Illumina Trueseq RNA prep kit problems

[cburke11]

Hello all,

I have recently gone through the Illumina Trueseq RNA sample prep kit with no success. Has anyone else had problems with this kit/what were any problems encountered?

A little bit more about what I did:

I extracted with Ambion's RNAqueous 4-PCR kit but do not have a bioanalyzer to assess total RNA quality prior to the protocol which I have heard can have a huge effect on yields. Instead I ran a 2% agarose gel in TAE and stained with Ethidium. I got one bright, non-smeared band (I am working on Lepidopteran species which have a "hidden-break" in the 28S band, so it migrates with the 18S band) for all of my samples, so I assumed the RNA was ok. I ran 10 uL of sample on this gel with RNA concentration being from .332-1.66 ug/mL.


Questions:

1) Would it have been better to run a denaturing RNA gel? I talked to Illumina tech support and they recommended this. Could I see denaturation better?

2) I standardized the total RNA put into the Trueseq protocol to 1 ug... should I have used more starting material? This is on the lower end of the recommended range of .01-4ug.

3) Our department just acquired a nanodrop so I speced the RNA that had been stored at -20C for 2 weeks since I tried the Truseq protocol. The 260/280 readings were from 1.25-2.03. I don't have a lot of experience with these machines but think the range I am looking for is 1.8-2.1 , does this mean that some of my samples are contaminated with excess salts/proteins/something else?

Any advice would be appreciated. I know crap in is crap out, so does it seem like my starting material is ok/are there any additional methods I can use to check it?

Thank you!

[Jon_Keats]

Finding a way to get a bioanalyzer run would be a good idea but in the absence try the denaturing gel only after nanodrop readings meeting the following requirements:
Concentration: 200-400 ng/ul (dilute to range)
260/280: >1.8 min better if >1.9
260/230: >1.9 min better if >2.0 (this one is important, tells if you have contaminating ethanol that will mess up the down stream steps)

[pmiguel]

260/230: >1.9 min better if >2.0 (this one is important, tells if you have contaminating ethanol that will mess up the down stream steps)

Hi Jon,
Ethanol doesn't absorb at 230nm itself, but acetate does. So if you are using sodium acetate as your precipitation salt, then it might show up at 230 nm. But that does not mean that the sample still has ethanol in it (although it might). It just means that the acetate salt precipitated and was not sufficiently resolubilized by the 70% EtOH washes.

Also, the first step of the TruSeq RNA kit is polyA RNA purifications. This is hybridization based, so I am not sure a little ethanol and/or sodium acetate would ruin this step.

--
Phillip

[pmiguel]

Hello all,

I have recently gone through the Illumina Trueseq RNA sample prep kit with no success.
[...]
Instead I ran a 2% agarose gel in TAE and stained with Ethidium. I got one bright, non-smeared band

Yes, we see that with many insect species. But what size was this band? You did run a molecular weight standard (ladder), right? If not, the single band might be genomic DNA for all we know...

I ran 10 uL of sample on this gel with RNA concentration being from .332-1.66 ug/mL.

Is this right? 1 ug/mL is 1 ng/uL. That would mean you needed 1 mL of your sample to get to 1 ug.

Questions:

1) Would it have been better to run a denaturing RNA gel? I talked to Illumina tech support and they recommended this. Could I see denaturation better?

Better, sure. But if you ran a ladder and your putative rRNA band was a reasonable size, then that should have sufficed.

2) I standardized the total RNA put into the Trueseq protocol to 1 ug... should I have used more starting material? This is on the lower end of the recommended range of .01-4ug.

If it really was 1 ug, not 1 ng (see above), then that should have been enough.

3) Our department just acquired a nanodrop so I speced the RNA that had been stored at -20C for 2 weeks since I tried the Truseq protocol. The 260/280 readings were from 1.25-2.03. I don't have a lot of experience with these machines but think the range I am looking for is 1.8-2.1 , does this mean that some of my samples are contaminated with excess salts/proteins/something else?

Any way for you to post the spectra for us? They will be saved on your nanodrop computer. Generally 260/280 below 1.8 or so may indicate protein contamination. But there are so many other possibilities that the metric is nearly useless in isolation.

Any advice would be appreciated. I know crap in is crap out, so does it seem like my starting material is ok/are there any additional methods I can use to check it?

We take an aliquot at each stage (where possible) of the protocol. That way if something goes wrong we can see what the sample looked like at many of the steps in the protocol. However without a bioanalyzer this may be of limited use to you.

Was your result no band at all after the final 15 cycles of PCR?

--
Phillip

[cburke11]

Reply to Questions

Yes, we see that with many insect species. But what size was this band? You did run a molecular weight standard (ladder), right? If not, the single band might be genomic DNA for all we know...

That is a really good point. I ran a DNA ladder and the band ran at 700 bps but I heard that DNA and RNA run at different rates. Guess I will splurge for a RNA ladder! But I think this would eliminate the possibility of genomic DNA as I would be seeing a really high molecular weight band?

Is this right? 1 ug/mL is 1 ng/uL. That would mean you needed 1 mL of your sample to get to 1 ug.

I entered this wrong, the concentration of my samples of total RNA is in ug/uL. Good catch.

Any way for you to post the spectra for us? They will be saved on your nanodrop computer. Generally 260/280 below 1.8 or so may indicate protein contamination. But there are so many other possibilities that the metric is nearly useless in isolation.

I will try and post this today.

We take an aliquot at each stage (where possible) of the protocol. That way if something goes wrong we can see what the sample looked like at many of the steps in the protocol. However without a bioanalyzer this may be of limited use to you.

Was your result no band at all after the final 15 cycles of PCR?

We had our final samples coming out of the Truseq protocol analyzed on a bioanlyzer which we will unfortunately not have access to in the future. There is also a gel at the very end. I will attach that to this post.



Additionally, something that caught my eye is that when we looked at our sample concentrations on a Qubit via RNA assay we got a concentration reading of 1.66 ng/uL and when we used the nanodrop on the same sample we got a concentration reading of 44.6 ng/uL. The nanodrop reading was two weeks after the Qubit but I can't imagine RNA/DNA being created in the -20C in this timeperiod. Would this also indicate some sort of DNA/protein contamination?


Best,
Crista

[tboothby]

Additionally, something that caught my eye is that when we looked at our sample concentrations on a Qubit via RNA assay we got a concentration reading of 1.66 ng/uL and when we used the nanodrop on the same sample we got a concentration reading of 44.6 ng/uL. The nanodrop reading was two weeks after the Qubit but I can't imagine RNA/DNA being created in the -20C in this timeperiod. Would this also indicate some sort of DNA/protein contamination?

Just a heads up, different methods of determining concentration (Bioanalyzer, nanodrop, Qubit, qPCR) can give you very different readings. I have experienced this myself and heard about it from a lot of other people.

Cheers,
T

[pmiguel]

That is a really good point. I ran a DNA ladder and the band ran at 700 bps but I heard that DNA and RNA run at different rates. Guess I will splurge for a RNA ladder! But I think this would eliminate the possibility of genomic DNA as I would be seeing a really high molecular weight band?

Yes. And for normal gel electrophoresis the single stranded stuff runs at about 1/2 the size of double stranded stuff. (Which makes sense to me, because it is 1/2 the MW.)

[...]

Additionally, something that caught my eye is that when we looked at our sample concentrations on a Qubit via RNA assay we got a concentration reading of 1.66 ng/uL and when we used the nanodrop on the same sample we got a concentration reading of 44.6 ng/uL. The nanodrop reading was two weeks after the Qubit but I can't imagine RNA/DNA being created in the -20C in this timeperiod. Would this also indicate some sort of DNA/protein contamination?


Best,
Crista

An A260 reading is subject to the greatest number of confounding factors. In this case you may have substantial protein amounts in your samples. But it seems more likely that your samples contain a large amount of degraded DNA and/or RNA. Since it is the nucleotide bases that absorb the UV, degrading the strand they compose does nothing to their absorption. (Well, it may slightly increase it.) Did you do a DNAse digestion of your RNA?

Fluorimetric assays (eg, ribogreen) will generally not detect mono or oligonucleotides. So you may be starting with way less RNA than you think you are.

--
Phillip

[cburke11]

I had really low 260/230 readings, is this guanidine contamination and has this been know to affect the Truseq protocol?

The Ambion RNAqueous 4-PCR kit uses a guanidium-based lysis solution...

[pmiguel]

Guanidium salts are strong protein denaturatants, or so I have been led to believe. However, given that the first step of the TruSeq RNA prep involves a hybridization of oligo dT/ magnetic bead purification, it seems likely anything other than very high concentrations of inhibitory salts would be washed away.

I have seen cases where 230 nm absorbing ions (eg, high concentrations of acetate) were concentrated enough that the 260 nm reading was basically a shoulder off the 230 nm peak. That would explain why your fluorimetric reading was so much lower than your spectrophotometric reading.

If you presume the 1.66 ng/ul fluorimetric reading was correct, how much total RNA did you use?

--
Phillip

[cburke11]

I have been using 1 ug which is in the range that is specified to work in the Truseq protocol which is from .1-4ug total RNA.

How does the RNA bind to the beads? If there is a really strong denaturant would there be a protein adapter that is inactivated?

[pmiguel]

I have been using 1 ug which is in the range that is specified to work in the Truseq protocol which is from .1-4ug total RNA.

Ah, but that would be using the nanodrop concentration, right? If we presume your fluorimetric result was correct -- and fluorimetry tends to be much less confounded by contaminants that spectrophotometry -- then you actually only used about 37 ng of total RNA. Possibly too little to generate a library visible after 15 cycles of enrichment PCR.

How does the RNA bind to the beads? If there is a really strong denaturant would there be a protein adapter that is inactivated?

Well, I can't rule that possibility out, but the old saw says "if you hear hoof beats, suspect horses, not zebras". Since your spectrophotometry result was 27-fold higher than your (probably accurate) fluorimetry result, I would say "very little RNA" is the horse here.

Still, if you live near a zoo, or in certain parts of Africa, zebras hooves may be just as common as horse hooves. And, there is nothing to say you didn't get run over by both horses and zebras.

--
Phillip

[pmiguel]

How does the RNA bind to the beads?

Sorry, I did not answer this. The oligo dT is biotinylated and binds to the beads via a streptavidin molecule.

--
Phillip

[cburke11]

Ah, but that would be using the nanodrop concentration, right? If we presume your fluorimetric result was correct -- and fluorimetry tends to be much less confounded by contaminants that spectrophotometry -- then you actually only used about 37 ng of total RNA. Possibly too little to generate a library visible after 15 cycles of enrichment PCR.

The library was standardized using the floroumetry reading to 1ug.

[cburke11]

Still, if you live near a zoo, or in certain parts of Africa, zebras hooves may be just as common as horse hooves. And, there is nothing to say you didn't get run over by both horses and zebras.

I sure hope I don't get run over by both, one horse or zebra seems to be enough to be trampled by!!

[pmiguel]

I forgot to ask:

How long did you do the RNA fragmentation step for? What method did you use to heat the samples?

--
Phillip