Failing scale-up in insect cell culture

Hello everybody,

we have recently made a puzzling observation with some of our preparative-scale (1-2L) insect cell cultures and I had the hope that someone in this forum might have come across the same thing or be able to enlighten us as to the cause.

For some of our cultures the scale-up of expression from 100mL test-culture to 700mL culture volume (both HighFive) seems not to be working for unknown reasons. Expression, though very strong in the small scale, was lost completely in the larger scale cultures; although the same P3 virus stock had been used to infect both cultures.

Moreover, when taking said P3 virus to go back into the smaller scale culture, the expression recovered back to its original level. Expression is also fine when re-running the entire process of amplifying the virus and testing expression and TOH in the 100mL scale, starting from the P2 virus stock, which leads me to believe that the storage of the viral stocks is most likely not the cause of our problem.

Again, these observations occurred sporadic, affecting some of our projects repeatedly, whereas other cultures/infections that were processed in parallel (HighFive & Sf9) ‘behaved’ as expected.

 

Coming from 24well plates, we use flasks with and w/o baffles to culture 50-100mL in 250mL flasks, while the larger scale culture is grown at 700mL in a 2L flask.  

 

Any ideas or clues anyone ?

 

All the best,

Tim

Posted on 15-Feb-2013 13:06 CET
Tim Bergbrede

Hi Tim,

We have seen the same thing for a couple of constructs, and we were able to trace it to the volume-flask ratio.  We have 2 L baffled flasks and 2.8 L Fernbach flasks that we use for scale-up, and we realized when we put more than 400-500 mL in a 2 L flask we see reduction in expression for some constructs.  We have not really tested this systematically, but we had one construct that failed to express at all in 2 L flasks with about 600 mL and then worked fine in Fernbach flasks with 1 L.  We then realized that the other constructs that failed probably did so for the same reason. So now we only use the 2.8 L Fernbach flasks with 1 L volume for scale-up.  We had another construct that didn't even like that, and we had to go down to 600-700 mL in the Fernbach flasks to get expression. 

best,

Peggy

 

Posted on 15-Feb-2013 13:17 CET
Peggy Stolt-Bergner

Hi Tim,

I fully agree with Peggy. The lack of aeration will be the cause of loss of infection efficiency.

Fernbach bottles of 3 L  with an orange aeration cap are sold for Cell Culture. We reuse them although we sterilize the caps seperately from the bottles (both covered with Alu foil). We assemble them in the cleanbench again.

To improve the infection you could use BIIC stored virus according to the protocol from Wasilko et al. (Protein Expression and Purification 65 (2009) 122–132). This gives always reproducible infection and production experiments in large scale.

Sincerely yours Joop van den Heuvel

Posted on 15-Feb-2013 14:25 CET
Joop vandenHeuvel

Dear Peggy & Joop,

thanks a lot for your prompt replies!

In fact, the flask/culture volume ratio struck us as one of the things to look into immediately, although (as mentioned) we have never had these problems before despite using the same protocol.

So, in parallel to posting our problem and prompting some of you directly (thanks Linda!) we set up a small test series from 100mL to 800mL in 2L flasks – and cannot see a difference in the expression level between these culture conditions. That is to say, the expression is back to normal!

What we now fear might have caused these inexplicably sporadic failures of our large scale cultures just became visible today, as precipitates on the bottom of recently purchased, closed(!) media bottles. Looks a lot like bacterial growth!

Whether or not this was the underlying reason for our initial observation, as soon as we have verified that what we see now really is a contamination, we will let you know and add the provider and batch number to Peggy’s earlier thread.   

 

Cheers,

Tim

Posted on 18-Feb-2013 13:19 CET
Tim Bergbrede
Displaying 1-4 of 41