Hi Tobias, 

In the paper you enclosed, they used 10^8 cells as the starting material. However, sometimes it is not so necessary to grow so many cells for one 3C assay. I know labs using less cells (e.g. 10^7 cells) for 3C, so that less reagents are needed. 

Please find this Nature Protocol paper in the link below. In that paper, they use 10^7 cells and only 400U of restriction enzyme for one sample. However, in this paper they didn't give the detailed protocol for generating the 3C control template from the Bacmid, but you can find it in the Method paper in your link. 

Good luck!

http://www.ncbi.nlm.nih.gov/pubmed/17641637

Hi Tobias,

I also advise you to use the protocol that Fei Sun linked to. Normally, you should be able to digest chromatin from 10^7 cells by adding 2x 400 U of NlaIII (addition of enzyme, 4 hrs digestion, second addition of enzyme, digestion over night).

Is there a specific reason why you chose NlaIII? The enzyme is a frequent cutter (4-bp recognition site) which could give you very high resolution 3C data. The big downside is, that in my experience it is very difficult to get any PCR signal after a few kb's. If your promoter - enhancer are more than 10 kb away, I strongly advise to use a less frequent enzyme (6-bp recognition site, like EcoRI, BglII). If you do need the ultra-high resolution at further distance, I would advise you to look into 4C-seq:

http://www.ncbi.nlm.nih.gov/pubmed/22961246

Compared to a high-quality 3C setup (using molecular beacons/TaqMan probes) the costs are not that much different, whereas you will get much more information.

Good luck,

Daan

http://www.ncbi.nlm.nih.gov/pubmed/22961246

Dear Fei Sun, Dear Daan,

thanks a lot for your suggestions.

I chose this frequent cutter on purpose since I am trying to adapt 3C for a viral promoter/enhancer system which is very small in comparison to human promoters/enhancers (only 400bp)... hence I need a high resolution in order to detect the "short range interaction" specifically. But that also limits my choice of enzymes (at least when I want to use "natural" cleavage sites). NlaIII is perfect in that respect because it cuts once in between the sites of interest and then again outside the sites and multiple times throughout the viral genome (low background). The fact that I am working with a virus is probably also a chance to downscale the whole thing since I have many more then just two copies (alleles) per (infected) cell.

Thanks again!