Which result is correct-Western or RT-PCR - (Sep/15/2006 )
Hello, dear all : 
I got a weird results right now.
I want to check the expression of my target gene in 3 kind hepatocytes. When I use the antibody to check by Western and all of them show the expression of my target gene. And I think this signal is specific because I have positive control and also I use blocking peptide as negative control. However, when I use reverse transcription-PCR to check my target gene from mRNA level, only one of the hepatocytos and the positive control show the expression of my target gene(~1000bp), other two hepatocytes never show the correct size bands(~1000bp), but have small bands around 600bp, which is not the correct size I wanted. I changed PCR condition and even try the primers published, this two hepatocytes never show the correct results.
Now I am still not sure whether this two hepatocytes have my target gene or not. which results is correct, western or RT-PCR?
Does anyone experienced this kind problem or could you please give me some suggestion?
Thanks a lot! 
I have no experience with a problem exactly like the one described but I do have some experience in rearrangment in vivo of my introduced tandem array DNA.
In this situation I would believe the Reverse Transcription PCR as
1- a correct band size was obtained with one colony
2- work has been conducted to optimise conditions and no change has occured with the cell line giving the wrong band size.
(One question though, which is not apparent, in all those optimised situations, has the single "right" colony given the "right" answer?)
Western really only tell
1- that the the epitote the antibody recognises is present.
2- how much of that epitote (and there for protein which carries said epitote) is present
No where does a western require that the whole protein be present.
My believe is that your construct has suffered some form of rearrangment/deletion. One that renders the protein product truncated (ie the mRNA is now only 600bp rather then a full lenght 1000bp). This truncated protein is still being produced in the 2 mutant (wrong) cell line, and that truncated protein is sufficient for antibody recognition and binding by the western.
(Was there any shift in band distance in the western?)
In conclusion, I would say two cell lines are duds... Only the cell line with the possitive result for reverse transcription is correct.
But sequence the construct in your cells. The sequence data will tell you exactly what is going on.
Hopes this helps
P.S: Umm.... The gene's mRNA is not spliced into a 600bp, does it?
yes in that case you should see a band at different molecular size, smaller one.
If without any mistake that you doesn't care, I think that your results are reliable because you have repeated them times. There are some possible reseans: (i) mRNA alternative splicing may cause the 1000 bp fragment into 600 bp if the fragment you amplified covered the splicing region; (ii) It is known that mRNA degraded fasterly than protein, so it is possible that one can't detected some mRNA while the coded proteins by those mRNA are still detectable.
Hope this helps!
In this situation I would believe the Reverse Transcription PCR as
1- a correct band size was obtained with one colony
2- work has been conducted to optimise conditions and no change has occured with the cell line giving the wrong band size.
(One question though, which is not apparent, in all those optimised situations, has the single "right" colony given the "right" answer?)
Western really only tell
1- that the the epitote the antibody recognises is present.
2- how much of that epitote (and there for protein which carries said epitote) is present
No where does a western require that the whole protein be present.
My believe is that your construct has suffered some form of rearrangment/deletion. One that renders the protein product truncated (ie the mRNA is now only 600bp rather then a full lenght 1000bp). This truncated protein is still being produced in the 2 mutant (wrong) cell line, and that truncated protein is sufficient for antibody recognition and binding by the western.
(Was there any shift in band distance in the western?)
In conclusion, I would say two cell lines are duds... Only the cell line with the possitive result for reverse transcription is correct.
But sequence the construct in your cells. The sequence data will tell you exactly what is going on.
Hopes this helps
P.S: Umm.... The gene's mRNA is not spliced into a 600bp, does it?
Thanks, perneseblue, Kathy and zhongmindai !
I would like to give more details about my experimnet.
1. I want to check the endogenous expression of my target gene in these 3 hepatocytes(all of them I ordered from ATCC recently). I didnot transfect any constructs in to the hepatocytes.
2. My western results shows all 3 hepatocytes and positive control have the same size band.
3. The total RNA was prep. by RNAeasy kit from Qiagen. When I did RT-PCR, I also use GDPH as control and the control looks pretty good in all the samples, which means my total RNA should not have any problem.
4. I used two pairs of primers for RT-PCR, one is designed by myself, only the positvie control and NO.1 helpatocytes showed the correct size band. Then I used another pair of primer(published), and also only positive control and NO.1 hepatocytes showed the correct size band.
I just could not believe this gene's mRNA have a alternative splicing in two of the three hepatocytes. If it is, probably I can start a new project to study the alternative splicing of my target gene.
I have no experience with a problem exactly like the one described but I do have some experience in rearrangment in vivo of my introduced tandem array DNA.
In this situation I would believe the Reverse Transcription PCR as
1- a correct band size was obtained with one colony
2- work has been conducted to optimise conditions and no change has occured with the cell line giving the wrong band size.
(One question though, which is not apparent, in all those optimised situations, has the single "right" colony given the "right" answer?)
Western really only tell
1- that the the epitote the antibody recognises is present.
2- how much of that epitote (and there for protein which carries said epitote) is present
No where does a western require that the whole protein be present.
My believe is that your construct has suffered some form of rearrangment/deletion. One that renders the protein product truncated (ie the mRNA is now only 600bp rather then a full lenght 1000bp). This truncated protein is still being produced in the 2 mutant (wrong) cell line, and that truncated protein is sufficient for antibody recognition and binding by the western.
(Was there any shift in band distance in the western?)
In conclusion, I would say two cell lines are duds... Only the cell line with the possitive result for reverse transcription is correct.
But sequence the construct in your cells. The sequence data will tell you exactly what is going on.
Hopes this helps
P.S: Umm.... The gene's mRNA is not spliced into a 600bp, does it?
Thanks, perneseblue, Kathy and zhongmindai !
I would like to give more details about my experimnet.
1. I want to check the endogenous expression of my target gene in these 3 hepatocytes(all of them I ordered from ATCC recently). I didnot transfect any constructs in to the hepatocytes.
2. My western results shows all 3 hepatocytes and positive control have the same size band.
3. The total RNA was prep. by RNAeasy kit from Qiagen. When I did RT-PCR, I also use GDPH as control and the control looks pretty good in all the samples, which means my total RNA should not have any problem.
4. I used two pairs of primers for RT-PCR, one is designed by myself, only the positvie control and NO.1 helpatocytes showed the correct size band. Then I used another pair of primer(published), and also only positive control and NO.1 hepatocytes showed the correct size band.
I just could not believe this gene's mRNA have a alternative splicing in two of the three hepatocytes. If it is, probably I can start a new project to study the alternative splicing of my target gene.
If you have alternative splicing, you should have a protein with a diferent size compared to the No.1, while your protein has the same size. Or do you also see another band? Is your WB clean or you have multiple bands? Maybe you could attach a picture of your WB (full membrane) and tell us which is no.1, 2 and 3.
You could have a smaller amount of correct size RNA in the two cell lines (and therefore you are not able to detect it??) but the translation rate could be higher due to several reasons... Have you tried to increase the amount of starting material to detect the gene of interest? Do you have at least a faint band corresponding to 1000bp together with the 600bp one?