Your FFPE Questions, Answered
In a recent webinar, you asked our experts, Eva Haenssler, Ph.D. Associate Director of New Nucleic Acid Solutions and Martin Schlumpberger, Ph.D., Director of Exosomes and ccfDNA Product Development for more information about formalin-fixed, paraffin-embedded (FFPE) samples and how QIAGEN products can help improve quality and quantity of nucleic material for downstream applications when derived from FFPE samples. Here are the answers.
1. Can FFPE and cryo-preserved samples be used interchangeably? Will they show the same quality for downstream applications?
Schlumpberger: No. Because of the formaldehyde involved in formalin-fixed samples, FFPE samples will likely perform quite a bit worse than the cryo-preserved samples. However, whatever precautions you take to make FFPE samples work better, such as looking for short PCR amplicons, these can also be used on cryosections. If you look for longer amplicons, however, you will see a difference. FFPE samples don’t amplify well.
2. How do you address the variable yield and quality of FFPE samples?
Haenssler: Some factors cannot be influenced. If a sample is stored at elevated temperatures or the fixing process has not been ideal, there’s nothing you can do. You cannot reverse these effects. You can, however, use an optimized protocol. QIAGEN’s protocols include careful deparaffinization using QIAGEN’s Deparaffinization Solution, Proteinase K digest and crosslink removal steps optimized for DNA and RNA.
3. How does the QIAGEN Deparaffinization Solution work?
Schlumpberger: While we cannot release the exact chemical composition for the Deparaffinization Solution, it is designed to dissolve the paraffin and keep it liquid so that it won’t interfere with the steps that come after. It is essential that it penetrates the tissue so that all of the paraffin is removed, allowing aqueous buffers to interact with the samples.
4. After deparaffinization, do you need to remove it, or can you leave the deparaffinization solution there? Will it interfere with anything?
Schlumpberger: It shouldn’t affect anything, although it may make some steps easier if you remove the Deparaffinization Solution before recovering your lysate. Usually, the steps for deparaffinization and lysis work as follows:
After deparaffinization, you will have the deparaffinization solution with paraffin as the top layer. If you have too much sample or too much paraffin, it may become waxy or solid again. In those cases, you may need to use more of the deparaffinization solution. Keep in mind you shouldn’t use too many sections in an isolation, to not overload your procedure.
Once complete, you will have the lysate phase with the tissue flake usually at the bottom. After that comes the Proteinase K treatment, and what happens then will depend on whether you’re doing RNA or DNA protocols. If you do a limited proteinase K treatment for RNA, some tissue fragments remain that still contain DNA and even after you perform a longer PK treatment for DNA solubilization, there may still be something left. Unless you conduct an AllPrep (AllPrep DNA/RNA FFPE Kit), you may disregard the remaining tissue. It should no longer contain any of the desired nucleic acid.
5. Can RNA and DNA extracted from FFPE samples be used for next-generation sequencing (NGS)? Does quality matter?
Haenssler: The general answer to the first part of this question is yes, you can use the RNA and DNA from FFPE samples, but it depends on the sample and the NGS application you’re looking to use. For targeted panels, for instance, the amplicon length and quality of nucleic acid from FFPE samples is often okay. However, if you have severely degraded samples, you might not be able to amplify your targets and your library prep.
QIAseq panels for RNA-seq, DNA-seq or Methyl-seq targeted sequencing utilize single primer extension (SPE) to get complete and uniform coverage from difficult samples like FFPE or Liquid Biopsy and unique molecular index (UMI) technologies to minimize false positives by removing PCR artifacts during analysis. We offer more details on NGS applications, workflows, discoveries, related kits and more on our NGS content page. Remember, the size and quality of nucleic acid fragment from FFPE samples will vary and generally be of lower quality than nucleic acids from fresh or frozen samples.
6. Are there any special tricks or tips to improve RNA and DNA isolation when using your kit?
Schlumpberger: All the tips and tricks we know or that we figured out are part of the protocol. When we encounter something that makes it work even better, we try to include it in the protocol. There is no generic tip we hold back for situations like this. What I can say is that specifically aside from initial sample quality, the critical parts of the protocol are:
- Perform the Proteinase K treatment for exactly as long as noted in the protocol. Don’t extend it longer than necessary because it may affect the nucleic acid quality over time.
- Reach and hold the correct temperature during the crosslink removal step. E.g. in case of the RNA isolation protocol, it’s not sufficient to turn your heat block from 55 to 80°C and then start the timer for 15 minutes. You should make sure that your heat block has reached 80° and then start the incubation for 15 minutes. Especially for the RNA, crosslink removal is a balance between being as efficient as possible removing crosslinks while taking care to avoid fragmenting the RNA any more than it already is from the FFPE fixation and storage.
What we see for qPCR, in particular, is that too much crosslink removal starts having a negative effect on RNA integrity. However, crosslink removal also gives you a slightly better result on qPCR if done correctly, but it can be challenging to find the right balance.
7. How and when is the Uracil-N-Glycosylase (UNG) treatment performed? Is it an on-column treatment? Is there a specific protocol provided?
Haenssler: The UNG treatment is carried out within the sample preparation. We have the QIAamp DNA FFPE Advanced UNG kit that was recently launched. It contains the enzyme, together with the sample prep kit. After lysis and crosslink removal, there is a five-minute incubation period with UNG before loading the sample on the column. It’s part of the workflow. We also have a similar kit - the EZ1&2 DNA FFPE UNG Kit - for the EZ1 Advanced XL and the EZ2 Connect instruments.
8. Is it possible to isolate miRNA from FFPE? What is the difference between the RNeasy and miRNeasy FFPE Kits?
Schlumpberger: When we developed the first generation of RNeasy FFPE Kit during the very early days of miRNA. We developed a protocol to recover everything useful for qPCR with a goal of 70 nucleotides. That protocol already recovers quite a bit of the miRNA content, but then we realized that to get all of the miRNA, we need to treat the conditions a bit more and the result is the miRNA FFPE protocol. We are looking into creating more kits and protocols for miRNA in the future, however.
9. What are the initial steps involved in formalin fixation and paraffin embedding?
Schlumpberger: We as molecular biology researchers often have no control over these steps, as is the fixation is conducted according to a pathologist’s wishes and needs. But, even pathologists have guidelines that, unfortunately, they don’t always adhere to very strictly. In general, you don’t want the formalin treatment to go for more than 24 hours. If it is much longer than that, you get more crosslink formation and some crosslinks will become irreversible in that process. I don’t think the exact chemical reactions are clear, but this is something we observe very clearly. In practice if pathologists receive biopsy material on a Friday, it will probably be in formaldehyde until Monday. That’s one of the factors why we see human samples with much lower nucleic acid quality.
It would be best if you didn’t use any unbuffered formaldehyde, of course.
What is also important is the size of the tissue or biopsy you want to fix in formaldehyde. Formaldehyde penetrates tissue at a rate of about 1mm per hour for the first few millimeters. After 5 mm, it will slow down a lot. In the thinnest dimension, the general recommendation is that the tissue should not be more than 5 mm to make sure that the formaldehyde penetrates quickly enough to prevent autolysis of the tissue. If it’s thicker, you’ll have fixation on the outside and degraded goo in the middle. This won’t provide any decent quality nucleic acid anymore, especially RNA.
10. Regarding extraction of prokaryotes in FFPE tissue samples: Has anyone had any success with it?
Haenssler: Personally, no. And it’s quite challenging. You would probably have a low amount of prokaryotes in your tissue sample. Depending on the downstream application, if you lyse your background cell, you would have an excess of human/host DNA/RNA material. If you’re going for 16S sequencing or qPCR with a bacterial DNA/RNA target, you may have a chance, but it also depends on if you know what bacterial species you’re looking at. You may need to add extra lysis steps if they have a robust cell wall. This is definitely a challenging experiment. If you’re going with a modified protocol, you will have to ensure you have sufficient heating steps for bacterial cell lysis.
Schlumpberger: I concur. We also haven’t looked at it closely. Another factor could be the thickness of your sections. Usually, when you look at eukaryotic cells and have a section between 5-10 µm, this is around the thickness or a bit less than a thickness of a eukaryotic cell. The majority of the cells in your sample will be cut open already. That would make it very easy for the proteinase K to access everything and degrade all of the protein in your section. Bacteria are smaller and that part may still be necessary: to cut through the cell wall, which may be a limitation. The smaller the bacteria and the tougher the cell wall, the more you may need additional lysis steps.
11. Can you speak about qPCR inhibitors in FFPE samples?
Schlumpberger: The primary source of inhibitors in FFPE samples are crosslinks. They become modifications on the nucleic acid. Other inhibitors like tissue-specific inhibitors are often already removed in the embedding process when the tissue is dehydrated and perfused with paraffin. Removing the paraffin with another solvent will remove other remaining inhibitors, so usually, this will be less of a problem. We do not see many of these issues, but if you have strongly inhibitory components, such as e.g. stool present for whatever reason, that is very inhibitory and additional steps may be needed. We don’t have an FFPE protocol for removing this type of inhibitor at the moment.
12. Is DNA and RNA isolation with the AllPrep protocol available for the QIAcube Connect?
Haenssler: Yes. The handbook will provide you with information on how to download the protocol. For example, check out the AllPrep DNA/RNA Mini Kit.
13. How should samples be stored?
Schlumpberger: Generally, we recommend storing FFPE sections or blocks cold at 4° or -20°. At -20°, blocks can be used after ten years. Eventually, there will be a decline in integrity. Still, generally, if all the fixation was correctly done and blocks are stored safely, there will be a chance to get usable nucleic acid from them years later.
14. Can QIAGEN RNA stabilizers work with FFPE samples?
Schlumpberger: We have RNA stabilization reagents like RNAprotect Tissue Reagent and the AllPrep Tissue Reagent, but they’re used for non-fixed tissues – so without formalin fixation. And of course, you can store non-fixed tissues in RNA protect tissue reagent for at least a year at - 20°. You could probably store it up to 5 years at that temperature, but if it’s going to be more than a couple of years, I recommend moving it to -80°. Generally, those can be used for quite a long time as long as they are kept frozen throughout. And while RNA stabilization reagents keep RNA intact, they’re not designed to preserve tissue morphology, like formalin does. The PAXgene Tissue System from PreAnalytix preserves both tissue morphology and nucleic acids.
15. Is it possible to recover extracellular vesicles from FFPE samples?
Schlumpberger: I wouldn’t know how to do it. Since the vesicles would be crosslinked with the proteins I’m very skeptical that they could be separated from the rest of the section without destroying them. Even when you’re embedding them in paraffin, the sample is dehydrated and in those steps, I would assume the vesicles are enveloped and the phospholipids are being dissolved and extracted. So there would be no vesicle anymore. You see the cell membrane, so it may still be there, but I still don’t see how you would get intact vesicles into the solution in the first place. So we haven’t tried it and I don’t believe that it’s possible.
16. Does QIAGEN offer a UNG treatment?
Schlumpberger: Yes, we have kits that include it. Individual kits will describe at what point UNG treatment is performed, although, generally it’s after the crosslink removal step.
17. Can Proteinase K digested lysate be stored? How should it be stored? Will it help?
Schlumpberger: For RNA, add the binding buffer as well because it helps keep RNA denatured. After that, it would be safe to freeze the lysate when it’s ready for binding to column or beads.
Haenssler: Speaking for DNA, it’s not beneficial, but it also doesn’t harm anything to do the Proteinase K digestion overnight if that fits in your schedule. It doesn’t have a positive impact, but it also doesn’t damage anything.
Schlumpberger: You might see a bit more fragmentation for RNA, but it’s similar and you may get more DNA solubilized. The RNeasy FFPE DNA removal is pretty robust and will deal with that.
Haenssler: In general, for DNA extraction, there are stopping points, like after crosslink removal, you can store your lysate at 4-8° or freeze it for up to 4 weeks.