Friday, March 20, 2015

Genetically Engineered Color Changing Flowers, Do they exist?

I am a pretty skeptical person. That is what they taught me during my Ph.D. at UChicago, "Question Everything". That can be good but also bad sometimes (look like a dick for putting down someone's pet project). Anyways, there is company called Revolution Bioengineering that started an Indiegogo campaign up about creating Color Changing Flowers using genetic engineering. Of course I am skeptical. The glowing plant project was supposed to deliver flowers that glow 7 months ago and still there is nothing, yikes!

What I get from their Indiegogo and the answers below is that they are planning to have two products ones that change color once during budding and others that "Change color on Demand"

It just so happens that I met one of the founders at SXSW and asked him if he would do dueling blog posts with me and he agreed. Understand I have never met Nikolai before in my life and this is not preplanned or a publicity stunt. I have not invested in their Indiegogo because I am wary. I gave him this post beforehand so he could write up a response HERE and everyone can see the Scientific basis of the project. If needs be we will then post a second time to allow us both to have rebuttals and ask new questions. In the end we can see if it holds up to scrutiny. I am all about Scientific transparency, if you can claim you can do something you should be able to explain the mechanism of how you plan to do it. I think many of these Science projects people post on Indiegogo or Kickstarter are very vague and lack details. This helps them sell because the public is not usually concerned about that stuff but it makes me wary of over-hype and exaggeration of what is currently possible.



Questions(underlined and bold) and Nikolai/RevBio's Answers(italics)

Color changing is a complicated process, I have worked with engineering bacteria that change color and it takes a long period of time. Visible color change either through chromophores or proteins can take 24 hours and this is usually under special growth conditions such as decreased temperature or optimal nutrient availability. Bacteria are some of the easiest to engineer organisms. Once you take an organism out of a controlled environment or lab how realistic do you think it is to recapitulate the function?

This was something that we considered when we designed this project—Keira and I are two experienced plant synthetic biologists, and it was important for us to do something achievable and not overpromise.  Credibility is a problem in the biotech community.
To that end, we are playing around with the existing anthocyanin pathways—the enzymatic pathways native to petunias that produce the vibrant flower colors.  Petunias are already programmed to produce milliMolar quantities of anthocyanins in flower vacuoles, and we probably can’t do better than that with bioengineering.  (Well, we might be able to with a lot of iterative work, but we don’t have time or money for that right now..)
For the petunia we are crowdfunding to create—the color change on demand—we have an anthocyanin pathway that is broken at a certain early step resulting in a plant with all white flowers.  If we inducibly express the broken enzyme, we get purple/red flowers.
You bring up a very good point about function in real world conditions.  It happens all the time in plant biotechnology that your favorite transgenic works great in a growth chamber, in a laboratory, or in a greenhouse-- but when you take it outside to real world conditions, the trait disappears.  This phenomenon is common knowledge, and part of the reason that the “big guys” do about 6000 unique transformation events, and select the best from that group.  Plant biotechnology is still in the stone age as far as technology goes, and many things are still mysterious enough that it is just a numbers game to get a good plant out of a biotech project.
We don’t have the money or facilities to do 6000 unique transformations, but we will be doing a lot of them and selecting the best responders.


Do you have an exact plan of what you are going to modify and if so can you tell us?

In our color-change-on-demand petunias [anthocyanin 2]AN2 is the broken enzyme.  This is actually not an enzyme in the anthocyanin pathway, but instead a transcriptional activator of that whole pigment pathway.  Without it, the pigment-producing genes are not transcribed.
We are linking expression of AN2 to an ethanol-inducible transcriptional activator.  In our prototype flower, we have AN2 linked to a dexamethasone-inducible transcriptional activator, and that clearly works in a robust way.
https://www.youtube.com/watch?v=hOlO1Cu6E9I
However, we could never distribute a flower that requires dexamethasone to the public.  Or maybe we could, but very few people have Dex laying around their house that they can apply to their flowers, thus the changeover to the ethanol system.
We talk an awful lot about “petunia circadia” and “continuous color change”.  This is a different project, and will be more technically challenging, and is thus listed as a stretch goal on the campaign, not the main goal.
To make Petunia circadia we will be changing the pH within the flower vacuoles.  



I am a little bit confused, will the actual flowers that are growing change color or only new buds about to flower?

Color-change-on-demand : new buds
Petunia circadia : developed flowers
The color-change-on-demand flowers will typically bloom all white.  If you water them with a beer, or spray a flower bud with a beer (we present these beer applications as a possibility, we will need to test the resulting flowers to see what works and what doesn’t) the new flowers will bloom purple.
Once a petunia flower is open, the small molecule building blocks used to make anthocyanins will have been diverted for other uses.  Anthocyanins are also astonishingly stable molecules.  The flowers are locked into being that color, whether they are white or purple.
Petunia flowers last for about 5 days before they senesce, but petunias are prolific bloomers, and will continuously put out new blooms.  So you could change over the entire plant’s flower colors on a one week time scale.  



I assume you are basing your work off of: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1456866/ which seems reasonable but it is not mentioned in the paper how quickly the fading occurs, any idea?

This is for the petunia circadia.  I mentioned that that is a more ambitious project, and there is a lot of testing we would have to do along the path to creating this flower.  That paper you linked to demonstrates the pH change you can achieve through a knockout (roughly 0.5 pH units) and the corresponding color change of the flower.  That pH-color change phenomenon is described several different ways using several different mechanisms by that laboratory.
Circadian promoters form plants are likewise well described and used in synthetic systems (check out luciferase linked to a circadian promoter in arabidopsis here: https://www.youtube.com/watch?v=hOlO1Cu6E9I )
Linking pH change in the vacuole to a circadian promoter will be the challenge.  Will we be doing proton channel expression on a circadian promoter, or doing RNAi knockdowns of a channel on a circadian promoter?  Will the protein lifetime of a proton pump on the plant vacuole exceed the 12 hour timeline of a circadian expression system?  Will we need to stabilize/destabilize proteins/RNAs to get this to work?  There are many unknowns in this project and a lot more testing, measuring, and fiddling required to get petunia circadia to work.  This is definitely higher risk science than color-change on demand, but is still a totally reasonable and doable project.


Do you think any of these engineered processes will affect the normal growth and metabolism properties of the petunias?  

In all cases we will be having a selectable marker in our plants that will be constitutively expressed.  This will place a metabolic burden on the plants that their non-transformed cousins won’t have.  All the clever things we are adding to make the flowers change color will be expressed only using a flower-specific promoter—so metabolic burden will be limited to the reproductive organs, not the photosynthetic tissues.
With all the varieties of plants out there which have been developed with conventional breeding that have one or two genes knocked/overexpressed and still have vigorous growth habits, I don’t think it won’t be too hard to get a robustly-growing petunia out of this. 





If you have questions feel free to post them and I will send them to Nikolai next time around