Consider the case of the liverwort genome, which hints at some of the genetic
innovations that define land plants. Liverworts are among the earliest diverging
lineages of land plants and grow as flat, green photosynthetic cushions. They have probably lost some ancestral traits and acquired a mosaic of others. Other groups of liverworts include ‘leafy’ liverworts that have a more familiar morphology, with a stem and leaf-like extensions. The Marchantia genome has ca. 19 000 genes, a few thousand less than Chara. Some of these genes are shared with these green algae, others with vascular plants to the exclusion of green algae, and still others are uniquely its own. An expansion of the diversity of genes regulating the genome has been detected and these correspond with an increased diversification of developmental and environmental responses compared to green algae. Presumably these
are useful for coping with the wider range of environmental conditions found on
land compared to life in the water. The genome also has components of the flower-
ing plant-like machinery for plant hormones, genes encoding ‘novel’ biochemical
pathways to cope with the stress of bright sunlight, and components for cuticle
and wax biosynthesis to prevent water loss and protect against insects. These
genomic attributes are again consistent with plants gaining critical features of their biology during the evolutionary transition to land. There is also strong genomic
evidence of substantial gene transfer from fungi and bacteria, including genes
involved in the biosynthesis of chemical compounds called terpenes that contrib-
ute to oil bodies, defensive organelles unique to liverworts.
Genomes decoded a 51
One lesson to be drawn from analysing the genomes of bryophytes is that
duplicating and tinkering with bits of DNA inherited from algal ancestors can
generate the evolutionary novelty needed to build and operate a multicellular
land plant. Genes of algae, it seems, can usefully serve land plants with some evolutionary fiddling in how they are wired together. There is an important general
principle here: evolution works with what is to hand. As we have already seen,
some of the genetic material for the physiological adaptations needed for living
on land was likely already present in the algal ancestral lineage, with much of the extra genetic material needed for developmental innovations originating in the
ancestral land plant. We see the pattern repeated when vascular plants, with their hallmark internal plumbing connecting roots to leaves, originated.
Lycophytes (clubmosses, spikemosses, and quillworts) are the descendants
of early-evolving vascular plants and represent an interim step in evolution,
reproducing by spores instead of seeds. Decoding the genome of the spikemoss
Selaginella moellendorffii (Selaginellaceae)32 was the first step undertaken by a large consortium of scientists. Comparing it with the genomes of algae, moss, and
flowering plants was the next step. The results showed it took a trifling number of extra genes (about 500) to make a vascular land plant, as represented by spikemoss. The later evolution of a flowering plant required an extra 1350 genes. Put
another way, three times as many genes were required to invent a flowering plant
starting from a lycophyte baseline than were required to invent a vascular plant
starting with a bryophyte baseline.33 At least this is true for a spikemoss: coverage of species within particular lineages, whether algae, bryophytes, or lycophytes, is extremely limited in terms of genome sequencing, so we should be cautious about
drawing firm conclusions regarding the importance of changes in gene numbers.
More important is that these examples serve as model plant species and this rep-
resents one of the great successes in modern biology, allowing us to investigate a wide range of long-standing biological questions about evolutionary mechanisms
and processes.34 We can illustrate the point by looking at how plants ‘invented’
vascular tissues for transporting water and the lignin needed to strengthen these specialized water-conducting cells.
As we might expect, some of the extra genes found in the spikemoss genome
compared to the moss and liverwort genomes are involved in synthesizing vascu-
lar tissues. These tissues are comprised of specialized cells for transporting water (xylem) and sugars (phloem). Mosses lack xylem and phloem but do have bundles
52 a Genomes decoded
of simple hollow water-conducting cells and phloem-like cells whose develop-
ment is genetically controlled. Genes carrying the detailed instructions for vascular plants to make hollow, water-conducting xylem cells were discovered to be
closely related to those used by mosses for the formation of their conducting bundles.35 The liverwort Marchantia, which has no internal water-conducting cells, has a rudimentary network of the same genes. Charophytes possess a handful of
related genes whose function is uncertain. All of which suggests something
extraordinary. Vascular plants ‘learned’ to make their vascular tissue from ancestral pre-vascular land plants.
The appearance of vascular tissue presented its own challenges because these
specialized hollow cells, linked end-to-end, needed stiffening to prevent collapse and support the stem. An ability to synthesize the immensely strong and recalcitrant polymer lignin provided a solution. Lignin gave plants structural rigidity
and the ability to stand upright. Lignin explains why trees are woody and why
wood takes so long to decay. The extensive coal forests that developed through-
out the Carboniferous and Permian (ca. 323–252 million years ago) were com-
posed of giant arborescent relatives of clubmosses, as well as lignin-producing
ferns, horsetails, woody shrubs, and trees related to conifers. During burial in
sediments, the organic remains of the plants are slowly transformed into peat,
lignite, and then coal. Massive coal deposits formed in this way developed largely because of widespread waterlogged conditions inhibiting decay, thanks to a combination of climate and tectonics during the formation of the supercontinent
Pangaea. Towards the end of the Permo-Carboniferous, as the interval is known,
burial of organic matter slowed as the climate dried out, drawing to a close an
extraordinary era of coal formation.36 Hundreds of millions of years later, that
coal was to fuel our own industrial era. That, at least, is the widely accepted
standard view.
A complementary hypothesis emerged when David Hibbett of Clark University
sequenced the genomes of a wide variety of fungi and used this information to
reconstruct their history, and specifically when white rot fungi gained the meta-
bolic capacity to degrade lignin and unlock its concentrated carbon energy.
Hibbett’s group suggested that a diverse and versatile arsenal of lignin-digesting enzyme pathways, comparable to those of modern white rot fungi, may have arisen
late in the Permo-Carboniferous.37 Estimates of the range of dates are sensitive to
Genomes decoded a 53
the choice of fossils used to calibrate the molecular clock involved and have considerable uncertainty. Granted these uncertainties, Hibbett suggested that after
they had evolved, they assisted decomposition of the lignin-rich woody debris of
early forests and slowed coal formation.38 Intriguing fossils resembling modern
white rot occur in seed ferns from the Permian (260 million years ago), and
gymnosperms from the Triassic (230 million years ago), consistent with the infer-
ences from genomics that these fungi were on land at around the right time. But
debate highlights the challenges facing those looking to integrate a genomic per-
spective into Earth’s complex history.
The invention of lignin probably begins with the first land plants, not algae.40
To be clear though, the biosynthetic machinery for making chemical compounds
required to help multicellular plants cope with the challenges of a terrestrial existence is evolutionarily very ancient.41 The point is that occasional reports of discovery of lignin in charophyte algae have proved controversial and not yet
with stood further scrutiny. The moss Physcomitrella has a handful of genes for synthesizing the basic chemical scaffold of the stuff,42 whereas our freshwater green alga Chlamydomonas has none. The chemical complexity of lignin is such that even a scaffold has its uses for moss because it absorbs wavelengths of solar energy in the ultraviolet range, protecting emerging land plants. Components of the biochemical apparatus involved in its synthesis were likely present in ancestral freshwater algae, which used them to make the tough protective outer coating of
spores—sporopollenin.43 The spikemoss genome includes genes enabling it to
synthesize proper lignin to give its water-conducting cells structural rigidity. The steps by which the pathway for the synthesis of lignin was assembled are not yet
known; some of the genes may even have originated from symbiotic fungi and
bacteria.44 Other lineages of vascular plants, including ferns, gymnosperms,45 and flowering plants,46 invented the same type of lignin, but not always via the same biosynthetic pathways. Lignin is a rather wonderful molecular example of convergent evolution—the convergence of different evolutionary lineages on the
same, or a very similar, solution to a problem. Think of the independent evolution
54 a Genomes decoded
of wings for flight in bats and birds, and echolocation in dolphins and bats, which are classic examples from the animal kingdom.47
Lignin also makes trees possible. Once plants mastered the ability to synthesize
lignin, it was not too long, in an evolutionary sense, before trees evolved, as the rich Devonian record of fossil plant discoveries makes clear. In fact, the molecular basis for differences between woody perennials and annuals appears to be rather
small. Suppress the activity of only a couple of genes in Arabidopsis, for example, and you can extend its growth phase and make it produce wood.48 We can speculate that it simply was not that difficult for plant life to evolve perennial growth and the woody habit characteristic of trees. Indeed the tree habit has arisen
independently multiple times throughout plant evolution—eventually the
monocotyledons got in on the act (think of dragon trees and palms).
Another line of reasoning supporting this view comes from what happens on
islands when plants have free rein, as both Charles Darwin and Alfred Wallace
were well aware. They were among the first naturalists to report that, given the
chance, island-dwelling herbs soon exhibit a tendency towards a tree-like growth
habit. Astonishing evolutionary events on the Macaronesian islands, located in
the Atlantic between Europe and North Africa, have been unravelled by analysing
the DNA of plants belonging to the genus Sonchus, a member of the daisy family (Asteraceae). According to the DNA evidence, several species of tree on the islands, including Sonchus canariensis, evolved from the small, closely related herbaceous ancestral plant Sonchus asper, within a few million years of its appearance on Gran Canaria or Tenerife in the Canary Islands.49 One hypothesis is that extinctions in the Canaries following the Pliocene glaciation 2–3 million years ago created open habitats into which tree-like forms could evolve and diversify.
It is a similar story for those botanical glories on the desolate St Helena, an
isolated speck of a volcanic island where the fallen emperor Napoleon was exiled
for the rest of his life over 200 years ago: ‘gumwood’ ( Commidendrum) and ‘cabbage’ ( Melanodendron) trees.50 All species of Commidendrum and Melanodendron are threatened, and officially recognized as such by the International Union for
Nature Conservation. They survive today as small relict populations on cliffs or in fragmented patches of vegetation. As in the Macaronesian islands, the DNA of
these forest trees, which also belong to the Asteraceae, indicates they evolved
from a single common shrubby ancestor that arrived on the island, probably
from mainland South Africa. By fortuitously arriving on St Helena, these plants
Genomes decoded a 55
found the niche of forest trees unoccupied and soon exploited it, evolving into
diverse species of proper trees with trunks, spreading crowns, and woody rooting
systems. Trees are obviously admirable opportunists, and the staggering esti-
mated three trillion trees in forests worldwide today constitute over 80% of the
biomass on the continents, and harbour 50% of terrestrial biodiversity.51
As they support, shelter, or feed an enormous diversity of animals and mic robes, it’s no surprise that the genomes of trees such as the black cottonwood ( Populus),52
and the common fast-growing Australian Eucalyptus grandis,53 have large numbers of genes deployed for doing what a tree needs to do: fighting attacks from harmful organisms like fungi, bacteria, and insects. Protecting themselves from the relentless onslaught of attack by pests and pathogens requires synthesis of chemical
defence compounds. Eucalyptus, with its oil-containing leaves, releases numerous volatile compounds, synthesized by a large collection of more than 100 genes, for defence against a wide array of pests and pathogens; a moss, for comparison, has
just two, and a liverwort, only a few more. The genome of Eucalyptus also carries the signature of the unusual reproductive biology of this tree: its unopened flower buds are covered with a pixie hat, a strange feature reflected in its name— Eucalyptus derives from the Greek eu-, meaning well, and kaluptos, meaning covered. During the blooming process, the pixie hat falls off the vase-shaped floral buds, and the flowers that develop lack visible petals. Inside the genome, we find that genes
regulating the formation of elaborate flowers have gone, while genes for producing large amounts of pollen and seed have massively increased compared to other
flowering plants.
A recurring feature of the various genomes we have been considering is a gen-
eral trend of increasing gene numbers more or less along a gradient of increasing plant complexity moving from algae to trees.54 As if by magic, gene numbers
handily increase, and plants are free once again to systematically advance in
organized complexity. Where do all of these extra genes originate? Whole genome
duplications can provide one major source. During normal cell division, the par-
ent cell divides into two or more daughter cells, with each cell receiving the normal complement of chromosomes. Failure in the normal process of cell division
and chromosome duplication can lead to a genome doubling, with each chromo-
some having a twin. The process effectively creates a ‘spare’ genome, potentially enabling thousands of extra gene copies to be free to evolve new functions.
Genome doubling can also occur when the merging of chromosome sets of
56 a Genomes decoded
closely related species follows failure in cell division. This is the basis of hybridization. Organisms (plants or animals) possessing more than two complete sets of
chromosomes are known as polyploids and the scie
ntific investigation of poly-
ploids dates back over a century,55 to work by the distinguished botanists Hugo de Vries (1848–1935) and G. Ledyard Stebbins (1906–2000). The formation of polyploids is a generally accepted process of speciation, the evolution of new species.56
Only with the advent of new technologies in the era of genome biology did
scientists begin to document whole genome duplication events that were tens of
millions or hundreds of millions of years old. Detecting the remnants of ancient
genome duplications is a high art form, with the search for traces of genetic events that happened millions of years ago demanding computational approaches to
identify blocks of duplicated genes and the application of different techniques for dating the duplication events.57 Few ancient genome duplication events survive,
suggesting that they are often evolutionary dead ends, leading to genomic shock
and sterility. Occasionally, though, duplications can have a profound impact on
evolution by providing two copies of each gene for evolutionary forces to work
on.58 We shall learn how shortly, but for now we can marvel at the theoreticians
who delight in analysing the genome of Arabidopsis. It is worth unpacking a little of what they have found so far because it tells us about the fate of all those extra genes. Arabidopsis shows evidence for having undergone three whole genome duplication events of varying antiquity over the past 150 million years. Scientists have found that the first doubling created 17 193 duplicate genes with only ~4%
retained; the second doubling created 20 316 duplicate genes with ~14% retained;
and the third doubling created 24 351 duplicate genes with ~16% retained.59
The lesson from these numbers is obvious. Whole genome duplications are not
genetic stockpiling events, far from it. Most duplicated genes are quickly erased from the genome on a ‘use it or lose it’ basis. Expansion of gene numbers happens with the gradual accumulation of small numbers of surviving genes over time and
Making Eden Page 9