An anonymous indian tomato vendor in Chennai, Tamal Nadu. photo mckaysavage, flickr (click to see photo in it's original context)

First, since I didn’t explicitly state it in my previous post, the paper on the longer lasting tomatoes developed by India’s National Institute for Plant Genome Research didn’t report any data on how the RNAi knock-down tomatoes actually taste.* The tomatoes are nearly twice as firm as tomatoes in which these genes are NOT knocked down, so it’s possible they’d seem unpleasantly crunchy, I don’t know how doubling the firmness of a tomato translates into the feeling when a person bites into one.

On the other hand, if the tomatoes do turn out to be tasty and delicious, it’s quite possible the trait could be replicated without genetic engineering. And if that turns out to be true, it’s absolutely the approach anyone developing longer lasting farmers to Indian farmers, or farmers anywhere, should take (for why I’m saying this, check out the bit in bold further into this post). The synthetic microRNAs used in their experiments reduced gene expression by at least 99%, so, if it turns out that remained <1% isn’t playing a key role, the researchers at NIPGR have effectively created knock out lines for each of the two genes they were studying. Knocked out genes (genes so broken they don’t work anymore) has been a key part of genetics since before the word gene even existed. (Mendel in the 1850s and 1860s and Wilhelm Johannsen in 1905 respectively) With a couple of known targets, and a target phenotype that’s known to be worth the effort, creating tomatoes with “naturally” broken copies of the gene is possible and probably worth the effort to avoid the expense and controversy associated with trying to commercialize a new genetically engineered trait.

The mutant (left) and wildtype (right) tomatoes from mutant line e2615m1 from the mutant population cited below. Photo from the searchable database of identified mutant phenotypes at: http://zamir.sgn.cornell.edu/mutants/

Back in 2004 a paper** from two research groups in Israel described a saturation mutagenesis population*** of tomatoes created using EMS**** and fast neutron***** techniques. Screening 13,000 inbred lines for knock outs of either of these two genes would take a fair bit of time and money, but less than is involved trying to get approval of a genetically engineered trait (especially in India, where a political battle over their first genetically engineered food crop, an insect resistant breed of eggplant is still ongoing).

If it weren’t for (what I consider to be) irrational fears about genetic engineering and actions of people who exploit those fears, the arguments of speed and cost would instead rest with the genetically engineered RNAi knock downs already created in this study. Given the world we live in, and given there is an way to get the same benefits without genetic engineering in this particular case, getting the benefits of cheaper produce to people who could use the vitamins, and higher effective yields to farmers who could use the money must take priority. In the mean time people like you and I will just have to keep doing our best to combat that ignorance and fear so someday the deciding factor will be whatever technique is safest, fastest, and makes the most efficient use of scarce resources, not what people have, apparently arbitrarily, decided to natural or unnatural.

And as I said above, we don’t even know if the tomatoes are tasty, or if the NIPGR is or will be working on creating varieties of tomatoes with this trait for use by India’s farmers in the first place, so speculation on this paper may have gotten a bit too far ahead of itself.

*If they people who worked on the project are at all after my own heart, I’m sure they’ve tried the tomatoes for themselves, but subjective judgements like taste aren’t going to make it into a PNAS paper on fruit ripening (and its possible consuming genetically engineered tomatoes that haven’t been approved would technically be breaking the law in India, in which case the researchers would even less inclined to publicize any off the books tasting they did on their own.)

**Menda, N et al “In Silico Screening of a Saturated Mutation Library of Tomato” The Plant Journal DOI: 10.1111/j.1365-313X.2004.02088.x

***A mutagenesis population is a group of plant lines that have all been exposed so some mutating agent (see the footnotes on EMS and fast neutron). A saturation mutagenesis population is one where,  based on the number of lines of plants in the population and an estimate of the number of mutated genes in each line, a mutation of (almost) any gene in the genome will likely be found somewhere in the population. The exception being really genes that are so bad to lose that plants carry even one broken copy, or gene and pollen cells (which only have one copy to begin with), die without being able to reproduce.

****EMS stands for Ethyl methanesulfonate, a chemical that creates a specific kind of mutation in the genome of a plant (changing some Gs into Ts). Exposing plants to specific doses of this chemical creates (on average) predictable numbers of mutations in the genome.

*****Fast neutron mutagenesis is less common than EMS, likely because creating fast neutrons involves the use of radioactive substances (one source I found specifically cited uranium aluminum alloys). The advantage of fast neutron mutagenesis is that as the neutrons tear through cells they tend to rip away large chunks of DNA (hundreds or thousands of As, Cs, Ts, and Gs at once) which makes the resulting mutants more likely to completely lose the function of a gene (imagine using EMS to create mutations as changing a couple of key letters in a recipe, and using fast neutrons as tearing out a whole page of a cook book and throwing it away).

Author’s note: This would seem to be the week for vegetables I hated as a kid. Yesterday was onion, today tomato, if there’s a story about brinjal/eggplant in the next few days we’ll have hit all the big ones.

Ripening tomatoes. Photo: Photos_by_Lina, fickr (click to see photo in its original context)

I was recently pointed to an early publication paper that went up on the Proceedings of the National Academy of Sciences website on Monday, where a research group at India’s National Institute of Plant Genome Research describes two genes from tomato that, when knocked down by RNAi*, result in tomatoes that can remain ripe but not spoiled for up to three times as long as tomatoes where these two genes function normally.

Their approach targets specific genes involved in breaking down certain proteins found in the cell walls of tomatoes (in fact in the cell walls of all plants). Breaking down the cell wall is a key part of ripening in fruits (which the tomato is, botanically if not culinarily). Which makes sense if you’ll think about it for a moment. One of the traits we associate with ripening is getting softer, from bananas to peaches if it’s still crunchy when you bite into it, it wasn’t ripe. What makes plants stiff and crunchy? The strength of their cell walls. Since, unlike vegetables, fruits WANT to be eaten**, as they ripen they begin to break down their cell walls to make themselves more appealing to passing animals. Unfortunately, ripening and spoiling are, in a lot of ways, the same process. If fruits aren’t eaten when they become ripe, they continue to get softer, transitioning from delicious looking -> unappetizing -> inedible -> a puddle of mush on your kitchen counter.

Preventing ripening entirely is relatively easy, and there are plenty of known mutants in tomatoes and other species that never ripen (these naturally mutant tomatoes stay green and hard no matter how long you wait). But getting part of the way to ripeness but stopping before crossing the line into spoiled is a much less tractable problem.

A spoiled tomato. A rotting tomato is visible in the bottom left, but that's the result of the growth of microorganisms which is a more complex process. Cropped version of a photo from goldberg on flickr. Click to see the original photo on flickr.

To the non-cell wall biologist like me, one of the most attention grabbing parts of this paper was figure 3A, which simply shows photos of tomatoes that have been sitting at room temperature for 10, 20, and 45 days***. At ten days all the tomatoes look fine. By twenty days, the control (normal) tomatoes are shriveled. After 45 days sitting on the scientific equivalent of a kitchen counter the control tomatoes are basically brown balls of goo, while tomatoes with either of the two genes identified in this paper knocked down show no change in appearance over the same period of time. So what are these awesome genes?

Both genes studied in this paper are glycosyl hydrolases, a kind of enzyme that breaks the chemical bond holding a sugar to either another sugar or some other molecule, like a protein. Specifically the two genes, which are normally expressed in ripening tomatoes, each break specific kinds of sugar off of a specific kinds of protein found in the cell walls of plants. Plant cell walls are mostly made of hydrocarbon polymers like cellulose and lignin, but plants also use some structural proteins (usually less than 5% of the cell wall) and it is the sugars attached to these proteins that the glycosyl hydrolases studied here act upon.

This is where it gets scientifically cool. The prolonged ripe-but-not-spoiled state of the transgenic tomatoes they produced wasn’t simply a result of preventing the structural damage caused by the break down of the bonds between cell wall structural proteins and the sugars they’re connected to. Instead, when they looked at gene expression in plants where either of these two genes had been knocked out, they found that genes involved in breaking down cellulose, lignin and pectin (the main components of the cell wall) were also less expressed. The authors speculate that the kinds of sugars/carbohydrates these two genes break free from cell wall structural proteins actually serve as a signal to the plant to increase the production of all the other proteins needed to break down cell walls and in their transgenic plants, that signal never comes, letting the tomatoes stop ripening before the process leads to spoiling.

Tomatoes at a farmers market in NYC. photo: SeenyaRita, flickr (click to see photo in its original context)

The authors themselves point out the huge potential upside to reducing spoilage in the developing world. As much as 50% of produce is lost to spoilage between harvest and diner plate in the developing world. Reducing spoilage is one of those rare almost-a-free-lunch opportunities to increase the food supply without bring more land under the plow, or increasing the inputs (in the forms of fertilizer, pesticide, and all to often back-breaking manual labor).

At this point you may be thinking, haven’t we heard this story before? There are lots of differences between these tomatoes and the Flavr Savr tomato produced by Calgene in the 90s. Scientifically they come at the problem from very different angles, but rather than get into that let me point out two crucial practical differences:

1. The authors present data that the tomatoes with knocked down expression of either of these two genes are twice as firm as normal tomatoes of comparable ripeness. An important trait for transporting ripe tomatoes over any significant distance as illustrated in this segment of First Fruit talking about the Flavr Savr tomatoes of the 1990s:

The shipping test out of Mexico, however, proved to be yet another disaster. It was designed to test, not only whether the Flavr Savr gene would enable vine-ripened fruit to survive 2000 miles in a truck … The test results were clear before the vehicle came to a complete stop. Tomato puree seeped from the truck’s back end.

2. If these research leads to a commercializable fruit, it will likely be grown first in India, where, as described above, spoilage of produce is a major issue. In the United States, the Flavr Savr tomato had to go up against an existing system built on tomatoes that, without any genetic engineering, never ripen on their own, described in this way by MAT_kinase of TheScientistGardener:

Fresh market tomatoes, in nor cal, are all picked green and gassed with ethylene to force ripening (imperfectly). In the midatlantic, virtually all tomatoes have a natural gene mutation that prevents them from ever ripening completely in the first place. Either way, you end up with an inexpensive, pretty, red tomato that’s often hard and white on the inside. Heirloom varieties taste great, but are very susceptible to pests, have to be hand picked and turn to goo shortly after ripening.

When Pamela Ronald of Tomorrow’s Table talks about the development of transgenic crops, she points out that by 2015, it is projected that more than half of transgenic crop varieties will be produced by the national research labs of developing countries like India, China, and Brazil for they own farmers. If this paper is a sample of the sort of research such labs produce, 2015 should be a truly fascinating year for agriculture.

I shouldn’t have to say this, but there are currently no genetically engineered tomatoes on the market. For a short time in the 1990s Calgene sold the Flavr Savr tomato in California grocery stores, but they weren’t able make a profit doing so, so they stopped. The poor taste of most tomatoes for sale in the grocery store today is purely the result of conventional breeding (my post on the subject and Mat_kinase’s)

Meli, V., Ghosh, S., Prabha, T., Chakraborty, N., Chakraborty, S., & Datta, A. (2010). Enhancement of fruit shelf life by suppressing N-glycan processing enzymes Proceedings of the National Academy of Sciences DOI: 10.1073/pnas.0909329107

-The gene knocked down in the Flavr Savr tomato was Polygalacturonase.

-The two glycosyl hydrolase genes studied in this paper are alpha-mannosidase and beta-D-N-acetylhexosaminidase.

*Using RNAi means inserting a backwards version of part of a gene into a plant under a strong promoter (so the plant makes lots of RNA copies of the backwards bit.) Those backwards copies will bind to the RNA transcript of the actual gene, creating double stranded RNA. One of the main times a plant cell normally sees double stranded RNA is when it is being attacked by viruses (the genome is made of double stranded DNA and the RNA messages transcribed from the genome are single stranded), so making a double stranded copy of the a particular gene causes the plant to treat that gene itself like an invading virus and keep the protein that gene encodes for from being produced. (<– this is the simplified version of the story, this work actually uses synthetic microRNAs which are a much more refined version of the technique.)

**When a plant produces a sweet and tasty fruit in the wild, its goal is to attract some animal that will eat the fruit and carry the plants seeds to someplace new where the seeds can grow into new plants. Domestication has changed the rules of that bargain somewhat, as we artificially selected for bigger and tastier fruits, but fruiting plants still trade animals (us humans) food in exchange for having the seeds of their species distributed across whole fields by farmers, and have their growth protected and nurtured by human hands and human ingenuity.

***There’s also numerical data which is probably better science (the images only track two fruits of each type which I’m sure isn’t statistically significant), but the best scientific papers will include hooks like that image of unrooting tomatoes to draw the reader in long enough to read the exciting data itself.

There is a very useful resource on growing tomatoes made available by the Iowa State extension service. They estimate yields of 12,000-16,000 pounds of tomato per acre. A pound of tomatoes contains 86 calories. That means an acre of tomatoes yielding 14,000 pounds per year is creating 1.2 million calories of food per year or enough to feed 1 and 2/3 people for a year (at 2000 calories per day), the comparable figure of a midwestern cornfield is 29.5 people per acre. In the US we currently farm a bit over 400 million acres so, assuming that expanding tomato production to all land currently farmed wouldn’t result in much lower average yields (a big assumption considering tomatoes are probably grown on the land best suited for their production at the moment), we could, in fact, feed 657 million people (or approximate twice the current population of the US) by converting to tomato based agriculture. And this doesn’t consider greenhouse based production, which accounts for ~10% of current tomato production in the US and has markedly higher yields, although the cost of production is also higher.

According to my back of the envelope calculations, we could feed our entire population with tomatoes.

Caveats:

• I already mentioned the first one, much of US agricultural land may be unsuitable for tomato production.
• Since tomato production occurs only in the summer (outside of greenhouses and idilic climates like California) and tomatoes to not store well (unlike dried grains and beans) much of the year would be spent living on canned and dried tomatoes.
• I’m guessing conventional tomato production results in higher pesticide application rates than production of grains but I couldn’t find any numbers on the subject, so feel free to ignore that one.
• An acre of fresh tomatoes requires at least 200 hours of labor per growing season, with the growing season lasting ~4 months, 18 weeks (probably a lot less in some parts of the country). Assuming workers work only 40 hours weeks and using my calculations above we only need about half of our current 400 million acres of farmland to keep everyone in tomatoes year-round we can make the following calculation. To grow and harvest 200 million acres of tomatoes would require 55 million farm workers or 1 in six americans. But of course not all americans currently work. Our workforce stands at ~140 million people. Almost 40% of them would spend their summers working in tomato fields (working for minimum wage unless we’re willing to drive the price of tomatoes even higher but that’s my next point)
• The average retail price per pound of tomatoes in april of 2008 (most recent numbers I could find) was $1.77 per pound. To get 2000 calories today, each person in the US would have to 23 pounds of tomatoes a day (more than the average person in the US eats in a year right now), costing more than $40 a day equalling a $14,600 a year food budget. (For comparison, a person working full time earning the (new increased) minimum wage earns $15,080 a year.)

So in conclusion, yes we could feed our nation with tomatoes, but there would be some major unintended consequences. Just to be clear no one is suggesting we feed ourselves with nothing but tomatoes. I just thought it’d be an interesting calculation to make.

It does raise an interesting question. Since I’m pretty sure most vegetarians aren’t eating 23 pounds of fresh vegetables a day, can any vegetarians reading comment on what foods you get most of your calories from? Personally when I’m trying to cut back on my meat consumption, most of my energy comes from rice, bread, and beans. (With honorable mentions to peanut butter and cheese.) But I don’t know that my own experience is representative (in fact I’m pretty sure it isn’t).

If you enjoyed this post may I suggest:

How Viable is Local Food (there’s nothing wrong with trying to buy more from local farmers, in fact there’s a lot right, but it isn’t physically possible to feed everyone all the time with locally produced food without forcible relocating a whole lot of people)

All Flesh Is Grass (an acknowledgement of the grains, a small group of plants which enabled practically all of human civilization)

Driveway tomato garden. How much diversity do these plants contain?

Crops like tomatoes, even heirloom tomatoes, aren’t found in the wild. Domestication of crops usually involves only a relative handful of individual plants. Narrowing the species down to a few hundred (or possibly even a few dozen plants) means only a limited number of copies of each gene will be carried through and many of the variant copies of the genes present in the wild population won’t be included in that number. Keeping the population small for multiple generation reduces variability even more as by chance some rare version of genes in one generation won’t be passed to any of the offspring in the next.

Genetic bottlenecks happen in the animal world as well. Skin grafts between unrelated Cheetahs aren’t rejected because the animals are so genetically similar their immune system can’t distinguish the grafted skin as being different from its own skin. Even less fortunate are the tasmanian devils who have so little genetic diversity that they are being decimated by a transmissible cancer. After fighting with an infected devil, which has tumors on its face and neck, tiny bits of the cancer will get into an uninfected devil’s wounds, and since the immune system can’t distinguish the foreign cancer cells from the devil’s own cells, the cancer cells reproduce unchecked, the trait that makes normal cancers, produced by mutated versions of our own cells, so deadly. And the solution mentioned in the article, to save the species by protecting 200 individuals, while better than letting them all die, will sacrifice even more genetic variability by subjecting the already inbred devils to a new population (and genetic) bottleneck.

I’ve gone off-topic with cool biology.* most crops capture only a small slice of the genetic diversity present in their wild progenitors. Crops can even go through extra bottlenecks, as when a few potatoes, carried back from the new world, gave rise to most of the potatoes grown throughout Europe. Even less diversity, which would have given any plant breeders of the say even less to work with if they tried to develop a blight resistant potato.

To overcome those limitations, today breeders of many crops will hunt down the wild ancestors of that crop and do crosses to bring more of the wild species’ genetic diversity and look for valuable genetic variants that were lost during the original domestication hundreds or thousands of years ago, or new genetic variants that have emerged in the wild since.

It is these intentional introgressions of wild genetic material that make modern tomato breeds more genetically diverse than older tomato breeds such as heirlooms that drew from the more limited range of genetic variants available in tomatoes at the time of their initial breeding. h/t to TheScientistGardener for bringing that awesome fact to light.

*Come on, a disease that spreads when broken off pieces of external tumors get into open wounds is pretty cool, especially when you consider fighting in a crucial part of tasmania devils mating rituals. I do hope the devils can be saved though, Australia has lost enough cool species without the Tasmanian Devils going the way of the Tasmanian Tiger.

Photo of the last living Tasmanian Tiger. Died 1933.

This writer has eaten a genetically engineered Flavr Savr tomato, and while the taste can’t compare with a summer garden grown tomato (what can?), it was pretty darn good. The Flavr Savr wasn’t intended to compete with summer produce, but in the winter when the grocery stores only sell the hard, green bullets that bear no resemblance to a real tomato, the Flavr Savr should be appealing.

Source. The whole page was an interesting read, from a more optimistic time when GM tomatoes came with little pamphlets explaining their benefits and how they were created. The article mentions that Calgene (back then the face of genetic engineering, a title since passed to Monsanto), was working on a genetically engineered chocolate following the release of their tomatoes. But it was not to be.

Calgene waited several years for FDA approval, planting tomatoes and then plowing them when it became clear approval would not come in time for that season’s harvest (lossing money the whole time). When the tomatoes were finally approved, issues remained with production and distribution, as with any new product, with the result that there were never enough of the new tomatos to meet demand and when tomatoes were avaliable they were sold at a loss. Calgene burned through their venture capital funding and were eventually bought up by Monsanto for their work on cotton and canola. The flavr savr tomato disappeared into history.

*Which would have made sense as deadly nightshade is in same group of species as tomatoes, potatoes and eggplants.