For more details:

http://worldwidescience.org/topicpages/p/pcr-based+molecular+method.html

You have to do a DNAse treatment to get rid off the genomic DNA. We use the Turbo DNA-free,  from Ambion #1907 and we are able to remove all the DNA present in ur samples. You just have to follow the protocol included with the kit. After that you can store your RNA at  –80°C.

You might try homogenization in TRIzol, and phase separation with chloroform. DNA will migrate to the interphase - just avoid it when pipetting off the RNA containing aqueous phase. Alternatively, use phase lock gel to completely avoid DNA (I add 1/3 volume chloroform and 1/3 volume H2O relative to TRIzol in the phase lock gel tube - in results in a more reliable phase separation). 12000 x g 5min 4C. I mix the aqueous phase with 1/2 volume isopropanol and run through RNeasy column, then proceed with the washes and elution. I don't have DNA contamination visible on an agarose gel or with the Bioanalyzer. I also avoid DNase treatment because some protocols, especially those using heat, can degrade RNA. I use this RNA for qPCR.

TRIzol protocol found: here: http://tools.lifetechnologies.com/content/sfs/manuals/trizol_reagent.pdf

I'm not sure what lysis protocol you are using for the cells before loading the columns, anyway, a mild cytoplasmic lysis normally minimizes DNA contamination (typically one can use buffers such as 100 mM NaCl, 10 mM MgCl2, 50 mM Tris/HCl pH 7.4, 1% TritonX-100 and 40 u/ml of RNAseIn or RNAseOut).

If you have quite a lot of starting material (at least a confluent 10 cm dish) you can also use the Trizol extraction, as suggested, but beware of leaving quite some volume over the interphase, otherwise you will suck up the DNA without realizing... Also be aware that Trizol extraction leaves behind contamination from guanidine isothiocyanate, due to its insolubility in any medium except water, so you need to dilute well your RNA to minimize uncontrolled effects on your RT (that's the second reason why you need a lot of starting material).

In these cases the DNAse treatment should be more effective.

You can also go around gDNA contamination designing your qPCR amplicons spanning one or two introns, if your viral genes contain any...

Deep DNAse treatment with Turbo DNAse followed by phenol/chloroform extraction will work perfectly on your already extracted RNA samples. 

For the extraction use a mixture of acid phenol/chloroform/isoamyl alcohol. With acid phenol eventually remaining molecules of DNA will be separated from the RNAs contained into the water phase. This step complete the DNAse clean up with excellent results.

You could try to digest with RNase-free DNase I first, then phenol/chloroform to get rid of DNase.

One question back to you, sorry. How did you assess DNA contamination? By nanodrop or Qubit? Thanks!

I agree with Aditi Singh above. Digestion with Turbo DNAse  followed by concentration and extraction using PCI protocol.

You have to do a DNAse treatment of your samples to get rid off the genomic DNA .

 Qiagen RNAeasy mini kit  with Zymo R1015 RNA Clean & Concentrator

First of all i would like to know how you have detected that you are getting DNA contamination?

I have used TaKaRa MiniBEST Universal RNA Extraction Kit for isolation of total RNA which is having gDNA Eraser Spin Column (for remove the genome DNA) and RNA Spin Column (for binding RNA), which gives high purity of total RNA. for qPCR experiment strictly demand on the purity of RNA, genomic DNA digestion can be carried out by adding DNase I to the silica membrane of the Column. This two step removal of gDNA gives highly purified RNA for qPCR analysis.

if you are getting difficulty in 100% removal of gDNA contamination, the best way to exclude the the effect of DNA contamination by designing primer pairs (product at least >100bp as this will help differentiate product from primer-dimer on the melt curve) that span across an long intron from one exon to another which is long enough to be NOT amplified during the elongation step. You could track the dissociation curve, and make sure the sizes of the products by running on an agarose gel. If there is a significant contamination you would get a larger band with an added size of corresponding intron to your target amplicon size then you have reduce the extension time so that long product won't be able to amplify.

I always get good results with the DNase I, amp. grade, from Fermentas, # EN 0521 (I think it's now part of Life Technologies) using 1 unit DNase I per µg RNA and incubating for 30 min at 37°C. If I have more DNA in my sample I use 1,5 units DNase I and incubate for 1 hour. I routinely tested my DNAse-treated RNA with PCR primers not spanning an intron aftwerwards in a real-time PCR, and it was alsmost everytime clean.

Thanks to all of you guys! 

I have got a couple of tips that could help. As I said we have already tried DNAse I ( digestion on column and after RNA extraction) but still have residual DNA. We still detect DNA by PCR. 

I guess you know that Qiagen supplies kits with genomic-eliminator columns... that are very effective, especially when combined with Trizol pre-treatment of samples.

anyway: more than a few times genomic contamination is caused by troubles with sample preparation. are you sure that your cell lysis/harvesting protocol is suitable for avoiding downstream problems?

last, but not least: sometimes "difficult" sources imply probable genomic carry-over, hard to avoid even with the best practise. if this is your case, a smart primer design could help you to minimize gDNA interference in your results.

The part of the problem that we can't avoid is the absence of introns into viral genome. So, we need to use gene specific primers only.