Science AMA Series: I’m Sarah O’Connor, Professor and Project Leader in Biological Chemistry at the John Innes Centre in the UK. I do research to discover how plants produce medicines and to engineer strains of microbes and plants that produce these compounds. AMA!


Hi Reddit!

Plants produce hundreds of thousands of complex molecules called "natural products" that have many uses. Anti-cancer medication such as vinblastine and taxol, and the anti-malarial drugs artemisinin and quinine are all natural products that are produced by a plant. But we still don’t understand exactly how plants, such as the medicinal plant Madagascar periwinkle, make such powerful but complicated compounds. My research is about trying to find this out. If we can, it will be the first step towards making “new-to-nature” compounds with even better qualities such as reduced side effects.

I am on the Sense About Science Plant Science Panel, where you can ask your questions and get an answer from scientists. The Panel is made up of over 50 independent plant science researchers. You can ask questions to them on Twitter (@senseaboutsci #plantsci) Facebook or using their online form. Answers are sent back within a couple of days and posted online. The Panel has answered close to 400 questions over the last three years and it's a great way to cut through the noise around what can sometimes be a really polarised debate.

I will be back at 12 pm EDT (5 pm GMT, 9 am PST) to answer all your questions.

Insulin was the first protein produced on an industrial scale through recombinant DNA technology in e.coli - it was cheaper than using cow pancreas! Now we're engineering whole enzymatic biosynthetic pathways into bacteria and plants, and that's just wonderful. It's so cool to see the science fiction of the past become the technology of the present.

My question is: where do you see this netting out in 30 years. What incredibly fantastical vision do you see for the day you're bestowed the title of professor emeritus?


Insulin is an incredible story of how biotechnology can be used to improve peoples' lives. One cautionary note, in light of some of the comments above: the price of insulin has risen substantially over the years, despite the fact that the recombinant technology should be cheaper and is more environmentally friendly. The washington post did a great article on this. You can google insulin prices and find any number of news articles on this.

When I retire: first, I think that the speed at which we do things will be much much faster. Thanks to advances in sequencing, bioinformatics and gene manipulation, I bet that we will be able to identify entire metabolic pathways in weeks rather than years. I also think that we will overcome one of the key hurdles we face now, which is yield. We can produce tiny amounts of most anything we want to, once we put our minds to it. Its getting economic production scale that is a challenge. I hope that the advances in synthetic biology- such as generating new sensors- will allow us to use techniques like directed evolution to rapidly enhance the yield of a pathway produced by an organism. I also imagine that advances will come in how we screen the biological/medicianl activity of compounds. Right now we frequently screen compounds in a somewhat haphazard, random way and hope that we get a hit. As we learn more about human biology, I hope that we will learn how to tailor these screens such that we can get pharamcologically relevant hits much more quickly.

Hi Sarah and thank you for doing this AMA. When your research finally finds out the way to reproduce these “new-to-nature” compounds, do you expect pharmaceutical manufacturers to do anything they can to stop you from making these easily available for people, as it would limit the manufacturers' profits generated through already established means?


Thanks to hyperprolifitive, Maegor8, trolis_toll, PhoRoX for comments as well. This is a complex issue, and involves areas well outside my area of expertise. A couple of real world case studies are informative. One example is the story of the antimalarial artemisinin, produced by a plant but engineered to be produced in yeast, as developed by Keasling and coworkers in the early 2000's. A 2016 Nature article says "That such celebrated drugmaking technology — developed with the help of US$64 million from the Bill & Melinda Gates Foundation — stands idle illustrates the complicated web of economic forces that affects the market for malaria drugs. 'This is a perfect example of how a new manufacturing process becomes extremely hard to scale up when there is a complex ecosystem of players,'" (Nature, 2016, 530, 389–390). Essentially, the price of artemisinin stablized as the yeast-based production system was being developed. It takes so much time to develop a viable production system, and the economic and regulatory landscape can shift radically during this development process. Antibiotics are another example. After decades of great discoveries of new antibiotics, discovery of new compounds has slowed, while antibiotic-resistant strains of bacteria continue to rise. We have to come up with innovative solutions, and the pharma industry, with its need to stay in line with a viable business model, is having trouble. I am optimistic about the e increasing partnerships between industry and academics to try to address this problem. You can take a look at this roadmap for antibiotic discovery put out by Pew

Finally to directly answer raisedbyowls question, I genuinely think that the pharamceutical industry welcomes and and all new approaches for drug making. If our lab generated a process that could be adapted to industry scale, I expect that pharma would express interest in licensing the technology. The more options pharma have, the more chance they will be able to produce and sell the drug and generate a profit.

Does your study include the Fungi Kingdom as well? If so, what kinds of differences/similarities exist between the beneficial compounds that are created by the two types of organisms?


We have done a little bit of work on the ergot alkaloids (LSD is the most famous ergot alkaloids, but there are many with important medicinal uses such as ergotamine to treat headaches), which are produced by fungi, though this has always been in collaboration with others who are expert in fungal genetics. There is some overlap in the types of compounds that are made by bacteria, fungi and plants. Both fungi and plants make polyketide natural products for example, though I cant think of any examples where the fungal product is identical to the plant product. However, there are also many many differences among the natural product pathways of these organisms, and in general fungi produce a range of natural products that are not found in plants, and vice versa. Fungal natural product pathways, in many ways, can be a little bit easier to work with than plant pathways, because the genes of fungal pathways are nearly always located in a group together on the fungal DNA, which makes them easier to find. Also, fungal genomes are smaller, and most tend to grow more quickly than most plants. So, if both a plant and fungus made the same compound, I would try to elucidate the pathway in the fungus.

What is your opinion on marijuana?


Well, it is not clear how marijuana smoke impacts your lungs, but probably smoking anything is not very healthy. However, marijuana has been used to address a range of medical problems including nausea and pain, and legalization of marijuana is beoming more common. Additionally, marijuana contains some amazing compounds that can be used in purified form to treat a range of problems. For example, Sativex (a marijuana extract that contains two cannabinoids) and marinol (contains synthetic THC) are in use in the clinic. There are also a range of synthetic cannibinoids under development that may have other medical uses. Bottom line, the marijuana plant has provided us with chemicals that have amazing medicinal activities and may provide the starting point for new-to-nature versions.

What motivates you to do this AMA, is this something useful for your research? For example, are you able to value it for grants, as an outreach justification?



I probably could use this as an outreach justification in my grants, but I usually forget what outreach I have done by the time I write a grant. Mainly I do this because I use taxpayer money to fund my research and the public should have the opportunity to ask me about it.

Hi Sarah. Vectors are currently inserted into organisms such as bacteria in order to produce biosynthetic compounds, however, some studies are instead looking at compartmentalising production into plant organs such as trichomes. How feasible would this be in order to mass produce medicinal compounds, and would there be any advantages to this method?


This is a really interesting question. In general, I think we have a lot to do in the area of engineering plants- and obviously trichomes, which are small oil glands on the surface of leaves- are specific to plants. I really like some of the work that is being done in engineering trichomes. A key advantage in my opinion is that the compounds are produced in higher purity- there are fewer endogenous plant enzymes around to derivatize the compound that you are trying to get the plant to overexpress. The downside is that yields in tricomes could be limiting- you are limited to the surface of the leaf in terms of production volume. Also, trichomes may only be good at expressing oily, hydrophobic compounds, not more polar compounds. Incidentally, the idea of compartmentalization in engineering pathways is a very exciting one, and I have seen some recent work on localizing short pathways to other organelles such as the mitochondria with excellent results.

How do you find novel plants/compounds to look at? I spent some time with indigenous people on the Amazon and was always amazed at their knowledge of the local fauna; are people in your employ out searching for these sort of things in exotic places?


Honestly, finding target plants is one of the biggest hurdles. We study plants that someone has already done some initial studies on...which just sidesteps your question. Indigenous medicine is an incredibly rish source of knowledge. Artemensinin came from ancient chinese medicine, and the extracts of madagascar periwinkle, which contains the anticancer agent vincristine, was also identified by studying traditional medicine. We are trying to collaborate with a group in Brazil and I hope that we can help merge the fields of traditional medicine and biotechnology. The regulatory hurdles to collaborate and obtain this plant material are challenging however.

Hi Dr. O'Connor, thanks for the AMA!

What do you think is the most promising way to produce useful quantities of natural products for use? It seems like this is a field where many things are happening right now in bioengineering and synthetic biology, which technologies are you most excited about?

edit: typo


I am hugely excited about the use of 'heterologous systems' to produce natural products. In this approach the genetic pathway is transferred from the native producer to some other organism that is easy to genetically manipulate and grow (like yeast). This approach falls squarely in the fields of bioengineering and synthetic biology. That being said, we are still at the stage where the yields in these systems are often very low. I am hopeful that as we learn more about the chemistry and biology of both the producers and the hosts, we will be able to solve this problem.

Do you believe that as the years go on, we'll be able to breed plants to produce the substances we want them to? Or only more/stronger versions of the substances they already produce?

On that note, how do we find these plants? Just searching and testing?


Absolutely yes that we can breed plants to produce more of compounds that they already make. Commercial poppies used to produce morphine were selectively bred over many many years to produce high levels of morphine. Also (I hear) there is much more potent cannabis today than there was a few decades ago, and this is also due to selective breeding. However, some plants are not amenable to this type of breeding approach, for example, the yew tree that produces the anticancer agent taxol grows very slowly and a breeding program for this plant would be very difficult. To use breeding to introduce a new compound, you would have to be able to cross your plant with aspecies/cultivar that has the compound that you want. Such a species that can be bred to your plant of interest is often not available. As for your last question, yes we largely find these plants through searching and testing. Ill try to answer this more thoroughly in one of the other comments that asks this question.

Many of the botanically derived drugs have been used in traditional medicine by indigenous cultures for many generations. Often, they are not recognized or compensated for their discovery and experimentation. How do you address that in your work?


I have not had to address it so far, because we typically study compounds that are very well studied, widespread and from plants that are in the public domain. This is an important issue, and we need to work to make sure that researchers can get access to medicinally iportant plants while ensuring that the original discoverers get credit and compensation. Unfortunately, drug discovery is such a convoluted and lengthy process, this is complicated.

Prof. O'Connor! What do you think it would take to change the public stance on using GMO's in everyday life, so that projects like golden rice would actually be applied and put to good use?


Well, dont support Greenpeace as a start. Also, agri-companies need to be committed to making their GM products friendly and accessible to the farmers that use them. And we need better science education, so that people understand exactly what GM is, and can understand the risks and benefits for themselves.

Hi Sarah!

Is you plan to mimic the metabolic pathways you find in plants in bacteria/fungi for active compound synthesis?

Are there many plants left to screen for medicinal compounds?

As a chemistry student aspiring to do natural product synthesis, how do you think will that branch of chemistry evolve in terms of career opportunities?

Thank you for taking the time for answering this ama!


Yes, that is some of what we do: building pathways in other plants or yeast to make these compounds, using the enzymes from the producing plant. So many plants have not been studied- I estimate hundreds of thousands of plants have not been looked at carefully. The field of natural product synthesis is in transition. As someone entering that field it is very important to keep in mind the application/translation of what you do- the days of getting funded just to make a natural product to put it on the shelf are over. However, teaming up of natural product synthesis with biologists/med chemists can be hugely beneficial and has led to outstanding research. Also, the development of new synthetic methods is an exciting area, and one that is relevant to my field as well. For example, in the engineering of yeast to make artemesinin, the yeast actually makes artemesinic acid, and then a chemical process is used to transform artemesinic acid into artemesinin. This combination of bioengineering and synthetic chemistry is very powerful and I hope we continue to see more of it in the future.

Such a coincidence that you are hosting an AMA! I was just reading some of your work. I plan on going into a PhD on this topic.

  1. Why brought you into this field and what sort of research skills do you think fit it best?

  2. I've read many papers on exploring the metabolite pathways of a variety of alkaloids, terpenoids, and polyphenols. Do you think any of these types of compounds have a specific advantage in bioproduction over others?

  3. Finally, I just want to know, why aren't more scientists studying this field!? I find it so exciting and interesting. The idea of using plants and bacteria as little factories with huge potential to produce a huge variety of chemicals that we can use to improve crops, medicine and commodities is very clever.


I am really glad to hear that you want to get a PhD in this area! I originally started in chemistry, but you can approach this area from a variety of different backgrounds (biochemistry, molecular biology, plant biology, bioengineering). I was interested in enzyme mechanism, and went into plant metabolism bacuse of all of the amazing chemistry that these enzymes catalyze. I think that the type of compound needs to be carefuly matched to the type of bioproduction. Yeast is very good for producing some types of compounds, less so for others. Bacteria is better for others. I think that the field is relatively new, and that you will see a lot more people entering it over the next few years- hopefully you too.

Hi Sarah, Have you studied cannabis and the medical benefits of cannabinoids?


I have not- frankly the regulatory hurdles have been very high, and when I started to look at plants I focused on alkaloids, which did not have so many restrictions. An example of someone who does work on cannabinoids:

Prof. O'Connor!

Do you think we understand plant physiology and biochemistry well enough to stably introduce new metabolic pathways as opposed to transient transformation? Could you give me some examples? Of course, i m asking the question purely from the scientific point of view, setting aside morality/legality of such mutants.


The short answer is no I dont think we know enough to engineer in these metabolic pathways as effectively as we could. We are learning so much about regulation, compartmentalization and transport, and I imagine that all of these issues are important in these engineering studies. Trasnient transformation is more straightforward, because you only subject the plant to a few days of massive overexpression of your metabolic pathway of interest, and then it doesnt matter if the plant gets sick due to this. With stable trasnformation, the heterologous pathway has to be integrated into the whole metabolism of the plant and the plant health and viability must not be compromised.

Do you think "plantibodies" will ever take off?

(Antibodies produced by plants.)


We will see! In principle, expressing vaccines/antibodies in plants could solve a lot of problems. Its faster, no problems with egg allergens etc. There is a facility being built next door to the John Innes Centre, LeafSystems, which will attempt to pilot production of some of these vaccines in the plant Nicotiania benthiamia.

I understand that rain-forests are treasure troves of natural plant medicines. Do you have any ideas on how to help save what rain-forest this world has left?


Your question, which is an excellent one, makes me think I should have gone into public policy instead. I dont have any great ideas...

Are there any plant compounds that combat anxiety and depression?


As -42 said below, St John's Wort is the best example of a plant based treatment for depression. It is sometimes called the Prozac of the plant kingdom. The plant makes a compound called Hyperforin which is believed to be the most important active ingredient. It is believed to work by inhibition of serotonin reuptake.

In the US would these medicinal compounds fall under the scrutiny of the FDA like traditional medicine, or would they be outside the FDA's scope like herbal supplements?


If a compound is used in a clinical setting in purfied form it absolutely is regulated by the FDA or similar agency and will have gone through extensive safety and clinical trials. This is not true for herbal supplements, so be careful when you take these supplements.

Hello Dr. O'Connor!

What are your thoughts on the current controversy surrounding kratom (mitragyna speciosa)? The DEA has recently placed this plant and its alkaloids in a category as severe as heroin. This is despite the plant being safely used as a natural supplement for years across the globe.


It is true that kratom has been used for a long time and the regulation of kratom may be overkill. However, I also wish that the supplements market in general was better regulated. The compounds in kratom are powerful, and the consumer really has no way of knowing how carefully the herbal preparation was made. I hope that the compounds in kratom continue to be carefully and rigorously studied, so that like the cannabinoids in marijuana, they can be safely used in the clinic.

Have you done any work with cannabis as a cancer medication?


No we dont do any work with cannabis!

What kind of software do you use? Matlab? Any statistics software?

What kind of computing infrastructure, or is a laptop powerful enough?

Is there any kind of software which would make your research work easier?


Most of what we do can be run from a laptop. Sometimes when we analyze genome data we will use a compouter cluster. We use chemdraw, Geneious, excel...

Hi Sarah! How exactly do you engineer these microbes/plants to produce these chemicals effectively?


Good question. On a simple level, we take the genes from the plant, put them into special "expression vectors" specifically designed for the new host yeast or plant, and transform them all in. The new host can recognize these expression vectors and start to produce the protein that the genes (that we put in the expression vectors) encode. These proteins catalyze the reactions that make the chemical of interest. There is a lot of work being done to develop new more effective hosts and expression vectors.

Hi, do you know why are we still not able to create ecosystems. For example, why nobody grows pineapples in Sweden. Is that possible but too expensive? Is that not possible? If just expensive - why expensive and is there any possibility to make it cheaper than transportation?


Trying to understand one organism in isolation is really hard. trying to understand how one organism relates to and interacts with another is even harder. Trying to disocver, and then put in context, all of the bioloy of an ecosystem is really, really hard. But i hope we are getting there.

What do you think are promising compounds in plants that have yet to be analyzed? With what plants do you work normally? And what microbes?


There are so many plants that have not even been discovered, let alone analyzed. And then there are also so many plants for which a little bit of data has been reported- a few compounds and bioactivity assays have been reported for example- but no other work has been done. We have so much yet to discover from plant metabolism. We work with Madagascar periwinkle, and other alkaloid producing plants and we work with many mint plants that make interesting terpenes. We regularly express plant proteins in both E. coli and bakers yeast.

What's the next best thing that you have high hopes for with what will make an difference?


I can think of a lot of things...but staying within the bounds of natural products, I would say: ready access to effective drugs, and a discovery/development platform that isnt hampered by any economic/marketing issues!

How much of a concern is it to researchers, that species of plants in rainforests, are brought to extinction by human efforts.

Before any properties of these species can documented?

Sorry, if it sounds like a silly question. My Geography teacher brought it up once, and I always think about it


It is a concern to me, but I struggle to come up with a good way to address this issue.

What is your opinion in cannabis? Do you think, as a scientist, should it be legalized? Do the medicinal pros counter weight the addictive cons?


Good question- see answer to whompy410 above

Despite the existence of synthetic cannabinoids (Marinol and Cesamet) plus the availability of Sativex (a marijuana extract), many patients and some physicians still use the dried flower of cannabis and other whole plant extracts because of an "entourage effect" between the different constituents of the plant. Do you believe that researchers will ever come up with a synthetic version that can mimic this entourage effect?


Technically yes: it is just a question of mixing the different isolated compounds. Determining the correct mixture and the correct ratio is the challenge. Clinical studies of small molecules are challenging- clinical studies of mixtures even more so. However, I agree that you are correct that we should be looking at synergistic effects of compounds.

Hi Sarah, thanks for the AMA. I have two questions.

The recent flurry of synbio papers making various natural products in microorganisms have reported abysmally low yields. One report suggested that commercial applications require over a thousand-fold improvement. Do you think we will see economical, commercially-viable microbial biosynthesis of natural products in our lifetimes?

My second question is related to the first. Plant genes necessary for biosynthesis are increasingly being patented. Do you think the developing IP landscape will help or hinder application of your type of research?


I alluded to this issue in one of the questions above- about what I hoep to see in the future. I think if we have better approaches for screening and evolving strains we will get there.

I have mixed feelings about the gene patenting. On the one hand, IP helps drive the industrial translation of research. That being said, I have never had any of the genes that we discovered patented. The IP people at the institutions where I have worked somply do not want to pay for a patent for a single gene that is not much use on its own.

How do you effectively screen the plentiful complex molecules for cures against the many ailments?


As hyperproliferative says, screening is usually a numbers game. I think we are getting a little more sophisticated- while we used to focus just on the sheer size of a library of chemicals to screen, we now more carefully consider how to effectively bias the chemical structures in the library to find a hit. This is a big issue, and as I said in one of the answers below, I hope that in the next few decades we will improve on how we approach the screening issue.

I saw a documentary about a guy who goes into the jungle to work with native peoples to learn and preserve their ways. How much do you work with these types of sources? And, are you concerned about losing this source of knowledge and resources too?


We do not do so much in this area, though we are trying to establish some links with Brazil. What you describe is incredibly important, but essentially a full time job on its own.

How likely is it that a funny little plant with a range of a few square kilometres in the Karoo could have dramatic life-saving properties. I am thinking of a common conservationist argument.


If you mean that if we fail to conserve our natural plant resources, we will lose out on some valuable medicines, I completely agree. How to best address this, I am not sure.

Additional Assets


This article and its reviews are distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and redistribution in any medium, provided that the original author and source are credited.