Protein monosaccharide analysis not quantifying correctly

[JJG]

I have been put in charge of establishing equivalency between two Dionex instruments by taking a method we regularly run on one instrument, and repeating it on a second rarely used system. I have done this a few times, and I get calibration curves which are acceptable (r^2 values >0.995, 5 points) on each instrument, but when I quantify my samples, the nmol amounts for my compounds of interest are different by ~40%.

The samples I am injecting is protein which is acid hydrolyzed by 4.5M TFA to monosaccharides. The hydrosylates are analyzed using a Dionex ion exchange chromatography system with a Pulsed Amperometric Detector (DX500 model). The peaks of interest are glucosamine and mannose. A Dionex CarboPac PA20 column and same guard column are used, and the peaks are eluted isocratically with 10mM NaOH.

My concern right now is that I don't seem to be able to quantify correctly any sample that started with protein in it. I have tried injecting protein sample with a spiked amount, along with a protein sample without the spiked about, and also a placebo with the spiked amount (n=5). The placebo with the spiked amount, which contains no protein, quantitates correctly as expected (1nmol). But, when I add the nmol amounts for the the placebo spiked (1nmol) and the protein sample not spiked (1.22nmol), the amounts do not equal the protein spiked sample amount (1.76nmol). When I do this on two instruments, I get similar results, where the placebo spiked is the same nmol amount, but the protein samples are about 40% different between each instrument.

Does anyone have any insight as to whether the protein would have any effect on the column over time, or if the TFA which is evaporated and reconstituted with water would have any effect? Would the protein have any effect in the detector cell if one were cleaner than the other? I know the column and instrument are good for analyzing sugars, but do they work fine when injecting protein along with the sugars? Please let me know if anything jumps out as the culprit. Thanks!

[mmadson53]

You are quite an astute observer. It turns out that if protein and carobhydrate are not separated before hydrolysis a significant amount of carbohydrate could be lost due to the formation of amadori compounds with partially hydrolyzed amino acids. Dionex has a method that separates N- and O- linked oligosaccharides from proteins using their ON GUARD II cartridges which are about $2-4 per cartridge. I suggest separating the carbohydrate from protein somehow if Dionex's method is not used.

[JJG]

Thanks for your response. The samples I am hydrolysing are protein, and I am trying to measure the carbohydrates on the protein. The hydrolysis is to remove the carbohydrates from the protein. Does the chance for formation of amadori compounds exist still during the hydrolysis if no carbohydrates are in solution at the beginning of the reaction? It seems the ON GUARD II cartridges were for separting protein and carbohydrate in solution before hydrolysis.

[mmadson53]

Good question. Yes you may get appreciable amadori compound formation even starting only with glycoprotein. The sugars are relased over time as well as some amino acid, thus producing amadori compounds. The ON GUARD II cartridges are used in conjunction with PNGase F release of N-linked oligosaccharides and ammonium hydroxide release of O-linked oligosaccharide. Of course it's a bit more complicated than that but that's essentially what one does. If one uses this protocol one will have in two separate vials N-linked oligosaccharides and O-linked oligosaccharide alditols of which one can further analye by hydrolysis and chromatography.