size exclusn..hw to reduce tailing.. proteins..silica matrix

[alicee!]

hi all.. i m working on sec with proteins..
i've tried acetate buffer.. phosphate buffer.. both with and without salt.. with and without 10% ethanol.. my protein gives a lot of tailing.. this is on a sillica matrix tsk columns.. any suggestions??

[HW Mueller]

Did you also play with pH?
We found that sample solvent is also extremely important in work with antibodies on TSK-Gel Super SW3000. Best results were obtained with phosphate buffered saline (0.15M phosph., 0.15M NaCl, pH 7.2) as mobile phase and as protein solvent.
Hopefully you didn´t use any perfluoro acids in your mobile phases, you may find some earlier chains were our results are mentioned.

[tom jupille]

Proteins on silica-based GFC columns are very touchy about conditions. Different proteins "prefer" different conditions.

HW Mueller's suggestions are an excellent general starting point. From there, try increasing the NaCl concentration in 100 mM increments up to about 600 mM and see if that has any effect. then pick the best NaCl concentration and tweak the pH (in increments of 0.5 units at first).

You might also want to try a polymer-based column.

[Uwe Neue]

How old is the column?

[alicee!]

i have tried these..
TSK- SWxl3000 with phosphate buffered saline (0.15M phosph., 0.15M NaCl, pH 7.2)
also, TSK- SWxl3000 with acetate buffer pH 5.2,
both give tailing and low recovery.

Yes.. I do see differences with the no. of injections in columns: one with 145 injections gives 30% recovery the other with 60 injections gives around 60%.
How long should the column be used? What to do to better this? tailing up to 1.6 is obtained.

Thanks!

[Uwe Neue]

The use of both columns is rather low for explaining problems.
How do you measure recovery?

[Kostas Petritis]

What are the dimensions of the columns and how much are you loading on them? In your other post you mention that the recovery is low when your injection amount is low (see below in the link)?


http://www.sepsci.com/chromforum/viewtopic.php?t=8974

[alicee!]

The use of both columns is rather low for explaining problems.
How do you measure recovery?

yess.. how many injections does yr sec column survive generally?
so.. like my two posts say.. there are two problems..
first.. tailing.. even if i use a fresh column.. my tailing factor does come to some 1.5 for my protein.. for the tsk column.. and buffers describd above.. so what r the preferred conditions for my run? how do i better this?

secondly.. with the 1.5 tailing methods.. what i do is load a say 30 microgram protein (diluted in the mobile phase), and my purpose is to detect aggregates which are up to 2% of this. so hwat i do is also inject a 2% amount i.e 0.6 microgram of the standard protein.. to be able to calcultae and demonstrate column/protein recovery as part of system suitability. if i consider that the area obtained for 30 mcg is 100%, and see how much 2% should give, eg.. 50,000 for 30 mcg, 2% should giv 1000 area, what i get is some 500.. so say its a 50% recovery.. (these are not the actual values i m getting) but they r near these! how do i solve this. and this recovery does vary with the number of injections on the column.. higher the number lower the recovery.. i.e i don't get a dose response/linearity thruout the range in my methods..!

wat say?!

[HW Mueller]

This could primarily be a matter of integration. The peaks are relatively broad in SEC, they disappear relatively fast in noise as you lower the injection amount.
Also, the volume you inject should be around 5µl.

How is the response when you inject without a column (same flow rate)?

Furthermore, you may have to "live" with some tailing, which, of course, contributes to bad integration at low levels.

[alicee!]

This could primarily be a matter of integration. The peaks are relatively broad in SEC, they disappear relatively fast in noise as you lower the injection amount.
Also, the volume you inject should be around 5µl.

How is the response when you inject without a column (same flow rate)?

Furthermore, you may have to "live" with some tailing, which, of course, contributes to bad integration at low levels.

i use some 7.8*78 dimension column.. dont remember exactly but near this.. and what i do is make a 0.3 mg/ml solution from around 0.6 mg/ml in mobile phase.. and inject 100 microlitres of it.. u recommend injecting lower volumes? isnt dilution in mobile phase prefered for better recovery? do i need higher concentrations of the starting solution? i can get that.. wat say!? thanksss!!

[HW Mueller]

The 5µL is for a SW column, you can put considerably more, the optimum can be found faster than the time needed for this discussion. Also I would suggest to try the suggestions made so far and telling us about the results.

[Uwe Neue]

You make a 100 microL injection for the 30 microg injection. Do you make a 2 microL injection for the determination of the peak area for the 2% value?

If you dilute the sample 50 fold, a possible source of loss is the sample vial. Proteins love to stick to glass...

It is not clear if the tailing is from silanols or from other features of the packing. For silanols, an increase in the concentration of the salt would help. If it is due to other interactions, you need to sort them out one by one. One possibility is hydrophobic interaction, which does not go away with the addition of salt, but would get worse. If this is a possibility, you can add some 5% or 10% methanol. Don't know what this will do to your assay of the aggregates, but it will help sorting out the cause of the tailing.

[alicee!]

You make a 100 microL injection for the 30 microg injection. Do you make a 2 microL injection for the determination of the peak area for the 2% value?

If you dilute the sample 50 fold, a possible source of loss is the sample vial. Proteins love to stick to glass...

It is not clear if the tailing is from silanols or from other features of the packing. For silanols, an increase in the concentration of the salt would help. If it is due to other interactions, you need to sort them out one by one. One possibility is hydrophobic interaction, which does not go away with the addition of salt, but would get worse. If this is a possibility, you can add some 5% or 10% methanol. Don't know what this will do to your assay of the aggregates, but it will help sorting out the cause of the tailing.

i dilute 50-fold yes.. so i wud try lower injection volumes.. and i have gone from 100-200 mM salt.. i wud try higher.. i have tried higher pH 5.2 to 7.2 i cant go higher there.. and i have tried 10%ethanol.. the tailing increases.. indicating hydrophobic interaction not being a cause. also.. i have been trying acetate and phosphate buffers for these conditions.. wat else?

[hplcuser]

(0.15M phosph., 0.15M NaCl, pH 7.2) is very common. You can increase salt if the column allows (Sepax gets up to 2M salt) which gives better elution with higher salt. http://www.sepax-tech.com/SEC.php

Hope this helps!