We were pleased with the turnout of our April 2020 webinar, “Simple Microbiome Analysis with the Shoreline Complete™ StrainID™ Kit“. Mark Driscoll, Ph.D., Shoreline’s Co-Founder and Chief Science Officer, walked attendees through the simple workflow of the Shoreline Complete™ StrainID™ kit, the first DNA sequencing assay to comprehensively, rapidly, and inexpensively identify members of the microbiome down to the strain level.
Many of those in attendance came prepared with thoughtful, interesting questions. In fact, the Q&A portion of the webinar took as much as half the time of Dr. Driscoll’s presentation! For those who would like a review or simply couldn’t attend, we present them here below the recap.
Mark presented a number of case study data sets from customer applications of the StrainID kit. The first data set was from Dr. Daniel Rosenberg’s lab at the University of Connecticut Health Center, Center for Molecular Oncology. Dr. Rosenberg is a Professor of Cancer Biology and Director of the Colon Cancer Prevention Program. In one of his studies, Dr. Rosenberg is investigating the role of the microbiome in the susceptibility to colonic ulcerations in mice. Use of the StrainID kit and its resulting data output have provided his lab with key actionable information at the strain level for gut microbiome differences between different mouse types that may help explain disease susceptibility.
Mark also discussed E. coli studies in humans done by Dr. Bo-Young Hong in Dr. George Weinstock’s lab at The Jackson Laboratory for Genomic Medicine, Department of Microbial Genetics. In this study, the StrainID amplicon differentiated between pathogenic and non-pathogenic strains of E. coli.
Still another E. coli study was done in collaboration between Dr. Adam Matson and Dr. Joerg Graf at Connecticut Children’s Medical Center and UConn, respectively. Pulling together Dr. Matson’s clinical expertise as a pediatric physician and Dr. Graf’s microbiome research expertise, the pair leveraged our StrainID kit and SBanalyzer™ software to study the longitudinal microbiomes of twin pairs treated in separate hospitals. Data generated using the StrainID kit allowed their research teams to identify novel (i.e., previously unsequenced) E. coli strains and observed that longitudinal samples taken weeks apart yielded amplicon sequence variants (ASVs) with identical sequences.
StrainID has the potential to be extremely useful in the microbiome field. In Mark’s presentation, he showed how the StrainID kit helped to expand exploration of how microbiome changes affect colonic inflammation, how it can identify novel bacterial strains, and how it can be used for tracking known and unknown E. coli strains in a hospital environment.
Q: Does your kit work on all sample types?
Mark Driscoll: The kit has such a low input requirement that some of the more challenging samples, such as skin swabs, tend to work very well. That’s because there isn’t a lot of sample handling and loss of DNA at multiple steps. There are basically two protocols. The first protocol is for samples like fecal material (people have tried putting in human infant and adult, fish, snail, mouse), sand, whole flies, fly guts, and cheese, where a spin step is needed to get debris removed. The second protocol is for samples like skin, spit, or some liquids that do not have debris and don’t need a spin step.
Q: How could this kit be modified to look at the microbiome from salt water?
MD: The approach would be relatively straightforward: salt water is a liquid input, so just drop 50 µL into the kit and go.
Q: Will your kit work on plant microbiomes?
MD: Plant microbiomes are sometimes easy and sometimes complicated. If you have microbes that are inaccessible because they are on the inside of some woody stem, you’re going to have to get access to that. If you want the microbes on the surface of a root, it’s simple—just put it into the kit like you would a fecal sample.
Q: How extensible is your workflow to “non-standard” systems? For example, we study the microbiome of ticks so sample input is going to be an issue.
MD: We have done a few studies on ticks (nothing published). You would probably do the tick microbiome similar to how we do the Drosophila microbiome. Essentially, you could use a whole tick: just squish it up and put it in our kit.
Q: Can metagenomics studies be achieved by these kits?
MD: For metagenomics, you need double-stranded DNA and this lysis is really tuned for PCR in that DNA will be single-stranded. If you are looking for a metagenomics-compatible kit, we have a different DNA prep kit in development for that. We have a unique-yet-gentler bead beater system called the Shoreline Breaker™ that will yield high-molecular-weight double-stranded DNA. The instrument is small in size, benchtop, and inexpensive. We plan to release this new instrument and sample prep kit product in Q4 of 2020. Please contact us for additional information.
Q: Have you considered even longer amplicons than the StrainID amplicon? Is there much strain variation a little bit further out?
MD: Yes! Based on this, we are doing some work right now on getting longer amplicons. It’s doable, so stay tuned because there is some interesting stuff coming up!
Q: Resolution of mapping depends on a good quality database. How is Athena™ different from existing databases?
MD: Chances are, at the level of resolution of the StrainID kit, you’re going to find a tremendous number of new strains that just won’t be in the database. So [for strains not in the database] what the database serves to do is mapping as best you can to known strains [in the database] and it’s going to say it’s unknown at the strain level or species level. This is especially true for fly and mouse microbiomes where almost nothing has been sequenced.
We have curated this [Athena] database so that when you get a call, it’s going to be for a really well-characterized genome, typically sequenced on a PacBio system, pulled out from the existing databases, and with all the tRNA genes and ribosomal RNA genes placed inside the 16s cassettes. When you map that, you get as high a resolution as possible. However, you’re still going to get, even in humans, a tremendous number of strains that just aren’t sequenced. So whatever database you have, I guarantee it isn’t big enough.
Q: Do you have any data on relationship between 16S-ITS-23S amplicon variability and pathogenicity? For example, does high amplicon variability correlate with pathogenicity, or vice versa?
MD: There is nothing in terms of specific toxin genes that we pull up with this kit. This is simply a 16S-23S name tag, if you will, for whatever the strain is. It’s still possible that certain bacteria could have lost their pathogenicity. We’re really identifying the microbe by its name and its sequence, not saying anything about its pathogenicity. You need to know what the rest of the genome looks like in order to determine that and that’s not part of what this assay is. It just tells you about the 16S gene.
Q: What is the total number of strains in the Athena database? How often do you update it?
MD: The list of strains is ever growing. There’s something like 80,000 in there now (NOTE: as of 20 April 2020). We update it whenever we update the software. It’s probably every six months that we have a new build of the Athena database. The software itself lets you map to both the SILVA database and the Athena database, so you are able to do both of those and compare results.
Q: Can you comment on the accuracy of the PacBio Sequencing platform?
MD: Because the reads themselves are a circular consensus (ccs), what you can see is that it’s accurate enough for 2500 bases that you are able to get a super high accuracy mapping for each individual read. Because the reads are near perfect, the errors you’re looking at mostly are PCR polymerase errors which is low enough (a few bases in 1000) that by the time you use the ccs reads to generate ASVs, it’s going to give you ASVs that are highly accurate so you can discriminate between strains in 1 or 2 bases. [For more information on the accuracy of the PacBio platform, please visit this link]
Q: If I find novel strains, can I get them into the Athena database?
MD: That’s something we’re working on right now! It is possible for you to build a custom database in which you can enter your own strains and have your version of Athena that you’re mapping to. It’s not automated at this point, but we are building a tool so you will be able to put them in yourself. Stay tuned!
Q: Your lysis procedure, used for long read DNA, seems less harsh compared to standard bead beating. Does your method influence the diversity, since many gram positives are not lysed?
MD: Our current method of lysis yields far more Firmicutes than other methods we compared it to, including bead beading, without losing gram negatives. We wanted to eliminate bead beating because the more you get gram positives, the more you will destroy DNA. This is an unacceptable tradeoff to us. So our lysis now is something that does not involve bead beating, enables really low input, yields high molecular weight DNA, and gets a lot more Firmicutes than we’re used to getting, without losing the gram negatives.
Q: How many samples can you multiplex with your StrainID kit?
MD: We offer a 96-sample and a 16-sample kit. The barcodes for each are unique, allowing you to combine the two and multiplex up to 112 samples in one run.
Q: Are your kits available for purchase outside the US?
MD: Yes! We work with our global distributors for interested parties outside of the U.S. Please visit our website’s international distributor page to see if we have a distributor in your region. If not, send us an email at firstname.lastname@example.org.
Thank you to everyone who attended and participated in our webinar. For more information on the Shoreline Complete™ StrainID™ kit, Shoreline Wave™, SBanalyzer™ software with Athena™ database, and all of our products, visit our shop.