Hi, deal all. We are trying to detect a secreted protein, about 20-35kDa, in the cell culture. What we used was the Amicom ultra-10kDa. We rinsed the filter with ddH20, centrifuged cuture supernatant for 60min at 5,000g, and finally got a 30-fold concentrated viscous supernatant(500ul). We took 20ul out, mixed with SDS-loading buffer and boiled them. However, after boiling, the mixture was too dense to load in gel. We looked for solutions but found people treated their samples in same way. We really wonder wether their concentrated samples were viscous and how they loaded sample. Actually, we aslo tried to dilute the recovery and then boiled it. This time we could load gel, but nothing could be detected, except the recombinant protein control. We don't know what's wrong with preparation. - -
Thanks a lot.
Did you filter the culture media and/or give it a high speed centrifugation prior to concentrating to make sure that any cells that may be present were removed?
Have you measured your protein concentration by e.g Bradford or BCA assay so you can have an idea of how much protein you are loading on the gel?
30x concentration does seem a bit excessive, your protein could have crashed out of solution if it was too concentrated, especially if it is perhaps not maximally stable in culture media (which is not really designed to buffer proteins). A better idea may be to concentrate the media somewhat and top it up with a physiological buffer that your protein is likely stable in. Do this concentration then top up a few times until the media has been exchanged for the buffer. If you can't/don't want to measure the protein concentration, can I suggest perhaps taking a sample of the concentrate from the concentrator every 10 minutes or so while it is in the centrifuge and preparing an SDS-PAGE sample for each? That way you can load a series of different amounts of protein on your gel and see which one is optimal.
Dear Herui Wang
How do you preparate the cell surnatat before load it in the Amicon 10KDa?
E.g : did you centrifuge at high speed? and do you filtrate the surnatant at 0,.2uM before load it the 10KDa amicon?
Also addiction of DNAse could be usefull but since you are running a SDS-page and DNAse can give you an additional bands, and you have to consider it in your data analisys and/or add a DNAse amount that is not visible in SDS-page
best regards
Manuele
When it comes to viscous protein samples, I would guess
1. the protein is too much
2. the denaturing process has to be adjusted
For suspect 1., you can try to run a dilution sets, e.g. 0.5X, 0.1X, 0.01X, and denature them with the same condition.
For suspect 2., you can try boiling longer, e.g. 15 mins, or try to denature with lower temperature like 70 degree celsius 15 mins, or the combination of both time and temperature.
Good luck with you experiments!
Dear Samuel Davis
Thanks for your suggestions.
We did a centrifugation before filter. It's about 2,000 rpm for 10min. After filter, the proten samles became quite viscous and the concentration was around 70ug/ul (BCA assay). We would try taking samples out every 10 mins as your suggest. And we are thinking about centrifuging twice: 100kDa first and the passage for 10kDa. We want to get rid of the large proteins in samples.
Dear Yi-Chun Cheng
We would try longer boiling and lower temperature. Thank you for your suggestion
Hi! It looks like DNA made your samples viscous. Did you monitor live-dead cells when you exposed them to a serum free medium? If they were fine, then try to centrifuge at no more than 500g 5-10mins to pellet all live floating cells, when taking out supernatants leave some volume in the tube (for 15ml tube I leave 300-500 ul). Then proceed with high speed centrifugation and again leave some volume above the pellet. 30-50x is a good concentration factor for broad cell types (for hepatocytes it can be used lower). There is no need to get rid of larger proteins. As it was mentioned above by @Samuel Davis, you should measure total protein concentration (1÷10 dilution of sample aliquote might be necessary to fit in linear range on a calibration).
Hope this helps!
- Badges
- Science method