14 replies to this topic

[tube]

Hello everybody! This is my first visit here and I have a big problem, so if anybody can help me, please do so... It is important!
Imagine that you have 2 samples: one is DNA isolated from tumor, and the other one is DNA isolated from blood. There is a resonable doubt that the person who isolated both DNA, mixed the eppendorf tubes. So now-what I think is DNA from the blood, could also be DNA from the tumor. There is no way to get "fresh" samples and isolate DNA again...
So, my question would be: is there any way I can test what*s inside my tube?    I*M desperate...

Thanks.

Edited by tube, 23 February 2009 - 06:55 AM.

[perneseblue]

Hmm... interesting problem that should never ever happen. I am sorry but you will have to take this as a painful lesson. Always have very good labels.  

My simple answer is no.

A more complex answer is maybe. Do you know anything about the origins of the tissue. Do they come from the same individual? Different individuals of different sex? If the two samples come from different sexes, it should be possible to PCR amplify the SRY gene which is found only on the Y chromosome.

Some cancer do have polyploidy of some of its chromosomes. If you do know which chromosomes have been duplicated it is possible (with a bit of difficulty) to design a PCR to check for an increase in gene copy number. (Provided also if you know the normal variation in the copy number of that gene in the population).

There are probably biochemical assays to detect for specific protein residue... but I think the DNA extraction would have scrubbed it from the sample. And that isn't my field.

Perhaps, I hope somebody here has a better answer.

[merlav]

From dna...I don't think that there is a way to tell the difference betwen samples.  Thats why is very importat not only to label really well, but also to check and match the labels with the contents and work carefully.  I extract the dna/rna from tissue and from blood in different moments.  I take a day for tumor tissue, another for mucosa tissue and another for blood and a few samples per assay for example I do 10-12 tissue samples/assay and no more than 24 for blood.  Just for curiosity what type of protocol was use for each extraction?

[aimikins]

is this a homework question?  it sounds like a tricky homework question...

and yeah, I agree with everything said by perneseblue.  

the primary difference I can think of would be that you should have lots more DNA in the tissue sample (rbc's don't have a nucleus) but without more information that certainly can't be guaranteed. it would work if your protocol involed a very specific # or cells or something like that.

there are other ways to test if you know the nature of the tissue. are they methylated in different ways based on tissue specificity? you'd really just have to know more about the differences in the samples, beyond that one is from blood and the other from a tumor

[AquaPlasmid]

If you blood DNA prep is not so pure, you might be able to compare the hemoglubin OD maximum (I forget the number) of the two samples, or there are card for detecting Hb and you spot some DNA solution on it to see if the color change. Further, depending on what tumor tissue it is, some has tumor specific biomarker even both samples were from the same patient. You'll find it out, or at least try your luck, if desperate....

[tube]

Thank you all for your answers…if you can think of any other way or trick, please let me know.
I know it sound like a stupid mistake-to mix tubes, but…it can happen to anyone. Unfortunately, it happened to my lab partner, and he feels pretty bad, so I*m not trying to make him feel even worse, instead we look around for a solution.

Here are some of the answers to the questions that you have asked me:

1. Do you know anything about the origins of the tissue?  Do they come from the same individual?  Tissue=brain tumor (metastasis to be precisely, primary tumor was found in lungs). Tissue and blood come from the same person (male)

2. what type of protocol was use for each extraction?  Method of isolation by salting out the proteins is faster, less aggressive, and uses less dangerous chemicals than the method of phenol and chloroform. We did comparison of these two methods-and statistic analysis showed that the two methods gave equally good results. So, our lab decided to use method of salting out, because of the reasons above mentioned.

3. this a homework question? No, unfortumately, no. This is a real problem.

4. The nature of the tissue…are they methylated in different ways based on tissue specificity? I don*t know… Our lab only gets informations I already mentioned (Answer no.1).

Remark about Hb OD gave me some idea…Perhaps, simply be  measuring concentration of DNA in samples we can get some indication…as you*ve stated out there is much more DNA in the tissue, so the DNA concentration of that sample should be  higher…but this is not proof…

[merlav]

Yes salting out is a good protocol I used for the tissues.  The question was because I use 2 types of extraction protocol, the blood with qiagen dneasy mini and salting out for tissue, if I got a problem like yours I only need to centrifuge the samples usually the biggest pellet will be  the dna from tissue.  But in this case this will not necesarily true.

[perneseblue]

I have an idea.
What kind of test are you going to do with the DNA sample? How about getting another blood sample from the same individual. Do the test with all three samples. The odd sample out is the cancer sample.

And for an extra test, extract DNA from sample using the same volume of blood used originally. Find out how much DNA is present in the final sample. See if the value obtained is similar to any one of the two unknown samples. This will give a little weight to your guess.

[tube]

There is no way we can get more blood...Blood is usually taken during operation. And since it*s been a while, I don*t know even if the patient is still alive. Even so, it wouldn*t be so ethical to ask for more blood...
There must be another way...

[perneseblue]

why kind of test are you going to do with the DNA sample? Is it specific?

Could a comparison with a second blood DNA sample work?

[mdfenko]

could you probe for tumor markers?

[tube]

perneseblue, on Feb 25 2009, 03:05 PM, said:

why kind of test are you going to do with the DNA sample? Is it specific?

Could a comparison with a second blood DNA sample work?

PCR of one gene, and then RFLP (restriction fragment lenght polymorphism). We are searching for LOH (loss of heterozygosity). I*m afraid if we get something, we wouldn*t be able to tell if it is really LOH or something else...(it could also be mutation or something else). There is no use of doing anything untill we find out what is what...
A second DNA? No, comparison is always done with a tumor and blood DNA of the same patient.

[tube]

mdfenko, on Feb 25 2009, 04:21 PM, said:

could you probe for tumor markers?

Can you please give some example of how it is done?

[NemomeN]

It's a long shot, but the blood sample should contain genomic DNA that's undergone some somatic rearrangement at the Ig locus (antibody and TCR rearrangements).  Simple in theory, but finding out where is a bit more problematic as it can happen at many different places and in different combinations including heavy chains and light chains.  However, getting certain PCR combinations can give you an idea if you see shifting bands.  Here's a reference...they were looking at lymphomas, but the principle should still be the same.  Am J Pathol. 1995 September; 147(3): 806–814.

PMCID: PMC1870987
Copyright notice
Detection of clonal lambda light chain gene rearrangements in frozen and paraffin-embedded tissues by polymerase chain reaction.
R. Küppers, K. Willenbrock, K. Rajewsky, and M. L. Hansmann
Institute for Genetics, University of Cologne, Germany.

The other "long shot" has been mentioned before and it would be helpful to know what kind of brain tumor the DNA came from...glioma or other kind and doing some kind of quantitative assay to look for genomic amplifications or deletions (such as LOH, microsatellite instabilities, etc)

[tube]

Ok, thanks!  

This is great...I feel like Sherlock Holmes...We*ll try to solve a problem and use some of the suggested options.