Sword Fern Endophyte Sequencing, Fall 2024

aug 19. Dylan's Contract and Work Plan
Document - 19 august 2024

aug 19. Initial Timeline (extracted from pdf)
  1. Sample collection: August 26 – 31
  2. Sample preparation: September 1 – 6•
  3. Sequencing: September 7 – October 31
  4. Draft report: November 15
  5. Final report: December 15

aug 30. Ready to Begin
Email from Dylan (30 aug 2024)
Alright, I can confirm that I will have all necessary supplies to move forward with the sword fern / hemlock sequencing project by the end of next week. Paul, which date works best for you to do field work? I would prefer Tuesday 9/10 so that I can immediately get started on lab work the following day. If you'd like to go out sooner, Friday 9/6 would also be fine.

sep 25. Dylan Reports Adapter Problem
FYI - I needed to order a different adapter for a piece of lab equipment before proceeding with DNA extraction, so I put the samples in the -20C freezer last week after completing an initial step in the surface sterilization process. I should be able to resume lab work next week.

oct 4. Dylan Reports Extraction Problems
I've encountered some issues with the quality of the final extracts so it looks like I will need to spend some time optimizing the DNA extraction workflow. I was hoping it would be a more streamlined process of following the protocol, but troubleshooting things like this is not uncommon in molecular biology.
Background
I've been working with a small number of test samples from extra sword fern pinnae that I collected to make sure that we're getting good results before using the kit on the actual samples. I measured the spectra of the DNA extracts and calculated the 260/280 and 260/230 purity ratios. The values that I got were below the acceptable range which indicates potential contamination. The concern is that these contaminants could be PCR inhibitors.
Findings
The 260/280 ratios for DNA should ideally be above 1.8. I had values range from 1.60 to 1.86, which indicates potential contamination from proteins or phenols. Proteins would be inherited from the tissue samples. Phenols could be from the plant tissue and/or from reagents that carried over from earlier steps in the workflow.

The 260/230 ratios should ideally be around 2.0, but I had values as low as 0.56 and 1.06. This is likely caused by contamination from salts, proteins or cellulose.

Total DNA yield was adequate: 30 to 50 ng/ul. It should be 5 ng/ul at a minimum.

I think the poor quality is most likely the result of me overloading the reagents with more plant tissue than you're supposed to use. I was trying to carefully follow Younginger's protocol, and his methods clearly indicate that he extracted DNA from the entire pinna. But that meant that I was adding 3 to 12 times more tissue than is specified in the current Qiagen kit protocol (50 to 100 mg). This may be a slightly different protocol than what Younginger used, which has since been discontinued / updated.

Another contributing factor is probably the unique biochemistry of sword fern leaves - old perennial fronds that are probably rich in phenols and complex organic compounds. DNA extraction from younger or softer plant tissue is generally thought to be easier.

Next Steps
I'd like to do some further troubleshooting to optimize the protocol. This will involve varying the amount of tissue used, and potentially augmenting other steps in the protocol. For example, the kit comes with a special solution that is completely optional - it reduces overall yield but improves quality. If protocol optimization hasn't brought the purity ratios back up to acceptable levels, I can also try some diagnostic PCR. If I'm able to create PCR products using the same primers as Dalhousie, it doesn't matter much what the 260/280 or 260/230 values are. If it works, it works.

In terms of timeline, this probably looks like doing optimization work in October and extracting DNA / submitting samples for sequencing in early to mid November.

Please let me know what you think of this plan or if you have any questions.

oct 11. Brief Project Update for FoSP meeting
Current status is the same as last week. I think it's possible that workflow optimization could continue into November. The current queue at Dalhousie University is 1.3 months long, so if we submitted samples for sequencing at the beginning of December, we would likely get results in mid January. Analysis and reporting on my end would take a few more weeks, pushing out the end of the project to late January / early February.

oct 16. Hannah McConnell offers comments
I work more frequently with fern RNA extractions as opposed to DNA extractions, but I also struggle significantly with less than ideal 260/280 and 260/230 values, I expect for many of the same reasons - there are a lot of secondary compounds in the ferns that tend to make extractions fairly challenging. I would expect that to be the bigger issue than too much tissue, although it would depend on exactly what kit was used. However, with an adequate DNA yield, as Dylan has indicated he's been getting, I wouldn't be too concerned; the best way I have found to manage contamination like that is to use pretty heavily diluted samples for PCR, which given the DNA yield, should be a viable option here.

oct 21: Dylan declines
"I don't expect to need assistance from Hannah”

Meeting / planning / engaging with these 3rd parties is outside the scope of my contract. It would also significantly delay my lab work. At this point, I don't have any specific questions for these people, so it seems a bit unnecessary to spend time engaging with them. However, if it's important to FoSP that I bring these folks into the project, I can certainly make that happen. Would you like me to invoice at my hourly rate for that work?

Otherwise, I plan on continuing the troubleshooting steps that I outlined in my previous project update on 10/4, which Paul Talbert seemed to approve of in his response to that email.

oct 21. pshannon summary sent to Dylan

oct 24. more from Hannah
I think without looking at the actual nanodrop plots I wouldn't say I'm 100% sure, but from his description, I'm pretty confident that Dylan can consider himself successful. I do have a protocol I've used for RNA, but I expect I got much lower yields from my RNA extractions, from looking at notes from a former technician in our lab, who did more of the DNA extractions than I did, she usually ended up with pretty high concentrations, around 100 ng/microliter, and would dilute down to 20-30 ng, which is close to what Dylan already has - I might send just one sample at a 1/2 and a 1/10 dilution and see if one of those comes out better than the other, and then dilute the other samples accordingly.

Dylan's Protocol
Protocol is under the Resources tab

oct 28: Paul Talbert Discusses Protocols
Hi Dylan:

Sorry to be slow to respond. I'm not sure where you are in your trouble-shooting, but I see a number of differences in the DNeasy Plant Mini Kit that Younginger used and the DNeasy Plant Pro kit you used, mostly having to do with somewhat different buffers. Those seem unlikely to be the source of purity problems to me. I think you are probably correct that overloading the tissue is the likely reason for impurities. The other step I would look at carefully is the tissue homogenization and disruption.

You say that the Younginger "methods clearly indicate that he extracted DNA from the entire pinna." but I didn't find this to be entirely clear. He says that he surface-sterilized pinnae and placed them in tubes for disruption of "fresh tissue" in a Tissuelyzer II bead mill, "immediately preceding DNA extraction... according to the manufacturer's instructions." Both the DNeasy Plant Mini and DNeasy Plant Pro kits specify 100mg or less of wet tissue, so he may have used less than an entire pinna. Regardless of whether he extracted DNA from the entire pinna or from just 100 mg, I think I would try testing DNA prepared from <100mg of tissue as recommended to see if that solves your problem. It's possible that the Pro Kit may be more sensitive to the quantity of material than the Mini Kit.

The Pro kit offers 3 methods of homogenization and I would guess that this is a critical step for clean preps. The mini Kit recommends a Tissuelyzer II, Tissue Ruptor, or a mortar and pestle, but Younginger says he used the Tissuelyzer II. I haven't used these kits so I am not sure how this step works, and you don't mention which method you used. The Tissuelyzer II and Powerlyzer 24 Homogenizer appear to be a separate pieces of (probably expensive) equipment. Do you have access to one of these, or did you use the vortex method? Grinding tissue with a mortar and pestle is used in a lot of different DNA preps, but you'd need liquid nitrogen and this is usually used for larger quantities (grams, not milligrams) of tissue than just an individual pinna, so I'm not sure that would be a good choice even if it were recommended in the Pro kit.

You also mention that you could do a PCR test to see if you get products (of the expected size). If you have the primers, this seems like a quick one-day way of seeing if your samples are sufficiently pure.

Calling Qiagen customer service wouldn't hurt either. Sometimes they have no better idea than you do , but other times they may know how to help, or at least be a sounding board for your ideas of how to trouble shoot. They probably have the best information about differences between the two kits, homogenization methods, and tissue quantities.

Let me know if I can be of help.

oct 28: Liz V's questions for Paul Talbert
That's a good detailed response you gave to Dylan. I hope he replies to you.

Will you clarify something for me? (just me, I took the rest off the address list)

I never have done this sort of work, so although I can follow the biochemistry of a protocol, I don't know the overall steps, or exactly what Dylan is supposed to do himself, before the product from Dylan gets sent somewhere else.

Nov 4: Dylan replies to Paul Talbert
Thanks Paul,

I'm back from my conference and ready to make more progress on this project. Could you please clarify - are you all interested in organizing a meeting with one of those other biologists that Paul Shannon was in touch with? I don't want to waste time troubleshooting the extraction protocol if our approach is going to potentially change after that meeting.

Nov 4: Liz to Dylan

From my point of view, the suggestion of talking with Hannah from the di Stilio lab was to save you time in optimizing your protocol. That lab has been extracting fern DNA for some time.

But --- it's possible you will be able to optimize by one or two more runs through your own protocol, so consulting the other lab may not be necessary.

I believe Paul Talbert will or has been in touch with you about talking via Zoom, to offer help and see where you are at this moment in the process(es),

So, I'll wait to hear more from Paul T and from you if I'm included.

Nov 13: email to board from Paul Shannon

I mentioned this at the meeting on Saturday, but after discussion, and (I think) some general agreement, I failed to make a formal motion. Which could be this:

We ask Paul Talbert to specify for Dylan, approximately this:

  1. what a validated DNA extraction from fern tissue looks like. This will include successful amplification of fungal ITS-28S rRNA, bacterial 16S rRNA, and oomycetes 28S rRNA from generic fern tissue. Dylan appears to already be working on this, which is all to the good. Making it a formal requirement will be wise.
  2. How much DNA, with what QC statistics, to send to Dalhousie.
  3. If Dylan accomplishes these two tasks, with good quantities and good QC, he proceeds with the next steps: sequencing at Dalhousie, taxonomic analyses of the result.
  4. If he is unsuccessful, we pay him a modest amount for the work he has done, cancel his contract (assuming we can) and turn to Marianne and Joey at WSU, or the UW eDNA group for a fresh approach.
I propose this for board discussion and a vote. Which probably needs to wait for the December 14th meeting, since I failed to get this on 11/9. Perhaps Paul Talbert can inform Dylan, a bit more formally than he already has, that these are our expectations? Perhaps we do not need a vote.

This is a fair amount of fuss, for which I apologize. It is worthwhile, in some form, however: this check on the extraction process raises the odds of a successful sequencing project.

Nov 17: Paul Shannon's followup email
Dear FoSP board,

May I return this to your attention? Only Chris Hollinger has responded thus far.

As board members we have the commonsense obligation, and likely also the fiduciary responsibility, to manage the sword fern endophyte sequencing contract with care. 25% of our bank account is involved.

You may reject my proposal. Or suggest modifications. All that is fine, of course. Let democracy - our collective views - work this out.

Please let us know your position!

Nov 20: Dylan reports to Paul Talbert
Last week he tried DNA preps again using only 100 mg of fern pinna as recommended in the Qiagen kit, and he used the special solution to help get rid of phenolics. Not surprisingly, since there was less input material, he got a lower but still adequate yield. However, the purity ratios are still off, actually a bit worse. It is hard to know how much of a problem this is until we get the primers later this month and try amplifying the target sequences, but meanwhile he plans to contact Brett Younginger, the only person we know who has extracted DNA from sword fern, and Qiagen, who makes the kits used by Dylan and Brett (and thousands of other plant molecular biologists) to see if they have any advice. This is what I would recommend as well. We don't seem to know exactly when the primers will arrive, but before the end of the month. Dylan had no objection to handing the project off to someone else if he cant make progress.

Dec 8: Dylan's monthly report
Yes, primers arrived last week. The lab manager and I also inspected the nanodrop to verify that it is providing accurate readings. It had 260/230 values in the expected range using an independent DNA standard, so that suggests that it is definitely our challenging sword fern samples, and not the machine.

I didn't hear back from Brett Younginger, but I'm not too surprised given how busy he is these days. I did talk to Qiagen customer service though. They had several encouraging suggestions:

The person that I talked to said that the low 260/230 values might not be very meaningful, especially at relatively low yields such as this. She also thought it was less likely to be polyphenol contamination, and more likely to be residual ethanol. In any case, she suggested trying out PCR and if it works, it works.

I've also reached out to the sequencing vendor to check what their quality control requirements are. Even if we can show that we can do PCR, they still might not accept the DNA extracts (I wouldn't if I was them). It's possible they will require us to send purified PCR products, in which case I will need to order a PCR cleanup kit. These are fairly expensive, so I would like to first see if the suggestions from Qiagen improve the 260/280 and 260/230 values.

I'm hoping to put in another lab day the week of December 16, but I'm pretty slammed with end of year reporting right now, so I might not be able to do more lab troubleshooting until early January when my schedule opens up.

Feb 4, 2025: Dylan's last email, contract cancellation offered Hello Paul,

Last week, I ran two PCR programs designed to carefully follow the vendor's workflow. In this case, I used their same primers to target the 16S region. I also included a rigorous positive control from a bacteria mock community standard and a negative control.

Unfortunately, I'm having trouble imaging the gels, so I can't confirm that the PCR worked yet. I think it's a problem with the DNA dyes that I purchased for this project. The reason that I'm confident it's a problem with the dye is that I can barely see the DNA ladder, even when I load the gel with 6x the concentration that I normally use.

So my next step is to order a different DNA stain and try again.

I realize this is probably not the level of progress that you were hoping for. If you prefer, we can cancel the rest of the contract. However, I thought you should know that I will be continuing this troubleshooting regardless, because I also have samples from Mercer Island to sequence.

Please let me know if you have any questions or suggestions.

Yours,

Dylan Mendenhall
Founder | Ecologist
Haven Ecology and Research LLC
Email: dylan@havenecology.com
Cell: 206.225.4277

Feb 8, 2025: Contract suspended [This note is reconstructed from memory by Paul Shannon, 2 jun 2025]
At the February 8th FoSP meeting, the board voted to suspend Dylan's project, due to lack of progress, missed milestones, his lukewarm (at best) pursuit of solutions to obstacles he has encountered.