Clean up prior to analysis?

[SMC]

Dear Colleagues,

The sample: Hydrophilic small-molecule compound in blood plasma.

The task: Separation from the biological macromolecules prior to analysis - preferably using an online setup, i.e. direct injection of plasma and using e.g. precolumn(s) and switch valve.

The literature: Several approaches to this issue (including offline procedures), e.g. protein precipitation, SPE, LLE, TFC, RAM, SPS, monoliths….

Anyone experienced who could advise?


Thanks in advance and best regards,

smc

[Kostas Petritis]

How hydrophilic is it your molecule (i.e. is it retained at all at any C18 type of stationary phase)? Does it have any ionizable function on it?

[Hassel Hoff]

Of all the methods you listed, the only two online are SPE and monoliths. SPE might be a little tough to optimize with this type of clean-up. I have seen a paper by someone in GSK using monoliths for direct inject of plasma. Here's the reference:

Rapid Communications in Mass Spec.
2001; 15: 986-993
Plumb, R., Dear, G., Mallett, D., Ayrton, J.

Here they use Chromolith from Merck. The Onyx from Phenomenex will give you the same performance.

-HF

[Uwe Neue]

You may check out the following publication: C. R. Mallet, J. R. Mazzeo, U. D. Neue, Evaluation of several solid-phase extraction tandem mass spectrometry on-line configurations for high-throughput analysis of acidic and basic drugs from rat plasma. Rapid Communications in Mass Spectrometry 2001, 15, 1075-1083.

Essentially, the rat plasma samples were diluted with internal standard, the pH adjusted as needed, and then injected at high flow 4 mL/min onto a short 2.1 x 20 mm 25 micron Oasis column of the correct type for the analyte. The proteins were eluted from the cartridge under these circumstances. The analytes were subsequently eluted from the cartridges with a gradient at suitable pH, both with and without focussing on a second column. Typical run times were 3 minutes. RSD 5%.

I'll be glad to answer questions about the technique used, including expansions of the technique with mixed-mode ion-exchangers etc.

[Kostas Petritis]

Hassel,

RAM (Resticted Access Material) is also online. These materials take care of the proteins of the sample and Merck provides either a C18 or SCX version of the RAM material.

If SMC provides more information about his/her molecule it would helpful suggesting something more specific...

[SMC]

Thanks to all for your prompt replies!!

Kostas: In fact there is no molecule as yet. The idea is to perform a metabolic study, where a pharmaceutical compound (possibly having amines or carboxylic groups) is transformed into a usually more hydrophilic compound during the metabolic process, e.g. by oxidation.
Does Merck really provide a SCX material? Is it based on the ADS approach?

Hassel Hoff: I reckon apart from precipitation all of the methods can be automised (perhaps the precipitation as well…). Anyway, laziness lies in my nature , and so the plan is to place a pre-column online into my LC-system.

Uwe: Very interesting paper indeed!! I like especially the use of pH in order to “lipophiliseâ€

[Uwe Neue]

There is no turbulent flow in such columns. The real trick is to keep the residence time small enough such that large molecules won't interact with the packing. The proteins are washed out at 4 mL/min, then the elution of the analytes are is accomplished at low flow rates that are compatible with the MS detection system and/or the analytical chromatography, if it is added afterwards.

The elution of ionizable samples from the mixed-mode ion-exchangers is carried out in the same way as the method using hydrophobic packings. You load the sample at a pH where the analyte (and internal standard) is ionized and retained via ion-exchange. You wash out the proteins at high flow, then you can wash out hydrophobic components and oppositely charged contaminantswith methanol at the same pH. Then you convert back to aqueous, change the pH and elute in a gradient, either forward without an analytical column or backward with an analytical column. The elution protocol is carried out at MS compatible flow rates, while all wash steps are carried out at high flow rates.

[SMC]

So, let us see whether I got this mixed mode right:
Given two basic compounds say, Nortriptyline and Quinine with an aqueous pKa >10 and ~8.5, respectively; i.e. optimal loading at pH <6.5 (actually - pH lower to avoid elution effects during the wash with MeOH resulting from %B-based pH/pKa shifts and perhaps to facilitate the choice of buffer).

Presumably, the drugs remain at the cation X-change moieties. At high flow rate, the proteins are supposed to leave the column unadsorbed (in spite of the protonated sites the proteins might possess[b]??[/b]). Wash with buffered MeOH.

Now I start my %B-gradient (at pH >12[b]??[/b]). However, I will have a protonation problem with drugs when doing electrospray ionisation (ESI). Also, the drugs may show long retention in their unprotonated form. Which buffer should be used with the chemically resistant column at high pH[b]?[/b]

For bases, perhaps the approach (low pH elution) you have described in your recommended paper is more appropriate for LC-ESI-MS.

Thanks Uwe.


smc

[james little]

We just used protein precipitaton and injected above the pellet in the 96 well plate.

http://littledomain.com/james/files/matrix_effects.htm

Need to be keenly aware of matrix ionization, see chapter and other general information in..

H. Mei in: W. Korfmacher (Ed.), Using Mass Spectrometry for Drug Metabolism
Studies, CRC Press, Boca Raton, FL, 2005, p. 103 (Chapter 4).

[Uwe Neue]

You got the principle. We use formic acid for acidic pH and either ammonia or ammonium bicarbonate for the basic pH.

With respect to MS ionization of bases at alkaline pH: in our experience there is a decline of the signal, but the signal does not go away. I believe that this has something to do with the ionization in the gas phase. The worst decline that we have seen was about 60% compared to acidic conditions. This is much better than ion-supression due to interferences, and most importantly, it is reproducible. However, if you want to avoid this completely, you can use additional steps to do the elution from the main column under acidic conditions.

I have to disagree with James Little: in all our experiments with all comparisons of different sample preparation techniques, we have found inferior results due to ion suppression when using protein precipitation. Of course, more sophisticated cleanup schemes, either off-line or on-line, are a bit more work, but they deliver good results. I actually personally prefer off-line techniques, but we have developed successful generic on-line techniques as well.

[SMC]

Uwe,

Uwe wrote:
The worst decline that we have seen was about 60% compared to acidic conditions. This is much better than ion-supression due to interferences, and most importantly, it is reproducible.

Yes – I have seen protonation of bases at alkaline conditions, but just as you said: with reduced sensitivity (signal suppression due to presence of ammonia!?). The problem arises for very small concentrations where sensitivity is the decisive factor.

Uwe wrote:
However, if you want to avoid this completely, you can use additional steps to do the elution from the main column under acidic conditions.

Mmmh… I cannot trap on an additional enrichment column due to %B, a split flow injecting acidic solution would mean putting a lot of acid into the system, which again leads to signal suppression…so I would not know how to do that? Could you comment please?

James:
If drugs are strongly bound to blood proteins I guess I will loose much of it because of co-precipitation. Anyway, I am going to dive into the side you suggested - thanks


Regards,

smc

[HW Mueller]

It seems that one can not get around rechewing things, so:
SMC, one can prevent co-precipitation by using organic solvents, or acid with some organic solvent added, or some chaotropes.
I didn´t see any mention of ultrafiltration which is very simple with microcentrifuge assemblies. (Don´t always want to mention this old paper, but it is an example which indicates that off-line ultrafiltration + two chromatographic steps is about euqal to an on-line 3-step method using restricted access as first step in many respects: J Chrom B, 678, 137 (1996))

Uwe, are you talking about acid precipitation? It seems that using ACN, instead, will get rid of many interfering ions, besides most of the proteins.

[Uwe Neue]

SMC: sorry for the late response, but summer is vacation time for some of us.

The simplest thought is the post-column addition of acid to get you back to an acidic pH. Run the sample cleanup as discussed, the separation at alkaline pH, add post-column a sufficient amount of a suitable, non-suppressing acid such as formic acid, possibly dissolved in your organic modifier, and you will have the best solution possible.

As a brief estimate: assume that you loose 5x in sensitivity at alkaline pH, the change back to acidic pH with a 1:1 dilution will get you a gain of 2.5x in sensitivity, potentially more if you consider the gain in ionization in the presence of a larger amount of organic solvent.

[chhubert]

Uwe,
Do you have any information or literature regarding the use of mixed mode cartridge for on-line clean-up and using alkali for elution of trapped analytes and direct analysis by LC-MS. Or did you try it in your lab before and find it okay? Thanks.

[Uwe Neue]

Many application examples can be found in the very thick Oasis application booklet available online at Waters.com. Most applications are off-line applications. On-line applications can be found in C. R. Mallet, Z. Lu, J. Mazzeo and U. Neue, Rapid Communications in Mass Spectrometry 2002; 16: 805-813 and C. R. Mallet, J. R. Mazzeo, U. Neue, Rapid Communications in Mass Spectrometry 2001; 15: 1075-1083. If your question is more specifically toward the MS response of basic drugs at alkaline pH, you may look at Y.-F. Cheng, Z. Lu, U. Neue, Rapid Communications in Mass Spectrometry 2001;15: 141-151, where we used an XTerra MS C18 column with 10 mM ammonium bicarbonate at pH 9 and ammonia at pH 10.5 for basic drugs.