Making Eden
Page 17
by Bergmann’s group. A few years before this triumph, her team had undertaken
gene-swap experiments, like those we encountered previously in relation to roots
and leaves (Chapter Four), transferring moss genes SMF1 or S MF2 into Arabidopsis lines lacking copies of their own stomatal master genes ( SPCH, MUTE, or FAMA).
Gas valves a 109
They found that under guidance of SMF1 but not SMF2 Arabidopsis mutants developed leaves with stomata.34
Although this proved to be an exciting step forward, it was not yet the whole
story. Around the time we discovered that a single moss gene was the forerunner
to one of the master genes discovered in Arabidopsis, we learned that distinguished professor Ralf Reski’s research group at the University of Freiburg, Germany,
had been carrying out similar experiments investigating the role of a gene called SCRM in stomatal formation in moss. Reski’s obsession with moss has, amongst other things, seen him sequence its genome and start up a biotechnology company
utilizing moss bioreactors for the production of pharmaceuticals. Reski’s experi-
ments had shown that Physcomitrella without SCRM developed normal-looking sporophytes that lacked stomata. We soon set up a Skype call and after summa rizing the findings of our recently completed experiments to each other, decided to join forces and work together collaboratively. It turned out to be a wise strategic
decision .
We knew that two genes ( SMF1 and SCRM) are essential for triggering stomatal development in moss and this raised the question: what mechanism links
their action together? The explanation comes, once again, from what we know
about how things operate in our model flowering plant, Arabidopsis. In
Arabidopsis, SPCH, MUTE, and FAMA are unable to work alone. Instead, they work in partnership with SCRM. What actually happens is that the proteins coded for by SPCH, MUTE, or FAMA are produced at different times and places and chemically bind to the more widely expressed protein coded for by SCRM.
On binding together, these proteins regulate the gene expression patterns guid-
ing stomatal formation.35 The question then becomes whether this partnership
is ancient: does the moss SCRM protein interact with its SMF1 protein and lead
to stomata? The short answer is yes, with the physical binding of the SMF1 and
SCRM proteins activating stomatal development in moss in a manner analo-
gous to how they work in flowering plants.36 In fact, it turns out that nature had already done these sorts of genetic manipulations for us, and reached the same
conclusion in a lineage of flowering plants—the seagrasses. Around 70–60
million years ago, the genome of the marine eelgrass, Zostera marina, evolutionarily lost its stomatal genes ( SPCH, MUTE, FAMA, and SCRM2) and with them its stomata, structures that simply are not required for living a submerged
lifestyle.37
110 a Gas valv es
These experiments with moss established that a pair of ancestral transcription
factor genes ( SMF1 and SCRM) formed the core genetic machinery essential for stomatal development. Closely related genes are also present in a hornwort
genome, the only other bryophyte lineage that makes stomata, although we have
still to sort out the details of how they regulate the molecular pathways leading to stomata.38 So this core genetic module for stomatal development existed in the
last common ancestor of mosses, hornworts, and vascular plants, and this takes
us closer to understanding the earliest events in stomatal evolution. We can also tentatively say something about how the genes enabling this early land-plant
ancestor to build stomata might have worked.39 The first sporophytes of tiny
ancestral land plants that lived hundreds of millions of years ago probably oper-
ated with a pair of genes closely related to SMF1 and SCRM that had multiple functions. The SMF1-like gene probably possessed MUTE- and FAMA-like activity as the minimum requirement for producing stomata in partnership with SCRM.
In early land plants, FAMA-like activity may have specified guard cell identity and driven specific differentiation programmes leading to stomatal formation.
MUTE-like activity then may have ensured that the guard cells matured into stomata. Here again is the extraordinary power of evo-devo in connecting molecular
events in cells of living plants to ancient fossil structures found in long-extinct plants preserved in the rock record.
When it comes to stomatal evolution and the genetic trio of SPCH, MUTE, and FAMA, the story continued until, some 350 million years later on, evolutionary rewiring of closely related genes produced a fundamental innovation in grasses—
a new type of physiologically improved ‘super stomata’.40 The stomata of most
plants consist of a pair of kidney-shaped guard cells surrounding a central pore, but grasses evolved a second type composed of four cells—two dumb-bell shaped
guard cells each flanked by linear subsidiary cells. Dumb-bell stomata are thought to be a customized version of their earlier-evolving kidney-shaped cousins. They
tend to have faster responses to changing environmental conditions, such as a
drying soil, thanks to the rapid transport of ions and metabolites between the
guard cells and adjacent subsidiary cells41 and the efficient mechanics of their
operation.42 This means they can snap shut faster to reduce water escaping in
transpiration more effectively.
Advanced approaches recently led to the discovery that MUTE in a wheat-
like grass species called Brachypodium regulates stomatal and subsidiary cell
Gas valves a 111
development , but that the protein coded for by the MUTE gene is larger than in Arabidopsis and can move between cells (hence it is known as ‘mobile MUTE’).43
Mutant plants lacking mobile MUTE had two-celled stomata instead of the four-
celled usual type. In comparison to the doctored mutants, the wild type of
Brachypodium grew faster and had stomata that opened wider, and responded more rapidly to changing light intensity, than the mutants with ‘normal’ two-celled stomata. Most importantly, in growth experiments, two-celled mutants grew poorly,
producing a third less biomass than normal plants. These exciting findings pro-
vide compelling evidence that speedy grass stomata help maximize the pro ductiv-
ity of grasses under changing environmental conditions. An added bonus is that
they support the speculative idea that dumb-bell stomata helped the remorse-
lessly successful diversification and spread of grasses during the most recent era of Earth history, the Cenozoic (the past 65 million years), as the global climate dried and cooled.44
Genes for stomatal development represent only one part of the ‘air-breathing’
toolkit needed by land plants. Additional toolkits are required for taking strategic decisions about how to space them apart properly for optimal efficiency in regulating the exchange of gases across the leaf surface.45 If you look at the arrangement of stomata on leaves under the microscope, it quickly becomes evident that
they are beautifully ordered so that no single stoma is too close to its neighbour, each being separated by at least one epidermal cell. This spacing ensures the pore can open and close properly because, as the guard cells swell up, they elbow adjacent epidermal cells out of the way, gaining mechanical leverage in the process.
Cells are, therefore, constantly in communication with each other, discussing
who is going to do what by secreting small signalling compounds (called pep-
tides) coded for by EPF genes. This molecular dialogue is a form of chemical Twitter, communicating in ‘words’ of 140 amino acids or less, and it operates
before SCPH/ MUTE/ FAMA get
to work, to hold them back and make decisions about when it is time to make a stoma.46 Different versions of these signalling
molecules are produced in differing amounts by neighbouring cells. Once syn-
thesized, they diffuse outwards through cells and tissues, crowding on to special receptors poking out of the surfaces of neighbouring cells.47 Decisions to make
stomata are taken by the majority vote of the peptide molecules; when those
synthesized by one set out-compete those of another on the receptor complex,
it is time to advance toward making guard cells.48
112 a Gas valv es
Fossils of early land plants from Silurian and Devonian rocks tell us that this is an ancient mechanism because the stomata scattered over the surfaces of these
simple vascular land-plant stems appear to obey the one-cell spacing rule, with
little evidence of clustering. The components of the stomatal patterning toolkit
used by these plants are inferred from those present in moss.49 Moss has genes
coding for the EPF peptides, receptor complexes, and so on, and if you switch off the EPF ‘Twitter feed’, it develops numerous clusters of stomata breaking the one-cell spacing rule. The moss stomatal patterning toolkit is simplified and wired up differently from the complex version found in flowering plants, but analysis of the evolutionary history of the genes shows the debt they owe to early land plants.
So far, in this chapter, we have seen that early land plants had at least two
closely linked genetic toolkits for building stomata, the air-breathing hardware.
One set decides when and where stomata are made (i.e., they specify patterning)
and the other triggers the specialized development of guard cells (i.e., develop-
ment). Stomata also require metabolic ‘software’ for making those ‘water-for-carbon trader’ decisions Berry mentioned. Flowering plants have fantastically complex
control systems regulating stomatal opening and closing in response to things
like light and drought. Decisions about when to open and close stomatal pores
as the plant experiences constantly fluctuating environmental conditions are
taken in response to chemical and environmental signals. Roots that detect dry-
ing soils, for example, release a mobile chemical messenger of drought in the
form of the hormone abscisic acid,50 usually abbreviated to ABA. The name is a
historical legacy;51 ABA was originally discovered to have a function in shed-
ding (abscising) leaves. The mechanism by which this triggers stomatal closure
is as follows: upon sensing ABA, guard cells extrude negatively charged ions,
causing the membrane-spanning potassium channels to open and export
potassium. This causes guard cells to shrink, closing the stomatal pore, and
thereby helping to alleviate the shortage of water and turn off ABA signalling. It provides plants with a hormonal feedback system facilitating the conservation
of water.52
Faced with the double penalty of shallow soils prone to drying out, and a lack
of forest to cast relief from the Sun’s glare, early vascular land plants would have needed an ability to fine-tune transpiration by sensing a drying soil. This raises an interesting fundamental question: how did these plants go about the crucial business of limiting water loss necessary to succeed on land? Historically, the first hint
Gas valves a 113
came from experimental work reported by botanists from Cornell University dat-
ing to the early 1970s. Working with a moss (rather than a flowering plant) called Funaria hygrometrica, they were able to show that it closed its stomata in response to the application of ABA.53 Intrigued, we repeated this experiment nearly fifty
years later to check if this really was the case. Sure enough, exposing the stomata of two moss species, including Funaria for good measure, to an increasing dose of ABA caused the pores to close progressively.54 Taken at face value, these observations suggest that an extant early land-plant lineage (moss) had an operational
core ABA signalling and response pathway for tuning water loss by transpiration.
In fact, some of the genes involved are ancient and may have been originally used for desiccation tolerance before being co-opted for regulating stomatal functioning in early land plants.55
Encouraged by our simple experiments with moss stomata, we investigated the
function of one of the key genes, called OST1, which sits in the middle of a metabolic pathway triggering stomatal closure by ABA in flowering plants.56 Repeat
the ABA experiments with moss lines lacking OST1 and the stomatal closure response is lost. Introduce moss OST1 into Arabidopsis mutants lacking their own copy and it restores the normal stomata closure responses to ABA. These genetic
manipulations show that moss and vascular plants share a common functional
component of the toolkit for responding to drought. A further detail is that OST1
works by activating ion-pumping channels in guard cell walls necessary for sto-
matal closure. The gene coding for a protein that makes the membrane-spanning
ion channel is called SLAC1. Fascinating work has shown that OST1 can activate the SLAC1 ion channels responsible for the export of negatively charged ions to
close stomata in mosses, lycophytes, and seed plants (groups with stomata), but
not those from groups lacking stomata—liverworts and charophyte green algae.57
We also know why this is the case. The protein coded for by SLAC1 in liverworts and green algae has small differences in its amino acid sequence and this prevents it from being recognized and activated by the OST1 protein.
Others have added greatly to the overall picture by elegantly mapping the
presence of genes and proteins related to the perception and metabolic signalling of ABA in numerous species of green plants, from chlorophyte algae through
charophyte algae to land plants ranging from mosses and liverworts to a diversity of seed plants.58 Such findings are often presented as coloured heat maps, formed by a matrix of rows and columns. Each horizontal row in the matrix represents a
114 a Gas valv es
species belonging to a particular lineage and each vertical column represents a
gene or its protein sequence involved in ABA metabolism. The blocks in the
resulting matrix are shaded from pale yellow to deep red, depending on their
similarity to the best-documented genes and proteins in Arabidopsis. The greater the similarity, the hotter the matrix glows.
At a glance, these sorts of analyses provide heat maps offering a snapshot of
over a billion years of evolution and as you cast your eye over it, two things are immediately apparent. First off, you see an overall increase in the number of genes relating to ABA metabolism moving from algae to seed plants. Second, you see
that the core ABA signalling pathway was established during the transition from
an aquatic to a terrestrial environment over 400 million years during the origin
of land plants, with the metabolism gaining in complexity as seed plant lineages
radiated into drier late-Palaeozoic environments.59 Experiments confirm the
functioning of the pathways, as opposed to the presence/absence of genes, during
the evolutionary transition to seed plants.60 It is a compelling approach, combining molecular and physiological tools to sort out the evolutionary history of an
essential adaptation of terrestrial plant life.
But we need to return to mosses, because their stomata may hold the key to
understanding the function of stomata in the first complex vascular land plants.
The question of function intrigued those Cornell botanists too, and they noted a
vital clue: as the small moss capsules ripen through ageing,
the responsiveness of their stomata to environmental stimuli declined until most of them appear to be
‘locked’ open as they developed thick rigid walls.61 There can be few explanations for encouraging an uncontrolled loss of water from ageing capsules (sporophytes)
and top of the list is reproduction. Second on the list is the idea that pores locked open might contribute to the transpirational pull of water and nutrients up into
the capsule.62 With the sporophyte structure attached to its gametophyte parent
plant, it is not clear how important this proposed effect might be. Aiding repro-
duction is implicated because after spores develop inside the capsule, open sto-
mata may facilitate drying of the capsule, a process causing its progressive contraction and gradual release and dispersal of spores.63 Of course, if this idea is correct, the mutant mosses lacking stomata mentioned earlier could be used to test it. So we did, and this is exactly what we found. Mosses lacking stomata showed delayed
maturation of their capsules and release of spores compared to normal mosses with stomata, as anticipated if stomata in this case functioned to promote water loss.64
Gas valves a 115
Applied to understanding the functioning of Silurian–Devonian vascular land
plants, these ideas suggest that the basic genetic toolkits for building and operating stomata provided them with a means of both staying hydrated and generating
a transpiration stream for accessing soil water and dissolved nutrients when it was safe to do so. Later, during the reproductive phase, the stomata might have
switched roles and promoted transpiration to cause release of spores from
sporangia, so facilitating colonization by the next generation. This emerging picture of stomatal function in early land-plant sporophytes ultimately links stoma-
tal action to reproductive success.
The story from the DNA records of modern plants is a telling one: the origin of
stomata in the last common ancestor of mosses, hornworts, and vascular plants
~500 million years ago coincided with the evolution of a core genetic module
enabling ABA metabolism signalling that links limited water supply to limiting
water loss via stomata. Another way to think about the significance of this new