Download On Systems Thinking, Systems Biology, and the

Scientific Correspondence
On Systems Thinking, Systems Biology, and the in
Silico Plant
Graeme L. Hammer, Thomas R. Sinclair*, Scott C. Chapman, and Erik van Oosterom
Agricultural Production Systems Research Unit, School of Land and Food Sciences, The University of
Queensland, St. Lucia, Queensland, Australia (G.L.H., E.v.O.); Agricultural Production Systems Research
Unit, Queensland Department of Primary Industries, Toowoomba, Queensland, Australia (G.L.H.);
Agronomy Physiology Laboratory, United States Department of Agriculture-Agricultural Research Service,
University of Florida, Gainesville, Florida (T.R.S.); and Commonwealth Scientific and Industrial Research
Organization Plant Industry, St. Lucia, Queensland, Australia (S.C.C.)
The recent summary report of a Department of Energy Workshop on Plant Systems Biology (P.V. Minorsky [2003] Plant
Physiol 132: 404–409) offered a welcomed advocacy for systems analysis as essential in understanding plant development,
growth, and production. The goal of the Workshop was to consider methods for relating the results of molecular research
to real-world challenges in plant production for increased food supplies, alternative energy sources, and environmental
improvement. The rather surprising feature of this report, however, was that the Workshop largely overlooked the rich
history of plant systems analysis extending over nearly 40 years (Sinclair and Seligman, 1996) that has considered exactly
those challenges targeted by the Workshop. Past systems research has explored and incorporated biochemical and
physiological knowledge into plant simulation models from a number of perspectives. The research has resulted in
considerable understanding and insight about how to simulate plant systems and the relative contribution of various factors
in influencing plant production. These past activities have contributed directly to research focused on solving the problems
of increasing biomass production and crop yields. These modeling approaches are also now providing an avenue to enhance
integration of molecular genetic technologies in plant improvement (Hammer et al., 2002).
We suggest that the future requires a broader view
of plant systems biology than the one recommended
by the Workshop so that research effectively connects
emerging capabilities at the molecular and cellular
level to whole-plant improvement and to performance at the crop level. This is certainly true if the
objective of the systems modeling is to predict complex responses in plant behavior to meet the realworld challenges outlined by the Workshop. Lessons
learned over the past decades of plant systems analysis are that exact, careful definition of the system
and careful integration across scales of biological
organization are absolutely critical. In fact, the
progress on simulation of plant systems using biochemical and physiological information has been explored in two recent symposia in which we participated, one held in Europe in 2001 by the European
Society of Agronomy (Donatelli et al., 2002) and the
other held in 2000 in the United States by the Crop
Science Society of America and American Society of
Agronomy (Weiss, 2003). Papers presented at these
symposia explored the opportunities and methods
for systems analysis and modeling to provide the link
between genome scale molecular biology and plant
improvement.
A system can be generally defined as a network of
interacting elements receiving certain inputs and
* Corresponding author; e-mail [email protected]; fax 352–392–
6139.
http://www.plantphysiol.org/cgi/doi/10.1104/pp.103.034827.
producing certain outputs. Quantitative models are
generated as tools to aid understanding and prediction of system output in response to the environment
and system inputs. Models are simplified representations of system dynamics, usually in mathematical
form. A key challenge is defining the scale of operation within the system to be simulated. In plants, a
system can be described equally at the level of gene
action, biochemical pathway, an organelle, a cell, an
organ, a whole plant, or a community of plants (crop
or ecosystem). Defining the system boundary has
proven to be an absolutely critical step in systems
thinking. Orders of magnitude of complexity are introduced as we move from molecular scales to whole
organisms. Although system function can be viewed
and modeled at all of these scales, experience has
shown that only two or three scales are desirable in
effectively simulating any particular system.
The Department of Energy Workshop (Minorsky,
2003) suggested that the very core of systems biology
is the goal of being able to model a living organism
and that the focus should be on the structure and
dynamics of gene function. Although this aligns with
our notion of systems thinking, the implicit assumption in the discussion was that the molecular or genome scale provided the starting point to address
real-world problems. Although we accept this as a
possible starting point, we suggest it is not the only
one, and probably not the best one, if the target is to
design and engineer improved plants.
Although there are examples of success in the use
of genome-wide strategies to understand the control
from on
June 14, 2017 - Published
by www.plantphysiol.org
Plant Physiology, March 2004, Vol.Downloaded
134, pp. 909–911,
www.plantphysiol.org
© 2004
American Society of Plant Biologists
Copyright © 2004 American Society of Plant Biologists. All rights reserved.
909
Scientific Correspondence
systems of specific components of development that
are highly conserved, for example segment polarity
genes of fruitfly (Drosophila melanogaster; von Dassow
et al., 2000) and endomesoderm specification in the
sea urchin (Strongylocentrotus purpuratus) embryo
(Davidson et al., 2002), it is much more difficult to
discern the controls of overall plant growth, which is
critical in crop yield formation. The difficulty in anticipating the controls of growth, i.e. mass accumulation, were highlighted in a study on yeast (Saccharomyces cerevisiae) growth response (Giaever et al.,
2002) that included a nearly complete collection of
gene deletion mutants (approximately 6,000 genes). It
was discovered that only 7% of the genes that exhibited an increase in mRNA expression under optimal
growth conditions were actually involved in decreased growth under those conditions when the
genes were deleted. Ideker et al. (2001) had more
success in obtaining alteration of yeast growth as a
result of gene mutations by focusing on the single
process of Gal utilization, but they still found a relatively high level of posttranscriptional control. We
suggest that a starting point focused on structure and
dynamics of organism function will likely be advantageous and can inform targeting of relevant process
and component studies. We discuss this “top-down”
approach further below.
A molecular “bottom-up” approach to understanding plant growth and development is, on its own,
likely to flounder in complexity. Morandini and
Salamini (2003) highlighted the complexities associated with distributed control of fluxes in plant metabolic pathways. They noted the difficulties this complexity engenders in predicting overall system
dynamics from the enzyme level and referred to
similar problems experienced by fermentation engineers, even when modeling simple yeast reactions
within closely controlled artificial incubators. These
difficulties suggest that a focus on higher order control may be more appropriate, especially when the
goal is to understand overall plant performance.
Plant growth is influenced by the interplay of the
continually changing environment with complex
gene networks (containing many more genes than
yeast), and the functional consequences usually operate at higher levels of biological organization, i.e.
organ and whole-plant development and growth.
Successful adaptive strategies for an organism are
related to the flow of information across levels of
organization in a manner that allows the plant to best
adjust to the prevailing environment. It is the organism level of organization at which selection and evolutionary fitness are effected, although the successful
adaptive strategies must be contained in the genetic
make-up.
Plants are complex adaptive systems. The functioning of such systems is best understood by exploring
how they handle information (for discussion, see
Gell-Mann, 1994). A plant acquires information about
910
its environment and its interaction with that environment and uses that information to dictate its adaptive
responses over time scales of seconds to months or
years. The underlying processes (Gell-Mann’s “schemata”) generate consequences for the plant phenotype. These processes are encoded in networks of
genes and transcription factors that control signaling
and communication processes, and these in turn initiate the growth and development responses of organs. Selection pressures will ultimately identify the
mix of genetic schemata giving rise to the most superior phenotype. However, the genetic schemata
likely contain numerous, redundant networks. They
have become increasingly complex as modifications
have accumulated through evolution via chance
modifications of existing networks. This notion of
evolution as tinkering suggests that selective pressures operating on the organism led to retention of
useful, rather than purposeful, biochemical, and molecular changes (for discussion, see Jacob, 1977). As a
consequence, many insights at the genome scale are
likely to be uninformative at the scale of organism
growth, as found in yeast by Giaever et al. (2002).
Hence, in relation to plant systems biology, it makes
sense to think much more about function at the organism scale as the entry point for considering consequences on processes at lower levels of biological
organization, i.e. a “top-down” approach.
Significant endeavors in the field of whole-plant
modeling are now being directed at understanding
genetic regulation and aiding crop improvement (for
review, see Hammer et al., 2002; for examples, see
Cooper et al., 2002; Chapman et al., 2003). Crop models with generic approaches to underlying physiological processes and software structure to support scientific investigation (Wang et al., 2002) provide a
means to link phenotype and genotype—the in silico
or virtual plant. This approach has utilized a “topdown” approach to dissect the physiological basis of
adaptive traits and their control at the whole-plant
level and, thus, provide the understanding to support predictive modeling. The models must quantify
the functional controls that drive plant adaptive responses. This is akin to “modeling plant hormone
action without modeling the hormones” (de Wit and
Penning de Vries, 1983). It is necessary to connect
across levels of biological organization. For example,
Sinclair and Horie (1989) derived simple and robust
relationships for key phenomena at the crop level by
integrating the functional plant biological processes
at the lower, explanatory organizational level. This
provides the basis for the dialectic and linkage between “top-down” and “bottom-up” approaches. We
suggest that this dialectic is required if new molecular capabilities and tools are to have significant impact on improvement of crop plants beyond single
gene traits. Snape (2001) also has suggested that combining the “forward” and “reverse” approaches will
allow us to get comprehensive knowledge of the
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2004 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 134, 2004
Scientific Correspondence
biology of agronomic traits at the physiological, biochemical, and molecular levels so that the “circuitry”
of crop plants can be elucidated.
There is evidence emerging that engaging a “topdown” approach will be critical in linking molecular
genetics to plant improvement. Tardieu (2003) outlines use of an ecophysiological model of plant water
use to explore the genomics of tolerance to water
deficit. The model contained both physical and control equations. Parameterization of the latter represented coordinated genotypic responses that quantified a “meta-mechanism” at a higher level of
organization. This could then be connected to gene
regulatory networks or quantitative trait loci. Studies
by Reymond et al. (2003) on responses of maize (Zea
mays) leaf growth to temperature and water deficit,
Leon et al. (2001) on photoperiod responses in sunflower (Helianthus annuus), and Borrell et al. (2001) on
the staygreen trait in sorghum (Sorghum bicolor) have
pursued physiological dissection and quantitative
analysis of the phenotype as a means to predict responses of genotypes. They elucidated Tardieu’s
“meta-mechanisms,” Gell-Mann’s “schemata,” or
Snape’s “circuitry” in a manner that facilitates linkage to more detailed understanding of genetic regulation. These studies offer approaches for crossing
scales of biological organization and illustrate the
potential for spanning the gap between genome scale
molecular biology and functional control at the organism scale.
Katagiri (2003) discussed in the same issue of Plant
Physiology as the Workshop report a “top-down” approach and the need to link “top-down” and
“bottom-up” approaches. However, his “top” was
recognizing coherent patterns in microarray-based
gene expression profiling. This approach still remains a long way from the plant level. Perhaps,
though, this can still provide an initial point of linkage between the worlds of molecular physiologists
and whole-plant physiologists. This possibility is reinforced by studies on genetic controls and biochemical signals operating over short time periods within
hypocotyls (Nemhauser et al., 2003) or young plants
(Thum et al., 2003). Groups of coregulated genes
reflect Gell-Mann’s “schemata,” Tardieu’s “metamechanisms,” or Snape’s “circuitry.” We propose that
developing models that mimic plant functional biology is the key to identifying these critical control
factors. With this type of quantitative understanding
of control at the organ or organism level, experimenters would be able to better target and simplify their
large combinatorial experiments. Hence, we suggest
that supporting and investing in the dialectic across
biological levels of organization is of far greater importance to progress in crop improvement than suggestions arising from the Department of Energy Workshop
(Minorsky, 2003) on enhancing cross-disciplinary activities within the confines of plant molecular biology.
Only by understanding the links across all biological
Plant Physiol. Vol. 134, 2004
levels will the “in silico plant” have a real chance of
germinating and flourishing in support of the design
of improved plants to meet real-world challenges.
Received October 15, 2003; returned for revision October 29, 2003; accepted
December 7, 2003.
LITERATURE CITED
Borrell AK, Hammer GL, van Oosterom E (2001) Staygreen: a consequence
of the balance between supply and demand for nitrogen during grainfilling? Ann Appl Biol 138: 91–95
Chapman SC, Cooper M, Podlich DW, Hammer GL (2003) Evaluating plant
breeding strategies by simulating gene action and dryland environment
effects. Agron J 95: 99–113
Cooper M, Chapman SC, Podlich DW, Hammer GL (2002) The GP problem:
quantifying gene-to-phenotype relationships. In Silico Biol 2: 151–164
Davidson EH, Rast JP, Oliveri P, Ransick A, Calestani C, Yuh C-H,
Minokawa T, Amore G, Hinman V, Arenas-Mena C et al. (2002) A
genomic regulatory network for development. Science 295: 1669–1678
de Wit CT, Penning de Vries FWT (1983) Crop growth models without
hormones. Neth J Agric Sci 31: 313–323
Donatelli M, Van Ittersum MK, Bindi M, Porter JR (2002) Modelling
cropping systems–highlights of the symposium and preface to the special
issues. Eur J Agron 18: 1–11
Gell-Mann M (1994) The Quark and the Jaguar: Adventures in the Simple
and the Complex. Abacus, London
Giaever G, Chu AM, Ni L, Connelly C, Riles L, Veronneau S, Dow S,
Lucau-Danlla A, Anderson K, Andre B et al. (2002) Functional profiling
of the Saccharomyces cerevisiae genome. Nature 418: 387–391
Hammer GL, Kropff MJ, Sinclair TR, Porter JR (2002) Future contributions
of crop modeling: from heuristics and supporting decision making to
understanding genetic regulation and aiding crop improvement. Eur J
Agron 18: 15–31
Ideker T, Thorsson V, Ranish JA, Christmas R, Buhler J, Eng JK, Bumgarner R, Goodlett DR, Aebersold R, Hood L (2001) Integrated genomic
and proteomic analyses of a systematically perturbed metabolic network.
Science 292: 929–934
Jacob F (1977) Evolution and tinkering. Science 196: 1161–1166
Katagiri F (2003) Attacking complex problems with the power of systems
biology. Plant Physiol 132: 417–419
Leon AJ, Lee M, Andrade FH (2001) Quantitative trait loci for growing
degree days to flowering and photoperiod response in sunflower (Helianthus annuus L.). Theor Appl Genet 102: 497–503
Minorsky PV (2003) Achieving the in silico plant: systems biology and the
future of plant biological research. Plant Physiol 132: 404–409
Morandini P, Salamini F (2003) Plant biotechnology and breeding: allied for
years to come. Trends Plant Sci 8: 70–75
Nemhauser JL, Maloof JN, Chory J (2003) Building integrated models of
plant growth and development. Plant Physiol 132: 436:439
Reymond M, Muller B, Leonardi A, Charcosset A, Tardieu F (2003) Combining quantitative trait loci analysis and an ecophysiological model to
analyse the genetic variability of the responses of maize leaf growth to
temperature and water deficit. Plant Physiol 131: 664–675
Sinclair TR, Horie T (1989) Leaf nitrogen, photosynthesis, and crop radiation use efficiency: a review. Crop Sci 29: 90–98
Sinclair TR, Seligman NG (1996) Crop modelling: from infancy to maturity.
Agron J 88: 698–704
Snape J (2001) The influence of genetics on future crop production strategies: from traits to genes, and genes to traits. Ann Appl Biol 138: 203–206
Tardieu F (2003) Virtual plants: modelling as a tool for the genomics of
tolerance to water deficit. Trends Plant Sci 8: 9–14
Thum KE, Shasha DE, Lejay LV, Coruzzi GM (2003) Light- and carbonsignaling pathways: modeling circuits of interactions. Plant Physiol 132:
440–452
von Dassow G, Meir E, Munro EM, Odel GM (2000) The segment polarity
network is a robust developmental module. Nature 406: 188–192
Weiss A (2003) Symposium introduction. Agron J 95: 1–3
Wang E, Robertson MJ, Hammer GL, Carberry PS, Holzworth D, Meinke
H, Chapman SC, Hargreaves JNG, Huth NI, McLean G (2002) Development of a generic crop model template in the cropping system model
APSIM. Eur J Agron 18: 121–140
Downloaded from on June 14, 2017 - Published by www.plantphysiol.org
Copyright © 2004 American Society of Plant Biologists. All rights reserved.
911