Hi, I'm developing an LC-MS method for the quantification of 15 human milk oligosaccharides but I'm having trouble separating two isomers (LNT, lacto-N-tetraose and LNnT, lacto-N-neotetraose). I want a good resolution between them, affecting as little as possible the other oligosaccharides.
I'm using as mobile phase water and acetonitrile, both with 0.1% formic acid and a Hypercarb column (100x2.1mm, 3u, Thermo Scientific). The LC gradient is 0-12%B in 21 min; 12-40%B in 32 min; 40-100%B in 37 min; 100%B in 42 min; 0%B in 43 min, keeping at 0%B up to 55 min. I've tried several changes in the chromatography, but the best I got is in the image attached.
Does anyone have an ideia what I could try?
You can use the technique of chiral functionalized polymeric membranes is a very promising technique
I would try a softer gradient: instead of reaching 40%B in 32 min, I would test 40% in 50min (for example). As you have the LC coupled on-line to the MS, you can also test to include an additive in the mobile phase such as citric acid, that is not going to interfeere in ionization but can help in the resolution between both isomers.
I would try to leave the gradient isocratic for a period from 22 min.
For instance 0-13% in 21 mins; leave it at 13% for another 10 mins(31 mins) and see how it goes or 0-12% 21 mins; leave it @ 12% for another 10 mins. Play around this concentration allowing a zero gradient for about 10 mins. Modify as appropriate
Hello, thanks for the help!
Rosana, I tried to soften the gradient in several ways. I've used 1.5; 1.0 and 0.5%B/min and what I got were two peaks even more united... I don't understand why.
Ken, in the attached chromatogram I did exactly what you suggested, kept the isocratic gradient of 12%B for a few minutes, but I didn't get a complete resolution. However, it was the best result I obtained so far.
Miloudi, honestly I don't understand about chiral membranes. I don't know if it applies in human milk oligosaccharide analysis, because I didn't find any reference in the literature.
After much effort in optimizing the method, I got a great sensitivity and resolution of other 5 groups of isomers. The only problem is the separation of LNT and LNnT, which I expected to manage with little change in the gradient, but nothing seems to work ...
You may want to replace 0.1%formic acid with 0.05% TFA. The greater viscosity of methanol may also be an advantage. May I know how long you kept the 12%B after 21 minutes? I had a similar case with plant constituents and I had to flatten the gradient for over 10 minutes to get quantifiable resolution.
I kept 12%B for 5 minutes. The peaks came out around 24.5 min. I'll try both 0.05% TFA and citric acid next time...
In my oipnion, TFA is not a good idea as it will suppress ionization and so, you will decrease sensitivity at the mass spectrometer. I know that citric acid is a good option when analyzing by LC-MS because I have tested for increasing signal sensitivity in complex mixtures. In addition, it may help to resolve both signals, by binding to acidic sites (silanol groups) of the column.
I suggest you try reducing the formic acid concentration. Since this phase has weak anion exchange character, a lower pH should increase the interaction with the phase.
Hi Karina! I have to say that we are actually dealing with almost the same method and same column, maybe we are following the same paper. And what you got on LnNT and LNT is also the best I could get so far...
But that is not my major problem. Are you also testing sialylated HMO in the same run? Because by using 0.1% formic acid in ACN and water as mobile phase, I got really bad resolution on sialylated HMOs, something that I would not call peaks. But from the literature, they got good results. Do you have similar observation as me, or it works quite well for you?
Besides, I encountered unacceptable shift in retention times of neutral HMOs, no matter how long I equilibrate the column. It would be very nice if you could share some related experience.
Last but not least, good luck for both of us.
Hi Ginny! Thank you, I was very happy when I read your message. It's nice to know there's someone else in the same situation as me. It will be great to share our experiences.
Yes, I am testing neutral and sialylated HMOs in the same run. I am also using water and acetonitrile with 0.1% formic acid. With the gradient that I described above, I have obtained a great resolution between the sialylated. They elute from 24 minutes, with the increase in the proportion of acetonitrile. The latest peaks are slightly wider than the others, but still have less than a minute length. However, there's something interesting going on when I analyze samples. In some runs, I notice an "extra" peak between the 6-SL and 3-SL. In the beginning, I thought it was an unknown isomer, but in other runs, using the same method and samples, the extra peak joins 6-SL in a single peak. I'm still trying to understand this variability...
Regarding the retention times, they are quite stable to the neutral, varying between 0.1 and 0.2 min. I notice a greater variation with the sialylated, between 0.6 and 0.8 min. Perhaps washing the column will solve this problem.
Besides, what helped me a lot to get narrower peaks was to connect the column directly to the MS and to replace the reducing agent.
I will send you my email address, and maybe we can help each other if you want.
Good luck for both of us!
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Column