[MattNeedsFat]

Hello everyone,

Before I explain to you my problem, let me first say how much I appreciate you helping out in advance. I will try to participate in other topics besides my own, but I have a biomedical sciences background and therefore am not that knowledgeable in chemistry, hence my difficulties with my current problem.

Here's what I try to do: in order to be able to investigate the metabolism of brown adipose tissue, I want to radiolabel a triglyceride in a way that it's injectable in mice and men and possible to track in situ through PET scans.

I already have a method for developing 18F-FTHA, a fatty acid that can't be broken down due to the sulfur and is trapped in the brown adipose tissue, and has a 18 fluor radioactive isotope attached to it. I also have a successful method to create chylomicrons with triglycerides and other fat molecules like cholesterol esters and phospholipids etc. My problems lie with the synthesis of the radioactive triglyceride.

Why not just add a diglyceride to a fatty acid and wait, you say? Well, the point is, that 18F has a d1/2 of 109 minutes. Thus, ideally, I want a high yield within 2 hours.

This is what I've tried thus far: I've added oleic acid with thionyl chloride (SOCl₂). I've added 5mg to 0.5 ml acetonitrile (ACN), for this amount is somewhat representative of the amount of 18F-FTHA we start with. I've used several concentrations of thionyl chloride, from 2% to 0.1% of total volume.

I've tried basically everything to let this perform well: I've incubated for 0, 5, 15 and 30 minutes, I've let the ACN and SOCl₂ evaporate in a 100°C oil bath in a reaction vial, I've filtrated with water and ACN to remove the thionyl chloride, and all possible combinations of these. Unfortunately, I don't have access to a mass spectrometer, so the only method I can test the aforementioned methods with is a HPLC before and after I've incubated the oleyl chloride with dioleine.

After about 100 HPLC runs, I used the following method: same concentration oleic acid, 2% thionyl chloride, let it react in an open vial at 100°C so the produced HCl and H₂SO₄ can escape (I assumed they were volatile enough to escape the solution at 100°C). After the ACN and thionyl chloride have evaporated, I add additional ACN to dissolve the oleyl chloride, and add dissolved dioleine in isomolar conditions.

This seemed to work great with no radioactivity. However, turns out, when I use 18F-FTHA instead of oleic acid, about 60% of the 18F is released from the FTHA and I get a startlingly low yield.

Thus: I need another way to let this triglyceride synthesis happen, that is not so aggressive as thionyl chloride and does not affect my 18F bond to FTHA. However, it must work fast enough to produce enough triglycerides in 1-2 hours. Can you perhaps spot any problems in my method, or have any other ideas? Any help is much appreciated! Thanks in advance!

With kind regards,
Matthew, student at Maastricht University, Netherlands

[Dan]

I'd just like to clarify a few things for myself and others - you present this not quite as a chemist would, so I just want to check I am following you.

1. Abbreviations - are these correct?

FTHA = 14-[18F]fluoro-6-thia-heptadecanoic acid
diolein = 1,3-Di(cis-9-octadecenoyl)glycerol

2. So the reactions you are attempting are:

i) Formation of the acyl chloride from FTHA (or oleic acid) with SOCl₂
ii) Condensation of the acyl chloride with a secondary alcohol (1,3-diolein) to form a labelled trigylceride

----

My suggestions:

I've filtrated with water and ACN to remove the thionyl chloride

let it react in an open vial at 100°C so the produced HCl and H2SO4 can escape

If you add water to your acyl chloride you will hydrolyse it back to the starting acid. You must work under anhydrous conditions (i.e. anhydrous MeCN). Using an open vial may even be detrimental due to atmospheric moisture. H₂SO₄ is not produced, it's SO₂.

This seemed to work great with no radioactivity.

How much evidence do you have that the reaction worked?

A couple of other things:

1. It would be better to use ¹⁹F-FTHA (i.e. non-radiolabelled) as the model, not oleic acid. The functionality in FTHA is, I would say, substantially different to that in oleic acid. This will allow you to develop a procedure that you know for certain will transfer to the radiolabelled version.

2. The method of heating your compounds to dryness in open flasks at 100 °C instinctively seems like a bad idea. It is much safer (in terms of decomposition) to use vacuum and lower temperature. My gut feeling is that the FTHA is destroyed at this stage (but I am guessing with very little evidence).

3. Since you do not use any bases to neutralise the HCl formed in both steps, you may be getting unwanted reactions caused by the highly acidic conditions. A base, commonly triethylamine or a carbonate, is often added to "mop up" HCl, especially in the condensation step.

about 60% of the 18F is released from the FTHA

How did you determine this and what do you mean by "released"?

[Enthalpy]

I have doubts about H₂SO₄ being volatile at +100°C. Under 1atm it boils at +337°C.

[Furanone]

Just curious if you've tried transesterification with a lipase enzyme instead of the thionyl chloride? It may be able to catalyze the reaction so it is faster than the 109 minutes and increase your yield, plus it is very selective so shouldn't damage your 18F-FTHA.

[MattNeedsFat]

I'd just like to clarify a few things for myself and others - you present this not quite as a chemist would, so I just want to check I am following you.

1. Abbreviations - are these correct?

FTHA = 14-[18F]fluoro-6-thia-heptadecanoic acid
diolein = 1,3-Di(cis-9-octadecenoyl)glycerol

2. So the reactions you are attempting are:

i) Formation of the acyl chloride from FTHA (or oleic acid) with SOCl₂
ii) Condensation of the acyl chloride with a secondary alcohol (1,3-diolein) to form a labelled trigylceride

Yes, this is all correct, my apologies for being unclear.

My suggestions:

I've filtrated with water and ACN to remove the thionyl chloride

let it react in an open vial at 100°C so the produced HCl and H2SO4 can escape

If you add water to your acyl chloride you will hydrolyse it back to the starting acid. You must work under anhydrous conditions (i.e. anhydrous MeCN). Using an open vial may even be detrimental due to atmospheric moisture. H₂SO₄ is not produced, it's SO₂.

I realize that water will reverse the reaction. However, my supervisor doubts we can, in our laboratory conditions, completely remove water from our stock ACN and the FTHA. We use ACN because it can be injected in mice and humans and is readily available, so I rather keep this reaction in ACN (also, is MeCN safe to inject in humans, or can you easily remove it after triglyceride synthesis?)

This seemed to work great with no radioactivity.

How much evidence do you have that the reaction worked?

I use HPLC under certain conditions that completely isolate the peak of oleic acid, dioleine and trioleine. I've checked all these separately, so I have a reference point as to where I expect my trioleine peak should be. I have yet to analyze the readings at different wave lengths (I only compared UV and radioactivity), however, so I cannot say for sure that the peak I have is actually trioleine.

A couple of other things:

1. It would be better to use ¹⁹F-FTHA (i.e. non-radiolabelled) as the model, not oleic acid. The functionality in FTHA is, I would say, substantially different to that in oleic acid. This will allow you to develop a procedure that you know for certain will transfer to the radiolabelled version.

2. The method of heating your compounds to dryness in open flasks at 100 °C instinctively seems like a bad idea. It is much safer (in terms of decomposition) to use vacuum and lower temperature. My gut feeling is that the FTHA is destroyed at this stage (but I am guessing with very little evidence).

3. Since you do not use any bases to neutralise the HCl formed in both steps, you may be getting unwanted reactions caused by the highly acidic conditions. A base, commonly triethylamine or a carbonate, is often added to "mop up" HCl, especially in the condensation step.

Thank you very much! 1. I use oleic acid because my lab has it in complete abundance, so I can experiment limitlessly. I also agree that I should try it with 19F-FTHA, but unless we really need it to get the results we want, I don't think my supervisor will let me buy it.
2. I am sorry, but I don't understand what you mean with vacuuming? And what temperature and incubation time would you suggest? And after that?
3. I have done one experiment with triethylamine, which didn't give any triglyceride. I will try it again with different conditions to be more sure, as you're not the first to suggest this.

about 60% of the 18F is released from the FTHA

How did you determine this and what do you mean by "released"?

Yes this is a bit too hastily concluded, sorry. When I use my acyl chloride synthesis method, without radioactivity, I have one peak which is clearly the fatty acid. However, when using 18F-FTHA and comparing the UV signal with the radioactivity signal, I see two radioactive peaks. One peak with a high amount of radioactivity had emerged before the fatty acid peak. I start with slightly hydrophilic conditions (70% ACN, 30% water), so an earlier peak means something more hydrophilic than the fatty acid has been created. I concluded that it was free 18F, but since the bond between C and F should be pretty strong, I can also imagine it is a piece of the acyl chloride, cut by thionyl chloride. My guess is that thionyl chloride cuts the FTHA at the sulfur. I am not sure how I can determine this (we don't have a MS, but I can try at a different department), but if it is cut at the sulfur, I can also try another radioactive fatty acid without one (BMIPP comes to mind).

Again, thank you very much!

[MattNeedsFat]

Just curious if you've tried transesterification with a lipase enzyme instead of the thionyl chloride? It may be able to catalyze the reaction so it is faster than the 109 minutes and increase your yield, plus it is very selective so shouldn't damage your 18F-FTHA.

Thank you very much, but I'm confused: as far as I know, lipases cut down triglycerides into glycerol and fatty acids, so exactly the opposite of what I'm trying to do

[Furanone]

I do not have much experience with enzymatic interesterification, but my understanding was that the reaction with lipases can go in the opposite direction by shifting the equilibrium of reactants. For example adding in glycerol and fatty acids with lipase should synthesize triglycerides, while adding in lipase with triglycerides with an excess of hydroxyl groups should create fatty acids & glycerol.

I commonly run fatty acid methyl ester analysis using GC with sodium methoxide as the catalyst (which is instead chemical interestification) so this gives an excess of methoxide to catalyze fatty acid hydrolysis of  TAGs, and methylate them then precipitate the glycerol with oxalic acid (Christie Method, 1982). I know enzymatic interesterification is used to make the trans-free margarines (Becel) by blending a highly saturated fat (ie palm) with a highly unsaturated oil (ie soybean, canola) to obtain a rearrangement that gives the desired melting properties of the margarine.

I was only asking if you have possibly tried this route (enzymatic interesterification) or looked into it more as an option. Perhaps it is would be possible to use DAGs with excess 18F-FTHA with the enzyme in the right environment (with no excess hydroxyls so no methanol or water) where the reaction could take place, fast enough for your needs, and much gentler/specific not damaging the 18F-FTHA. A common enzyme used is immobilized Lipozyme (from Novozymes) on a resin you could mix in then recover. Enzyme is stable up to 70-80 C to increase reaction rate, and if necessary turn a fat into an oil. Anyway just an idea, maybe worth looking into.

http://www.sigmaaldrich.com/catalog/product/sigma/62350?lang=en&region=CA

[Dan]

I realize that water will reverse the reaction. However, my supervisor doubts we can, in our laboratory conditions, completely remove water from our stock ACN and the FTHA.

You could buy anhydrous MeCN.

Actually, a thought crossed my mind. I did some work with benzoxazoles a couple of years ago and I made some of these by condensation with imidates. I used to make the imidates by stirring an alkyl nitrile with an alcohol in dry ethereal HCl. Given this, I wonder if using MeCN as the solvent to generate the acyl chloride with SOCl₂ may actually produce imidate anhydrides of this type:



Just a thought, I have not checked the literature, but the imidate anhydride may react similarly in the esterification anyway perhaps it doesn't matter so much.

is MeCN safe to inject in humans

I've no idea if there is a safe limit.

I use HPLC under certain conditions that completely isolate the peak of oleic acid, dioleine and trioleine. I've checked all these separately, so I have a reference point as to where I expect my trioleine peak should be. I have yet to analyze the readings at different wave lengths (I only compared UV and radioactivity), however, so I cannot say for sure that the peak I have is actually trioleine.

Ok, while I agree that it is probably the right stuff, it is obviously a good idea to check by MS at least (especially before you work with live animals - depending on where you are in the world, it could be illegal to inject what is more or less an unknown compound into a live animal)

Thank you very much! 1. I use oleic acid because my lab has it in complete abundance, so I can experiment limitlessly. I also agree that I should try it with 19F-FTHA, but unless we really need it to get the results we want, I don't think my supervisor will let me buy it.

It looks like you might. So far you know that oleic acid is a poor model, I think you need a better one.

2. I am sorry, but I don't understand what you mean with vacuuming? And what temperature and incubation time would you suggest? And after that?

I mean apply a vacuum. Boiling point is related to pressure, if you reduce the pressure, the volatile materials will evaporate at lower temperature. I actually removed 30 mL of MeCN from one of my samples this week at 40 °C under reduced pressure on a rotary evaporator - it is a standard procedure in chemistry labs.

When I use my acyl chloride synthesis method, without radioactivity, I have one peak which is clearly the fatty acid. However, when using 18F-FTHA and comparing the UV signal with the radioactivity signal, I see two radioactive peaks

OK, so you have a side product.

I concluded that it was free 18F, but since the bond between C and F should be pretty strong, I can also imagine it is a piece of the acyl chloride, cut by thionyl chloride. My guess is that thionyl chloride cuts the FTHA at the sulfur. I am not sure how I can determine this (we don't have a MS, but I can try at a different department), if it is cut at the sulfur, I can also try another radioactive fatty acid without one (BMIPP comes to mind).

Please define abbreviations - BMIPP = 15-(4-iodophenyl)-3-methylpentadecanoic acid?
Most of us here are not familiar with biochemical abbreviations.

In order to solve the problem, you need to know what happened. I think you are going to have to investigate options for characterisation of your products. This is the way a synthetic chemist would normally approach the problem: adapt the conditions to product. If the conditions result in undesired reactivity, we make modifications to avoid these pathways - but you can't do that unless you understand which pathways you are trying to avoid. Switching to an alternative fatty acid is an option - this would be adapting the product to the conditions, but again, unless you know why FTHA doesn't work it could be a waste of time. That said, I think it is most likely that the thioether linkage in FTHA is the root of the problem, and BMIPP looks more robust. You have the added advantage of the longer half life of ¹²³I as well. Does it otherwise fit the needs of your application? I guess you can't use PET imaging, presumably FTHA was first choice for a reason (I know essentially nothing about imaging).

[MattNeedsFat]

I realize that water will reverse the reaction. However, my supervisor doubts we can, in our laboratory conditions, completely remove water from our stock ACN and the FTHA.

You could buy anhydrous MeCN.

Actually, a thought crossed my mind. I did some work with benzoxazoles a couple of years ago and I made some of these by condensation with imidates. I used to make the imidates by stirring an alkyl nitrile with an alcohol in dry ethereal HCl. Given this, I wonder if using MeCN as the solvent to generate the acyl chloride with SOCl₂ may actually produce imidate anhydrides of this type:



Just a thought, I have not checked the literature, but the imidate anhydride may react similarly in the esterification anyway perhaps it doesn't matter so much.

is MeCN safe to inject in humans

I've no idea if there is a safe limit.

I use HPLC under certain conditions that completely isolate the peak of oleic acid, dioleine and trioleine. I've checked all these separately, so I have a reference point as to where I expect my trioleine peak should be. I have yet to analyze the readings at different wave lengths (I only compared UV and radioactivity), however, so I cannot say for sure that the peak I have is actually trioleine.

Ok, while I agree that it is probably the right stuff, it is obviously a good idea to check by MS at least (especially before you work with live animals - depending on where you are in the world, it could be illegal to inject what is more or less an unknown compound into a live animal)

Thank you very much! 1. I use oleic acid because my lab has it in complete abundance, so I can experiment limitlessly. I also agree that I should try it with 19F-FTHA, but unless we really need it to get the results we want, I don't think my supervisor will let me buy it.

It looks like you might. So far you know that oleic acid is a poor model, I think you need a better one.

2. I am sorry, but I don't understand what you mean with vacuuming? And what temperature and incubation time would you suggest? And after that?

I mean apply a vacuum. Boiling point is related to pressure, if you reduce the pressure, the volatile materials will evaporate at lower temperature. I actually removed 30 mL of MeCN from one of my samples this week at 40 °C under reduced pressure on a rotary evaporator - it is a standard procedure in chemistry labs.

When I use my acyl chloride synthesis method, without radioactivity, I have one peak which is clearly the fatty acid. However, when using 18F-FTHA and comparing the UV signal with the radioactivity signal, I see two radioactive peaks

OK, so you have a side product.

I concluded that it was free 18F, but since the bond between C and F should be pretty strong, I can also imagine it is a piece of the acyl chloride, cut by thionyl chloride. My guess is that thionyl chloride cuts the FTHA at the sulfur. I am not sure how I can determine this (we don't have a MS, but I can try at a different department), if it is cut at the sulfur, I can also try another radioactive fatty acid without one (BMIPP comes to mind).

Please define abbreviations - BMIPP = 15-(4-iodophenyl)-3-methylpentadecanoic acid?
Most of us here are not familiar with biochemical abbreviations.

In order to solve the problem, you need to know what happened. I think you are going to have to investigate options for characterisation of your products. This is the way a synthetic chemist would normally approach the problem: adapt the conditions to product. If the conditions result in undesired reactivity, we make modifications to avoid these pathways - but you can't do that unless you understand which pathways you are trying to avoid. Switching to an alternative fatty acid is an option - this would be adapting the product to the conditions, but again, unless you know why FTHA doesn't work it could be a waste of time. That said, I think it is most likely that the thioether linkage in FTHA is the root of the problem, and BMIPP looks more robust. You have the added advantage of the longer half life of ¹²³I as well. Does it otherwise fit the needs of your application? I guess you can't use PET imaging, presumably FTHA was first choice for a reason (I know essentially nothing about imaging).

Sorry about the abbreviations, I forget sometimes. FTHA would be ideal because the thio-ether bond actually makes it unable to catalyze in brown adipose tissue, so it stacks and gives more signal. Also, 18F has a nice half time that gives enough signal to reliably image. I123 not only radiates less and gives less signal, it also radiates longer, which is a problem if you want to inject it in humans. But hey, if FTHA doesn't work, it doesn't work.

Here's what I want to try:
- I've bought oxalyl chloride, which is less aggressive than thionyl chloride and may leave the thio-ether bond intact.
- I already have BMIPP, so I'll try to radiolabel it next week and see what it does.
- Meanwhile, I want to try your acyl chloride methods by including vacuum and triethylamine. So I'll see if I can create a vacuum (so far, I've only used closed reaction vials with a standard screw lid and heat it up to 70°C before evaporating it in 100°C). If I can't create a vacuum, should I just leave the thionyl chloride and triethylamine in the vial and add dissolved dioleine?

[Dan]

- Meanwhile, I want to try your acyl chloride methods by including vacuum and triethylamine. So I'll see if I can create a vacuum (so far, I've only used closed reaction vials with a standard screw lid and heat it up to 70°C before evaporating it in 100°C). If I can't create a vacuum, should I just leave the thionyl chloride and triethylamine in the vial and add dissolved dioleine?

Just to clarify, conduct the reaction as normal but remove the volatiles by applying vacuum after the reaction has finished. If you don't have a vacuum pump and don't want to invest in one, you can use a water aspirator.

You probably need to remove the SOCl₂ and SO₂ before attempting esterification.

I have always used oxalyl chloride (with catalytic dimethylformamide [DMF]) to make acyl chlorides. It has the nice advantage that you can see when it is complete because CO₂ evolution stops. I've only made amides this way (not esters) and usually do as follows:

To a solution of the acid (x mmol) in dichloromethane [DCM] at 0 °C was added oxalyl chloride (x mmol) followed by DMF (0.2x mmol). The reaction is allowed to stir at room temperature [RT] until gas evolution stops (usually 30-90 mins). The mixture is then transferred via syringe to a vessel containing the nucleophile (x mmol) and triethylamine (3x mmol) in DCM at 0 °C and then allowed to stir at RT.

You will need to purify this crude mixture to remove the triethylammonium chloride formed, excess triethylamine and the DMF catalyst. They can all be removed by dilute HCl aqueous extraction from an organic solvent. The organic solvent must be immiscible with water, i.e. not MeCN. I guess you would need to find a solvent that is either safe to inject or easy to remove under vacuum (diethyl ether, for example) but can be tolerated in traces when injected. Or use a different purification method - there are undoubtedly faster methods than organic/aqueous extraction currently used for preparing PET tracers, probably LC, but I don't know that field at all.