Damn QPCR

[mark-ER]

---

Members do not see this ad.

I HATE real-time quantitative RT-PCR. For real. Anyways, here is my question:

I get my RNA, measure concentration in triplicate by OD (spec) and then dilute out accordingly. Then I perform RT and run my samples on icycler (biorad) with SYBR Green. My control is 18s RNA. Now, since I normalize my RNA and assume that RT should not change anything, how come I'm seeing huge (2-7 fold) increase in 18sRNA levels between my control and experimental conditions? Shouldn't it not change, given I standardize RNA concentration prior to RT and most (95%+) of the RNA measured by spec is ribosomal RNA? What the hell am I missing here? Anybody else ran into this problem?

 

[ninebillion]

I HATE real-time quantitative RT-PCR. For real. Anyways, here is my question:

I get my RNA, measure concentration in triplicate by OD (spec) and then dilute out accordingly. Then I perform RT and run my samples on icycler (biorad) with SYBR Green. My control is 18s RNA. Now, since I normalize my RNA and assume that RT should not change anything, how come I'm seeing huge (2-7 fold) increase in 18sRNA levels between my control and experimental conditions? Shouldn't it not change, given I standardize RNA concentration prior to RT and most (95%+) of the RNA measured by spec is ribosomal RNA? What the hell am I missing here? Anybody else ran into this problem?

I'm not an expert, but you may be getting an unexpected result because the SYBR Green is nonspecific. Perhaps you should design some specific oligonucleotide probes (dye-quencher combinations) that will only bind the 18s RNA cDNA.

 

[Mitro]

I have experienced what you?re explaining once, it was because I had the annealing temperature for the ?normalization? primers was a couple degrees too high. I lowered it, and got much more reasonable data. Having only used SyBr Green, Ninebillion has a valid point, you should always do the dissociation curve at the end of your PCR. This'll let you know if you have a non-specific product, which could also explain why you're getting weird results. Good luck.

-M

 

[AndyMD]

it's been a while since I did anything with QPCR, but one of my personal conclusions by the end of it was that SYBR might not be as reliable as sequence-specific primers tethered to fluorescent probes ... it sounds like the primers you're using in your 18s amplification are binding non-specifically, and making it appear (by SYBR) as if the starting quantities were 2-7x larger because the net amount of DNA amplified each cycle is due to 18s + some other unknown amplification events. What someone said before about checking the dissociation curve sounds right, as it should show you whether or not you have other products in the mix... or run a gel of the rxn product. If anything, perhaps the annealing temperature is too low for the control primers, which would increase the liklihood of them binding non-specifically. If it's not that, I'd be really perplexed... let us know if you figure it out.

p.s. I hate QPCR too! I think the reason I could never obtain any reliable results when I was doing it for a summer was because the damn fiberoptics on the QPCR machine kept getting contaminated with the fluorescent probes, which would throw off the readings from well to well. And even if you make sure to seal those bitches tightly, it won't do anything to prevent some other random user from using the machine and contaminating it. Haha I never got a clear explanation for why my experiments were so inconsistent though... anyways, have fun now.

 

[Fixed Gear]

I can't share in your lamentations, as I enjoy qPCR and have had good consistent results. Here's my $0.02
1. Spec'ing RNA only leads to trouble (as you are having). Instead, run 1 ul of your sample on a simple 1% agarose gel with EtBr, running in tandem with an RNA of Known concentration. Remember, Spec will pick up degraded RNA, free nucleotides and all sorts of crap that will lead to a faulty measurement of your true RNA concentration.

2. Try a different normalizing gene/probe. I think that 18s has Cts too low (i.e., amplification too early) to be really happy with it, so I stick to using Gapdh.

3. I'd be happy to give you the sequence of the primers I use for Gapdh if you'd like. I got them from another investigator who got them from another investigator that has been doing qPCR since I was in diapers. Just PM me.

However, I think that if you ran your diluted/normalized RNAs on a gel, you would see tremendous variation leading to the 2-7 cycle disparity you're getting. That's a solid place to start, in my opinion.