Arabidopsis as a model

Dr. R. A. Savidge savidge at
Fri Apr 28 14:20:35 EST 1995

Yesterday I sought clarification as to why so much research effort has been
and is being put into Arabidopsis, a "weed" species.  I received quite a
number of responses that may be of interest to others, so I'll summarize
them below and provide some additional comments [in brackets].  First,
however, I would like to thank Andrew Bishop, Anne Britt, Gregory
Copenhaver, Robertson McClung, Brian Osborne, Dan Vernon, Thomas Warne,
Ruth Wilson, and some anonymous respondees for their time and help.  Rob
McClung's and Thomas Warne's comments were posted to all three lists, so
they need not be repeated.

1]  I had expressed the views, common heard from students, that "the
additional chromosomes in most crop plants surely have something to do with
their being useful to man" and that "research into a genetically limited
non-crop weed is, therefore, simplistically misguided and a waste of money."

Some of the responses to this were:

I)  If Arabidopsis researchers use `crop improvement' as a justification
for their research, they are stretching the point.  The research will
improve our understanding of the systems underlying vascular plant biology,
which will support the next generation of development in crop improvement.

II) Chris Sommerville and co-workers just succeeded in introducing a gene
for plastic production into Arabidopsis so its status as a weed may be

III)  A concerted effort making use of a combined genetic/molecular
strategy in a model system [i.e., Arabidopsis] will identify key processes
of plant development, and will identify the genes that control those
processes.  All plants have, by and large, the same genes, and most will be
identified most efficiently by focussing on a model system.

[I think it is reasonable to assume that all plants have some genes in
common, another way of saying that the present diversity in the plant
kingdom sprang from earlier unity.  But on this same theme, one can argue
that the present diversity - which is a major factor determining the
usefulness of plants - could be due to acquisition of new genes and/or
discarding of existing genes, e.g.  through mutation, transposition,
vectors, etc.  In addition, the evolutionary concept of `reduction'
may well be relevant to Arabidopsis.  Arabidopsis appears as a
relatively simple model system for vascular plants; however, it likely
had it origins in genetically more complex beginnings.  The important
question for many is, will the genes most important for improving crop
productivity be overlooked because the selected model system lacks them?]

  IV) Arabidopsis will not tell us all we need to know about plants.
But the benefits of focussing on a solidly established genetic/molecular
model system for defining basic biological processes and their genetic
control is inarguable.

[Those biological processes considered "basic" for Arabidopsis may be basic
to many other plant species but, nevertheless, may have little or no
relevance to those aspects of plant morphogenesis underlying crop

V) It is simplistically misguided to believe that extra DNA
contains the information important for crop utility.  This DNA is mostly
repetitive junk.

VI) Evidence based on reassociation kinetics, cDNA sequencing,
genome structure and other work indicates that Arabidopsis has a comparable
number of genes to other plants, including crop plants.

VII) Chromosome number or genome size doesn't correlate well with any
particular phenotypic or yield traits.

[I am not convinced about the correctness of responses V, VI and VII.  To
my knowledge, it is fairly well established that the mass of DNA in the
diploid genome is inversely correlated with growth rate for both annual and
perennial plants.  The genetic explanation for this correlation remains
wanting.  Exceptions to this generalization have also been noted, so more
research clearly is needed.  In addition, we know virtually nothing about
the control of many types of cellular differentiation in higher plants.
It has been estimated that a thousand different genes must participate in
cell wall biosynthesis alone, and we don't even know for sure yet how
cellulose is made.  Is junk DNA really the DNA controlling all the aspects
of plant biology that remain to be resolved?]

VIII) Almost no one who works on Arabidopsis has any intrinsic interest in
the organism; it is after all "a weed".  However, many researchers are
interested in what is often referred to as `basic research'.  This type of
research would include investigations into the nature of fundamental
biological mechanisms such as transcription, DNA replication, modes of
inheritance, epigenetic mechanisms, translation, gene regulation, and much


2]  I had suggested that "projecting genetic findings made with Arabidopsis
to other plant species is only valid upward through the evolutionary
tree...and that the research is unlikely to have any value outside of the

Responses to this were:

I) Disease resistance genes have been isolated from Arabidopsis, tobacco,
flax and tomato in the last year.  All these genes are very similar,
suggesting related mechanisms in resistance.  The resistance pathways are
similar, but from a research standpoint Arabidopsis is going to be leading
the field.

II) You may be right [about Arabidopsis research being of minimal
usefulness when projected down to more primitive dicots], but are these
primitive dicots useful as crops?

[If we include woody plants as crops, and indeed many are cropped, then the
answer is clearly `yes'.]

III) I think the term "primitive" is misleading when applied to organisms
which exist today.  It's true that algae, mosses, ferns, horsetails, cycads
and gymnosperms first appeared on the earth at earlier times than
flowering plants.  But are the modern representatives of these groups
really primitive?

[Leaves identical with those of Ginkgo biloba occur in the fossil record as
far back as the Triassic (some 200 million years ago), and there are a
number of other similar examples.  I think it is fair to say that
gymnosperms and many of the lower vascular plants are truly primitive.  The
angiosperms, e.g.  Arabidopsis, are actually the more troubling because of
the uncertainty over their origins.  If the angiosperms arose
independently, then even basic research findings with Arabidopsis may be
of minimal usefulness in relation to more primitive taxa.  Arabidopsis
research could be very useful in helping to know better how angiosperms fit
into the evolutionary tree, e.g.  by comparing DNA homologies between it
and other species.]

3]  My third point, a logical `question' from thoughtful students, was that
"if the goal is to understand aspects of plant physiology that are general
to the majority of the species in the plant kingdom, an evolutionarily much
simpler system should become the model."

Responses to this were:

I)  Arabidopsis has several characteristics that make it a good model
organism for higher plants.  It has a very compact genome (5 small
chromosomes) and  a rapid generation time (8 weeks from germination to seed
set) allowing several generations to be studied in a year.  It self
fertilzes naturally but can be made to cross.  Being small, a few hundred
seedlings can be analyzed in a single petri dish.  There is a vast bank of
seed stocks, DNA, mutants and resources available, and Arabidopsis can
easily be made to express foreign genes.  Few if any agronomically useful
organisms present the complete package that Arabidopsis provides.  If you
want to study fundamental mechanisms, A. thaliana is a powerful tool.  If
you want to study how to build a better tomato or make a pine tree grow
fast, then other tools will be of much greater value.

II)  There is plenty of research going on in other plants too, when that
plant is the appropriate model system.  Just a year or two ago, USDA
required that an investigator explain why he or she is doing a project
using arabidopsis rather than a crop plant.  Now, however, panels often ask
investigators to explain why they are NOT using Arabidopsis.

III)  Much of basic plant metabolism, DNA enzymology, photosynthesis, and
cell biology could be better addressed by further development of
Chlamydomonas as a model system.  However, you'd still need a "higher
plant" model for developmental biology.

IV)  If one wants to learn about things which all plants do, then any
member of the plant kingdom can serve as a model.

V)  The fact that they are studying a dicot reflects the degree to which
they are motivated by _applied_ concerns.  A horsetail may be simpler, so
maybe we have to conclude students of Arabidopsis are interested in

[Thank you all, once again!]

   Rod Savidge, PhD                 |         E-mail: savidge at
   Faculty of Forestry and         \|/
      Environmental Management  \   |   /     Phone:  (506) 453-4919
   University of New Brunswick  _\/ | \/_
   Fredericton, NB CANADA          \|/        Fax:    (506) 453-3538
   E3B 6C2                          |

More information about the Plantbio mailing list