|
Plant Physiol, January 2001, Vol. 125, pp. 168-171
Plants as Factories for Technical Materials1
Chris R.
Somerville* and
Dario
Bonetta
Carnegie Institution of Washington, 260 Panama Street, Stanford,
California 94305; and Department of Biological Sciences, Stanford
University, Stanford, California 94305
 |
INTRODUCTION |
Until the latter part of the 19th
century, humans were largely dependent upon contemporaneous biological
sources for the production of all organic materials. Plants and animals
provided the only sources of fibers, coatings, lubricants, solvents,
dyes, waxes, fillers, and insulation, fragrances, detergents, sizing,
leather, wood, paper, rubber, and many other types of materials. As
recently as 1930, 30% of industrial organic chemicals were derived
from plants (13). The discovery of extensive petroleum reserves and advances in chemistry and petroleum engineering resulted in a major
shift to reliance on fossil sources of organic feedstocks and the
development of materials such as inexpensive plastics, with properties
that could not be duplicated by abundantly available natural materials.
Nevertheless, many important materials are still derived from plants
and animals. Wood, cork, paper, and leather remain ubiquitous. Cotton,
ramie, hemp, flax, sisal, wool, and silk are also important sources of
fiber for many applications. Rubber from natural latex is still the
only material that can be used to produce tires that will reliably
withstand the forces associated with airplane landings. Linseed oil is
still used to make paint, although linoleum is no longer produced. Thus
for many applications, biological sources can still be used to produce materials on the scale necessary to meet the needs of populous industrialized nations.
The traditional strategy for using plants has been to modify, by
breeding and selection, a species that produces something useful so
that it is suited to our needs. Many of the important food species such
as maize or the many varieties of Brassica oleracea no
longer bear much resemblance to wild progenitors because of strong
selection for useful traits. Attempts have been made to improve jojoba
for production of wax esters and guayule for latex production and
species such as crambe, meadowfoam, Euphorbia lagascae, Lesquerella fendleri, and Cuphea sp. for
technically useful oils (5). However, these initiatives have met with
limited success because of an inability to develop lines with
acceptable production, quality, and agronomic properties. Except for
relatively small-scale production of jojoba and guayule, there does not
appear to have been a newly created field or plantation crop for
production of technical materials during the last century. Several
species that produced useful materials such as kapok, which was used
for applications such as waterproof fiber filling in life vests, have
declined because of the labor costs associated with harvesting the materials.
Interest in the possibility of using genetically engineered
plants as factories seems to have several motivations (19); in the
short term, it would be desirable to diversify crop production by
producing high value technical materials in crop plants and create
potentially large new markets for excess agricultural production. In
the longer term, the concept addresses a widely held social goal of
developing more sustainable and environmentally benign methods of
meeting our needs for materials that are currently produced by
chemical synthesis from declining petroleum or coal feedstocks. In
addition, it is possible to envision the production in plants of novel
biologically inspired materials, with properties not easily simulated
through chemical synthesis.
Lingering concern about the consequences of the 1973 oil embargo appear
to have been the motivation for the first article that clearly
described the concept of using genetic engineering to produce an
industrial product in plants. Even before the first paper describing
the production of a transgenic plant had appeared (24), Melvin Calvin
(1) explicitly outlined the basic steps by which Euphorbia
lathyrus could be engineered to produce industrial quantities of
sesquiterpenes. This paper and related work from Calvin and colleagues
established the concept of engineering plants as factories, but because
of economic considerations, the work on E. lathyrus was
discontinued. In retrospect it was unrealistic to expect that the first
applications of genetic engineering would be directed toward creation
of a new crop for a nonexistent market. Calvin's ideas remain
interesting, but until the price of petroleum increases substantially
there will not be sufficient economic incentive to attempt the
engineering of E. lathyrus. Perhaps the major lesson from
Calvin's work was the necessity of minimizing the threshold for the
introduction of a new crop. In practice this has focused attention on
using genetic engineering to make incremental changes in plants that
are already grown on a large scale.
| |
MODIFYING STARCH AND OIL |
To utilize plants as a source of organic molecules, the proposed
products must accumulate to levels that justify the costs associated
with growing and processing the plant material (6). For low
Mr water-soluble compounds of industrial
utility, the osmotic effects of high level solute accumulation place
practical limits on how much can be accumulated. In general,
microorganisms are the preferred vehicles for production of
water-soluble small molecules such as organic acids and amino acids
because they frequently secrete the compounds into the media. By
contrast, high Mr soluble compounds such as
proteins or water insoluble compounds such as starch, oil, terpenoids,
polyalkanoates, and cell wall polysaccharides are not significant
osmolytes and can, in principle, be produced at very high
concentrations in some types of plant cells.
Because starch and oil are already produced from plants in very large
amounts and have a diverse range of industrial applications, some of
the first applications of genetic engineering were directed toward
modifying starch structure and oil fatty acid composition for food and
non-food uses. There is a vast literature concerning chemical
modification of starch and the many non-food uses for the resulting
polymers (23). The use of starch as a plastic and as a component of
plastics began in the 1970s and a number of products of this type are
commercially produced. Although the ideal starch properties for polymer
applications do not seem to have been explicitly defined as yet, it is
apparent that factors such as the amylose/amylopectin ratio and the
size of granules has a significant influence on the properties of
plastic made from starch (25). Amylose ethers have mechanical
properties similar to those of high-density polyethylene. If amylose
was available in large quantities at prices similar to starch,
amylose-ethers could replace polyethylene and polystyrene in
applications where biodegradability was desirable (25). Because of
progress in cloning and understanding the genes involved in starch
biosynthesis, it is just a matter of time before a type of starch that
has been optimized for polymer applications is available from
transgenic plants (4, 17). Because of the magnitude of demand for
polymers, this will have a major effect on the economics of agriculture.
The idea of creating new oil crops by domestication of plants that
accumulate unusual fatty acids seems to have been most vigorously
pursued by a group of chemists at the U.S. Department of Agriculture
northern Regional Research Center in Peoria Illinois (5). Beginning in
the 1960s, they conducted an analysis of the fatty acid composition of
the seed oil of approximately 8,000 species. These studies showed that
higher plants collectively produce more than 500 kinds of fatty acids,
which differ by chain length and the presence of double bonds,
hydroxyls, epoxy groups, acetylenes, conjugated double bonds,
cyclopropanes and cyclopropenes, furan groups, and cyclopentenes, to
name a few (13, 22; Fig. 1). As thinking
about the uses of recombinant DNA methods spread in the plant biology
community in the early 1980s, the idea of using genetic engineering to
transfer the biosynthetic capability to produce novel fatty acids into
crop plants spread very quickly among the small community of people
interested in plant lipids (9). The basic assumption was, and is, that
if industrially useful fatty acids could be produced in plants in large
amounts and reasonably high purity, they would be useful starting
materials for the synthesis of a wide variety of materials, including
polymers, elastomers, plasticizers, paints and coatings, detergents,
and hydraulic and lubricating fluids (but see 6). By way of
illustration of how fatty acids are used as synthons, the steps
required to convert a plant lipid to nylon is shown in Figure
2.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 1.
Some examples of the chemical diversity of
plant seed fatty acids (20). 1, A component of stillingia oil from
Sapium sebiferum; 2, crepenynic acid from Crepis
foetida; 3, sterculynic acid from Sterculia alata; 4, gorlic acid, from the family Flacourtiaceae; 5, vernolic acid from
Vernonia galamensis; 6, ricinoleic acid from Ricinus
communis; and 7, a furan-containing fatty acid from
Exocarpus cupressiformis.
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2.
Production of nylon 6,10 from ricinoleic acid, the
principal component of castor bean oil. By continuously removing water
produced in the condensation reaction, polymerization progresses at a
high conversion rate due to Le Chatelier's principle. Details of the
synthesis are available at www.psrc.usm.edu/macrog/nysyn.htm.
|
|
At the time that the idea of producing industrially useful fatty acids
in transgenic plants emerged in the mid-1980s, no gene involved in
plant lipid metabolism had been cloned and no enzyme purified to
homogeneity. It was not until the early 1990s that genes for key
enzymes began to become available (14). As these genes have been used
to create plants with novel oils, a new problem has emerged; the
expression of a new biosynthetic capability frequently does not lead to
more than 5% to 20% of the total oil being composed of the desired
compound (13). Thus there is now heightened interest in understanding
what factors control the amount of particular fatty acids. In view of
the fact that some plants such as Ricinus communis (Castor)
produce oil that is >80% composed of a single fatty acid, there is a
biological precedent for achieving the quality of oil that would
be acceptable to the chemical industry. However, to date the only
example of a transgenic producing more than about 50% of a novel fatty
acid is the high lauric acid canola developed by Calgene (10).
In addition to the problem of increasing the purity of desirable fatty
acids, the field crops used for oil production accumulate oil in seeds
along with relatively large amounts of protein that is used for food
and feed. It would be useful to uncouple the production of oil and
protein so that the value of the oil was not dependent on the value of
the protein. This may be particularly important for production of
useful nonedible fatty acids, which may prove to be unacceptable as
post-processing carryover even at low levels in seed proteins used for
food and feed. More importantly, field crops such as potato, cassava,
and sugar beet, which produce large amounts of starch, but relatively
small amounts of protein, are much more productive than plants that
produce oil. Therefore, it would be useful to learn how to modify these
plants so that they accumulate oil rather than starch. Since these
plants are used for production of roots and tubers, but can be
propagated by seed, it should be possible to produce materials in the
roots and tubers of these species that would otherwise interfere with the germination of seeds. This is obviously an ambitious goal. However,
the recent identification of a mutant of Arabidopsis in which part of
the root becomes filled with oil suggests that it may eventually be
possible to accomplish this goal by altering the cellular identity of
root or tuber cells so that they take on the identity of oil producing
cell types (12).
| |
INTRODUCTION OF NOVEL CAPABILITIES |
Nature is the palette of genetic engineering. Thus the evolution
of ideas about what might be produced in plants is rooted in knowledge
of what is already produced by various organisms. One of the earliest
examples of producing a new material in plants was the production of
polyhydroxybutyrate (PHB) in Arabidopsis (16). At the time this work
began, a copolymer of PHB and polyhydroxyvalerate was commercially
produced by Imperial Chemical Industries via bacterial fermentation and
sold under the trade name BioPol. It is unfortunate that BioPol is
expensive to produce compared with plastics derived from petroleum and
is, therefore, not in widespread use. The idea behind the plant work
was that since starch and sugar could be produced in plants at costs
well below the cost of commodity plastics, it might be possible to
produce PHB and related polyhydroxyalkanoates at similarly low costs.
The first evidence for production of PHB in plants was obtained by
modifying two genes from the bacterium Alcaligenes eutrophus
so that the polypeptides were produced in the plant cytoplasm (16). The
resulting plants were tiny and produced only traces of PHB. However,
when the genes were restructured so that the polypeptides were targeted to the chloroplasts, the plants exhibited normal growth and development and plants with as much as 14% of dry weight of PHB were obtained (11). The plastic produced was brittle and not of suitable quality for
most applications. More recently, scientists at Monsanto have produced
transgenics that produce a copolymer of hydroxybutyrate and
hydroxyvalerate, raising the possibility that it will eventually be
possible to produce a useful material (18). However, because of the
possibility that very high levels of production in chloroplasts will
disrupt their function, it may be necessary to learn how to increase
the number of chloroplasts to obtain higher levels of production.
The variety of fibers produced in nature greatly exceeds the relatively
small number of synthetic polymers that are used for fibers. Before
synthetic polymers such as nylon and polyesters became widely
available, some efforts to produce man-made alternatives to silk
involved the production of silk-like fibers from seed proteins such as
zein from corn, arachin from peanuts, and casein from soybean (8). For
example, the production of a fiber based on zein (Vicara) reached an
output of 10 million pounds by the mid-1950s. Preceding the production
of Vicara, the Ford Motor Company had developed a soybean casein fiber
for use in automobile upholstery in the late 1930s. However, these
natural protein fibers suffered from a number of deficiencies and were
quickly replaced by synthetic fibers. More recently, silk-like proteins
have been produced in bacteria through genetic engineering, and
consequently, interest in protein based materials is resurgent. This
interest is attributable to the technical advantages that certain
protein polymers have over synthetic ones. Unlike the random
monomer arrangements, which typify synthetic polymers, the ordered
blocks of amino acids found in certain structural proteins impart a
greater degree of crystallinity to the polymer. Thus collagen and
keratin have more in common with the new highly oriented superfibers
such as Kevlar and thermotropic polyesters than with conventional plastics.
In the early 1980s it was discovered that the structural properties of
most fibrous proteins could by mimicked by polypeptides consisting of
short amino acid repeats and that structural blocks derived from
natural proteins could be used to design artificial genes (2). Since
this time, polymer compositions that mimic silk, laminin, elastin,
collagen, and keratin have been produced in E. coli. It has
also become apparent that the structural blocks from different natural
proteins can be combined to have properties that are not inherent in
the synthetic block or the natural protein. A potentially large number
of unique polymers can be produced using this principle and more than
20 patents are based on this idea. Despite high expression of some
polymers in bacteria, commercial production of protein polymers has yet
to be achieved. The expression of an elastin-like polymer in plants
(3), albeit at low levels, represents a first step toward the
development of a low cost production system that may bring this thread
of ideas to fruition. It has generally been difficult to produce
transgenic plants that accumulate more than a few percent of total
protein from an introduced gene (15). However, the successful
commercial production of avidin in maize (7) provides an example of
progress toward realizing the goal of producing industrial proteins in plants.
| |
CONCLUSIONS |
The genetic engineering of plants for production of industrial
materials is still in the conceptual stage. A limited number of
specific goals have been defined and there have been some promising preliminary results. The results to date suggest that as in any engineering project, broad theoretical knowledge of the entire system
is required to ensure progress. It seems likely that progress will be
dependent on concurrent initiatives to understand the basis of cellular
differentiation, morphogenesis, metabolism, and many other aspects of
growth and development that have little direct connection with
metabolic engineering. Progress toward understanding the function of
all genes in Arabidopsis and other angiosperms (21) will greatly
facilitate progress toward the development of a true engineering
discipline in which it is possible to envision a goal and predict the
outcomes of making specific changes in the genetic composition of a plant.
| |
FOOTNOTES |
1
This work was supported in part by the U.S.
Department of Energy (DOE-FG02-00ER20133).
*
Corresponding author; e-mail crs{at}andrew2.stanford.edu; fax
650-325-6857.
| |
LITERATURE CITED |
-
Calvin M
(1983)
Science
219: 24-26
[Abstract/Free Full Text]
-
Ferrari FA, Cappello J
(1997)
In
K McGrath, D Kaplan, eds, Protein Based Materials. Birkhäuser, Boston, pp 37-60
-
Guda C, Lee SB, Daniell H
(2000)
Plant Cell Rep
19: 257-262
[CrossRef]
-
Heyer AG, Lloyd JR, Kossmann J
(1998)
Curr Opin Biotechnol
10: 169-174
-
Hirsinger F
(1989)
In
G Röbbelen, RK Downey, A Ashri, eds, Oil Crops of the World. McGraw Hill, New York, pp 518-533
-
Hitz B
(1999)
Curr Opin Plant Biol
2: 135-138
[Medline]
-
Hood EE, Witcher DR, Maddock S, Meyer T, Baszczynski C, Bailey M, Flynn P, Register J, Marshall L, Bond D, Kulisek E
(1997)
Mol Breed
3: 291-306
-
Hudson SM
(1997)
In
K McGrath, D Kaplan, eds, Protein Based Materials. Birkhäuser, Boston, pp 313-337
-
Knauf V
(1987)
Trends Biotechnol
5: 40-47
-
Murphy DJ
(1999)
Curr Opin Biotechnol
10: 175-180
[CrossRef][Web of Science][Medline]
-
Nawrath C, Poirier Y, Somerville CR
(1994)
Proc Natl Acad Sci USA
91: 12760-12764
[Abstract/Free Full Text]
-
Ogas J, Cheng JC, Sung R, Somerville CR
(1997)
Science
277: 91-94
[Abstract/Free Full Text]
-
Ohlrogge J
(1999)
Curr Opin Plant Biol
2: 121-122
[Medline]
-
Ohlrogge J, Browse J
(1995)
Plant Cell
7: 957-970
[CrossRef][Web of Science][Medline]
-
Owen RL, Pen J
(1996)
Transgenic Plants: A Production System for Industrial and Pharmaceutical Proteins Wiley, New York
-
Poirier YP, Dennis DE, Klomparens K, Somerville CR
(1992)
Science
256: 520-523
[Abstract/Free Full Text]
-
Riesmeier J, Kossmann J, Trethewey R, Heyer A, Landschutze V, Willmitzer L
(1998)
Polymer Degrad Stabil
59: 383-386
[CrossRef]
-
Slater S, Mitsky TA, Houmiel KL, Hao M, Reiser SE, Taylor NB, Tran M, Valentin HE, Rodriguez DJ, Stone DA, Padgette SR, Kishore G, Gruys KJ
(1999)
Nature Biotechnol
17: 1011-1016
[CrossRef][Web of Science][Medline]
-
Somerville CR
(1993)
Phil Trans Roy Soc Lond B
342: 251-257
-
Somerville CR, Browse J
(1991)
Science
252: 80-87
[Abstract/Free Full Text]
-
Somerville CR, Somerville SC
(1999)
Science
285: 380-383
[Abstract/Free Full Text]
-
van de Loo F, Fox B, Somerville CR
(1993)
In
T Moore, ed, Plant Lipids. CRC Press, Boca Raton, FL, pp 91-126
-
van Soest JJG, Vliegenthart JFG
(1997)
Trends Biotechnol
15: 208-213
[Medline]
-
Zambryski P, Joos H, Genetello C, Leemans J, van Montagu M, Schell J
(1983)
EMBO J
2: 2143-2150
[Web of Science][Medline]
-
Zhao WY, Kloczkowski A, Mark JE, Erman B
(1998)
Chem Material
10: 804-811
[CrossRef]
© 2001 American Society of Plant Physiologists
This article has been cited by other articles:
|
|
|
|
 
Y. Madoka, K.-I. Tomizawa, J. Mizoi, I. Nishida, Y. Nagano, and Y. Sasaki
Chloroplast Transformation with Modified accD Operon Increases Acetyl-CoA Carboxylase and Causes Extension of Leaf Longevity and Increase in Seed Yield in Tobacco
Plant Cell Physiol.,
December 15, 2002;
43(12):
1518 - 1525.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Hong, N. Datla, D. W. Reed, P. S. Covello, S. L. MacKenzie, and X. Qiu
High-Level Production of gamma -Linolenic Acid in Brassica juncea Using a Delta 6 Desaturase from Pythium irregulare
Plant Physiology,
May 1, 2002;
129(1):
354 - 362.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
TOP
INTRODUCTION
MODIFYING STARCH AND OIL