New possibilities for drought tolerance

An Arabidopsis stomate showing two guard cells exhibiting green fluorescent protein and native chloroplast (red) fluorescence. via Wikipedia.

This image is an extreme closeup of a stomate (singular, the plural form is stomata). These two cells, called guard cells, control the plant’s respiration: how much carbon dioxide gets in and how much oxygen and water vapor gets out. The control isn’t very good, though. Most plants just have their stomata open all day every day so they can pull in lots of CO2 to use during photosynthesis to make sugar. And that means a lot of water, painstakingly pulled up from the soil, through the roots, gets lost. If stomata could be more selective, only opening when more CO2 was needed for photosynthesis, then water could be conserved.
An enzyme called carbonic anhydrase raises the levels of CO2 in chloroplasts so the plant can make plenty of sugar. It does this by converting CO2 from its storage form carbonic acid back to it’s useable form: CO2 + H2O ⇌ H2CO3.
Carbonic anhydrase also appears in the guard cells, where it controls the opening and closing of stomata.
Julian Schroeder, Professor of Biology at UC, San Diego hypothesized that more carbonic anhydrase in the guard cells would place tighter control over opening and closing. His group tried shutting off the carbonic anhydrase gene in the stomata of a little plant called Arabidopsis. Those plants were unable to respond to increased CO2 concentrations in the air, remaining open all day. They also tried expressing additional copies of the carbonic anhydrase gene in the stomata. Those plants closed their stomata when water was scarce. This makes sense – carbonic anhdrase needs water to function, so it can’t function when water’s not around.
Honghong Hu, a postdoctoral research working on the project, said in the press release Newly Identified Enzymes Help Plants Sense and Respond to Elevated Carbon Dioxide and Could Lead to Water-wise Crops: “The guard cells respond to CO2 more vigorously. For every molecule of CO2 they take in, they lose 44 percent less water.”
This research, Carbonic anhydrases are upstream regulators of CO2-controlled stomatal movements in guard cells, published in January 2010, indicates that increasing the number of carbonic anhydrase genes in the stomata could potentially decrease the water lost through stomata in crops. The implications for drought prone regions are obvious. Plants could need less water and could hold on to the water they have longer. It won’t be plug and play, though. As stated in the press release, water that evaporates from stomata cools the plants just like water evaporating from our pores cools us. Increased expression of carbonic anhydrase will have to be tested to determine its effects on plants in high temperature environments.
ResearchBlogging.orgHu H, Boisson-Dernier A, Israelsson-Nordström M, Böhmer M, Xue S, Ries A, Godoski J, Kuhn JM, & Schroeder JI (2010). Carbonic anhydrases are upstream regulators of CO2-controlled stomatal movements in guard cells. Nature cell biology, 12 (1) PMID: 20010812
Thanks to @RivenCactus for bringing this research to my attention by Tweeting a link to the TreeHugger article Newly Discovered Enzyme Could Create Crops That Thrive in Dry, High CO2 Conditions.
If a gene like this was used to make crops more drought tolerant, could it spread to weeds and make weeds weedier?
Yes and no.
If there was a sexually compatible wild relative or weed species growing nearby the drought tolerant crop, it is possible that weed/crop hybrids could include the gene. Sexual compatibility means that the weed not only has to be a fairly close relative to the crop but also means that they have to be pollinated by the same method, have pollen shed at the same time, not have any incompatibility genes, etc. In the United States, there are few weed species that are sexually compatible with crop species, but there are some. In these cases, farmers can use the same sort of strategies to reduce gene flow that they would use to avoid spread of a conventionally bred trait.
If gene flow does happen, the gene will only be present in the weed population at low levels, unless the gene makes the weeds that have it able to outcompete weeds that don’t have it. See Escape! Crop-Specific Gene Flow to Wild Relatives and Those naughty plants! on Biofortified for more discussion of gene flow.

Written by Anastasia Bodnar

Anastasia Bodnar is a science communicator and science policy expert with a PhD in plant genetics and sustainable agriculture from Iowa State University. Anastasia has had various risk analysis roles in US government and military service. She serves as BFI's Director of Policy and as Co-Executive Editor of the Biofortified Blog.


  1. Not sure why you included a NO with the YES in the question of weeds weedier.
    You yourself write “If gene flow does happen, the gene will only be present in the weed population at low levels, unless the gene makes the weeds that have it able to outcompete weeds that don’t have it”
    – It would be logical to think that the drought tolerant trait in wild individuals would indeed help them outcompete the other wild plants in a populations that lack it. As such I’d venture a guess that this would be one transgenic trait that would survive in wild populations quite well – indefinitely in fact as it would clearly be neutral, and likely beneficial.
    “Clearly, crop alleles can persist for many generations following a single hybridization event, and crop-wild hybrids may recover wild-type fitness in later generations. Thus, beneficial or neutral transgenes that recombine independently of deleterious crop alleles may spread and persist indefinitely (Snow et al., 2010).”
    So all I see is a YES. Unless by NO you were suggesting that there would need to be a potential wild population to spread the gene to, in which case, well of course. So this trait seems likely to flow easily and with persistence in certain families (Brassicaceae, for example, of any type would be problematic given the wild weeds in that family). But for crops like soy, corn, cotton – I see little problem.
    Good article, but a bit sloppy with the weed thing. Not sure if you are trying to appease the anti-gmo crowd or the pro with that Yes and No. Would have been nice to have you just come out and say – “this trait would likely spread in the presence of sexually compatible wild relatives and as such deregulation of such crops would need to consider the potential impact on such crops.”
    Leave the politics to the politicians.

    1. James, thanks for stopping by and commenting. Perhaps I should have said “no, but in some cases, yes”.
      For most crops in most places, there are no sexually compatible relatives = no chance of gene flow making for weedier weeds. While there are exceptions, we can look at the locations where various crop species originate and predict that those places are going to be more likely to have wild relatives for those crops while other locations are less likely to have them.
      In cases where there are sexually compatible wild relatives, farmers can use various measures to prevent gene flow such as having flowering time that does not coincide with the wild relative’s flowering time. Those measures aren’t perfect though, so yes, there is a chance of gene flow in these cases – and in those cases the chance of gene flow would complicate deregulation.
      I don’t think that’s politics or sloppy, it’s looking at facts. I probably shouldn’t even have included that part in this post, but I wanted to anticipate any commenters worrying about gene flow.

    2. I get the impression that most people think the Roundup resistance being found in weeds is due to the actual gene from the crops, when in reality the weeds are just evolving their own resistance. So I think it’s important to emphasize that that sort of gene flow CAN NOT HAPPEN unless the weeds are sexually compatible with the crop.

  2. Hm, not convinced this would necessarily produce “tighter control” when overexpressed – do we have good reason to believe that control by CO2 is a major limiting step in WUE? Is it not likely that inducing “tighter control” may well be detrimental to plants when not suffering drought conditions – it may be a good thing(tm) to reduce stomatal opening when water conditions are tight – as you preserve water, but if you close your stomata in the expectation that conditions are going to be tight, and then they aren’t, you’ve essentially reduced your productivity – instantaneous WUE may be positively impacted, but if a plant takes in a lot less CO2 over the course of a season it will suffer from a yield drag at the end of the season unless it experiences water stress – a tough balance to strike.

    1. I think we have good reason to think that water loss through stomata is a major factor in WUE.
      Overexpressing carbonic anhydrase, if I understand it correctly, allows the plant to take in the same amount of CO2 while releasing less water (uber simplification). Yield (if we can call it yield in Arabidopsis) was not affected. From the text of the paper:
      We then investigated whether overexpression of βCA1 or βCA4 in guard cells of wild-type plants can modulate intact plant gas exchange. Four randomly selected independent lines overexpressing βCA1 or βCA4 in the wild-type background (Supplementary Information, Fig. S8b) showed a reduced stomatal conductance at all [CO2] tested (Fig. 4c; Supplementary Information, Fig. S8c–e). Interestingly, substantial increases in the instantaneous water use efficiency (WUE) of all analysed guard cell-targeted overexpression lines were consistently found, with an average increase of 44% at ambient [CO2] (Fig. 4d), whereas CO2 assimilation rates were not significantly altered under the imposed growth conditions (Fig. 4e). All overexpression lines showed reduced fresh weight loss from excised leaves compared with wild-type plants (Supplementary Information, Fig. S8f), consistent with a reduced stomatal conductance at ambient [CO2] in all βCA-overexpressing lines (Fig. 4c; Supplementary Information, Fig. S8c–e).
      No whole plant phenotypic growth differences and no reduction in total plant dry weight (growth penalty) were observed in βCA-overexpressing lines compared with wild-type plants under limited-watering or well-watered conditions (Supplementary Information, Fig. S8a, g, h). Guard cell-targeted overexpression lines showed slightly lower than average stomatal densities (–13%) and stomatal indices (–14%) compared with parallel-grown wild-type plants (Supplementary Information, Fig. S9; P = 0.0312 to 0.0523 for overexpression compared with wild-type lines, Student’s t-test). The ‘single cell spacing phenotype’ was not violated in ca1 ca4 mutant and βCA-overexpression plants leaves28, 29. Therefore, guard cell-targeted overexpression of βCA is sufficient to modulate CO2 regulation of stomatal conductance and may provide an approach for improving the water use efficiency of C3 plants.

  3. Yield wasn’t looked at in Arabidopsis – I don’t see anything other than dry weights of seedlings – the non-significance of the double mutant in either well watered or drought conditions (although there does appear to be a trend towards loss of DW in well watered conditions) makes me wonder how much impact the overexpressor is likely to have – it may be that over short periods of time a decline isn’t visible, but I do wonder how the plants can, over a long period of time (rather than looking at instantaneous WUE) assimilate the same amount of CO2 if they spend more time with closed stomata – if the time period is short
    enough it may just not be in the detectable range.
    I totally agree that stomatal water loss is very important in WUE – it’s just hard to decouple that from the role of stomata in allowing gas exchange to drive CO2 assimilation, I’ve seen similar results with other genes, where stomatal control is implicated, performance in drought is improved – but generally with a negative effect on performance if there is no drought applied (and generally the magnitude of the improvement under drought is equivalent to the magnitude of the failure under well watered conditions) – the results seem to indicate that this may not be the case here, and it’d be awesome if that was right, but it’d also be cool to see how exactly the same amount of CO2 is assimilated when it truly appears that there should be less CO2 available – and also cool to see how this impacts metrics more meaningful to yield in the model – reproductive stage dry weight, seed set, seed yield etc – although the apparent increase in WUE with no change in dry matter accumulation would no doubt be awesome for alfalfa etc.

    1. I agree that the dry weight is a good indication that this would probably work in a crop like alfalfa, but I think there’s a good chance that seed yield would be maintained even in low water conditions as well. Since it’s in Arabidopsis, of course it’s just proof-of-concept at this point. Next they need to test in soybean or something to see if the results are consistent.

      1. Given experience of dry and fresh weights complete inability to correlate with yield in transgenics I don’t share your optimism in that respect, I also wonder to what extent light level impacted the results – 75umols looks low to me (unless I’m messing up my light levels etc – used to working at 1000+ and still being disappointed that this is lower than you’d see in the field) – what’s the general consensus on what’s needed to provide optimal “field” conditions for Arabidopsis (ie the photosynthetic machinery is working hell for leather) could this be a case of dry matter accumulation being such that the plant can’t possibly use all of the available CO2 anyway – in which case closing the stomata would be a great idea, whereas in relevant light conditions – one big disappointment in a lot of corn work I’ve read recently is the spectacularly low light levels that get used (400umols of light to me is practically a carbon starvation level for corn), I wonder if this also applies to arabidopsis, and if so to what extent (probably less so as C4 plants tend to require higher light intensity)

        1. lol well, I have no experience with yield testing so I probably shouldn’t be reaching anyway. Thanks for correcting me. Hm I didn’t notice that the light levels were so low (I’ve done very little incubator work). I hope the researchers are working on a plant that’s more comparable to crops in more normal conditions.

          1. I’ve started looking at growth conditions in papers now practically before I skim the figures – doesn’t necessarily invalidate any of the results just tempers how much meaning you can pull out of them.
            Also on yield – I think (though don’t know) that dry matter accumulation generally correlates pretty well with yield, but this has always fallen apart in transgenics – although you only have to be wrong once to change the world.

          2. Me either – its near incomprehensible to me why a plant that does amazingly awesome things in terms of getting bigger than everything around it then has a yield that you’d expect from a 1950’s hybrid – whereas in general (afaik) when breeding for improved yield you can generally take improved mass accumulation (be that dry matter, fresh weight, height) and use it as a pretty fair basis for judging how yield will look.
            One possibility is that a plant which has a transgene operating during vegetative growth which improves vegetative growth will have the same transgene operating during reproductive stages – increasing vegetative growth, which isn’t a good thing.
            However, if I could explain (to any degree of certainty) the lack of correlation for transgenics which generally appears to be the case for non-transgenics, I’d be making a lot more money than I am right now.

  4. It will be very seldom that a crop will have genes which favor weediness. Green Revolution wheat had improved yield but less biomass — the trick was the semi-dwarf trait, which in the wild would be best described as “stunted”.
    Other crop traits such as dehiscence do not survive well in the wild. Generally speaking, a good crop makes for a lousy weed. A WUE trait would flow with other other traits that would doom it as a weed. Perhaps after many generations in the wild under natural evolutionary pressures the WUE trait would single itself out, but we’re talking a long time here.

  5. Agree with Eric that “It will be very seldom that a crop will have genes which favor weediness.” The record has countless examples, perhaps corn being the primary. But unclear why positive water use efficiency would necessarily “flow with other traits that would doom it as a weed”? And why even bring up dehiscence? Of course most traits that favor crop production don’t do well in wild, but water utilization? What species (plant or animal) would not favor the incorporation of such a trait (unless it had a negative response to normal or excessive water conditions)? Perhaps I’m missing something here, if so, please let me know.
    Anastasia: Never address your writing to the worry-worts. Why bother? That last section weakened your writing. The way it’s included after references and ‘thanks’ is particularly awkward.
    Say Hi to Kendall for me and tell him to get a new earring.

    1. James,
      When ‘outcrossing’ occurs, genes do not move between plants as individual entities. They move in ‘bundles’, so to speak, which is why conventional breeding is so difficult.
      So for instance, a crop with the WUE trait would also have delayed/no dehiscence, that being the tendency to scatter seeds at maturity. Dehiscence is a valuable trait in a weed, because it’s an essential part of propagation. Dehiscence in a crop makes it a non-crop — at maturity, the seeds are scattered, making harvest well-nigh impossible.
      There are other deleterious traits in crops that would outcross in the gene-bundle. The odds of natural selection preserving one particular transgene out of the entire background are staggering.

  6. The tough thing about getting any WUE trait to work is getting it to work without the negative response in normal water conditions – my guess is that this may well be relatively species dependant rather than a one size fits all type of affair, so my guess would be that a trait made to work in one species (through promoter shennanigans and possibly even protein engineering) may not be optimal for another species – assuming this is the case it would be unlikely to see the trait prosper in weed species (regardless of associated traits linked closely enough to drag it down anyway) unless it retained some activity in drought conditions, and drought conditions prevailed for more than a few generations (which may be the case in some areas the WUE trait would be designed for with climate change in mind)

Comments are closed.