Navid

Check the standard addition method.

Regards

G

its possible by HPLC/UV to purify any compounds , However, LC-MS is destructive

Good luck.

Yes! Quantitative extraction is a primary requirement. Try as suggested a 'spike' experiment, and calculate your %Recovery. This is the reason I prefer a 'dilute and shoot' method.

If you have the standard, then first you have to check it in LC-UVD with recovery test as Bruce said in earlier comment. You may optimize it by using PDA detector for better %recovery.

If your recovery is not at acceptable range, then you may customize the clean up technique for better recovery. LC-MS for the unknown determination with details parameter.

Most likely you may have matrix suppression. You can check by put the same amount of standard in solvent and in your blank sample. You should have the same response if you have no matrix suppression. If you see less or no response in blank spike, it means you have matrix suppression and need to do sample cleanup to eliminate matrix such as protein and phospholipids.

Thank you all for such nice advises. Thats why discussion is very important and helps to solve problems in science.

As most of you advised, the first step for me to solve this problem is to do a matrix effect analysis also known as standard addition in which we spike matrix with standard compound of interest and analyze them to check for interferences; and then calculate recovery. I will do that and if find a major interference, i will share with you all in this thread and will open up the discussion for finding ways of an efficient extraction protocol to get rid of matrix effect.

Thank you all once again.

You need to spike at the level that you can see the analyte response in the sample matrix, otherwise you will have no data to compare.

I am planning to try low, medium and high concentration based on the LLOQ determined for the hormones in water. Thank you so much for the valuable advises.

Water is fine. But also spike plasma too. It gives you an opportunity to evaluate interferences as well.

What is you sample preparation protocol?

ResearchGate seems to be a good place for this sort of discussion, even if you can't divulge all the details. Biomedical researchers have to distil the bits of knowledge and lab-lore they need from more than half a century of research in some of the most difficult areas of analytical science. I don't know what suitable up-to-date textbooks are available. However, one thing that's been frequently re-discovered over the decades is that you can never determine a trace component by injecting plasma or serum directly into a column or mass spectrometer!

Recently, I had the misfortune to review a very long-standing project, probably involving steroids, where decades of confusion, controversy and worse can be attributed to consistently inappropriate choices regarding analytical methods (10.13140/RG.2.2.10992.35845).

It may be worth reading a review that emphasises the continuing importance of GC-MS in steroid research: Journal of Steroid Biochemistry & Molecular Biology 121 (2010) 496–504. However, this doesn't really discuss the major difficulty posed by matrix interference in LC-MS, particularly with neutral analytes. While GC-MS has retained its status as a major analytical technique, it is not always available in biomedical labs. Unfortunately, outsourcing the analyses or method development to specialised facilities having the necessary experience is not usually an option in this field. You could get ideas from the primary analytical literature, though it may take a specialist in steroid analysis to evaluate the quality of published methods.

While some kind of sample preparation is inevitable, it's likely that this could be simplified by using the LC column-switching technique known as 'heart-cutting'. Basically, you switch the eluant from the first column to the second only while the target analyte is being eluted. An important point, particularly for LC method development, is that the 'heartcut advantage' applies even when the columns and mobile phases are identical. Column switching is as old as column chromatography. It's becoming routine at last and you might consider a retrofit for your equipment.

The heart-cutting method is an alternative to the manual SPE method if the target analytes are in the narrow range window. I did it many times in the past for one analyte in the matrix. However, if the three hormones are in a wide window, the benefit of hear-cutting will be diminished. You may have to do more than one cuts. The hard part is to find two columns with different selectivity and are compatible with the mobile phase you plan to use. The idea is after cutting from the first column to the second column, the analytes have to be retained at the front end of the column before being eluted by another pump. This technique will require trial and error and need lots of expertise and hardware. Before going to this route, I would explore the severity of the matrix effect and work with easy matrix removal such as solvent precipitation (1:3 sample:acetonitrile), use special SPE from Waters (Oasis Prime) or Phenomenex to get rid of protein and phospholipids. You will need a fit-to-purpose method. How many sample you will analyze in the project, how low you want to go and so on. I am a lazy chemist and also like to do the easy one first. Yet, not do any cleanup on the biological sample is way too extreme.

Dear Drs. Chamkasem and Lee,

Thank you so much for taking time out from your busy schedule to provide such valuable advises to my problem here. It turned out to be a very insightful discussion. I would like to thank all other scientists who contributed towards the discussion.

I am extracting the biological samples with acetone (serum:aceton in 1:3 ratio). After exteaction, i am freeze drying the supernatant and resuspending the dried exteact in 20%acetonitrile with 0.1% formic acid. I am using acidic solvent sumytem in a shimadzu hplc coupled with a ab sciex 3200 mass spectrometer. The hormones are being detected in positive ms mode with a c8 column on reverse phase chromatography. And the peaks for three hormones are very well spreaded with a ms retention time of 10.49 min, 13.32 min and 15.80 min in a 20 mins method.

As suggested, i am preparing to check matrix effect first and result of which i will be discussing in this thread to determine the appropriate sample clean-up protocol.

I thank you all very much for your kind support and vaulable advises.

Sincerely,

Navid

I forgot to recall that the standard way of dealing with uncertain MS responses has always been to use a stable-isotopically labelled internal standard.

If these aren't available commercially, it may (depending on the importance of the project) be worth paying for bespoke syntheses. Somebody in your lab or another department may have enough chemistry experience to apply a published synthetic or labelling method. You may be able to buy a labelled precursor. Often you can incorporate deuterium by simple exchange or reduction reactions (e.g. hydroxy compounds, cardenolides, carboxylic acids with alpha hydrogens). Your compounds don't need to be chemically pure because you can standardise a stock solution against the unlabelled reference compound.

Deuterated compounds don't co-elute exactly and for critical work one uses a combination of carbon-13, oxygen-18 nitrogen-15 etc.

You may need to extend to run to clean the column after the last peak is out (15 min). The run time of 20 min is too short unless you did not tell me you have column equilibration added after 20 min. You may have matrix from the first sample showing up in the second sample.

That’s anither thing to correct then. My method run time is 20 mins including column equilibration. I will extend the run time to allow wash and equilibration between samples. Thank you.

At this point we are not thinking about using an internal standard because of the priority of the project. If the matrix effect can be checked and sorted out, it might diminish requirement of using an internal standard for this. Personally, i always prefer to use internal standard cause it can help answering lots if question and provides a layer of control on the data set.

Generally, you can matrix match the sample by diluting it 1/10 with the mobile phase and then just increase the injection volume.

With the dirty sample like this, I would be surprised if you do not need IS to correct for matrix effect (which it may be the cause of not seeing the peak in the first place). I would not add the IS at the beginning of the extraction because it is expensive and the most important thing is it may hide the poor recovery due to improper extraction procedure. If you have only 10% extractability, your IS will be 10% as native and your "apparent recovery" will be 100%. If you drop half of the sample on the floor, you will get 100% recovery because IS will correct for that.

Since the matrix effect occurs in the ion source you could, at the expense of analysis time, do 3 successive injections of representative unknowns, using the automatic injector to co-inject successively a blank solvent and two suitable concentrations of standard in the blank solvent. That way, you can check approximately the linearity starting at the amount corresponding to the lowest unkown amount. Since you have experimental conditions expected to deplete the hormones, the only remaining uncertainty will be the accuracy of the measured value of the lowest amount encountered, *provided* you account for any effect of the experimental manipulation on the matrix (lipid concentrations...).

I remain strongly convinced, however, that this kind of routine, even preliminary, LC-MS should always be done using two columns (or cartridge + column) with a common mobile phase. With a single column the ion source will be exposed to the salts, amino acids, etc in the partially aqueous extract you are using; the resulting deposits will lead to irregular responses. (Various alternative extraction methods have been published.) Also, the column that's connected to the mass spectrometer should be kept as clean as possible; lipids and so on are not easy to separate from steroids and they don't elute as nice sharp peaks.

Thank you Dr. Chamkasem and Dr. Lee for your valuable suggestions. I would be soon done with some of the advises you have given which will answer a lot of questions and make next steps easier. I also thought of doing cartridge purification of serum samples over C18 materials to see if that helps to remove salts and any other interfering materials if present in serum, but did not proceed with it as adviser ordered to inject without any cleanup of serum samples. I cannot thank you all enough for your such kind help, time and insightful supervision. Thank you so much.

Here's the reason you need to do sample cleanup. Use these protocols to get rid of protein and phospholipids from your serum. If the direct injection works, these products would not be in the market.

Thank you so much Dr. Chamkasem.

Dr. Chamkasem, thank you for sharing these two documents. It has lots of useful information. Thank you so much.

Remember, you can search anything on the internet. By doing it, it will train you how to solve the problem by yourself.

Dear Navid,

as we are doing pesticide analysis from frech food every day, we are facing exactly the same problems. What you can do is to add aliquots of defined solutions to your animal samples to see which effect the so called „matrix„ has. Matrix effects can be Peak shifts or ion suppression for inatance. What you can simply do to reduce these effects is to dilute your animal sample with eluent.

Best,

Tobias

Dear Dr. Becker, thank you for your valuable input. This discussion has been so much helpful. Ince again, thanks a lot.