Problem encountered when dosing my gDNA by spectrophotometry - Never experienced such thing before! (Jun/22/2009 )

Hi all,

My colleagues and I are experiencing a new problem when measuring the OD of our gDNA samples for a new experiment. This gDNA was extracted from blood using the Flexigene kit (Qiagen) between May and Dec 2007. We store our samples at a concentration of around 1 µg/µL (in reduced TE) at 4oC (which is what we call our DNA bank). Since then, we used these DNAs for genotyping on GeneChip arrays (Affymetrix) - between the time they were extracted and March 2008 - and we used them also for telomere length determination between Sep 2008 and Jan 2009. For all these experiments, we had to take the ODs and visualize the DNA on agarose gels many times. We always had good results, nice and non-degraded DNA. Now, we want to use the same samples for methylation studies. The problem is that this time, we cannot obtain good OD measurements to know the exact amount of DNA we use for sonication. The background (320nm) and the A280nm are very high, giving us low A260/A280 ratios. Plus, when we dilute the DNA to reach the ideal concentration, the spectrophotometer gives us aberrant results (for example: if I do a 1:2 dilution, the concentration will stay the same as the concentrated one or will be even higher...). This happened with our BIOTEK Ultrospec 3100pro AND our BIOTEK µQuant plate reader. I did a calibration test for both and the two spectrophotometers are accurate.

So I looked over the internet to see what could be wrong. I verified the pH of the TE solutions we use to dilute and dose DNA (=7,5). We tought it could be hemoglobin contamination (even though we didn't have any problem with our samples for the previous experiments) but I found that hemoglobin absorbs at 403nm - we don't read at this wavelength. I measured the absorbance at 230nm (apparently, the contaminants absorb more at 230nm than 320nm, which I didn't know). The A260/A230 ratios were over 3 for half of the samples, even though ALL the samples have a low A260/A280 ratio. My colleague reprecipitated a part of the samples with sodium acetate and ethanol, washed it twice, air-dried it and resuspended it again in reduced TE, but it didn't solve the problem: A320nm and A280nm were still very high (plus it seems we lost a lot of DNA in the process). In 25 years of working with DNA, she never encountered a problem like this when measuring gDNA concentrations.

I just ran a 1% agarose gel to see the quality of the diluted samples (and of the DNA from the bank as well) and the bands are clear. However, the bands of the diluted DNA are lighter in intensity than we expect when comparing with an Affymetrix's manufactured control DNA (I loaded on the gel what I thought was the same quantity of DNA for all samples and control DNA using the last concentration I measured with the spectrophotometer...)

All that being said, if any of you know what might be the problem with our samples or met a similar problem in the past, please let us know what we could do to solve this issue and be able to move on with the next step of our experiment!

Thank you!
zyka

Too much protein contamination possibly!

biotechnica on Jun 22 2009, 03:29 PM said:

Having high A280, we thought about it. However, what we cannot explain is why all of a sudden, our DNAs would be contaminated with proteins when initially (after extraction, protease treatment and several washes in ethanol) they were not. We just looked back at our past experiments results (microarrays QC and telomere length results) and they were very good. I assume that if we had had protein contamination, there would have been interference and the experiments would have failed...

you may have something growing in your sample (happens a lot when storing at 4C). that would account for an increase in 280 and other contaminants.

you may be able to clear it by re-extracting with phenol:chloroform then precipitating with etoh:acetate.

hmmm... I hope not!! But this is actually something that crossed my mind. I wasn't really sure if it could be possible but apparently, it is a common problem... I'll re-extract with phenol-chloroform the DNA from 2-3 samples to test this hypothesis.

On the other hand, we discussed the issue more today in the lab and we realized that trying to read a concentration as low as 25 ng/µL with the spectrophotometer is leading us under the detection limit, which could explaining the weird readings. What's your toughts on that?

Thank you!

zyka on Jun 23 2009, 04:28 PM said:

what were the readings before? if they were of the same order and everything worked well before then i would think that it is not the problem.

mdfenko on Jun 25 2009, 08:00 AM said:

We used to work with a DNA collection of 50 ng/µL. For this new experiment, we wanted to have the DNA diluted even more so it is the first time we work with a dilution as low as 25 ng/µL.

zyka on Jun 25 2009, 10:13 AM said:

then check the dna at the original conditions and see if is the same as in the past. if so then do your dilutions and assume they are correct. don't bother trying to read them or you can try a different method to quantify the dna.