A possible solution may be using ZIC-HILIC pipette tips or small diameter HILIC column (if you have an offline microHPLC) to fractionate your peptides into several fractions. Boersema et all have shown very good orthogonality to subsequent RP separations (http://www.ncbi.nlm.nih.gov/pubmed/21443167). Alternatively, High-pH RP (C18) (as described by Gilar et al. (http://www.ncbi.nlm.nih.gov/pubmed/16224963) can be used as first offline dimension, but be careful that your C18 material is compatible with pH 10.

Good luck!

The most common approaches for the first dimension of MudPIT peptide separation are using HPLC with a fraction collector -- either strong cation exchange column (SCX) with a salt gradient in the mobile phase, or a reverse-phase column with a high-pH mobile phase gradient. Either way, plan to collect lots of fractions at first, and you can estimate which fractions to combine to get an even mass of peptides per fraction just based on the UV-vis chromatogram, and clean up the fractions using reverse phase solid phase extraction before running them on LC-MS/MS. You should be able to find lots of literature on either of these approaches (SCX vs high pH RP), e.g. columns and gradients that work well. Any university proteomics facility will do this on a regular basis, and you may ask them for advice as well, depending on what equipment you have available. Five micrograms of total protein is a pretty small sample, but you should be able to work with it.

Stefan,

Thanks for the prompt reply. I use a SCX column as the additional mode of separation in my standard MudPIT experiment (when mgs of sample are provided). I had not thought of ZIC-HILIC. May be worth checking out. I don't have a microLC at present though.

The use of small SCX cartidges from Optimize Technologies together with a syringe pump make a useful alternative to micro-LC. The 40u particle sizes make for low back pressure. Using 25% acteonitrile, 10mM KH2PO4 pH3.0 and stepping KCl from 0 - 1000mM (40, 60, 80, 100, 150, 200, etc) gives a nice sample cleanup and some separation to reduce complexity.

I've done something similar to what John Haley mentioned above using Applied Biosciences hand cartridge (used to be cataloged with "ICAT Accessories", I think it's still there, part number 4326687) with accompanying SCX cartridge insert, switching the elluent KCl concentration in a hand syringe, though usually only for simple samples with 3 or 4 fractions. Depending on the sample complexity, and the research question, 3 or 4 fractions may be all you need, making that a good option.

Besides previous procedures regarding HILIC, SCX and basic RP I might suggest off-gel electrophoresis or SDS-free PAGE (Journal of Proteome Research 2008, 7, 2427–2434). Both of them works very fine. Off-gel electrophoresis is may be more straightforward for direct sample analysis via LC-MS, meanwhile SDS-free PAGE needs peptide diffusion into solution and desalting before LC-MS. Every procedure has pros and cons. Cheers!!!

If you had access, liquid pahse IEF (OFF-GEL) electrosphoresis is an excellent high resolution technique for a first dimension. But any high yield technique (orthognal) that reduces the complexity of your sample, even if it isn't high resolution (eg pipette tip methods) will greatly increase your coverage over direct injection.

In my opinion, gel-based approach (in-gel or OFF-gel) has the advantage of reproducibility and providing extra information for data validation (peptide pI). But please note that when compared with SCX, IEF approach requires higher peptide load for protein ID (Slebos et al., JPR, 2008, 7, 5286). If you are working with 5 ug of sample, it may have a big impact on your experiment.

Great suggestions and perspectives. Yes, this is the challenge (extremely dilute samples with next to nothing to work with). Too often we push back on our colleagues to just give us more material but that is not always feasible for variety of reasons. We could concentrate and go directly for SDS-gel separation first and do in-gel digest. It's not a bad option but there's the dynamic range issue and I'm interested in pushing our MudPIT approach to get a true profiling experiment. A few things I would try to do: minimize the transfers, tubing, handling, and steps as much as possible to avoid loss and contamination (disposable cartridge is attactive); keep buffers mass spec friendly, and keep volumes as low as possible. With regards to fractionation, 4 or 5 fractions from SEX or ZIC-HILIC may not be too bad. After all, we've got to balance reducing the complexity with loss of sample. But, are tip or u-spin methods the way to go if you don't have microfractionation nanoLC? Can these be reproducible? In the community, I'm not sure if there's a preference for one type of sorbent over another.

Both tip or u-spin method belongs to the SPE category, and SPE is not good in fractionation because of the large particle size. You will have lots of diffusion and your peptide may appear in two or more fractions, i.e. inefficient fractionation. It is not a common approach, but I think still works.

It is a good point to minimize transfer and handling. In that case, may I suggest you to try using single shot LC-MS/MS for protein ID. You may refer to Matthias Mann paper published in MCP for more details. Of course we will not be able to achieve the same data quality (near complete yeast proteome coverage in one shot), but I think it is more applicable on your experiment than 2D-LC. After all, it is the most minimalistic approach available.