'Deep Gene' and
'Deep Time'
Evolving
collaborations parse the plant family tree
By Barry
A. Palevitz
Amid
last month's hoopla over the human genome
sequence and what it says about humans, plant
biologists announced two new efforts aimed at
a firmer understanding of plant
evolution--who is related to whom and how--a
discipline better known as systematics.
Constructing evolutionary family trees is
harder than investigating personal
genealogies--biologists don't have the
equivalent of birth registrations or family
bibles to consult. Fossils tell them what
ancient plants use to look like, but placing
them in context with living organisms is
difficult at best. Even the systematics of
existing plants can be contentious, as
researchers disagree on lumping plants
together or splitting them apart in search of
the most natural taxonomy.
Scientists
liken constructing phylogenetic trees to
tracing all the branches and trunks of a real
tree, like an oak, with only characteristics
of its outermost twigs to go on. That's
because present day organisms are the sole
survivors--called "terminals" by
systematists--of multiple, diverging
lineages. However daunting the process,
researchers have made breathtaking progress
in the last 20 years, thanks to gene
sequencing. According to University of
Georgia systematist David Giannasi,
"it was a case of technology catching up
with theory." By comparing DNA sequences
such as those encoding ribosomal RNA and
chloroplast proteins, systematists redrew
large chunks of the plant taxonomic map.
A good
example of the redefining process is found in
the milkweeds, which taxonomists
traditionally placed in a family called the
Asclepiadaceae. They also thought the
milkweeds were allied with a second family,
the Apocynaceae. But based on molecular data,
"the Asclepiadaceae nests within the
Apocynaceae," says Giannasi, "so we
now know they should be lumped
together." The same is true for the
mints, thought to be in their own family just
a few years ago but now grouped with the
verbenas.
Researchers
have also clarified some of the most basal
groups in the plant family tree. They now
know that a previously obscure New Caledonian
shrub called Amborella is sister to all other
flowering plants, or Angiosperms, with water
lilies branching off the evolutionary trunk
at the same level or just above.1
They also think the gnetales, previously
considered flowering plant allies, are
probably more closely related to pines, in
the Gymnosperms.2 And horsetails
and whisk ferns, once thought to relic
descendents of early land plants, now seem
more closely tied to the true ferns.3
Feds Fertilize
Interactions
One of
the key ingredients in systematists' recipe
for success was cooperation and
communication. Thanks to joint funding
starting in 1994 from the U.S. Department of
Agriculture, Department of Energy, and
National Science Foundation, a consortium of
researchers called the Green Plant Phylogeny
Research Coordination Group, or Deep Green,
pooled ideas and resources in a joint plan of
attack. Machi Dilworth, head of NSF's
Division of Biological Infrastructure,
thinks, "Deep Green was one of the very
visible success stories" of the three
agency effort. "With a little support
they were able to come together and
accomplish major scientific
achievements."
NSF was
so impressed with the collaborative approach,
it decided to fund "Research
Coordination Networks" (RCNs) serving
all areas of the biological sciences. Like
Deep Green, the grants foster communication
and collaboration between scientists, but
don't directly cover research costs funded by
other programs. Two of the RCNs are scions of
Deep Green.
Systematists Dip
Into Genomics
In one
of the team projects, called Deep Gene,
systematists join forces with molecular
biologists working on entire genomes like
those of Arabidopsis and rice.4,5
By tracing suites of genes that govern
processes such as flower development, they
hope to clarify mechanisms governing major
evolutionary changes, including new
biochemical pathways and the appearance of
complex morphological characters. Sequencing
also uncovers large-scale genomic changes
including chromosomal rearrangements, which
can be invaluable in defining plant
relationships. Likewise, evolution depends on
alteration in spatial and temporal controls
governing gene activity--when and where genes
turn on and off. The new RCN hopes to
discover how gene regulation changed in the
evolution of various plant groups.
Tolerance
toward desiccation is a good example of how
traits may have appeared and disappeared
during evolution. The first plants to occupy
dry land faced a big problem compared to
their aquatic ancestors: an uncertain supply
of water. Mosses, for example, grow in moist
environments but also suffer periodic drying.
That's why they require biochemical
mechanisms that allow them to survive dry
periods. When larger vascular plants arose,
with roots and a plumbing system to extract
water from the soil and move it long
distances, desiccation tolerance became less
important. But it reappeared later on in seed
plants, which remove water from tissues
surrounding young embryos in preparation for
dormancy.
According
to Deep Gene principal investigator Brent
Mishler of the University of California
at Berkeley--and a veteran of Deep
Green--"around 80 genes are involved in
desiccation tolerance in mosses. When
desiccation re-evolved in seeds, some of
these genes were reused." Mishler would
like to know how such changes in gene
regulation arose during major evolutionary
events. Mishler chaired a symposium on Deep
Green at the annual meeting of the American
Association for the Advancement of Science,
February 15-20, in San Francisco.
Daphne
Preuss, molecular biologist at the
University of Chicago and Deep Gene co-PI,
says she brings to the table "the tools
and techniques of high throughput, big scale
biology." Still, in a true collaboration
everybody benefits. With Deep Gene,
genomicists like Preuss want to advance their
own projects. In her case, that means
figuring out how centromeres work.
Centromeres are DNA sequences located where
chromosomes attach to spindle fibers during
mitosis and meiosis. Preuss has dissected
centromeric DNA in Arabidopsis but
knows that "the sequences are very
diverse from organism to organism." The
question is, "how did these differences
evolve, and what key components are important
for centromere function?" Adds
Preuss," I want insight from looking at
conservation through evolution."
Preuss
admits that "this is expensive work, so
every decision counts. We're now making key
decisions as to which species to look at
next. We're looking to people in
phylogenetics to help." Mishler sees
other practical benefits from Deep Gene.
"Can we use the information for
agriculturally important plants that aren't
desiccation tolerant?" he asks. By
guiding researchers to promising sources,
Deep Gene can also "predict useful
chemicals for pharmacology," says
Mishler. That makes University of Georgia's
Giannasi smile because older studies
comparing the chemical composition of
plants--including substances such as
terpenoids--predicted changes cemented by
more recent gene sequencing projects.
"The secondary chemistry was there, but
nobody trusted it," comments Giannasi.
Fossils and
Morphology Join the Fray
Doug and
Pamela Soltis of Washington State
University in Pullman lead another RCN called
"Deep Time." Having done much of
the gene sequencing for Deep Green, the
Soltis' want to superimpose other kinds of
information on their phylogenetic trees, and
in the process add the dimension of time to
key points in plant evolution.
Years
before systematists accessed gene sequences,
they relied on other information in the form
of morphological, anatomical and chemical
characters. While valuable, such characters
can be misleading. For example, a structural
trait shared by two groups could have arisen
by convergent evolution rather than common
ancestry (though the same applies to DNA
sequences). Moreover, the number of
structural characters applicable to
phylogenetic analysis is limited; DNA
sequences, on the other hand, are far more
useful since the average protein encoding
sequence contains 1,000-2,000 characters, or
nucleotides. That's why they turned to genes.
But the
tide may be changing again, at least a
little. The Deep Time RCN will arrange plants
according to a "morphological
matrix" of characters, but
"constrain the taxa to conform to the
DNA-based topology already available, and in
which we have good confidence at this
point," say Pam and Doug Soltis. They'll
then "conduct a phylogenetic analysis of
the morphological matrix with fossils
included." The trick will be to pick
characters from existing plants that also
apply to fossils. Despite the fact that
"fossils have rarely been integrated in
a phylogenetic context for any group,"
the Soltis' are hopeful. Since dates are
available for many of the fossils, their
inclusion adds a time factor to the
phylogenetic tree--systematists can assign
dates to key branch points. They'll also
integrate data from molecular clocks governed
by mutations. "It's sort of like the
movie Back to the Future,'' note the
Soltis', "Having the timing of a key
event in the past nailed down is critical in
understanding what has occurred to produce
what we see in the present."
The
Soltis' also wax philosophical about the
collaboration: "We spent a decade in the
area of systematics largely focused on
molecules. There is a wealth of information
in nonDNA characters such as morphology and
anatomy, and we can't lose expertise in these
areas."
Problems?
Cooperation is the Key
Deep
Gene and Deep Time researchers realize that
reaching their goals may not be easy.
According to the Soltis', "two big
issues are missing data and the combinability
of molecular and morphological data
sets." Mishler agrees: "We don't
know entirely how to do it. Theory hasn't
kept pace--it's dealt mostly with sequence
data." Researchers hope the latest
collaborations will foster development of new
methods to tackle such problems. Mishler sees
promise. "The RCN will help us. Even a
small amount of data from these other sources
can improve phylogenetic trees" and
eventually "lead to more research
funding." The depth of cooperation is
all the more impressive because deep Gene and
Deep Time will interact.
The
"Deep" projects testify to the
importance of collaboration in modern
research. According to Doug Soltis, "the
cooperative nature of botanists has really
turned the tide in the past decade."
Mishler agrees that "research would have
gone on, but it would not have made the
progress it did." Preuss taps federal
agencies for greasing the skids. "Some
of these things are initiated by granting
incentives, so I think it's wise. It's good
to stir the pot and mix people
together." Adds Machi Dilworth of NSF,
"we would like to foster communication
among scientists, to advance science through
collaboration and coordination."
Barry
A. Palevitz (palevitz@dogwood.botany.uga.edu)
is a contributing editor to The
Scientist.
References
1. B.A.
Palevitz, "Discovering
relatives in the flowering plant family
tree," The Scientist, 13[24]:12,
Dec. 6, 1999.
2. L.M.
Bowe et al., "Phylogeny of seed plants
based on all three genomic compartments:
extant gymnosperms are monophyletic and
Gnetales' closest relatives are
conifers," Proceedings of the
National Academy of Sciences, 97:4092-97,
April 11, 2000.
3. K.M.
Pryer et al., "Horsetails and ferns are
a monophyletic group and the closest living
relatives to seed plants," Nature, 409:618-22,
Feb. 2, 2001.
4. B.A.
Palevitz, "Arabidopsis
genome. Completed project opens new doors for
plant biologists," The Scientist,
15[1]:1, Jan. 8, 2001.
5. B.A.
Palevitz, "Rice
genome gets a boost," The Scientist, 14[9]:1,
May 1, 2000.
