Enzyme and Microbial Technology 30 (2002) 279 –283 www.elsevier.

com/locate/enzmictec

From green plants to industrial enzymes

Elizabeth E. Hood*
ProdiGene, 101 Gateway Blvd., Suite 100, College Station, TX 77845, USA

Abstract
Transgenic plants provide a viable technology for producing industrial proteins. Advantages of plant production systems include low cost
of goods, stable protein storage in seeds, ease and speed of scale-up and the possibility of direct addition of plant material to industrial
processes. Using transgenic plants for production of proteins includes several steps designed to result in high expression of foreign
proteins—from gene manipulation to breeding. Characteristics of four protein examples are presented —␤ glucuronidase, avidin, laccase
and trypsin. These proteins represent a range of molecular weights, activities, and localizations, demonstrating the versatility of the system.
The benefits of transgenic plant technology for industrial enzyme production include replacement of chemical processes that cause
environmental pollution. © 2002 Elsevier Science Inc. All rights reserved.

1. Introduction remain soluble. Fungi produce glycoproteins that can be

Proteins for industrial applications include those used as
purification and diagnostic tools as well as enzymes for
industrial processes that may be as diverse as food process-
ing and paper and pulp bleaching. Enzymes were initially
obtained as natural products from animal, plant and/or mi-
crobial tissues. Originally, the largest number of such en-
zymes were obtained from plant and animal sources, but as
microbial fermentation became more cost-effective, this
system provided the bulk of commercial enzymes from
natural sources. However, natural sources often have disad-
vantages as source material for enzyme production for a
variety of reasons including limitations in the amount of that
material, geographical availability of that material, and cost
of the material. Therefore, foreign protein production sys-
tems were developed as sources of important industrial
products. Xenogenic protein production has been accom-
plished in bacteria, fungi, cultured animal cells, transgenic
animals and of highest interest here, plants [1–3]. Bacteria
and fungi are relatively simple systems; however, these
microbial systems require a large capital outlay initially for
fermentation equipment. Bacteria may efficiently synthesize
and secrete proteins and enzymes that are not glycosylated.
In many cases, this is acceptable. However, if glycosylation
is required, the bacterial system is not appropriate. The
bacterial-sourced proteins may be easily purified if they
* Corresponding author. Tel.: ⫹1-979-690-8537; fax: ⫹1-979-690-
9527.
E-mail address: eehood@prodigene.com (E.E. Hood).
secreted and are relatively easy to purify. However, in some
cases, mis-folding of proteins in bacteria and hyper-glyco-
sylation of proteins in fungi may occur [4 –7], making
alternative production systems necessary.
Animal cell culture and transgenic animal production of
foreign proteins are currently being explored in academic
and industrial laboratories. The major advantage of these
systems is that the sugars at the glycosylation sites more
nearly resemble those of the native protein in animals.
These systems are expensive and increasing the size of the
herd of transgenic animals is quite slow. Therefore, trans-
genic animals and cells are most often only cost effective
for high value proteins such as pharmaceuticals in which
sialic acid is required.
2. The advantages of plant production systems
Transgenic plants provide a viable technology for pro-
ducing protein products [1,2,8 –10]. They have many ad-
vantages including low cost of production, stability of pro-
tein products in storage tissues such as seeds, ease and speed
of scale-up and the possibility of direct addition of plant
material to industrial processes. In addition, plant systems
have expressed proteins with integrity across a wide range
of conformations and molecular weights, including proteins
such as trypsin and a laccase isozyme (see below) that were
not successfully expressed in fermentation systems.
The production cost of enzymes from any source in-
cludes several factors such as raw materials, processing and

0141-0229/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.
PII: S 0 1 4 1 - 0 2 2 9 ( 0 1 ) 0 0 5 0 2 - 6
280 E.E. Hood / Enzyme and Microbial Technology 30 (2002) 279 –283
possibly purification. For an enzyme that requires purifica-
tion, the raw material can be a minor part of the overall cost
of the product—in some cases as little as 10% depending on
purity. If a plant extract and fungal fermentation broth offer
a protein at the same concentration, then no difference is
seen in the final cost of production. However, if an extract
is 10 fold more concentrated, the cost advantage for a
partially purified product is significantly better, because
processing costs are directly related to concentration of the
starting material. The advantage that plants have in this
regard is that seed can offer proteins at a much higher
concentration than fermentation broth in the initial extract
because of the expression levels that can be achieved. This
will not be true for all enzymes, of course, but particularly
for those enzymes that do not express or are poorly ex-
pressed in fermentation cultures. We have current examples
of proteins expressed at 1% of dry weight. Moreover, the
technology for plant expression is very young and the po-
tential for improvement is very large.
An additional advantage of a seed-based production sys-
tem is the ability to apply the product directly to industrial
processes, minimizing handling and enzyme manipulation
and preparation. Direct addition of transgenic plant material
to industrial processes, including those in the food industry,
is not done today, but is probable in the USA. When using
an edible seed product such as maize as the carrier of a
transgenic protein, the question of direct addition to a food
or food process can be addressed as a food additive. If
defined as a food additive or established as GRAS (gener-
ally recognized as safe), then the products to be added are
regulated but the proteins would not have to be purified
away from the genetic material producing those products.
Precedence in this regard has been set by approval of the
direct addition of genetically modified yeasts that can be
used in bread, beer, wine and cheese, among other products.
In other industries where the transgenic material will not be
consumed as food, direct addition of the transgenic corn
flour may be advantageous for cost as well as synergy with
the process. An example of this could be amylases for starch
break down. The enzymes would be produced in the same
corn grain used for the production of ethanol.
The ideal crop for molecular farming has several distinct
characteristics. One of the most important requirements of
industrial enzymes is that they are inexpensive. In this
regard, stable storage of the protein in production material is
Fig. 1. Representative vector used to transform a plant with a foreign gene, using the Agrobacterium tumefaciens system. Gene transfer to the plant occurs
directionally from the right border (RB) and ends at the left border (LB) although the LB is usually less precise than the RB. Because transfer is directional,
the selectable marker gene, in this case the “pat” gene for herbicide resistance, is placed downstream of the gene of interest.
Table 1
Yield of target protein per generational increase in seed yield
Generation Yield of target protein
@ 200⫻ @ 1000⫻
0 1g 1g
1 2000 g 1 kg
2 40 kg 1000 kg
3 8000 kg 106 kg
an advantage. An additional direct influence on cost is the
ability to achieve high expression of the protein in the
production material. Plants are particularly advantageous
here, as several crops have promoters, targeting sequences
and other molecular components available that allow high
expression to occur.
Another characteristic of an ideal crop is one that has a
large-scale production capability already established. This
is true for major crops such as corn, soybeans, canola and
alfalfa. Scale-up time for protein production from identified
transgenic lines is short and inexpensive, mostly involving
planting of increased acreage. For example, corn can in-
crease anywhere from 200 to 1000-fold per generation de-
pending on the quality of the hybrid material (Table 1). The
protein yield potential from such a system is obvious.
Finally, plant-produced products are safe to humans be-
cause of their lack of human pathogens.
3. The process of plant production of xenogenic
proteins
Plant production of proteins involves many steps. The
gene itself may require engineering so that the signals are
recognized by the plant transcription and translation ma-
chinery. The signals required include a promoter, a subcel-
lular targeting signal, and a terminator (Fig. 1). The vector
of choice depends on the transformation protocol being
employed. ProdiGene is currently using Agrobacterium tu-
mefaciens to transform maize, so binary vectors are used.
For biolistic transformation, E. coli vectors can be used.
For plant systems to compete with production by fer-
mentation of similar products, the expression levels must be
high enough, usually 0.01– 0.1% of tissue weight, to justify
their use from an economic standpoint. Thus, the need for
 E.E. Hood / Enzyme and Microbial Technology 30 (2002) 279 –283 281

Table 2
Examples of industrial proteins produced by ProdiGene in transgenic maize

Gene Transformation Copy Promoter Target T1 high Tn ear

method # seed bulk

Avidina (17 kDa) Biolistics 2 Constitutive CW 100 ng/mg 2 ␮g/mg (T8)

GUSb (68 kDa) Biolistics 1 Constitutive Cyto 20 ng/mg 200 ng/mg (T5)
Trypsin (24 kDa) Agrobacterium ? Seed preferred CW 300 ng/mg 50 ng/mg (T2)
Laccase (63 kDa) Agrobacterium ? Seed preferred CW 35 ng/mg 65 ng/mg (T4)
a
Reference 17.
b
Reference 18.

high expression drives technology development in this in-
dustry, where it takes many forms. New promoters are being
isolated from all plant systems being utilized today. These
promoters take advantage of tissue-type specificity as well
as preferred sinks in the plant or seed. Targeting experi-
ments designed to take advantage of different subcellular
compartments to accumulate protein are being conducted as
well. In many cases, genes are resynthesized to reflect
codon usage in the host species, as well as to minimize
sequences that may destabilize the RNA in its new host.
When combined, these parameters can have a major positive
influence on expression of the foreign gene.
The expression vector is introduced into Agrobacterium
by electroporation [11] and the resulting strain is used to
transfer genes to maize immature embryos [12]. Embryos
are cultured to recover transgenic events and regenerated to
recover plants [13]. Seed from To plants is produced in the
greenhouse, using pollen donors that are selected inbreds
from our genetics program. Analytical assays are developed
for each protein —preferably ELISAs or enzyme assays,
and western blots are performed on a subset of samples to
confirm protein integrity and concentration estimates from
activity assays. Protein purification and characterization are
performed on transgenic events selected for commercializa-
tion. These events are also analyzed at the molecular level
for integrity of the insert.
Because we employ HiII maize lines [13] for transfor-
mation, ProdiGene’s genetics group back-crosses our trans-
genic lines for several generations into inbred lines that
when combined, produce high yielding hybrids. Yield per
acre of bushels of seed and of protein are critical to attain
the cost targets required for industrial enzymes. Thus ex-
perimentation with multiple germplasms may reveal lines,
e.g. those with higher oil content, that are more conducive
to high expression and high yield per acre in the presence of
the transgene. ProdiGene has an active breeding program to
achieve these goals.
Grain is produced for protein purification and process
engineering experiments at various points during the pro-
cess of generating high yielding hybrids. Strict adherence to
USDA guidelines is followed for growing of transgenic
grain. ProdiGene’s confinement system is documented with
grower contracts and Standard Operating Procedures
(SOPs). Processing of the grain produces whole grain or
fractionated flour that can either be directly used in an
industrial process, or extracted and formulated into a final
product. Often costs can be recovered through sales of unused
grain fractions i.e. through by by-product credits [14].
4. Examples of industrial proteins from transgenic
plants
ProdiGene has expressed and produced four industrial
proteins in transgenic maize (Table 2). The transformation

Table 3
Physical characteristics of recombinant GUS and avidin from maize

Biochemical Native E. coli Maize-derived Egg white Maize-derived

properties GUS GUS avidin avidin

Molecular weight 68,000 Da 68,000 Da 17,700 16,800

Km 0.21 ⫾ 0.04 nM 0.19 ⫾ 0.05 nM N.A. N.A.
Binding stoichiometry N.A. N.A. Binds one biotin per Binds one biotin per
subunit subunit
Vmax 3.2 ⫻ 105 ⫾ 3.3 ⫻ 104 nmoles/hr 1.5 ⫻ 105 ⫾ 3.8 ⫻ 104 nmoles/hr N.A. N.A.
Isoelectric point 4.8–5.0 4.8–5.0 10 10
Ki N.T. N.T. 3.16 ␮M 3.34 ␮M
Heat stable Yes Yes ? ?
Antigenic similarity Identical Identical Identical Identical
Glycosylated No No Yes Yes
N-terminal sequence Native Identical except for initial methionine Native Identical

N.A. ⫽ not applicable.

N.T. ⫽ not tested.
282 E.E. Hood / Enzyme and Microbial Technology 30 (2002) 279 –283

Table 4 from less than the high T1 seed, to 20-fold greater than the
Comparison of physical characteristics of laccase from two sources single high T1 seed (Table 2). This is correlated with the
Property Fungal Maize-derived number of backcross generations in the material. More
laccase laccase backcrosses yield selections with higher expression.
Kinetics Similar Similar To the extent they have been tested, the recombinant
PI 5–7 5–7 proteins are functionally equivalent to the protein from
N-terminal sequence Native Identical native sources [8] (Tables 3 and 4). This is crucial if the
Glycosylated Yes Yes proteins are to be used for their activity. A slight difference
Molecular weight 66 and 55 kDa 63 and 59 kDa
in molecular weight is seen between avidin (Table 3), lac-
case (Table 4) and trypsin (Fig. 2B) derived from maize
versus their native sources. For avidin, this is due to differ-
method used to produce GUS and avidin events was biolis- ences in glycosylation, which is probably also the case for
tics, while Agrobacterium-mediated transformation was laccase. Trypsin differences in molecular weight are still
used for trypsin and laccase events. Each gene is present in under investigation.
one or very few copies per genome. The expression level in Laccase is a blue copper oxidase with applications in the
the highest T1 seed ranged from 20 ng per mg seed dry wood products and textile industries, depending on enzyme
weight for GUS up to 300 for trypsin, and is not correlated properties. We have expressed the gene for the laccase 1
with promoter or targeting sequence. Through breeding and isozyme from Trametes versicolor [15] at commercial lev-
selection, these levels are present in an ear bulk at values els in transgenic maize. Maize seeds with high expression

Fig. 2. Western blot of maize seed extracts containing either laccase (A) or bovine trypsin (B). Molecular weight markers are as indicated. Control seed extract
and control protein standards are as indicated. The fungal laccase was purified from a recombinant Aspergillus strain containing the laccase I gene from
Trametes versicolor. The trypsin and trypsinogen standards were purchased from Sigma Chemical Co. Protein gels were electroblotted onto PVDF
membranes, blocked with dry milk in PBST, incubated with dilutions of primary polyclonal antibodies and horseradish peroxidase conjugated secondary
antibodies. Detection was with chemiluminescent substrate. Primary antibodies were prepared in rabbits that were pre-screened for a lack of serum
cross-reactivity with corn seed proteins. LCG: a laccase expressing corn line in which a seed-preferred promoter is driving expression; TRF: a bovine
trypsinogen expressing corn line in which a seed-preferred promoter is driving expression.
 E.E. Hood / Enzyme and Microbial Technology 30 (2002) 279 –283
levels show two bands at approximately 63 and 59 kDa (Fig.
2A) with several smaller bands that react with the antibod-
ies. The partially purified enzyme from maize has identical
physical characteristics to the fungal-produced enzyme (Ta-
ble 4). We are currently back-crossing high-expressing lines
into parity-yielding germplasm. This particular isozyme is
not well-suited to the textiles industry as a bleaching agent
because of its activity profile. New commercial applications
may now be possible in the wood products industry because
of laccase enzyme availability from transgenic maize.
Protease production in transgenic plants is best accom-
plished through zymogen production, preferably in a seed-
specific manner. This concept is protected by U.S. Patent
#6,087,558. Using this method, ProdiGene has produced
transgenic lines in maize that express trypsin at high levels
in single T1 seed and bulk ears from the T2 generation
(Table 1). Even though seed extracts contain 2–3 bands as
detected on a western blot (Fig. 2B), N-terminal sequence
indicates that the bands all represent active trypsins, and not
the zymogen form. The enzyme is active when extracted
from seed, though it was translated as a zymogen (data not
shown), presumably because tryspinogen is self-activating
[16]. The genetic background of the maize inbreds in which
trypsin is expressed has an impact on our ability to estimate
its total amount, most likely due to the variation in level of
corn trypsin inhibitor in these lines.
5. Conclusion
Paradigms now exist for production of industrial proteins
in plants. We have reached a level of expression (0.02–
0.2% of dry weight) that allows cost effective use of these
proteins in industrial processes. Proteins for large-scale pro-
cesses, demanding large amounts of enzyme are logical
targets for plant production, because increasing supply is
fast and inexpensive. In these cases, plants may out-com-
pete fermentation investments.
Transgenic plant technology has been under recent scru-
tiny. Much of the discussion about transgenic plants has
centered around crops for food. The application of trans-
genic plants to industrial processes avoids their direct ad-
dition to the human food chain, and may offer benefits by
providing replacement technology for polluting industries.
An example of this would be the use of laccase as a glue in
the wood products industry or as a bleach for paper, i.e.
replacing formaldehyde or chlorite, respectively. As more
products from transgenic plant technology come to market,
especially including human pharmaceuticals and vaccines,
public perception of the benefit of such technology may be
improved.
283
References
[1] Hood E, Howard J. Protein products from transgenic plants. Agro-
Food-Industry Hi-Tech 3 1999;10:35– 6.
[2] Hood EE, Jilka JM. Plant-based production of xenogenic proteins.
Current Opinion in Biotechnology 1999;10:382– 6.
[3] Jilka JJ, Hood EE, Dose R, Howard J. The benefits of proteins
produced in transgenic plants. AgBiotechNet 1999;1(9):1– 4.
[4] Archer DB. Enzyme production by recombinant Aspergillus. In: Mu-
rooka Y, Amanaka T, editors. Recombinant microbes for industrial
and agricultural applications. New York: Marcel Dekker , 1994. p.
373–93.
[5] Georgiou G, Bowden GA. Inclusion body formation and the recovery
of aggregated recombinant proteins. In: Ho C, Prokop A, Bajpaj R,
editors. Recombinant DNA technology and applications. New York:
McGraw-Hill, 1991. p. 333–356.
[6] Harashima S. Heterologous protein production by yeast host-vector
systems. Bioproc Technol 1994;19:137–58.
[7] Marino MM. Expression of heterologous proteins in yeast. In: Ha CS,
Prokop A, Bajpaj R, editors. Recombinant DNA technology and
applications. New York: McGraw-Hill, 1991. p. 29 – 65.
[8] Hood EE, Kusnadi A, Nikolov Z, Howard JA. Molecular farming of
industrial proteins from transgenic maize. Chemicals via Higher Plant
Bioengineering 1999, p.127– 47.
[9] Van Rooijen GJH, Moloney M. Plant seed oil-bodies as carriers for
foreign proteins. Biotechnology 1995;13:72–7.
[10] Fischer R, Emans N. Molecular farming of pharmaceutical proteins.
Transgenic Research 2000;9(4 –5):279 –99.
[11] Cangelosi GA, Best EA, Martinetti G, Nester EW. Genetic analysis of
Agrobacterium. Methods in Enzymology 1991;204:384 –97.
[12] Ishida Y, Saito H, Ohta S, Hiei Y, Komari T, Kumashiro T. High
efficiency transformation of maize (Zea mays L.) mediated by
Agrobacterium tumefaciens. Plant Molecular Biology 1996;14:745–
50.
[13] Armstrong CL, Green CE, Phillips RL. Development and availability
of germplasm with high Type II culture formation response. Maize
Genet Coop Newsletter 1991;65:92–3.
[14] Kusnadi A, Nikolov Z, Howard J. Production of recombinant proteins
in transgenic plants: practical considerations. Biotech Bioeng 1997;
56:473– 84.
[15] Ong E, Brent W, Pollock R, Smith M. Cloning and sequence analysis
of two laccases complementary DNAs from the lignolytic basidio-
mycete Trametes versicolor. Gene 1997;196:113–9.
[16] Kunitz M, Northrop JH. Isolation from beef pancreas of crystalline
trypsinogen, trypsin, a trypsin inhibitor, and an inhibitor-trypsin com-
pound. J Gen Physiol 1936;19:991–1007.
[17] Hood EE, Witcher DR, Maddock S, Meyer T, Baszczynski C, Bailey
M, Flynn P, Register J, Marshall L, Bond D, Kulisek E, Kusnadi A,
Evangelista R, Nikolov Z, Wooge C, Mehigh RJ, Hernan R, Kappel
WK, Ritland D, Li CP, Howard JA. Commercial production of avidin
from transgenic maize: characterization of transformant, production,
processing, extraction and purification. Molecular Breeding 1997;3:
291–306.
[18] Witcher DR, Hood EE, Peterson D, Bailey M, Bond D, Kusnadi A,
Evangelista R, Nikolov Z, Wooge C, Mehigh R, Kappel W, Register
JC, Howard JA. Commercial production of ␤-glucuronidase (GUS):
A model system for the production of proteins in plants. Molecular
Breeding 1998;4:301–12.