Chromatography Forum

Hi there,

I have been observing some weird things on my SEC results and wandering if anybody has some experience:

Method:
2 TSKgel superSW3000 columns in series + guard column
Isocratic elution, 0.35 ml/min, PBS, pH 7.4
30 min run for each sample
10 mg/ml mAb, 10 ul inj

stability samples, so some are clean mAb with little aggregation, some are stressed, with 20% aggregration. all samples were filtered through 0.45um before inj.

reference standards were injected every 10 samples in any sequence, from same vial. a typical sequence runs for 20 hours. samples were kept at 4 degrees C in autosampler.

Weird things:
1. for the reference standards, the aggregation peak (right before the main peak) started from about 5%, the main peak started from about 92% . in the seq, the main peak kept decreasing and the aggregation peak kept increasing, even for 3 consecutive inj (at the beginning and end). but total peak area (including all aggregation, main peak and smaller species peaks) kept decreasing.

2. for some stressed samples, the total peak area were even larger than their unstressed controls, protein gain?

this mAb should be stable at 4C, at least not aggregating in hours.

Is there anything in the method that could have caused these results?

Thanks a lot!

Your results do not surprise me at all, Welcome to the screwey world of SEC-HPLC !

You are seeing the column adsorbing the HMW material. I have to condition my column every time I use it. I am now using repeated injections of a "Stability sample" with increased HMW material until I get a reproducable HMW peak. This peak keeps increasing for up to 10 injections in a row. Also - this seems to be product specific, different MAb's behave differently.

I would love to get a good discussion of this going on this forum.

I posted a question to the forum a few months ago, discussing this issue.
Here is a copy of my previous posting. I recommend reading the paper about "Arginine Buffers" unfortunately it is not available as a free download. see citation below.

I have not had a chance to try Arginine Buffers yet, I am in QC and have alot of data generated using my current SOP.

I will comment that I am using the 0.1M Na-SO4, 0.1M Na-PO4, pH 6.7 mobile phase suggested by Tosoh Bioscience. I was getting significant tailing with PBS (only for some MAb's - not others).

Rande

######################################


Posted: Wed Nov 29, 2006 8:33 pm
Post subject: Pre-Treatment of analytical TSK-G300SWxl SE-HPLC column?


Friends,

I have been using TSK-G300SWxl (Tosoh Bioscience) columns for SE-HPLC of proteins, especially MAb's, for many years now. One recurrent problem has been the non-specific binding of some proteins - especially HMW peaks of aggregated IgG. When I have to "break in" a new column, it takes a long time and many injections until I can get reproducible results (% aggregate) with my reference standard.

(as shown in the D. Ejima, et al paper, the problem is not limited to this silica based column.)

Below are 3 different references that describe the problem and deal with it in different ways:

Who else has had this problem and how do you deal with it?

Thanx for your help ... all comments, suggestions or "food for thought" is welcome!


http://www.tosohbiosep.com/NR/rdonlyres ... SWXL01.pdf

http://www.tosohbiosep.com/NR/rdonlyres ... erSW01.pdf

Arginine as an effective additive in gel permeation chromatography
Journal of Chromatography A, Volume 1094, Issues 1-2, 11 November 2005, Pages 49-55
Daisuke Ejima, Ryosuke Yumioka, Tsutomu Arakawa and Kouhei Tsumoto
_________________
Rande

Rande,

Thanks a lot for your input. You might be right about the adsorption. However, what puzzled me is this happens in almost every sequence I ran, regardless of how new the columns are. I did change guard column quite often. In addition, like I said a typical sequence has 40 injections, this behavior was there even at the end of the sequences. One thing is, after each seqence, I flush system with water then PBS+0.05% NaN3.

I will try the BSA approach and the arginine buffer and post the results as soon as I have some data.

Your results do not surprise me at all, Welcome to the screwey world of SEC-HPLC !

You are seeing the column adsorbing the HMW material. I have to condition my column every time I use it. I am now using repeated injections of a "Stability sample" with increased HMW material until I get a reproducable HMW peak. This peak keeps increasing for up to 10 injections in a row. Also - this seems to be product specific, different MAb's behave differently.

I would love to get a good discussion of this going on this forum.

I posted a question to the forum a few months ago, discussing this issue.
Here is a copy of my previous posting. I recommend reading the paper about "Arginine Buffers" unfortunately it is not available as a free download. see citation below.

I have not had a chance to try Arginine Buffers yet, I am in QC and have alot of data generated using my current SOP.

I will comment that I am using the 0.1M Na-SO4, 0.1M Na-PO4, pH 6.7 mobile phase suggested by Tosoh Bioscience. I was getting significant tailing with PBS (only for some MAb's - not others).

Rande

######################################


Posted: Wed Nov 29, 2006 8:33 pm
Post subject: Pre-Treatment of analytical TSK-G300SWxl SE-HPLC column?


Friends,

I have been using TSK-G300SWxl (Tosoh Bioscience) columns for SE-HPLC of proteins, especially MAb's, for many years now. One recurrent problem has been the non-specific binding of some proteins - especially HMW peaks of aggregated IgG. When I have to "break in" a new column, it takes a long time and many injections until I can get reproducible results (% aggregate) with my reference standard.

(as shown in the D. Ejima, et al paper, the problem is not limited to this silica based column.)

Below are 3 different references that describe the problem and deal with it in different ways:

Who else has had this problem and how do you deal with it?

Thanx for your help ... all comments, suggestions or "food for thought" is welcome!


http://www.tosohbiosep.com/NR/rdonlyres ... SWXL01.pdf

http://www.tosohbiosep.com/NR/rdonlyres ... erSW01.pdf

Arginine as an effective additive in gel permeation chromatography
Journal of Chromatography A, Volume 1094, Issues 1-2, 11 November 2005, Pages 49-55
Daisuke Ejima, Ryosuke Yumioka, Tsutomu Arakawa and Kouhei Tsumoto
_________________
Rande

Rocnex 1, what % changes in total amounts did you see?
Some background: It is some years ago now that I first posted problems with such a column after using perfluoro acids in the mobile phase (peak shape, no answer yet). This was semiquantitative work so I don´t recall seeing stark changes in peak areas, though ~5% to ~10% changes I would have considered completely normal, considering the peak shape variations (and resulting integration difficulties). I would also expect that changes in aggregation should be expected unless you happen to have found a solvent (mobile phase), and other conditions, which perfectly stabilizes these. Most likely, though, a solvent which stabilizes the aggregate will give bad SEC (as an example: The chromatography may benefit from chaotropic agents, but these will likely reverse the aggregation).
Maybe someone has seen something quantitative on this: Proteins don´t always follow the Beer-Lambert Law.
One can make an extreme statement, namely that one uses the wrong mobile phase if a large quantitation problem is due to column absorption, but as mentioned, to find a good one when aggregates are involved....a very good problem.
Finally: For what reason are you looking at the aggregation?

Rocnex 1, what % changes in total amounts did you see?
Some background: It is some years ago now that I first posted problems with such a column after using perfluoro acids in the mobile phase (peak shape, no answer yet). This was semiquantitative work so I don´t recall seeing stark changes in peak areas, though ~5% to ~10% changes I would have considered completely normal, considering the peak shape variations (and resulting integration difficulties). I would also expect that changes in aggregation should be expected unless you happen to have found a solvent (mobile phase), and other conditions, which perfectly stabilizes these. Most likely, though, a solvent which stabilizes the aggregate will give bad SEC (as an example: The chromatography may benefit from chaotropic agents, but these will likely reverse the aggregation).
Maybe someone has seen something quantitative on this: Proteins don´t always follow the Beer-Lambert Law.
One can make an extreme statement, namely that one uses the wrong mobile phase if a large quantitation problem is due to column absorption, but as mentioned, to find a good one when aggregates are involved....a very good problem.
Finally: For what reason are you looking at the aggregation?

I have evaluated a number of silica type columns for mabs and found that some are more affected by NaCl than others. Biosuite 250 showed no effects with or without salt added to pH 6.8 phosphate 0.1M, but other columns showed increased or decreased binding. I had no luck with TSK due to its poor tailing and not as good selectivity as Zorbax GF250 or Biosuite 250 for my applications. Might try phosphate buffer only to see if you have a more reproducible response. I routinely monitor aggregates in the 0.5% range and have had good front to back of runs performance using columns above.

Hello HW and Oldtimer,

Thank you for your inputs.

The total peak area decreased by around 6% between the first and the last of 10 QC injections in the last seq (not consecutive injections: 3 at begn, 1 every 10 samples and 3 at the end). It's not a big big change, but the decreasing trend over the injections concerned me, there is no up and downs, just constantly decreasing.

Actually this decereasing in PA is not the biggest concern, because it didn't happen in every sequence, or not that obvious. The constant increasing in aggreate peak is my problem. Again, the increase might not be a big number, but it's a trend.

I am looking at the aggregate because we are doing stability study on these mAb samples and aggregate is one of the major stability issue. We quantitate amount of aggregate in the samples over time/temperature to check if the mAb is stable.

Regading the mobile phase, PBS is 10 mM phosphate with ~150 mM NaCl. I didn't try 100 mM phosphate buffer only. One colleague of mine who is working in analytical mentioned higher salt often helps separation with this column. I haven't done extensive try on this with my mAb yet.

The best results with my mab and fab was with a mobile phase AND sample solvent which was 150 mM in both phosphate and NaCL (high ionic strength PBS). It seems that you will not get around playing more with the mobile phase, as well as with the solvent used to dissolve your mab. If you dissolve your mab in a solvent which tends to favor absorption on the column you are just out of luck.
Regarding aggregation, I can´t see any alternative to finding a solvent in which it does not take place. There is enourmous literature on mab, including free material from manufacturers. My recommendation is to check into this for more ideas.