How to define RNA quality for Drosophila RNA using Agilent Bioanalyzer?

[patkrat]

Hi all,

I extracted RNA from Drosophila heads, and want to do RNAseq on them. I used Agilent's Bioanalyzer to assess RNA quality, and found an expected double peak at the 18S region; I read that this is common in insects, and due to 28S denaturation, so I am not worried about that. However, I think that due to this double-peak all my RINs are N/A, and I want to ask Drosophilists or insect scientists with experience how they assess RNA quality for library preparation.

I attached an example sample to this thread; all my samples are similar, even though the RNA concentration varies greatly, and the attached sample is my highest concentration (~400 ng/ul); my lowest concentration was about 30 ng/ul, but, again, shows the same characteristic double-peak and no 'shoulder' in the fast region.


Best,

Pat

[pmiguel]

You won't get a RIN score until you clear the error (whatever it is) for the lane.

--
Phillip

[vandykitty]

Here is the answer I received from Agilent: Total RNA from several insect species will appear to be lacking 26/28S rRNA. This pattern can be explained by the cleavage of the 28S rRNA during its maturation in the cytoplasm [ Jordan,BR, et al., J. Mol. Biol 101(1):85-105 ]. In Drosophila, the doublet peak results from the processing of the 28S rRNA into 28Sa and 28Sb mature forms that migrate in a similar manner to the 18S rRNA. Unfortunately if that is the case, it won’t be possible to get an accurate RIN from these samples

[pmiguel]

Originally Posted by vandykitty

Here is the answer I received from Agilent: Total RNA from several insect species will appear to be lacking 26/28S rRNA. This pattern can be explained by the cleavage of the 28S rRNA during its maturation in the cytoplasm [ Jordan,BR, et al., J. Mol. Biol 101(1):85-105 ]. In Drosophila, the doublet peak results from the processing of the 28S rRNA into 28Sa and 28Sb mature forms that migrate in a similar manner to the 18S rRNA.

This is all true.

Originally Posted by vandykitty

Unfortunately if that is the case, it won’t be possible to get an accurate RIN from these samples

This isn't true. At least we typically see RIN scores from from cleaved large subunit rRNA samples like drosophila and they seem reasonably accurate to me.
However the red dot at the top of the OP's supplied image, above the lane, denotes an error. Until errors are cleared, no RIN will be generated.

--
Phillip

[vandykitty]

Thanks, I did have the same problem as the OP and had the error message that there was an unexpected signal in the fast region, so I changed that threshold, resulting in unexpected ribosomal ratio, so I changed that and the error cleared but it did not give me a re-calculated RIN. Do you know if there is something else I can change... because apparently the Agilent rep did not know how to help me troubleshoot.. thank you for any help you can provide.

[vandykitty]

Ok, I figured it out, responded too soon

[pmiguel]

Originally Posted by vandykitty

Thanks, I did have the same problem as the OP and had the error message that there was an unexpected signal in the fast region, so I changed that threshold, resulting in unexpected ribosomal ratio, so I changed that and the error cleared but it did not give me a re-calculated RIN. Do you know if there is something else I can change... because apparently the Agilent rep did not know how to help me troubleshoot.. thank you for any help you can provide.

Yes, that is also true -- generally if you don't get a RIN score right away the program will never calculate one.

--
Phillip

[Chinanicole]

I'm also isolating total RNA from drosophila heads and my quality check involves both a bioanalyzer run and 1% agarose gel. I look for signs of rRNA degradation (smearing on gel, weird peaks on bioA) instead of relying on the RIN score.

For the 1% gel, good quality RNA should have single large band around 1200-800 bp which corresponds to the 18S unit AND the 28S unit (in drosophila, 28S migrates with the 18S: https://www.thermofisher.com/us/en/h...rna-sizes.html). In degraded samples, the band intensity around the 18S region will be diminished AND there is a smear in the region. See attached image: the sample in lane 5 is degraded, while lanes 2,3,4 and 6 are of good quality.

On a bioanalzyer: good quality RNA you'll see two peaks around 2kb corresponding to rRNA and, if you are isolating total RNA, you will also see much smaller peaks concentrated around <200 nt and between 200 and 12kb. This is the RNA. Now compare a degraded RNA sample (sample 6) to a non-degraded sample (sample 7). See all the excess peaks - those are degraded rRNA fragments? In addition, the amount of rRNA expected for the amount of starting tissue was much lower (FU) in sample 6, again indicating degradation. Note that sample 6 is 50 heads, sample 7 is 20 heads.

[Chinanicole]

I would also go back to your data in the bioanalyzer software and zoom in on the FU axis in the electropherogram to properly visualize the small peaks corresponding to RNA like in the photos I showed above.

[nucacidhunter]

Originally Posted by Chinanicole

I'm also isolating total RNA from drosophila heads and my quality check involves both a bioanalyzer run and 1% agarose gel. I look for signs of rRNA degradation (smearing on gel, weird peaks on bioA) instead of relying on the RIN score.

For the 1% gel, good quality RNA should have single large band around 1200-800 bp which corresponds to the 18S unit AND the 28S unit (in drosophila, 28S migrates with the 18S: https://www.thermofisher.com/us/en/h...rna-sizes.html). In degraded samples, the band intensity around the 18S region will be diminished AND there is a smear in the region. See attached image: the sample in lane 5 is degraded, while lanes 2,3,4 and 6 are of good quality.

Loading on the gel is uneven making comparisons difficult. From the gel I think that sample in lane 4 is degraded and it is difficult to comment on lane 5 due to lower loading.

Edit: lane counting from samples not the ladder