my recent projects where we used high flow with 2.1 mm ID columns on a UHPLC, and were able to setup very fast 5 minute robust gradients and do scheduled SRM on up to 800 peptides, something you just can't do with nano. :)

http://www.sciencedirect.com/science/article/pii/S1096717614001141

Cheers,

-T

One can understand the tale of woes with nano-LC. The smaller the column id. the more perfection in tubing connections and cuts it requires to minimize extra-column band broadening. Unfortunately, many (non-chromatography) researchers ignore the contribution of extra-column effects in peak efficiency.

With narrower peaks and higher efficiency you get the detection sensitivity enhancement in MS. So your discussion question is still open - the choice will depend on the needs and whether the detector is mass or concentration depending detector.

Hi Farooq, thank for your answer! Efficiency is  probably t´not the best argument for nano LC either, as N is roughly the same for all column IDs (when packed with the same material and just as well packed). Capillary cuts can also be critical, and I have even come across instrument installers who havent realized how important this is to execute well. I agree about the sensitivity, but I am wondering if cap LC may be just as good when there is ample sample amounts.

I think if you are working in proteomics you have to use because there are large number of peptides need to separate with high resolution. So it will be better to use nanoLC. 

We must resarch!, becouse we are researchers, but sincery I have no idea Steven.

Hi Steven,

I believe the choice is governed  by the Mass-spectrometry rather than separation chromatography. Normal LC stream needs to be nebulized with Nitrogen or any other gas and using higher temperature, some peptides may not be so resistant to the temperature and may get dissociated. While for Nano LC flow since the flow is very less, you do not require any nebulization nor higher temperatures to Ionise the molecules, hence many proteomics researchers prefer to use it for Peptide samples.

Regards,

Saurabh

nanoSpray ionization (although that probably starts at

Well, above discussion is very much interesting regarding the question on why Nano LC (0.05-0.1 mm ID columns), the most contemporary LC.

I appreciate the point "separation chromatography" indicating by Mr. Saurabh, Yes! LC having basic separation chemistry for all the systems, whatever it is HPLC (ml/min), UPLC (ml/min), nano-LC (nl/min) or LC-MS. Lets compares reversed phase HPLC and UHPLC (Ultra high performance LC) or UPLC; In HPLC, pump pressure could be 40Mpa , on the contrary in UPLC pump pressure usually 100 Mpa. Is pressure everything? No, particle size and column length are also an important phenomena for fastest LC. In the case of UPLC, usual column length, internal dia, particle size could be 50 mm, 2.1 mm, 1.7 micron respectively, on the other-hand, in HPLC column length, internal dia, and particle siz could be 150 mm, 4 mm and 3-5 micron. I know you all know the story, but the point is why we are we choosing UPLC and nano-LC-MS. Answer is simple, to get faster results with great resolution, in addition, to avoid sensitive sample degradation time as well as more sample analyses per system per scientists.

In Sum, if your primary goal is speed then certainly (0.05-0.1 mm ID columns) but if the primary goal is separation you may consider capillary LC.

Hi Christof and Saurabh, thanks, the source itself is a big point here. It would be interesting to see any published comparisons with newer nano and regular sources. Really appreciate the comments on this.

Dear Saddam, thank you for your input! I am not sure that UPLC works better with 2.1 compared to 4.6? However, fast separations with silica and organic monolith columns can be great, and this are often easier to assemble in cap/nano bore columns. So in that sense, nano-bores perhaps have a greater flexibility for choosing fast separation column.

Dear Prof. Wilson, please find the attached file as an explanation of how the column i.d. differ from HPLC to UPLC.

Yes, I am agreeing with you, Nano LC columns are predominantly packed in fused-silica capillaries. Frits strong enough to withstand ultrahigh-pressure liquid chromatography (UHPLC) pressures are used to retain the stationary phase. The frit commonly used in standard-bore columns, a stainless steel mesh in a connection, will provide too much dead volume, thus impairing the separation efficiency. Therefore, frits in nano LC columns must be prepared inside the fused-silica capillary. Alternatives to packed columns are monolithic columns. In this type of column, a porous (silica or polymer) structure is formed throughout the column, eliminating the need for frits because the stationary phase is fixed to the column wall. These columns are predominantly used in proteomics for the analysis of extremely complex tryptic digests. 

Having done this test on E. coli long time ago with 3 different vendors (nano 75um vs micro flow on .5 mm ID columns), large scale discovery type proteomics studies which require global coverage and quantitation, microflow will not do. It's just not sensitive enough even if you load more, the coverage was still limited to what you can achieve with nano flow. When taking into account PTM's it's even more difficult where large starting material is required to enrich for precious low amounts of PTM with low stoichiometry (phospho, acetyl). For these purposes nano still reigns, however for targeted studies, where sample is not a limiting factor there are many benefits of using larger micro scale columns. You can make up for the lack of sensitivity with higher loading amounts. In our case we found a sweet spot with 2.1mm columns using a high flow source and high temperatures/sheath gas (depending on the source/vendor ). There is a very good paper on this you can read:

http://www.ncbi.nlm.nih.gov/pubmed/22547352

Also a shameless plug for one of my recent projects where we used high flow with 2.1 mm ID columns on a UHPLC, and were able to setup very fast 5 minute robust gradients and do scheduled SRM on up to 800 peptides, something you just can't do with nano. :)

http://www.sciencedirect.com/science/article/pii/S1096717614001141

Cheers,

-T

Hi Tanveer, thanks for the excellent answer, and congrats on your recent and awesome paper! I expect a Tuborg was in order when this method was completed!

Thank you all for such a great discussion and advice! I would also like to point to the discussion "Why do proteomics and metabolomics differ so much in instrumentation?" see link below, which touched on similar points.

Instrumentation is the same. Techniques are different because metabolites are small molecules that have very different characteristics, and peptides are relatively uniform in molecular structure. (i.e., 20 subunits).

Best,

marta

Very interesting discussion. We used nano for over 7 years and migrated to capillary about 3-4 months ago with the purchase of a new MS. I'm very happy with the results with the obvious advantage of capillary being much more robust than nano allowing us to run almost non-stop even with relatively dirty samples i.e. no desalting step prior MS. In terms of performance, yes need to load more but, given that the chromatography is good, only 3-4x times more than nano. We used to load 1-1.5ug on  a 75um x 150cm nano column and now load 5-7ug on a 300um x 150mm capillary column. The PWHM at those loads are around 20-30sec when doing a 120min gradient. In terms of IDs we obtain about the same (maybe 5% better with nano) but so far capillary analysis has been done on a 3um particle column whilst with nano used to use 2um particle columns. In terms of quantitation, because of the superior robustness of the capillary set-up reporducibility is much better  when doing large batches of samples. One disadvantage with capillary is that at the moment there are no columns longer than 15cm and thus can no run gradients much longer than 2hrs in order to get more IDs, an option available for nano with 25 and even 50cm columns. Another issue might be if doing 2DLC separations where the 3-4x more load per fraction required for capillary ends up with a lot of starting material required for the 1st dimension. 

May I ask what kind of capillary flow LC you have?

Hi Lukas, for our capLC systems, we use 5-20 µL/min (0.3-0.5 mm ID columns). We have recently landed on using capLC-MS for ample sample amounts, but use nanoLC systems for tiny sample amounts.

We use a Dionex nanoRSLC system and with a capillary flow selector (1.5-10ul/min flow range).

An addition to my previous comment above, we recently did a 2DLC run (high pH on 1st dimension). Loaded 25ug total protein on the 1st dimension and collected 10Fcs which gave a very nice signal on the 2nd dimension with a C18 0.3mm ID column (direct load).

Lukas, we've also been doing SWATH analysis on this capillary set-up and data looks good especially since we started doing direct load (pre-concentration is convenient i.e. no desalting but peak widths suffered a bit)

So the 2DLC was online?

Indeed, SWATH/DIA is particularly suited to be used with capillary flow and rather short gradients. That's one reason why we bought now a dedicated system for this workflow. Very interesting discussion.

Best,

Lukas

1st RP high pH dimension was done online indeed collecting fractions using the Dionex autosampler which doubles up as a fraction collector. We did vacuum dry each fraction before loading for 2nd D. You could bypass this step thou and automate the workflow if you use a smaller ID column for the 1st D. We used a 1mm but I reckon with a 0.3 or 0.5 ID column and a suitable flow selector (micro, 10-50ul/min) will be able to collect 10-15ul fractions that will be concentrated enough to get a good signal without having to dry for 2nd D (with a 20ul sample loop will be able to load the full volume of each fraction if needed). 

The 1st dimension high pH fractionation can also be taken completely offline at good automation. There is a high pH-RP fractionation kit from Pierce/TMO in spin column format, takes up to 100 microgram peptides and works well for us. We use it for building deeper libraries, as well as an alternative to e.g. SAX tips in SILAC/FASP workflows. Expensive if you have a lot of samples, however if you are working on a back-financing basis e.g. in a core facility and do the numbers then it works out.

Dear Steven,

You migt be interested in our paper published recently about conventional-flow proteomic analyses:

Article Conventional-Flow Liquid Chromatography-Mass Spectrometry fo...