Making Eden

by David Beerling

  THE ‘HAND OF HISTORY’ FAILS TO HALT

  ‘ THE TIDE OF PROGRESS’

  -

  Liam Dolan holds the Sherardian Chair of Botany at Oxford University, once

  occupied by the plant geneticist Cyril Darlington (1903–1981). Darlington dis-

  covered chromosomes, but his excursions into eugenics involved him in

  numerous disputes and isolated him from mainstream science.39 Gifted and

  controversial, Darlington was tall and handsome with film star looks in his

  younger days—think Errol Flynn and you get the idea. Darlington remorse-

  lessly disparaged his colleagues’ work if they dared to engage in any branch

  of botany other than his beloved genetics. He sniped at his distinguished

  ecological colleagues by paraphrasing Oscar Wilde with the scathing quip

  that ecology is the ‘pursuit of the incomprehensible by the incompetent’.40

  Darlington was fascinated with the Botanic Garden at Oxford and intro-

  duced genetic hybrids important for his chromosome studies. Search the

  grounds and you might find the date-plum tree ( Diospyros lotus) planted to com memorate his memory. In the twenty-first century, under pressure for

  space as facilities expand, Darlin gton’s wood-panelled office became swal-

  lowed up by modernity.

  Office pedigree in academia is undoubtedly a niche subject and we should

  not become distracted by it, except to note that a similar story unfolded for the office facing mine in Sheffield. The lineage of its distinguished former occupants reaches back to at least 1938 with the leading plant ecologist Roy

  Clapham,41 our erstwhile head of the Department of Botany (as was) in

  Sheffield for 25 years, who had arrived from Plant Sciences in Oxford. David

  Walker, who laboured on elucidating important biochemical details of

  photosynthesis, followed Clapham.42 Walker was an eccentric. He had an

  extra door built into his office wall to permit rapid escape into his adjacent lab should earnest students come looking for their tutor. The distinguished

  mycologist Sir David Read, who went on to become Vice President of the Royal

  Society, followed Walker. Walker’s escape hatch has since been covered over

  during its conversion into office space. Everywhere, it seems, the remnants of

  scientists past are disappearing.

  Ancient genes, new pl Ants a 81

  This is not the end of the line for Dolan’s RSL genes, however, because they also control the formation of specialized cells and structures in liverworts.43 As we

  have already seen, fossils suggest that liverworts are living representatives of

  an ancient plant lineage. They may have ancestral attributes representative of

  the early land plants colonizing terrestrial environments,44 although such infer-

  ences are contingent on resolving uncertainties in the evolutionary relationships between the bryophytes (hornworts, liverworts, and mosses) and vascular plants.45

  Nevertheless, the point is that the liverwort Marchantia polymorpha anchors itself into the ground with single-celled filamentous rhizoids whose development is

  under the control of RSL genes. These genes also function to make asexual multicellular reproductive propagules (called gemmae) held in splash cups on the

  upper surface of the plant, and multicellular slime cells, called papillae. Marchantia lines lacking RSL genes fail to develop rhizoids, have empty gemmae cups, and lack papillae (Figure 11). Conversely, engineer plants with a hyperactive RSL gene and the liverwort sprouts hairy rhizoids on both its upper and lower surfaces, and a super-abundance of gemmae. So it seems that RSL genes can control both the formation of unicellular structures, such as rhizoids for nutrient acquisition, and multicellular structures for reproduction. Liverwort RSL genes also function perfectly well when transferred to flowering plants lacking RSL genes, enabling them to grow hairy roots.

  The startling conclusion drawn from these discoveries is that RSL genes likely controlled the development of the first land-plant rooting systems. Indeed, they

  probably already existed in the charophyte algal ancestors of land plants, which, for good measure, also passed on mechanisms for taking up the essential inorganic nutrient phosphorus from soils and sediments.46 Certainly, the advantages

  hairy roots give to plant life when it comes to succeeding on land are not to be

  underestimated. Rye plants typically have around 14 billion root hairs, and these provide a surface area of over four hundred square metres for nutrient and water

  uptake.47 Multiply such benefits up as land floras sank their rooting systems

  into developing soils, and it is easy to imagine how root evolution must have

  played a crucial role in shaping the green future of Earth’s continents. We can

  go further than this, because the transition from freshwater to a terrestrial

  existence involved the evolution of a number of specialized cells, tissues, and

  organs required for survival and reproduction, and these are ancestral functions

  of RSL genes.

  82 a Ancient genes, new pl Ants

  ‘Normal’ liverwort

  Marchantia with mutated

  Marchantia with hyperactive

  Marchantia

  RSL genes fails to develop

  RSL genes develops rhizoids

  rhizoids and gemmae

  on its upper and lower

  surfaces and gemmae

  Gemma

  Rhizoid

  precursor

  cell

  Thallus

  Gemmae

  cup

  with

  gemmae

  Empty

  gemmae

  Rhizoids

  cup

  Normal root hairs

  Arabidopsis with

  Arabidopsis with an

  on the root of

  mutated RSL genes

  RSL gene from

  Arabidopsis

  fails to develop

  Marchantia develops

  root hairs

  normal root hairs

  Arabidopsis thaliana

  root

  Root

  hairs

  Figure 11 Ancestral functions of RSL genes in land plants as demonstrated by genetic manipulations with a liverwort ( Marchantia) and a flowering plant ( Arabidopsis).

  Dolan’s group are keen to exploit these discoveries by showing how fundamen-

  tal plant science research can make a wider contribution to society by tackling

  issues related to food security in a changing climate. Engineering crop plants with hairier roots means they might be better able to withstand the heat and drought

  of warmer future climates by taking up more water from the soil. The genomes of

  Ancient genes, new pl Ants a 83

  grasses such as rice, wheat, and maize also contain RSL genes that control root hair formation and function interchangeably with flowering plants.48 This implies that their function has remained similar (or conserved) for millions of years since the evolutionary split between monocots and dicots. Eventually it may be possible to manipulate the RSL genes in wheat and rice to develop super-hairy roots.49

  Such genetically ‘improved’ plants might grow bigger and produce higher yields,

  yet require fewer artificially provided nutrients because they are more effective at scavenging them from the soil.50 Similar arguments hold for water, bringing

  within reach improved agricultural food production with reduced demands for

  artificial fertilizers.

  We are starting to appreciate how evo-devo studies into the genetic workings of

  living plants shines a light on the genetic make-up of ancestral plants that have been extinct for hundreds of millions of years. Fossils in rocks archive the dramatic
diversification of plant life, and DNA connects us to that vanished world

  by explaining how the molecular events inside cells brought it about. There is

  one further important point uniting evidence from fossils and molecules that

  we need to unpack, because it helps explain a little more how tissues in the

  sporophyte life-cycle phase may have arisen.

  Recall that extant plants have two different life-history phases. For example,

  the moss gametophyte (green filaments) and sporophyte (spore capsules) or, in

  the case of ferns, the diminutive gametophyte (called a prothallus) and the large sporophyte with its arching fronds. Now, fossils suggest that, unlike these living examples, the tissues of both life-cycle phases in early vascular land plants may have been complex. To be sure, the two phases were not identical and the habit

  and size are poorly understood, but it seems the gametophyte and sporophyte

  bore greater similarity to each other than is the case for living species. They also developed a greater diversity of tissue types, including rooting structures, a vascular system, and microscopic pores called stomata. And this has far-reaching consequences for understanding plant evolution because it implies that hundreds of

  millions of years ago the genetically controlled developmental programmes for

  making gametophytes and sporophytes were similar.51 The evolution of an inde-

  pendent sporophyte, which dominates the world’s floras today, was made possible

  by the co-option and redeployment of an existing repertoire of developmental

  programmes that the gametophyte was already using to live as a fully independent

  84 a Ancient genes, new pl Ants

  plant. This hypothesis infers genetics from fossils and is supported by the findings from evo-devo we have been considering. Many ancient genes and gene regulatory

  networks discussed—from leaves, shoots, roots, and vascular tissues—were

  recruited from pre-existing networks that operated in the gametophyte generation.

  Still, comparatively speaking, important innovations like leaves and roots are

  simple structures. So what does evo-devo have to say about how plants pull off

  the phenomenal trick of developing two completely different generations within

  a single lifetime? Botanists call this kind of split life cycle the ‘alternation of generations’, and it means that a single genome has to code for two developmental

  programmes: one for building the sexual generation (a gametophyte plant) and

  one for building the asexual generation (a sporophyte plant). Until recently,

  geneticists and developmental biologists working on the evolution of life cycles

  were stumped when it came to explaining how plants managed to achieve this

  remarkable feat. Then a recent series of breakthrough discoveries, in a variety

  of organisms ranging from fungi and algae to moss and flowering plants, trans-

  formed our understanding by showing that they all share common genetic

  machinery controlling the transition from one developmental stage to the next.52

  Our understanding of land-plant life cycles comes mainly from research on

  moss, and it is a story about gene silencing to completely repress the gametophyte developmental programme in order to make a sporophyte and vice versa.53 The

  gene responsible for controlling the development of sporophytes in moss is called MKN6. It works by suppressing the developmental programme for making the

  green filamentous strands, i.e., the gametophyte generation.54 Put another way,

  MKN6 is a gene silencer, blocking gametophyte development to enable the sporophyte to develop instead. It is a similar story in ferns, although the details are not nearly as well understood. Ferns are the evolutionary intermediate lineage

  between mosses and seed plants. The fern Ceratopteris has four genes closely related to MKN6, activated only when the adult (sporophyte) generation is developing.55

  Together these observations suggest that the fern genes switch off the developmental programme of the gametophyte. Mosses and ferns seem to have adopted similar

  gene-silencing mechanisms for releasing the developmental programme to make

  the sporophyte stage of their life cycles. The switch is of considerable antiquity, dating back to the time of the last common ancestor of eukaryotes.56

  Now consider the opposite situation, the formation of the gametophyte gen-

  eration. This requires inactivation of the sporophyte developmental programme,

  Ancient genes, new pl Ants a 85

  this time through a set of mechanisms involving epigenetic gene silencing.57

  Epigenetics is a fashionable branch of science that does not involve modification to the DNA coding for genes but instead changes to how genes are expressed, or

  indeed, if they are expressed at all. In the real world, gene expression is not really an ‘on’ or ‘off ’ toggle switch; the levels of gene expression vary, like the analogue volume control of a radio. Epigenetic modifications allow plants to vary the

  volume control of genes and can involve chemical modification to the globular

  proteins around which DNA is wrapped inside the nucleus of a cell. It is actually rather misleading to think of the DNA double helix as snaking around inside a

  nucleus. In reality, it is wrapped tightly around protein globules that are stacked on top of each other, condensing the volume of space the DNA molecule occupies. When a cell divides, chemical modification to these globular proteins, called histones, can be passed on, so control of gene expression stays consistent from

  mother cell to daughter cell.58

  The details are complex but the bottom line is that two genes are involved in

  epigenetic silencing of the developmental programme responsible for building

  moss sporophytes (for the record, they are called FIE 59 and CLF 60). These genes code for a protein that catalyses the chemical modification of a particular target histone, which blocks access to the DNA at key points. Mutations in either gene

  cause the filamentous gametophyte to develop spontaneously sporophyte-like

  bodies without the usual need for fertilization of the egg cell by sperm.

  Physcomitrella mutants lacking CLF, for example, have gametophytes relieved of epigenetic ‘silencing’, which then develop sporophyte-like structures.61 Under

  laboratory conditions, these mutants can exhibit branching patterns reminis-

  cent of Cooksonia fossils.62 Components of the epigenetic machinery may have originated in algal ancestors, which then increased in complexity as plant life

  transitioned from water to land sometime in the Ordovician.

  We are starting to see how ancient genetic machinery enabled the evolution of

  leaves, roots, and complex life cycles that lay behind the great Devonian diversification in the kingdom of plants.63 But as plants with increasingly elaborate architectures sprang up and began photosynthesizing in the sunlight, all this potential for riotous growth meant they needed the ability to start making reliable decisions about when to grow, and in what direction. Chemical messengers bringing

  news from the outside world, in the form of hormones, are the answer. Synthesized in response to environmental cues, and transported between different parts of the

  86 a Ancient genes, new pl Ants

  plant, hormones tell shoots or roots when and where to grow. Foremost among

  these hormones is auxin. We have already touched on its role in leaf and root

  development, but it actually regulates almost everything in plants. Writing The Power of Movement in Plants in 1880, Charles and Francis Darwin postulated its existence nearly 50 years before the Dutch botanist Frits Went (1903–1990)64 named

  and chemically characterized auxin in 1926. Auxin is the mysterious ‘power’ of

  plants
in the title of their book.

  The role of auxin biology in orchestrating the evolving complexity of the

  vegetable world’s takeover of the land is far from understood. Most green lineages have genes for synthesizing it,65 but what is important for seeing further is knowing about the shared heritage of the genes involved in transporting auxin. Modern plants produce auxin in their shoot tips and transport it downwards from the

  shoot to the root via files of specialized cells in the vascular tissues. This establishes a directional flow of auxin and creates concentration gradients within the different parts of the plant, which are the key to its widespread effects. Embedded in the walls of the cells transporting auxin are special membrane-spanning proteins that regulate the flux of auxin into and out of cells. Genes make the proteins and we can trace their evolutionary history to see when they originated. Traced in this way, we find that genes coding for the proteins involved in transporting auxin are ancient, originating sometime before green plants made the transition to

  land.66 Exactly how they arose is unclear, but one idea is that the ancestral function of the transporter proteins was to regulate cellular metabolism by sequestering auxin inside cells.67 To do this, the proteins were highly localized within the membranes of special compartments located inside the cells of the immediate

  algal ancestors of land plants. At some point in time, an important mutation then changed the localization of these proteins, moving them to the cell wall. With that development, plants gained the ability to export auxin from cells and establish

  directional auxin transport, a necessary step for achieving its action-at-a-distance properties.68

  Directional auxin transport can be seen in trees of the early Carboniferous for-

  ests, which have not only been preserved as fossil wood, but also carved in stone at the Natural History Museum, London. Located in a fashionable corner of

  London’s South Kensington, the museum is a famous cathedral to nature, cele-

  brating the extraordinary former diversity of life on Earth. Built between 1873 and

  Ancient genes, new pl Ants a 87

  1880, the British architect Alfred Waterhouse (1830–1905) dreamt up the design

  and terracotta façade. Over 4 million visitors pass each year through the grand

  entrance set at the top of the broad sweeping stairway leading up to the main

  building. On reaching the top of steps, few visitors notice that Waterhouse