Hi all,

There are plenty of people that like to grow ethnobotanicals. Growing the correct ethnobotanical can be tricky however, since poor taxonomy and classification can lead to people growing plants that have little or no ethnobotanical use. This is often result of misidentification from plant and seed vendors and there is usually no secure way to go around this problem.

There are plenty of cases where ethnobotanical plants have inactive look-a-likes. There are plenty of people who are growing false Mimosa hostilis, false Psychotria virisis, or fake San Pedros/Peruvian torches. On the other hand, there are people who would like to ID acacias and various Phalaris grasses. The same goes for difficult-to-distinguish seeds like A.colubrina and A. Peregrina.

How do people go around this problem? Usually the plant/plant material is photographed and published on-line asking other people some help with ID. This is fine, however it cannot always be 100% correct and ambiguity often remains.

There is a way to id ethnobotaincals (without having to try them). It utilises the DNA barcoding technology; in the simplest of the terms, it involves taking a small sample from any part of the plant and sequencing a small piece of DNA from it. Taxonomists are building databases where each plant/animal/fungus etc. can be identified by virtue of a specific DNA sequence that is unique to a certain species but not others. This is very powerful tool and it is considered as the gold standard for identification purposes.

I an proposing to built a facility which will be involved in identification and taxonomic characterisation of plant material. It is going to be analogous to the Ecstasy Pill Database; if anyone is interested in ID-ing a plant, then he/she should send a small sample of the plant to the facility. The plant will be analysed and the taxonomic results will be given back to the person wishing to learn the correct species of his/her plant.

This is very straightforward procedure for molecular biologists to perform. The cost for such an analysis is roughly around £20/$30 for each sample. the cost is also very likely to drop in the future as technology tends to be getting cheaper. I am more than willing to undertake the task. Collaborations with other biologists that have access to labs would also be extremely useful and immensely helpful.

There is however a major problem before the project starts; not all plant species on earth have been "barcoded". The depositories on public databases do not contain barcoding data for all plants. Most notably, from the searches I did, there are no barcoding data for most of the ethnobotanicals. This will require some initial work to establish unique identifiers for the plants of interest. A major first goal of the project would therefore be to establish the standards (i.e. sequences that are unique to the plants of interest) against which all the samples analysed will be compared to. To resolve this gap, samples of true, tested, and correctly identified plants need to be donated and analysed to establish those standards.

But all these are perfectly doable. It should take me no more that 2 years to organise the facility (I am not currently in the position to have total freedom in the lab I work but this is just a matter of time to accomplish) and it will also take from 2 months to 1 year to "build" the standards database under which all future plant samples sent and analysed will be compared against.

What are your thoughts on this idea? Given that most of the funding for such project will be coming either from donations and/or from the people who wish to ID their samples, money would not be a problem.

This sounds like a great idea.

On this moment, me and my GF are working on an online plant-database for better identification of plants by different properties like stem, leaves, flowers, etc. With these plants will come more information like herbal properties/benefits and recipes. First we will only enter herbs but later on other plants will find their way into the database. We also know this will be a multiple year exercise and we just started.

Adding the DNA barcode and having a real-life link to a test facility would bring two worlds together.
If I can help in any way setting up this database please let me know. I'm quite exited about this idea.

UPDATE:
I have found unique identifier DNA sequences for the following:

Psychotria species (including Psychotria viridis)
Mimosa species (including Mimosa hostilis aka Mimosa tenuiflora)
Anadenanthera species (both colubrina and peregrina)
Phalaris species
Diplopterys cabrerana
Acacia species
Lophophora species

I have not been able to find unique identifier DNA sequences for:

Trichocereus/Echinopsis species
virola theidora
Banisteriopsis species

As a proof of principle this is one example of how species of the same genus can be identified:

Distinguishing A.peregrina from A.colubrina

For this distinction, 2 genes are available, the matK and the trnL gene. Their DNA sequences are different, as highlighted below:

matK gene

peregrina TTCAAATGTGCGGCTAAATCCTTCAGTGGTACGGAGTCAAATGTTGGAAAAGTCATTTAT
colubrina TTCAAATGTTCGGCTAAATCCTTCAGTGGTACGGAGTCAAATGTTGGAAAAGTCATTTAT

peregrina AATGGAAAATCTTATGAAAAAGCTTGATACAATAATTCCAATTATTCTTCTAATTAGATC
colubrina AATGGAAAATNTTATGAAAAAGCTTGATACAATAATTCCAATTATTCTTCTAATTAGATC

peregrina ATTGGCTAAAGCAAAATTTTGTAATGTATTAGGACATCCCATTAGTAAACCGGTCTGGGC
colubrina ATTGGCTAAAGCAAAATTTTGTAATGTATTAGGACATCCCATTAGTAAACCGGTCTGGGC

peregrina CGATTCATCCGATTTTGATATTATTGACCGATTTTTGCAGATATGCAGAGATCTTTCTCA
colubrina CGATTCATCCGATTTTGATATTATTGACCGATTTTTGCGGATATGCAGAGATCTTTCTCA

peregrina TTATTACAACGGATCCTCAAAAAAAAAGAGTTTGTATAGAATCAAATATATACTTCGGCT
colubrina TTATTACAACGGATCCTCAAAAAAAAAGAGTTTGTATAGAATCAAATATATACTTCGGCT


trnL gene

peregrina CTGTTTTCCGAAAACCAAGAAGAGTTCAGAAAGGGAGAATAAGGATAGGTGCAGAGACTC
colubrina CTGTTTTTCGAAAACCAAGAAGAGTTCAGAAAGGGAGAATAAGGATAGGTGCAGAGACTC

peregrina AACGGAAGCTGTTCTAACAAATGGAGTTGACGACATTTCTTTCGTTTCGTTAGTAAAGAT
colubrina AACGGAAGCTGTTCTAACAAATGGAGTTGACGACAT---------TTCGTTTCGTTAGTAAAGAT

peregrina TTCGTTTCGTTAGTAAAGAGATCCTTCCATCGAAACTCCAGAAAAGAAAGGATCAAGGAT
colubrina TTCGTTTCGTTAGTAAAGAGATCCTTCCATCGAAACTCCAGAAAAGAAAGGATCAAGGAT

peregrina GAGCATACGTATAC-GTACTGAAATACTATTTCTATTTCAATTGATTAGACCAGACAGAC
colubrina GAGCATACGTATACCGTACTGAAATACTATTTCTATTTCAATTGATTAGACCAGACAGAC


Having this information means that one has a reliable characreristic for comparing the two species and distinguishing between them. If someone sends me a sample of anadenanthera plant material (seed, leafm root, stem, etc) I will sequence it and I will be able to ID it accordingly.

In this particular example A.peregrina and A.colubrina are quite similar, yet distinguishable at the DNA level due to these subtle differences. Comparison of other species, say Lophophora diffusa vs Lophophora williamsii have much more pronounced differences at the DNA level.

That would be awesome. I am crap when it comes designing databases and web pages, so I'll ask for your help when the time comes! When the service is ready and available it would be nice to offer it through this database. And theoretically, since we already have some good DNA identifier sequences we can start the project pretty soon, that is in a year's time.

For the moment the only thing that can be added in your database is the highlighted differences in the DNA sequences among different species of the same genus, e.g. A.colubrina vs A.peregrina, or Phalaris aquatica vs P.canariensis vs P. arundinacea vs P. minor vs P. paradoxa etc)

But such an info would be useful only to people who can perform such analyses themselves. And the said people should already know how to track down such differences...!

srry, but a basic question...

how do they know that theese differences at the genetic sequence are at the level of specie and not of individual?

also, how did u got this information? You found it at a book, at the internet or you did the whole genetic sequence by yourself and found the difference?

I can find these differences because it's part of my job to be able to find such differences. But re to your question, they are found in public databases; this information is coming form the NCBI database.

And there are strong reasons to make sure that such differences are not due to individual genetic variations:

1) they are coming from consistent sequencing from many different individuals frm each species. At least 10 according to the Code of Practice of those who assign DNA consensus barcodes to a given species.

2) These sequences (just as the vast majority of sequences used for barcoding are coming from the chloroplast genomes of plants. Chloroplast genomes are much more conserved (i.e. they have very little variation among individuals). Chloroplast genomes have also lower mutation rate, so they are even less likely to result to genetic variations.

3) in this particular example, only a part of those two different genes is presented; it was chosen for presentation because it was a cluster of several differences in nucleotides grouped in this short sequence span. When the whole gene is aligned there are many more differences for making more accurate comparison.

4) When trying to assign a certain plant sample to a species, one is going to first look at all the hotspot differences. Those differences that distinguish, say, peregrina from colubrina. If just by looking those highlighted consensus the sample has, say >80% of the "peregrina" hotspot nucleotide differences, then it' sgoing to be a peregrina.

5) Peregrina and colubrina are notoriously similar, even in the DNA level. It is a difficult comparison but done correctly it can be very very accurate. It is better to analyse more than one gene (just like in this example) so that one gets a much better idea of the genetic profile of the plant sample and has the chance to compare many more differences.

Other species comparisons are much more easy. The comparison between Lophophora diffusa and Lophophora williamsii have so many more differences! It is going to be a clear-cut assignment of ID should a suspected sample gets analysed.

Hope that helps!

just out of curiosity, how small a sample can it be?

I remember this story about the PCR multiplying very small samples but dont know the numbers and neither how available this technology is to people such as yourself for example

One cell is enough for PCR. But it's easier if you have more. Anything from 100mg to 500mg of plant material is an enormous-close to insanely sufficient amount of plant material.

As far as PCRs are concerned, I do around 100 PCR reactions every day. I can do them with eyes closed and standing upside down.

While I agree DNA barcoding is useful for taxonimical purposes its not always useful for chemotaxonomic purposes. Plant with the same exact genes can produce different alkaloid profiles based on their environmental conditions. Much variation comes from this as well as genes. If one was to do what you want to do SWIM would recommend a chemical analysis along side.

Indeed. The project is only intended for taxonomical identification of plant material. Just like SWIM's FOAF is growing some Psychotria viridis. Very nice plant, but is it really a viridis? Hard to say if that is not sequenced. Information from the internet is never accurate enough to discriminate it from other species from the same genus. One needs some "true" viridis to compare it with.

Same goes with cacti. It's nice to grow the cactus you think it is san pedro, would be even much better if that was really san pedro, and not some inactive lookaline. Or if one decides to grow an Anadenanthera tree on his backyard. Would be nice to know what that is instead of waiting 10 years for it to grow and bioassay it.

Neat project. Some plants might give some headaches in extracting DNA. Maybe you should get samples of all the plant types you plan on prepping to work out your extraction protocols.

Especially dead plant material. DNA may be harder to get out in some cases depending on what fragments are broken apart etc. Live specimens would help in this regard.

That's not a problem when it comes to DNA sequencing. Seriously. I can extract DNA from any plant material, even from the 5 year old dried parsley in my kitchen. If only a couple of DNAs from the gene that I wish to sequence from the plant material are intact, that would be more than enough. People can sequence DNA from mummies and the Neatherdal genome was published last year.

And that's my kind of stuff, I personally find it easy as fuck to do and that's' why I propose to start the project. I do not see any serious problems with a said task!

do you have any idea how much you would charge ?

Yes, it should not be more than $20-30 per sample but prices are very likely to drop within the next years. The most costly part is the sequencing of the gene(s) that will be used for identification.

But I know the guy who's in control of the sequencing facility pretty well. We were getting plastered together two days ago. I may therefore be able to further cut costs down.

But time will show, this project is planned to start in a years' time. I will ask The Traveler to host this facility in his planned ethnobotanicals identification site.

I'm also a molecular biologist, and have experience specifically with plant DNA bar coding. Would love to be involved in this but as yet am not sure about the viability of using my university's facilities for such activity.

I guess my first question is one of validation. Anadenanthera is a good example of barcode data already being present and valid in that numerous individuals of each species were collected by trained botanists and rigorous published results were the outcome. Building on the database will be harder, however. If you want to add a Mimosa hostilis reference sequence to the database, how will you know it is actually Mimosa hostilis? Will you have access to herbarium vouchers? Will you collect from the wild?

The Trichs/Echinopsis are going to be very interesting. Given the level of hybridization you may well need many more loci. I can't wait to see those results, something I've fantasized about doing for so long.

Also, I gather you're aware of the current hold-up in the literature on plant DNA barcoding. There seems to be no set of loci that have yet been satisfactorily demonstrated to be a consistently good identifier among species. It is likely you'll need to screen some new species you add to the reference database for up to 4-5 candidate loci which will up your initial costs.

Good luck for this! It's something I'd love to be involved in...

Sweet, more people on board!

I have no problem to do such a thing in my lab, of course I am not in charge of the lab, but my supervisor's a pretty cool person and he knows me to be a bit of a lunatic. As long as there are no costs diverted from our current funds to this project and as long I take care of the logistics of this, everything's going to be fine.

And we're lucky that PubMed already contains information about plenty of species. I have been able to find at least 2 identifier sequences for the species of Mimosa genus
(just go here, then click on a species, then go to the Entrez Nucleotide record to browse the identifier sequences)

Which means that we already have the standards for quite a few ethnobotanicals to begin with. I plan to align the sequences Pubmed gives me within the species of a given Genus to see which one shows the most differences and can thus be used for unambiguous sequencing.

But on the other hand, adding more services (i.e. more genera) to the facility is going to be more difficult. Especially when it comes to Trichocerei.