Growing pains with SPME, GC-MS and standards

[spiderylu]

Hi all,

I'm developing a protocol to analyse headspace from a gram negative bacteria in log phase growth.

I've been asked to use an internal standard to reduce the variability and the analyst who runs the GC-MS and one or two others suggested Decane. I'm analysing 5ul bacteria on 3ml LB agar in a 20ml autosampler vial, adding the Decane around 10 minutes before the bacteria is added. We found that 10ul pure Decane almost saturates the column, so now my analyst thinks I should use just 0.1ul of standard. I believe that's pretty much impossible given the inaccuracy of pipetting such a small amount. I had the idea of diluting, in Hexane, but perhaps this will not help if I can only put 0.1ul of anything in there. My analyst thinks diluting will make no difference, my lab instructor thinks it will be fine and I'm stuck in the middle since I have very limited knowledge of this and almost no training in this method. I've done a bit of digging around, but I can't seem to find anything about using internal standards for this particular system.

Also, I noticed that one of the fibres has some part of it missing...ie I can see the wire it's attached to. Does this spell the end for that particular fibre? We always pierce the septum with the fibre retracted so it doesn't bend.

If anyone has as experience with this or suggestions, I'm all ears. Also, if there's any specific details which would help you answer my query then let me know.

Thanks!

Lue

[Peter Apps]

Wherher or not decane is suitable as an internal standard depends on what the compounds are that you are interested in. Decane would be appropriate for non-polar compounds like hydrocarbons, esters, terpenes etc.

You need to select the SPME fibre based on what you want to analyse.

I am concerned that adding internal standards will affect the bacteria - especially since you are adding twice as much decane as you have organisms.

What kind of variability are we talking about ? - biological systems are intrinsically variable. Have you established that the SPME and GC-MS steps are repeatable by running standards ? - there is no point adding IS to the cultures if the variability comes from the analytical step.

A fibre with its wire showing is dead !

Peter

[spiderylu]

Thanks for your feedback.

A few things: I found the amount of decane a bit worriesome, but I heeded to the chemist. We did run controls to see if there was an impact on the bacteria when we originally tested the IS - there wasn't a noticeable difference for bacterial volatiles in the presence of Decane. Most of the compounds coming off the bacteria are hydrocarbons, and a SPME has been used to measure the volatiles of this bacteria before, so I know we have the right fibre.

To be honest, I can't answer your question about variability. I was told that variability had been found previously and that they want to try using a standard to reduce it. I will try to clarify what exactly is meant by variability in this context and try to shed some light on it.

[rb6banjo]

Microbiological systems are substantially aqueous. I would think that the hydrophobic nature of decane would be a huge problem in this system. If I were you, I'd choose a more life-friendly standard like ethanol or methanol. Then you can dilute them in water. I assume you are just using the internal standard to improve the variability in your SPME analysis (i.e., normalize the responses of your analytes to something that is known). Maybe a higher fusel alcohol could be used if ethanol is one of your analytes.

Your SPME fiber is useless. Get a new one. It is a very sensitive technique for many analytes. You don't need much. Certainly 1 µL of decane in a 20 mL headspace vial will give you results that are way outside of anything remotely linear. Even 0.1 µL of neat decane is "over the top". This is why I say find an internal standard that is water soluble so that you can spike your IS in at a reasonable concentration.

[spiderylu]

This is great advice! Thanks I'll give ethanol a shot and see what happens.....

...and buy a new fibre!

[spiderylu]

Although on another point - won't ethanol mess with the bacteria? Doesn't ethanol cause problems with the lipid bilayer in cell membranes? Or if it's really watered down will it not make a difference?

[Peter Apps]

Do not use ethanol as an internal standard for hydrocarbon analytes - its partitioning between the aqueous phase and headspace, and between the gas phase and the fiber will be completely different to that of your analytes. Decane is a good choice, provided that it has a similar molecular weight to your analytes.

What hydrocarbons are produced by the bacteria ?

It would be a goos idea to verify repeatability yourself - do you have standards of the hydrocarbons that the bacteria produce.

Peter

[rb6banjo]

I don't know much about their system but I'm betting those hydrocarbons are present at pretty low levels. Your internal standard doesn't have to show the same partitioning behavior as your analytes. It just has to be reproducible. For this type of trace analysis, you have to add something that you can add reliably that won't swamp out everything else or drastically alter the competitive nature of the fiber. Addition of neat decane will swamp everything else out - especially in your very small sampling container. Also, anything else I can imagine in which decane is soluble, will also swamp your system. You will not be able to reliably add a decane-in-water standard because the solubilty of decane in water stinks. If it were me, I would choose something fairly water soluble for this water-based system. Then you can add it at low concentration (in the linear region of the SPME technique) in a solvent that is not going to particularly care for your fiber (minimize the competition effects of the fiber).

If you are new to SPME, there is a phenomenon called displacement that comes into play in most real systems. Generally, heavier things will displace smaller things at longer sampling times. Care must be taken to ensure that you have reproducible sampling times and I have found that adding the IS at concentrations that give similar responses to the analyte(s) concentrations gives the best results.

As mentioned above, you do need to make sure that whatever you add doesn't distort the metabolism of your bugs. That would also be true of the decane. Most living organisms I know about don't particularly care to be swamped in gasoline either.

Sampling in this heterogeneous type of system is some tricky business.

[Peter Apps]

Your internal standard doesn't have to show the same partitioning behavior as your analytes. It just has to be reproducible.

An internal standard is added for the specific purpose of correcting for sources of variability in analyte peak size. In order to do that it has to be affected in the same way as the analyte by whatever the source of the variability is. Only substances with similar physical and chemical properties will be affected similarly by sources of variation. The peak size of a low molecular weight, water soluble alcohol will not be affected in the same way as a medium chain alkane, therefore ethanol is not a good internal standard for a hydrocarbon analysis.

You do not want your IS peak area to be reproducible - it is supposed to vary in order to do its job. If you are just looking for reproducibility in the area for the IS, then all you need to do is to divide all the analyte peak areas by any arbitrarily chosen number.

Peter

[rb6banjo]

Not reproducibility in the actual response of the internal standard. The internal standard should be added in the same manner each time (same concentration). The analytes and internal standard are measured for a given set of conditions for the bug growth and headspace sampling. The responses for the analytes are calculated relative to the response for the added internal standard. If the concentration of the analytes go up relative to what the known internal standard response is, the ratio will increase. Likewise they'll decrease if the analytes go down relative to that internal standard addition. To me, it matters little what the nature of the IS is in relation to the analytes as long as the internal standard partitions into the sample (bugs, whatever you're analyzing) in a reproducible way.

I think we're confusing the newbies. I challenge them to try both ways and report back with their findings.

[AICMM]

spiderylu,

Can I ask a question here? Is this a dry plate or vial? Are you adding some kind of "broth" (forgive my ignorance of the biology, I am but a humble chromatographer)? If so, why not add just a skosh of decane to the broth and then, assuming you add the same amount of broth, you have your IS already in the mix. That is to say, do the appropriate dilution of the IS in the broth and then add the same aliquot of broth to all samples.

I would also not recommend decane, and I certainly would not recommend ethanol (lots of other issues here.) Bugs, if I am not mistaken, can make alkanes. Therefore, I would choose something more like iso-octane or perhaps 1-decene but, again, my ignorance of biology is showing.

Best regards,

AICMM

[spiderylu]

Hello all!

I'm glad I've sparked a bit of a debate about this. It makes me feel less of an idiot >.<

This are the details of the analytes detected with my bacteria from a previous SPME (using all the same equipment I have now). I hope this helps in defining what would be good. And yes, these are very trace.




I would be lying if I said I fully understood the GC-MS output, I covered things briefly as an undergraduate, but I probably need to brush up on what it all means.

It would be super wonderful to find something that was water soluble, as I believe this will be the best course of action - and I had thought about adding decane to the agar (ACIMM - 3ml solid LB agar), assuming that it won't be affected by autoclaving?

[rb6banjo]

spiderylu--

What is your GC column? The alcohols in your mix are tailing quite a bit. You might move to a more "alcohol friendly" column (wax phase). The newer wax phases (solgel wax by SGE) are kinder and gentler to mass spectrometers. It should improve your peak shapes for those active-hydrogen-containing materials. Even an "old-school" wax will be ok as long as you stay away from temperatures where the column starts to bleed quite a bit.

From your first post, it looked like you were adding the bacterial slurry to the growth medium in the vial to be analyzed. Is that correct? If you can find the right concentration of hydrocarbon (if you want to go that way) that you can make up in the growth medium, then you should be ok in using it. I was under the impression that you were trying to add neat internal standard after all of the contents of your sample were in the vial. I don't think you want to saturate your SPME fiber at the expense of what you're trying to determine.

You certainly don't want to alter how your bacteria cells are propagating by adding the wrong thing to their environment. I'm not a biologist either, just an analytical chemist (polymers, plastics, alcoholic beverages, grains, commercial gases, etc. are my typical fare).

[Peter Apps]

In post 5 the analytes were described as hydrocarbons - but nothing on the list is a hydrocarbon. You have a quite challenging range of molecular weights and polarities and I doubt that you will find an internal standard that will cover all of them - decane is certainly not suitable, and neither is ethanol.

ISs for SPME are an extra challenge because there are three steps where chemistry and molecular weight affect peak size; partition into the headspace (I assume that you are doing headpspace, if you are immersing the fibers that might account for the poor repeatability), partition into the fiber and desorption from the fiber.

So you need to establish a repeatable method that does not need an IS. First step is to make up some broth with target analytes at the concentrations that you expect in the samples and to run multiple replicates to see what the analytical repeatability is.

If you are growing the bugs on agar slopes in vials I very much doubt that when you take the samples your vial content is anywhere near homogenous with respect to analyte concentration - diffusion into the agar will be slow, diffusion into the headpsace will be rapid. This might account for some of the variability. A liquid medium would be much more tractable from an analytical point of view if you can persuade the bugs to grow in it.

Please keep us posted.

Peter

[spiderylu]

I apologise for the confusion - my lack of knowledge of chemistry let me down there!

According to the paper, the GC column is Rxi-5ms. That's as far as my knowledge goes. I will find out if I have any say at all in changing the column type. I have to take my samples to an analyst on campus in the chemistry department.

I think I have a good basis for a case to not use an internal standard. Would you be able to point me to any specific peer reviewed literature that I can use to back up my arguments? Or suggest some search terms for stuff? Or anything?

As for the collection, we are collecting headspace as the bacteria grows on the agar. We're specifically interested in the volatiles they emit whilst in growth phase - it's trying to replicate bacteria growing on carrion. Do you think that growing the bacteria in a broth and then analysing the broth would be better? Why would that be?

I really appreciate all the advice on this. It will take me a few weeks to get this done since I have finals coming out of my ears, but I hope to get going again with this.

You guys are great - I'd love to keep picking your brains