Most of the caterpillars were new to science
March 17, 2019 7:10 AM Subscribe
The Janzen–Connell hypothesis is an attractive but difficult-to-prove explanation for rainforest diversity: It's dangerous for rainforest trees to grow up near their parents, because that makes them vulnerable to the herbivores who are already crunching and munching on their parents. A team from the University of Utah has added some detail [paywalled] with a study in Panama. "“People may think of a jungle like it’s a giant salad bowl. It should be paradise for pests because they’re surrounded by leaves. But plants have an infinite number of defense combinations—half the weight of a young leaf is poison”.... Over five years, the researchers collected leaf samples in the field..... Using high performance liquid chromatography, they separated all of the distinct compounds inside the leaves.... [O]nly 4 percent of the Inga compounds were known to science."
I think there is something to the hypothesis that the major upside of a largely vegetarian diet is not just the displacement of high calorie low fibre food but that it has to do with plant-life's defensive poisons functioning as low level antibiotics and chemotherapy that nip diseases in the bud.
But I am also sure that is just some ill informed quackery on my part and someone here will set me straight.
posted by srboisvert at 9:04 AM on March 17, 2019 [1 favorite]
But I am also sure that is just some ill informed quackery on my part and someone here will set me straight.
posted by srboisvert at 9:04 AM on March 17, 2019 [1 favorite]
Set you straight? Nah, no conversiion therapy here. Later today soup from flowering broccolini, a deadly brew for the enemies of my biome.
posted by Oyéah at 10:37 AM on March 17, 2019
posted by Oyéah at 10:37 AM on March 17, 2019
"Metabolomics, a relatively new field of science […]"
That's a great line and I must use it myself. "JiA, have you any source for that?" "It's metabolomics, a relatively new field of science. You probably haven't heard of it."
posted by Joe in Australia at 3:40 PM on March 17, 2019 [1 favorite]
That's a great line and I must use it myself. "JiA, have you any source for that?" "It's metabolomics, a relatively new field of science. You probably haven't heard of it."
posted by Joe in Australia at 3:40 PM on March 17, 2019 [1 favorite]
I once heard a biologist say, "-omics is basically a suffix we put on complicated stuff we don't understand."
posted by clawsoon at 4:14 PM on March 17, 2019 [1 favorite]
posted by clawsoon at 4:14 PM on March 17, 2019 [1 favorite]
Metabolomics isn't all that new of a field. It's a matter of cost and throughput.
Major pharmaceutical companies already have massive small molecule libraries (self link).
The thing with the claim "[O]nly 4 percent of the Inga compounds were known to science." is that it's technically true but
1) the vast majority of those compounds don't really do anything (but are produced by plants because they might do something - or a recombination of the genes that are involved with their synthesis might yield molecules in their offspring that might do something, and
2) Once you discover a small molecule that does something interesting, the question is how to synthesize it in useful quantities? Many of these small molecules are traditionally very hard to make through pure chemical synthesis. Also, it might do something interesting but doesn't do it well enough to be useful, so you have to modify it. This can be complicated/ not-worth it.
The classic examples are acetylated salicylic acid - the acetylation makes it a more selective COX-2 inhibitor and less so COX-1 - making it safer and less prone to causing abdominal pain and diarhrea.
The other is heroin (acetylated morphine). Funny story - Merck's original intention was to acetylate codeine, but ended up with acetylated morphine. Acetylated codeine is a potent toxic side product of bad heroin production from opium feedstock.
Most of these low hanging fruit modifications have already been studied; more nuanced modifications of small molecules tend to be difficult, the moreso the more complicated the small molecule.
2.5) Due to the difficulty in chemically synthesizing potentially useful compounds, another approach has been to study the metabolic pathways that lead to the end-product creation, cloning out those genes, and essentially "fermenting" the molecule of choice. With advances in genetic engineering, you can make targeted mutations in the genes in order for the enzyme to perform a different chemical reaction (or make the reaction more efficient).
I worked on Daptomycin (Cubicin) - it's a lipopeptide (made of amino acids [many of them 'exotic'] and lipids) antibiotic and it's big and next to impossible to chemically synthesize.
The approach we took was to clone out all of the genes (which is basically one giant NRPS [non-ribosomal peptide synthase], and the enzymes that make those 'exotic' amino acids) involved in its biosynthesis and making targeted mutations to try to come up with different related compounds that might present better pharmacological profiles.
It's been almost 20 years, and no real breakthroughs that I'm aware. Cubicin is still the original daptomycin. There are 9 analogues that show antibacterial properties, but apparently not much better/ also comes with poorer pharmacological profiles.
posted by porpoise at 4:25 PM on March 17, 2019 [16 favorites]
Major pharmaceutical companies already have massive small molecule libraries (self link).
The thing with the claim "[O]nly 4 percent of the Inga compounds were known to science." is that it's technically true but
1) the vast majority of those compounds don't really do anything (but are produced by plants because they might do something - or a recombination of the genes that are involved with their synthesis might yield molecules in their offspring that might do something, and
2) Once you discover a small molecule that does something interesting, the question is how to synthesize it in useful quantities? Many of these small molecules are traditionally very hard to make through pure chemical synthesis. Also, it might do something interesting but doesn't do it well enough to be useful, so you have to modify it. This can be complicated/ not-worth it.
The classic examples are acetylated salicylic acid - the acetylation makes it a more selective COX-2 inhibitor and less so COX-1 - making it safer and less prone to causing abdominal pain and diarhrea.
The other is heroin (acetylated morphine). Funny story - Merck's original intention was to acetylate codeine, but ended up with acetylated morphine. Acetylated codeine is a potent toxic side product of bad heroin production from opium feedstock.
Most of these low hanging fruit modifications have already been studied; more nuanced modifications of small molecules tend to be difficult, the moreso the more complicated the small molecule.
2.5) Due to the difficulty in chemically synthesizing potentially useful compounds, another approach has been to study the metabolic pathways that lead to the end-product creation, cloning out those genes, and essentially "fermenting" the molecule of choice. With advances in genetic engineering, you can make targeted mutations in the genes in order for the enzyme to perform a different chemical reaction (or make the reaction more efficient).
I worked on Daptomycin (Cubicin) - it's a lipopeptide (made of amino acids [many of them 'exotic'] and lipids) antibiotic and it's big and next to impossible to chemically synthesize.
The approach we took was to clone out all of the genes (which is basically one giant NRPS [non-ribosomal peptide synthase], and the enzymes that make those 'exotic' amino acids) involved in its biosynthesis and making targeted mutations to try to come up with different related compounds that might present better pharmacological profiles.
It's been almost 20 years, and no real breakthroughs that I'm aware. Cubicin is still the original daptomycin. There are 9 analogues that show antibacterial properties, but apparently not much better/ also comes with poorer pharmacological profiles.
posted by porpoise at 4:25 PM on March 17, 2019 [16 favorites]
"-omics is basically a suffix we put on complicated stuff we don't understand."
Eh. It's just a easy way to talk about different fields. The misunderstanding here might be that people keep adding -omics to emerging fields, and thus relatively less is known about them.
Also, the new fields are also pretty heavily data-driven and older crusty types might misunderstand that as "complicated stuff we don't understand."
/sorry, feeling grumpy today
posted by porpoise at 4:28 PM on March 17, 2019 [1 favorite]
Eh. It's just a easy way to talk about different fields. The misunderstanding here might be that people keep adding -omics to emerging fields, and thus relatively less is known about them.
Also, the new fields are also pretty heavily data-driven and older crusty types might misunderstand that as "complicated stuff we don't understand."
/sorry, feeling grumpy today
posted by porpoise at 4:28 PM on March 17, 2019 [1 favorite]
The -omics line was said at a bioinformatics seminar, so I think they were in on the joke. :-)
posted by clawsoon at 4:35 PM on March 17, 2019 [1 favorite]
posted by clawsoon at 4:35 PM on March 17, 2019 [1 favorite]
Could the molecules be intrinsically interesting for the roles they play in herbivore-plant co-evolution, rather than for any agricultural/medical/industrial use they might have for us?
posted by clawsoon at 4:39 PM on March 17, 2019
posted by clawsoon at 4:39 PM on March 17, 2019
What I'm wondering is, how do you take a whole bunch of organic chemicals and identify them quickly/accurately enough to say that they're "new to science". Is this the sort of thing that can actually be done easily nowadays? Because I thought that it would take fractional distillation or chromatography to separate them, followed by a long process of crystallisation or spectrography and NMR and goodness knows what.
posted by Joe in Australia at 5:16 PM on March 17, 2019
posted by Joe in Australia at 5:16 PM on March 17, 2019
intrinsically interesting
Of course! There are lots of common organic chemical reactions. Lessons learned from the structure/biochemistry relationship of any given enzyme (that is involved in making small molecules) may yield important lessons in how to modify a different molecules in a different plant entirely.
For example, when I was studying NMDA receptors, I used lessons learned in targeted mutations in a similar receptor. Saved me tons of time having to do it from scratch myself, and proved a new mechanism, that was later found to be shared with a bunch of other related receptors.
The issue is - while I'm sure you could find as many people as wants to study something - who's paying for the research, and paying the researcher... and how much?
posted by porpoise at 5:17 PM on March 17, 2019
Of course! There are lots of common organic chemical reactions. Lessons learned from the structure/biochemistry relationship of any given enzyme (that is involved in making small molecules) may yield important lessons in how to modify a different molecules in a different plant entirely.
For example, when I was studying NMDA receptors, I used lessons learned in targeted mutations in a similar receptor. Saved me tons of time having to do it from scratch myself, and proved a new mechanism, that was later found to be shared with a bunch of other related receptors.
The issue is - while I'm sure you could find as many people as wants to study something - who's paying for the research, and paying the researcher... and how much?
posted by porpoise at 5:17 PM on March 17, 2019
"New to science" basically just means - this has not been chemically characterized and catalogued.
Small molecules are kind of like snowflakes - if a given molecules has a single methyl group be in a D rather than a L position - you might be able to get away with claiming "new to science."
As to how much time? Yeah, that's going to take a long time to get the molecular structure, much less a crystal.
They could also be overspeaking in that they have HPLC signatures that haven't been reported elsewhere, because they are reported even less often and are only reasonably useful between two very similar instruments, so much much lower bar to "new to science."
/reads story
Yeah, mostly the latter. "They used a mass spectrometer to determine the chemical structure of each compound, and established that compounds with similar structures were likely affecting herbivores in a similar way."
This is methodologically very very weak. They're lumping together general size and relative chemical content. But got to start somewhere.
posted by porpoise at 5:27 PM on March 17, 2019 [1 favorite]
Small molecules are kind of like snowflakes - if a given molecules has a single methyl group be in a D rather than a L position - you might be able to get away with claiming "new to science."
As to how much time? Yeah, that's going to take a long time to get the molecular structure, much less a crystal.
They could also be overspeaking in that they have HPLC signatures that haven't been reported elsewhere, because they are reported even less often and are only reasonably useful between two very similar instruments, so much much lower bar to "new to science."
/reads story
Yeah, mostly the latter. "They used a mass spectrometer to determine the chemical structure of each compound, and established that compounds with similar structures were likely affecting herbivores in a similar way."
This is methodologically very very weak. They're lumping together general size and relative chemical content. But got to start somewhere.
posted by porpoise at 5:27 PM on March 17, 2019 [1 favorite]
I once heard a biologist say, "-omics is basically a suffix we put on complicated stuff we don't understand."
Works for economics.
posted by Spathe Cadet at 5:31 PM on March 17, 2019 [4 favorites]
Works for economics.
posted by Spathe Cadet at 5:31 PM on March 17, 2019 [4 favorites]
Just wanted to point out - the " Over five years" link is a university press release disguised as university news.
Everything is going to be blown up to the hilt, and then maybe a liberty or two.
Researchers hate this shit, and I'm sure the claims made in their manuscripts for peer review are much much more reasonable.
/reads article
//shoot - I had to renew my alumnus access last week and did it on my work laptop; I can't remember my new password on my home desktop
///Can't read article =(
posted by porpoise at 6:23 PM on March 17, 2019 [1 favorite]
Everything is going to be blown up to the hilt, and then maybe a liberty or two.
Researchers hate this shit, and I'm sure the claims made in their manuscripts for peer review are much much more reasonable.
/reads article
//shoot - I had to renew my alumnus access last week and did it on my work laptop; I can't remember my new password on my home desktop
///Can't read article =(
posted by porpoise at 6:23 PM on March 17, 2019 [1 favorite]
I read the methods on the Science paper and (as per usual) the popsci story elides most of the actual methodology. Using only general size would indeed be very weak, but they used tandem mass spectrometry (LC-MS/MS) to collect fragmentation spectra of individual ions (which IS indicative of general chemical structure, to the point where a fragmentation spectra is somewhat of a chemical fingerprint for specific compounds) and then compared fragmentation spectra to indicate chemical similarity from compound to compound. Global Natural Products Social Molecular Networking was the framework they used to determine similarity, which is new to me (I have only worked on human metabolomics, not natural products) but seems pretty dang cool.
And yes, nobody really understands metabolomics, including the people doing the research. It's actually kind of a huge problem that standards in the field are light years behind genomics and even proteomics.
posted by permiechickie at 6:27 PM on March 17, 2019 [3 favorites]
And yes, nobody really understands metabolomics, including the people doing the research. It's actually kind of a huge problem that standards in the field are light years behind genomics and even proteomics.
posted by permiechickie at 6:27 PM on March 17, 2019 [3 favorites]
With respect; caveat - yes, nobody really understands metabolomics -
Yet.
One problem is the scattered knowledge and framework for sorting the knowledge. Settlement on standards (of data entry, and what kind data, and format).
Early days.
posted by porpoise at 7:04 PM on March 17, 2019
Yet.
One problem is the scattered knowledge and framework for sorting the knowledge. Settlement on standards (of data entry, and what kind data, and format).
Early days.
posted by porpoise at 7:04 PM on March 17, 2019
porpoise: Just wanted to point out - the " Over five years" link is a university press release disguised as university news.
This is something I struggle with for FPPs. Sometimes there's no independently written story about a study, just the press release. The study itself is interesting but either paywalled or difficult to read for a non-specialist. Then I have to decide whether to spend a few hours putting together a whole bunch of links and effectively writing my own article as the FPP (which means that most of the time it won't happen), or just post the press release - which usually has all the faults and benefits of pop-sci writing - as-is.
posted by clawsoon at 9:26 AM on March 18, 2019 [2 favorites]
This is something I struggle with for FPPs. Sometimes there's no independently written story about a study, just the press release. The study itself is interesting but either paywalled or difficult to read for a non-specialist. Then I have to decide whether to spend a few hours putting together a whole bunch of links and effectively writing my own article as the FPP (which means that most of the time it won't happen), or just post the press release - which usually has all the faults and benefits of pop-sci writing - as-is.
posted by clawsoon at 9:26 AM on March 18, 2019 [2 favorites]
I totally get it clawsoon, no insult was intended. Just wanted to point out obvious biases and limitations of unversity press releases. I'm borderline revolting against my company's IR- / marketing- influenced press releases/ investor newsletters.
Then there's the problem with "independent content expert" weighing in - how does one evaluate "independency" and "content" and "expert?" Is the reporter/ medium (ie., NYT, MeFi) responsible for educating their readers? Media can't be, but "should" they be?
To most of MeFi, they couldn't tell me apart from an industry shill except for potential reputation that no other MeFite has publicly called me out to the extent that it made be burn this account in lieu of new "fake" accounts.
Talking about practical stuff, maybe the practice of adding [source] after links with a=href fronts might help? Nothing that a mouseover doesn't reveal, doesn't work when urls are obfuscated, and a link from ***.harvard.edu*** might give many disingenuous links too much trust-capital.
I don't have any good ideas towards solutions - other than spending a lot more on public education to recruit the best teachers and to give them an environment where they are supported to succeed in education.
posted by porpoise at 9:46 PM on March 19, 2019
Then there's the problem with "independent content expert" weighing in - how does one evaluate "independency" and "content" and "expert?" Is the reporter/ medium (ie., NYT, MeFi) responsible for educating their readers? Media can't be, but "should" they be?
To most of MeFi, they couldn't tell me apart from an industry shill except for potential reputation that no other MeFite has publicly called me out to the extent that it made be burn this account in lieu of new "fake" accounts.
Talking about practical stuff, maybe the practice of adding [source] after links with a=href fronts might help? Nothing that a mouseover doesn't reveal, doesn't work when urls are obfuscated, and a link from ***.harvard.edu*** might give many disingenuous links too much trust-capital.
I don't have any good ideas towards solutions - other than spending a lot more on public education to recruit the best teachers and to give them an environment where they are supported to succeed in education.
posted by porpoise at 9:46 PM on March 19, 2019
« Older Unfolds in a very interesting way | Bee-Lieve it Or Not Newer »
This thread has been archived and is closed to new comments
posted by hoyland at 8:10 AM on March 17, 2019