DNA behaviour?

[DaveM]

Hello,

I've been noticing some (what I think) is odd behaviour of DNA on reverse phase. Running it through a C18 column appears to be OK in an 10mM ammonium acetate/methanol gradient. However, on changing to a C5 column, the majority of the material gets stuck in the guard column (thank goodness for guards). I say this because when I run the system with nothing in line, just a dead volume connector, I get a certain signal but putting a guard column in line reduces this signal by a factor of at least 10.

The DNA is oligonucleotides, from 6 - 300 bases.

Is this odd or am I missing something? It also happens in 0.1%TFA/ACN gradients.

Thanks,

DM

[Uwe Neue]

The packing in the guard column must match the packing in the analytical column. If this is not the case, you get junk - exactly as you observe.

Use a good C18 that has been proven to work for this application and purchase a guard column from the same material.

[DaveM]

Darn it, it's hard to get all the important information across.

The C5 and the C18 both had guard cartridges sold as guards for those particular media. The columns and the guards were from the same supplier in both cases. Can I name the suppliers?

[unmgvar]

DaveM

when you say you run with only a union, are you also making an injection?
the difference is of the sample signal right as you do an injection while there is only a union?

what is the pressure difference when you do this?
what is the pressure with the union?
are you using some sort of delaying tubing and pressure builder?

with the C-18 the signal is staying high as well just like the union?

also if your sample stays like you say on the guard then you should be seing increasing pressure effects, are they there? how big and how fast?

[DaveM]

Hi,

Yes I am making an injection with the union in place, and the comparison I'm making is between injector-guard-union, and injector-union.

The signal (I won't call it a peak as there is no chromatography happening) reduces substantially when the guard in in place.

There is no pressure difference before and after injection of the sample with the guard column in place, on a system pressure of 17bar with the union in place. There is no delay tubing.

On the C18, I do not lose the sample so I haven't done the same comparison with the guard in place and not in place.

I'm not seeing increased pressure effects but I have only injected the sample a couple of times as it is precious.

I hope this answers everything. I am also trying different buffer systems to see if the effect goes away. Thanks for any advice.

[unmgvar]

I would also do the same test with the C-18 guard only
can you tell us the result of such an experiment compare to only the union?

[carls]

Perhaps 20mM triethylamine acetate (wrt TEA) pH 5 will solve the problem on the C5.

[Christian Eckhoff]

Hi DaveM,

I have always believed in the "what goes in must come out" rule until I started working with DNA. Trying to develop an LC/MS method with a mixed mode RP/IE LC method I have learned that DNA oligos can stick to your column hard and never elute. A slight change in mobile phase conditions changed the scenario to almost no retention at all. In other words, it was impossible to find the sweet spot for mobile phase conditions where retention was just about right. I was not able to develop a rugged method and have given up on the idea for now. SAX is much easier as long as you are not using MS for detection.

You not only have to consider the polyanionic nature of DNA but also the large size of oligos that may be inappropriate for a small pore RP column designed for small molecules.