Glucosamine and chondroitin by HPLC.

[sksrinivas]

Hello All

Long time since I posted.

I have a problem and I need help. I'm trying to estimate glucosamine and chondroitin on the same column and I've failed miserably.

We have: Agilent 1100 quarternary system, RI and UV detectors, and only three column chemistries at hand - C8, C18 and NH2.

We have tried all three, with no success. The problem is both glucosamine and chondroitin are merrily soluble in water and it's just not possible to retain either of them, long enough to attain baseline separation.

Thus far, we've tried:

C8, C18 and NH2, with 100% ACN, and with phosphate buffers at pH 2, 7 and 8, with varying amounts of acetonitrile. No separation achieved. Co-elution within three minutes.

The official method for chondroitin uses size exclusion, with RI detection, while that for glucosamine uses a C8 with UV detection. But we've been given the task of separating both glucosamine and chondroitin on the same column.

Given our restrictions on the column chemistries available to us, my idea was to try and increase glucosamine's retention time on the column, by any means possible. I didn't expect chondroitin to be retained at all, since its mol weight is above 50KDa, and it won' t be retained on any 100A column. One would expect chondroitin to elute fast, and it does. Unfortunately, so does glucosamine, no matter what we try.

Ion-exchange chromatography is ruled out. We can't procure an electrochemical detector, just for one analysis.

So I need help.

Do I derivatise? Say, ninhydrin for glucosamine?

Ideas, anyone?

Thanks in advance.

[Noser222]

This method might help you for glucosamine, although they are dealing with the N-acetyl form.

http://www.uni-trier.de/uni/fb6/umweltc ... l-Chem.pdf

[Mark Tracy]

You can derivatize glucosamine with FMOC, OPA/thiol, Dansyl, Dabsyl, etc. Ninhydrin works only as a post-column reaction because it deaminates the analytes and you get the same derivative no matter what the source.

[sksrinivas]

Dear Noser: Thanks! I came across this paper yesterday during a google search, and we might try out the method this week. Thanks for your help!

Dear Mark: Thank you for your help. I thought I'd try ninhydrin, but it seems to be unsuitable for our analysis, since we don't have a post-column reactor. I've had some experience with OPA derivatisation techniques, and I wanted to use OPA. However, my client is against the use of mercaptans, so I guess I'll have to try a different derivative.

I think I'll try out dansyl or fmoc.

[HW Mueller]

My suggestion: If FMOC works forget about the others (dansyl,etc.).

[Dan]

sksrinivas,

There are two or three monographs in the USP for products containing glucosamine and another active. Those methods don't use derivatization. I don't have access to a USP, so I can't look up details.

Several years back, we tried one of the USP methods for one of our products. I can't can't give details because of the company confidentiality. What I can say is that we were not happy with the retention of the glucosamine using the USP method. We improved the retention (to a later time) of the glucosamine using a polar-embedded column.

So, my recommendation is to try a polar-embedded type of column.

If you do use a derivitization, one caution: you can get two isomers of the derived glucosamine. So, if you do pre-column derivitization, you will get two peaks for the glucosamine and you will have to total the peak areas. Or you can do post-column derivitization and get one peak for both isomers.

Regards,
Dan

[sksrinivas]

Dan

Thanks. Sorry for the delayed reply.

Just got back from the lab, and I'm happy to report we've got some good results with: NH2 column, and acidic buffer with an organic modifier.

Yes, we did use acidic buffers on an amino column.

Glucosamine alone and chondroitin alone do not present significant problems. We got a single peak for glucosamine alone using an amino column, and MeCN/water. Chondroitin alone can be done on a size-exclusion column and RI detection. Again, no major problem.

But here the challenge was, to get glucosamine and chondro both on the same column. No SEC, no ion ex, no ELSD. Just three columns available - NH2, C8, C18. Only one RI detector available, so no gradients either.

What do you do?

What we did was - we bent the rules. Thus, acid buffer on an amino column, and an organic modifier.

Can't share any more details, since we still have to validate our new method, and since we hope to publish.

Thanks all, for your inputs!

[Noser222]

Why do you say you "bent the rules"? I didn't know there was a rule against using acidic buffers on amino phase

[SIELC_Tech]

I have big doubts that you are eluting chondroitin from amino column. We did extensive study on the mixture of chondroitin and glucosamine and unless we used size exclusion/anion exclusion/cation exchange mechanisms we were unable to separate chondroitin from glucosamine. You can be misled by impurity in chondroitin. Try to inject chondroitin without column and measure peak area (with RI). Then inject it with amino column and measure peak areas again (if you have multiple peaks). In both cases (with column and without it) peak area should be the same (5% variation is okay).
I would try to confirm this by running LC/MS (may be not in your lab, but somewhere else).
Chondroitin has multiple sulfates which will interact very strongly with amino column. May be what ever you see is sodium (chondroitin sulfate is sodium salt). I will inject chondroitin into amino column and try to monitor it by ELSD. I will let you know.

[sksrinivas]

Don't want to prolong this thread ad nauseum. However...

SIELC_Tech: We did inject chondro and glucosamine, separately and together, without the column. It's a common practice in our lab, and usually the first thing we do. And we did get reproducible peak heights and areas. We're certain that what we're getting is chondroitin, no question.

Noser: I too didn't know there was a rule about amino columns and acidic buffers! But some people on the forum tell me there is...and I've been getting some strong personal email on this 'rule'.

Anyway, as I've been saying all along - we had certain restrictions. Only three columns available : C8, C18, NH2. Only one RI detector available. And one very anxious client waiting for a result.

Under the circs, we did what we could.

And, we still have to validate the method, and eventually publish (I hope so).

I won't be visiting this thread again, so I'd suggest we move on. Agreed?

[AA]

I have been working on a method that uses enzyme degradation of the choindroitin followed by GPC using a column for low molecular compounds that seperates both the glucosamine and the degraded choindroitin. Working well enough so far. Just my 2 cents, maybe it becomes something, maybe not. Science is fun that way.