SDS, Cell numbers, dilution - (Jul/28/2009 )
Hi guys, I've been reading this board and it's a wealth of information on this method. Thank you all for your help both in the past and in advance.
Like most people, I'm using the upstate kit to perform my first ChIPs, and like most people I'm running into problems. Don't even get me started on Upstate's fast-talking about how they give you enough reagents for 22 IP's but only enough for 10 lysis/sonication events.
I'm using LNCaP cells and searching for nuances in p53 acetylation at the p21 locus upon HDAC inhibitor treatment. LNCaP cells are very very picky, and they do not like to be too crowded (or too sparse for that matter) in the culture dish. I can fit 10e6 cells in a 10cm dish, possibly 2X10e6 if I'm pushing it. Scaling up, I can get 4.5X10e6 cells into a 15cm dish, but that's REALLY pushing the limits of these cells. If I'm talking 96 hour timepoints, I can't put any more than 10e6 into the plate or they overgrow and start dying - as you can imagine, this kind of screws with the p53 status of the cell.
But this relates to my problem: If I'm using 10e6 cells/dish/sonication, I don't have any capacity to dilute the SDS lysis buffers in ChIP dilution buffer. This means I'm heading into the IP steps at a full SDS concentration. Won't this interfere with the antibody binding?
The protocol they supply is built around 10e7 HeLa cells and they only give passing remarks on what to do in other cases, with none of those remarks actually providing any solutions.
Has anybody successfully and reliably performed ChIP on a sample without usng the ChIP dilution buffer? Does anyone have any suggestions here?
MunkySpunk on Jul 28 2009, 07:02 AM said:
Like most people, I'm using the upstate kit to perform my first ChIPs, and like most people I'm running into problems. Don't even get me started on Upstate's fast-talking about how they give you enough reagents for 22 IP's but only enough for 10 lysis/sonication events.
I'm using LNCaP cells and searching for nuances in p53 acetylation at the p21 locus upon HDAC inhibitor treatment. LNCaP cells are very very picky, and they do not like to be too crowded (or too sparse for that matter) in the culture dish. I can fit 10e6 cells in a 10cm dish, possibly 2X10e6 if I'm pushing it. Scaling up, I can get 4.5X10e6 cells into a 15cm dish, but that's REALLY pushing the limits of these cells. If I'm talking 96 hour timepoints, I can't put any more than 10e6 into the plate or they overgrow and start dying - as you can imagine, this kind of screws with the p53 status of the cell.
But this relates to my problem: If I'm using 10e6 cells/dish/sonication, I don't have any capacity to dilute the SDS lysis buffers in ChIP dilution buffer. This means I'm heading into the IP steps at a full SDS concentration. Won't this interfere with the antibody binding?
The protocol they supply is built around 10e7 HeLa cells and they only give passing remarks on what to do in other cases, with none of those remarks actually providing any solutions.
Has anybody successfully and reliably performed ChIP on a sample without usng the ChIP dilution buffer? Does anyone have any suggestions here?
You could always try a sonication buffer that doesn't have SDS. The ChIP method is pretty robust and will handle changes in buffers. The buffer we use is: 150 mM NaCl, 50 mM Tris-HCl (pH 7.5), 5 mM EDTA, 0.5% (vol/vol) NP-40 (a.k.a. IGEPAL), 1.0% (vol/vol) Triton X-100, plus protease and phosphatase inhibitors. Also, don't be afraid to use fewer cells. If you are using magnetic beads you could probably go down to 5 X10^5 and if using magnetic beads with the Fast ChIP method, probably down to 1 X 10^5 (this is what I have used).
What volume do you generally sonicate in and what's the density of the cells in said volume?
I'm using 5X10e5 in 500ul and I'm getting good sonication after 8X10s pulses. I'll obviously have to do a shearing assay to optimize the fragment size for the new buffer, but can you give me a ballpark figure? Does your lysis buffer tend to require more or less sonication than its SDS counterpart?
MunkySpunk on Jul 28 2009, 11:51 AM said:
I'm using 5X10e5 in 500ul and I'm getting good sonication after 8X10s pulses. I'll obviously have to do a shearing assay to optimize the fragment size for the new buffer, but can you give me a ballpark figure? Does your lysis buffer tend to require more or less sonication than its SDS counterpart?
I was doing about 5-10 X10e6 cells per ml. I never compared sonication in the buffer I use to buffer with SDS but I have heard that SDS can increase sonication efficiency. So it's possible you would need to increase your sonication time.
SDS definitely affects ChIP efficiency. When a ChIP is working, the SDS final concentration is not greater than ~0.3%.
i'm using a kit which recommends sonicating 1million cells in 50ul of lysis buffer. then dilute this to the number of cells you want to use per ChIP. On averge 150,000 cells per ChIP works really well.
hope this helps
If that's the case (and it sounds quite plausible), the SDS concentrations are probably leading to my ChIP inconsistencies.
The Triton and the Nonidet P-40 in the alternative lysis buffer won't interfere with antibody binding, will they?
As far as the Chelex method for eluting DNA: Does it give a better yield than Phenol/Ethanol/Carrier-RNA for elution? It's quite obviously faster and easier, but in my experience, that usually doesn't equate to better.
Thanks for all your help.
EDIT: Looking at the Fast ChIP protocol, two lysis steps are required - one for cell, followed by spin and aspirate, then another with the same buffer, for nuclear lysis I would assume.
As both these steps rely on the same buffer, does anyone have any strategies to minimize nuclear lysis with the first step so as to ensure I recover as much genomic DNA as possible (as opposed to aspirating it up after the first spin)? Is it a function of time? i.e. do I just pipette a few times and throw it directly into the 'fuge?
as for chelex...it's faster than phenol; however sometimes it's hard to keep the same amount cosistent from sample to sample. invitrogen's chip kit it uses magnetic beads for DNA recovery which helps with yield.
epi123 on Jul 31 2009, 07:03 AM said:
A number of colleagues have begun using Fast ChIP with magnetic beads (e.g. Dynabeads). They get the speed and the low background. I haven't tried it yet, as I am mostly doing 96-well based ChIP, but it seems to be working very well for them.
Can I assume you're referring to Flanagin et al.?
I'm pretty keen to move onto the 96-well plate format ChIP once I get my own stuff working reliably, and maybe rig up a plate-sized electromagnet out of an old steel license plate or something for the washes. Out of curiosity, how many cells do you plate in each well for this format, and how many PCR reactions worth of DNA does each well yield?
I just got done using the above mentioned lysis buffer on a new set of cells and doing a battery of sonications to re-optimize. I'll post on the results (Sonicator settings as a function of SDS or Triton buffers) once I run the gel.
the invitrogen ChIP kit not only as dynabeads for the immunoprecipiation step but also has different magnetic beads in the kit that are used for the DNA recovery of the ChIP DNA.