Hello, in my PhD thesis I have to use ChIPqPCR to study the differences in plant's chromatin (control vs infected). To do so, I use some antibodys that bound to widely recognised histone modifications (bouth activators and repressors) besides the unmodified histone as a positive control.
I have done several experiments and, in each one of them I have obtained very good signal (% of INPUT and fold enrichment). The problem is that I haven't seen any difference with any antibody (with and without relativization to a constitutive and/or the unmodified histone).
All samples have been previously tested by RTqPCR with positive results in the problem genes (later used in ChIP). I have used as a negative control an irrelevant antibody and no antibody, with similar results (no signal).
My theory is that I have been using too much chromatin as a consequence of my way to measure it:
I took a chromatin aliquot, decrosslink it, extract the DNA and measure it with nanodrop, calculating later the concentration of potential DNA (in µg/µL) of my chromatin.
With this information I add the amount of the equivalent to 25 µg of DNA per inmunoprecipitation (the amount recommended in the dynabeads protocols).
Normaly a use 50 µL of dynabeads per IP an 0,1 µg of antibody per µg of DNA.
I think that if I have been using too much chromatin, the antibodys or the magnetic dynabeads (the antibodies probably) get saturated with the target protein, not allowing me to see differences between control and problem.
So, I have been reading similar problems and I noticed that the common method of chromatin measuring is the Bradford assay. In general the papers I read don't delve into this topic and in my lab they teach me to measure the chromatin directly (making dilutions obviously) with nanodrop (using ODs at 260 nm and additionally extracting the DNA and measuring it, expecting an approximated quantity of 15 µg/OD, (which is, more or less, which usually get).
Am I measuring the concentration in a correct way?
Do I put too much chromatin?
Does it make sense my "too much chromatin" theory?
What is the relative amount of antibody do you use (respect to dynabeads and chromatin)?
Thank you very much and of course any help would be appreciated
Hi Oscar,
Nanodrop or spectrophotometric quantification of DNA is the correct, the concentration of the extracted DNA depends on the starting cell number, and the volume of the buffer DNA is dissolved in, so there is no set concentration like 15ug/OD!! As you would know, 1 OD (at 260 NM) equals 50ug/mL for DNA. Bradford's assay is for protein estimation, and its being used in the publications you read for measuring histone, I used Bradford's assay to measure the histone proteins.
May be you have worked this out but,
1.Have you tested ChIP QPCR primers efficient?
2. How far is the region from TSS of the gene of interest? may be you need to look at several positions upstream
3. Have your initial RT qPCR results shown clearly that the gene of interest shows variation in expression? if yes then you expect possible changes at the chromatin level, if you do not see changes in RTqPCR expression then probably there is no major change in the histones.
4. Have you checked whether the antibodies you are using detect the histone modifications by western blot? to rule out the problem with your antibodies
5. Preferably use expression of one gene with respect to % input.
May be you could repeat your ChIP experiments and keep these factors in mind. Hope this helps.
Regards,
Prasad.
Hi Oscar,
Nanodrop or spectrophotometric quantification of DNA is the correct, the concentration of the extracted DNA depends on the starting cell number, and the volume of the buffer DNA is dissolved in, so there is no set concentration like 15ug/OD!! As you would know, 1 OD (at 260 NM) equals 50ug/mL for DNA. Bradford's assay is for protein estimation, and its being used in the publications you read for measuring histone, I used Bradford's assay to measure the histone proteins.
May be you have worked this out but,
1.Have you tested ChIP QPCR primers efficient?
2. How far is the region from TSS of the gene of interest? may be you need to look at several positions upstream
3. Have your initial RT qPCR results shown clearly that the gene of interest shows variation in expression? if yes then you expect possible changes at the chromatin level, if you do not see changes in RTqPCR expression then probably there is no major change in the histones.
4. Have you checked whether the antibodies you are using detect the histone modifications by western blot? to rule out the problem with your antibodies
5. Preferably use expression of one gene with respect to % input.
May be you could repeat your ChIP experiments and keep these factors in mind. Hope this helps.
Regards,
Prasad.
Thank you for your answer Prasad, allow me to clarified what I mean with 15 µg/OD.:
15µg would be the amount of purified DNA that I would obtain by decrosslink and extracting (with a purification kit )1 OD of chromatin.
I always used 25µg of DNA so I use approx 1,666 ODs per IP
In several papers they talk of µg of DNA or Chromatin as a reference unit in the IP, but I'm suspecting now that they in fact, always measure proteins, not DNA.
1. In some cases yes, I tested the efficiency, but I din´t see diferences eihter.
2. I have prove several different positions upstream downstream and in the gene (I have around 5 oligo pairs per gen) specially in the regions successfully amplified in previous similar works. Generally I make a primer pair for every exon, UTR and one upstream UTR 5' (promoter region)
3. Yes, I choose those genes that showed a clear variation in their expression (10 or 100 times at least).
4. No I don´t, but looking my results in comparison with my irrelevant antibody and the no antibody sample, the results are very positive around 2-7% respect to input (depends on the antibody) and even more with the unmodified histone (up to 50%). I think, this results are too good, that's way I suspected an excess of chromatin may be the problem.
I'm not in my lab these days but I will keep in mind all your advices in the next try.
Do you recommend me to change my measurement method?
Thanks again!!
--
- Hi I guess you should keep a look Not only on the 280nm values as measure for for your prot content in the sample but also on 230nm for the presence of phenolic traces in your chrom DNA preps. Rtqpcr works lousy if phenol has been passed over in to the sample. Also too much of glycogen can lead to problems especially when DNA is precipitated in ethanol... for important ChIp samples I would recommend column based purifications.
- Hope this helps Ales
Hello Oscar
First, yes too much chromatin theory is right. But if you are getting no background signal with IgG or other Ab. then you can use higher amount of chromatin.
Amount of chromatin also depends on protein of interest. If you are looking for transcription factor like TBP or Pol II then you will need very little chromatin. on the other hand if you are looking at weak binding then you may need more chromatin.
Do you dilute chromatin when incubating with antibody? I would suggest diluting chromatin 1:10 (final volume 0.8 to 1ml). You can use dilution buffer containing tween and low conc. of SDS. Diluting chromatin while incubating with ab definitely helps.
Also check different washing buffers. Since these buffers are designed to wash off non-specific binding and strengthen specific ones, it may be helpful to try a different set of washing buffers.
Decreasing the DNA used in your qPCR reaction may help- you could try varying amounts to see what works best. I have had to use as low as 0.5ng DNA at times to get good readouts.
Hi Oscar,
You are right in guessing of too much chromatin for your ChIP as the reason of the failure in detection of any difference between samples. The logical explanation is to just think how many histone molecules may be there in the fragments of your chromatin compare to a single transcription factor bound to a TF binding site in that stretch????? which is usually one or two in a stretch of 500bp to 1kb long fragmented DNA. if you are doing any of the histone modification analyses then I will strongly recommend to dilute your sonicated DNA (make sure to dilute after sanitation with ChIP dilution buffer and not before) to somewhere 20-50 fold and make sure you have similar dilution done for the Input controls. Also use a DNA purification kit (Gel extraction or PCR purification kit from Qiagen etc.) for final purification of your immunoprecipiated DNA after reverse cross linking and elution from beads. I am sure you have designed the primers only for amplifying the fragments of150bp or less in size and not any larger size. Hope these suggestions help you to resolve the problem. Good luck with your ChIPPPPING.....
Titration of antibody is important. Excess of Ab may lead to non-specific immunoprecipitation products. In my experience, excess of chromatin or beads can increase your background signal. Chromatin and bead titration experiments should be performed until you find the correct conditions. I also check the amount and fragmentation pattern of chromatin before performing the ChIP. Nanodrop is fine for readings.
I don't think you can have too much chromatin. We routinely measure amount of chromatin with nanodrop by reading absorbance at 260 nm, but we first denature chromatin in 0,1M sodium hydroxide. It does not directly tell you how much DNA is in your sample, but gives fast and reliable measurement to compare different chromatin preps between each other. For a single IP we use 700 ug of chromatin (that correspond to about 200 ug of DNA). As you can see we use way more chromatin then you do, but we never have problems of getting non-specific signal everywhere for different antibodies.
I would first test how much AB you are using and second how much IPed DNA you use per qPCR.
Hello Oscar,
I feel on the same line of Rushita and Alexander. No way that there could be too much Chromatin. People used 200ug of Chromatin per IP and did not represent a problem.
to measure the chromatin concentration you may definitely use nanodrop and measure it directly by reading absorbance at 260 nm (first you will need to denature chromatin in 0,1M sodium hydroxide). This it will be good enough to be consistent between different chromatin preparations. However if you want to be very precise you should go through the classical DNA extraction (Decrosslinking, Proteinase K, RNAase and phenol extraction).
The problem though it could be in the amount of ChIPped material that you use for your PCRs. In general ChIPs with Histone Abs gives very good yields.
Some people resuspend the ChIPped material in a fixed volume and than use the same ul as template for the PCR. If the amount of enriched chromatin is high, the output it could be that CTs will result too low and differences between samples could be masked by that.
What we do in the lab is to measure ChIPped material with PicoGreen and dilute to the same concentration in order load the same amount in the PCR tube. Something around 0.5ng it works fine for us.
I hope it helps
I think most of suggestions are based on experiences based on a TRANSCRIPTION FACTOR binding to chromatin and people are going all gung-ho with the notion that chromatin can never exceed the limit of IPing ability and efficiency of a particular Ab..... they are totally missing the point that HISTONE MODIFICATIONS are all over the length of the genome and has lowest selectivity when it comes to differentiate between treated and untreated genome.... moreover the histones can have a bivalent modifications and the ratio between di-methylation to trimythylation of same Lys or different Lysines can have a very different outcome of promoter activity..... I will again repeat my suggestion to use diluted chromatin
I do agree with some of the points raised by Hasan. If you are using H3-core antibody and getting almost 50% percent of input immuno-precipitation, then the amount of immuno-precipitation is too much. The question of using too much chromatin may arise here. In this case, you should dilute (1:2 to 1:10) based on the kit or beads you are using for the exp.
Another important criteria for ChIP is the quality of antibody. Don't trust on it blankly. The antibody should be fresh and of ChIP quality. Regarding measurement of chromatin, bradford method can be followed and 25-100 ug of chromatin can be used based on the ChIP kit or beads you are using for the experiment.
Histone modifications are so abundant throughout the genome that when you perform ChIP-seq you need to obtain much less reads then for transcription factors/epigenetic regulators. Like Hasan mentioned, you can dilute your ChIP when you set up your plate, and perhaps you might be able to see more of a difference between your samples. You can also design primers as positive/negative controls-- for example if you ChIP H3K27me3, you can design primers on inactivated x chromosome (if your species contains).
Another simple suggestion is the number and stringency of your washes. In our lab, we wash with 2X with low salt, high salt, LiCl, and TE buffers prior to eluting. I have seen other protocols which use much less stringent conditions.
Good luck with your thesis
Thank you all for your answers and suggestions, sorry for the delay I'm on my vacation and I can hardly write in English...
I'm going to reply everyone in order:
Ales, I already use column based purifications although I had all these problems at the beginning of my thesis.
Shreyas, I don't understand why the absence of background signal indicates that more chromatin can be added, I think that if the incubation and washing buffers are stringent enough; the chromatin amount could be irrelevant.
I dilute chromatin always10 times precisely to avoid the negative effects of SDS in IP (lowing its concentration from 1% to 0,1%), the incubation buffer have triton-100 at 1%. All my buffers are very similar to the ones I saw in others works.
Rushita, when I have my DNA prepared for PCR, normally I cannot properly determine its concentration (I guess it is too low). So, I always resuspend in 100 uL and use a final concentration of 4% (approx 0,4 uL per reaction of 10uL). With this parameters, my Cps values vary between 25 and 29.
Hasan, my chromatin dilution factor is limited by the maximum dilution factor of magnetic protein G dynabeads (10 times according to Invitrogen), so using 50 uL, my top volume would be around 500uL, I don't know if this top dilution issue is very important, I just read it in the web protocol. Is for example the cause I don't just put 1mL in every IP.
My oligos are designed normally for a 70-150 bp amplicon although sometimes I use larger, until 300 (usually oligos successfully used from other works).
I really really hope you are rigth about Histone modifications, I am thinking to low from "my" 25 ug, to 5 ug (1/10) will you recomend me more dilution according to my results (more than 1/10 I mean)?
Miguel, is interesting your point, because I have been trying to calculate what is the "real" limit of AntiBody that beads can attach; according to Invitrogen's web, 0.25-0.3ug of Antibody/uL of dynabeads is the capacity. Usually people put approx 1-10uL (I will asume 5) of antibody per IP and the AB concentration is usually 1 ug/uL.... So, using 50uL of dynabeads and 5uL of antibody, the used binding capacity of the beads is only 33-40% (0,1ug of antibody/uL of dynabeads).
In summary, I thinks that I am using an excess of beads, and my limiting factor is the antibody.
About nanodriop, how do you measure the chromatin, directly, extracting the DNA..? With what absorbance do you read?
Alexander, thank you very much for your advice about denaturing chromatin. Out of curiosity, why you do not measure directly a chromatin dilution?
My problem, besides not getting differences between samples is that with my current conditions I do not have enough chromatin to put the same quantity that you. (as maximum I get 200 ug with a starting material of 3g)
The qPCR DNA quantity is the one I said to Rushita.
Carmelo, as I said to Rushita and Alexander, I use a fixed elution volume and measure the DNA for the qPCR as a % of that. For measuring the DNA I denature and decrosslinking by incubating ON at 65ºC with 0,2M NaCl; then I use a column kit.
Thank you for the info about picogreen, I didn't know it and may be extremely useful.
Suvendu, I'm worry about the variation in the % of input recovery, the 50% data, comes from my first successful IP in wich I tried two different concentration of AB anti H3 0,04 and 0,2 (ug AB/ug chromatin), the fist one gives me a mean of 21% and the second 37%. So, with 5 times more AB, I got 2 times more signal. What would be the explanation?
Jaclyn, my species is Arabidopsis thaliana but I have some RTqPCR confirmed inactivated genes and, I already use just that antibody but the results are exactly the same as with the other AB... I made some modifications (Making only one LiCl wash.) with the washes because with the last I lose some dynabeads.
What stringency conditions do you use?
Also, I wonder how much it matters the incubations order; I do first the AB plus crhomatin incubation ON and, then I do the AB-chromatin plus dynabeads incubation during 2-4 hours (both at 4ºC).
I think that's all, again thank you very much.
One question not addressed so far here is whether your chromatin fragmentation has been effective (or possibly too effective). If you reverse crosslink a sample of fresh chromatin and run it on a gel, what is the size distribution? If your fragments are all too big (or too small?) then this might be the reason that you cannot detect differences between samples. This information is also important to know to understand whether your qPCR assay fragment lengths are reasonable. If the chromatin fragments are small compared to the length of the qPCR product, the apparent amounts assessed by qPCR will be greatly reduced.
Hello David thanks for your answer, I admit it's one of my biggest concerns (beacause of the difficulty of solving). I'm using a standardized conditions, chosen after multiple tries and similar to other works:
Diagenoda Biorruptor ultrasonic bath. At medium power 2 Cycles of 5 minutes (30 sec. on/ 30 sec. off). Altogether they are 300 seconds of sonication. Max volumen 300µL in 1.5 mL tubes.
Just like my chromatin and DNA measurements, I do two checks, one with the native chromatin, and the other with the purified DNA extracted from it. And usually the DNA sizes are smaller and less extended.
Initially he size was between 100 and 2000bp, and normally I could see more intensity in the center (500-1500). But it much depends on the experiment. In the worst cases I get a great smear even in the pre-sonicated aliquot and I cannot differentiate any most intense area (specially in chromatin).
Can be the extraction method (in the DNA case) the cause of the smear in the pre-sonicated aliquot? I use a PCR purification column with isopropanol (both sonicated and pre-sonicated). In the chromatin case, I have no idea what may be the cause.
In your opinion, which is the best method to value the sonicated and pre-sonicated size? Native chromatin, denatured chromatin, or purified DNA?
We don't work on plants, so that may be an issue. But we just do the reverse crosslinking step and then load that sample directly on an agarose gel. So there is no purification step.
Which is approximately the amount of chromatin you put in the gel? I mean how much do you dilute it?
I made a new immunoprecipitation with less chromatin (1/5) and my results are strange.
I still get a good recovery of DNA, but my control-problem differences are the opposite to the expected.
I made a test with H3K4me3 (general histone modification for activation) in control and infected plants and I've got more signal in the control ones with genes that have been verified in those samples with RTqPCR to be more active in infected plants.
Maybe I have to increase even more the Antibody/Chromatin ratio... but, it may be possible that the free histones of the necrotic tissue may be competing with the chromatin for the antibodys?
And did you normally relativize the results of % of INPUT with a constitutive?
What would make the most sense to me would be to always show %input control vs. infected. You can certainly compare one or more constitutive genes by showing the % input for those genes along side. In this way, I think that you preserve some sense of the magnitude of the binding that would be lost with further "normalization". Have you done constitutive controls? That might provide a clue to the presence of any non-specific effects that are confounding your results.
The ideal fragmented size of sonicated chromatin is in between 200-700 bp. Make sure that your chromatin is properly sonicated and they are same in control and treatment. Intact chromatin may interfare with the immunoprecipitation.
You must standardize each step before the IP. Those quality control steps are very much essential.
David, I do see differences between samples. The problem I have is that when I compare the %INPUT of my constitutives, they are also different (the very same difference), so relativizing the numbers with the problem genes, the result is no differences.
Suvendu I attached a picture of one of my last electrophoresis to the post, the size range is larger than the one you said, but I think that I have not a problem with Intact chromatin... However, lately I've been getting strange results, my non-sonicated sample shows also a range instead of one big band.
Do you have any idea what could be the cause?
Yes, The quality of chromatin is compromised here. You should follow some sound method for chromatin isolation. I am not sure whether RNA got contaminated here or DNA got fragmented. You should get a intact band of gDNA after decrosslinking treatment of chromatin. The sonicated chromatin size should be between 200-700 bp with high intense fluorescence around 500 bp.
I hope your problem remain sit in this step. Best of luck!
I have used Epigentek kit for chromatin isolation. You can follow the step as mentioned there in the hand out. Most important you should get a white spot in the pellet during fourth step. It is based on sucrose density gradient centrifugation.
- Badges
- Science method