LIMITATIONS OF PENTOSE SUGAR CONVERSION IN
RECOMBINANT ZYMOMONAS MOBILIS AND METHODS TO
ADDRESS THESE LIMITATIONS
B.A., University of Colorado, Boulder, 1992
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
This thesis for the Master of Arts
Nathan Amis Danielson
has been approved
Linda K. Dixon
Danielson, Nathan Amis (M.A., Biology)
Limitations of Pentose Sugar Conversion in Recombinant Zymomonas Mobilis and
Methods to Address These Limitations
Thesis Directed by Professor Linda Dixon
The development of a recombinant Zymomonas mobilis that is able to produce
ethanol from the pentose sugars xylose and arabinose has greatly enhanced this
organisms capabilities as an ethanologen. However, pentose utilization in this
organism is much slower than that of glucose. This led the researchers to examine
whether it was sugar transport or the activity of the recombinant enzymes that was
the cause of sub-optimal sugar conversion. It was observed that arabinose
utilization was most influenced by poor transport, and that xylose consumption was
limited by xylose specific enzyme activity. Further efforts were made in an attempt
to find suitable enzymes that would be able to function well in Zymomonas mobilis
as this organism produces ethanol. This led to the identification of two enzymes
from Lactobacillus brevis, xylose isomerase and xylulokinase, which had
characteristics that were ideal for the assimilation of pentose sugars into glycolysis
in Z. mobilis.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
I would like to express my deep appreciation to my thesis co-advisors Drs. Min
Zhang and Linda Dixon. I am also grateful to Drs. Audesirk and Levy for their
assistance as thesis committee members. In addition I would like to thank my wife
Erin, who was always very supportive. Thank you all for your time and guidance.
Objectives of Research..............................18
2. MATERIAL AND METHODS...................................24
Preparation of Cell Free Extracts...................24
In Vitro Sugar Conversion by Cell Free Extracts.....26
3. RESULTS AND DISCUSSION.......................................28
Cell Free System....................................28
Xylose lsomerase Screening..........................42
Xylitol Production in Recombinant Zymomonas.........50
5. FUTURE WORK..................................................60
1.1 The Enter-Duodoroff Pathway and the pentose phosphoate pathway in
recombinant Zymomonas Mobilis.........................................13
1.2 In vivo sugar utilization by Z. mobilis 206C/pZB301 grown on pure
glucose:xylose:arabinose (30:30:20 g/1) at pH = 5.5, T = 31.5C. 10 mg/ml
total protein concentration...........................................17
3.1 In vitro sugar conversion by Z. mobilis 206C/pZB301 CFE at 30 C.
10 mg/ml total protein concentration. All sugars were initially 20
grams/liter, but there is some conversion before the 0 minute sample can be
3.2 In vitro arabinose conversion by Z. mobilis 206C/pZB301 at different
concentrations of ethanol at 30 C and 10 mg/ml protein concentration.32
3.3 Xylose conversion by Z. mobilis 206C/pZB301 CFE at different
concentrations of ethanol........................................33
3.4 In vitro conversion of glucose by Z. mobilis CFE at 30 C at different levels
of ethanol (v/v). Note that protein loading in this figure is lmg/ml, 1:10 of
normal. This low protein loading was done to facilitate measurement of
3.5 Effect of ethanol on E. coli xylose isomerase activity. Percent activity is
determined by assigning 100% activity to the 0% ethanol sample....37
3.6 Effect of xylitol on E. coli xylose isomerase activity. Percent activity is
determined by assigning 100% activity to the 0% xylitol sample....38
3.7 Effect of pH on E. coli xylose isomerase activity. Percent activity is
determined by assigning 100% activity to the 0% pH 7.0 Sample.....38
3.8 Effect of ethanol on E. coli xylulokinase. Percent activity determined by
assigning 100% activity to the 0% ethanol sample......................39
3.9 Effect of pH on E. coli xylulokinase. Percent activity determined by
assigning 100% activity to the pH 7.0 sample...........................40
3.10 Activity of xylose isomerase from different organisms at different pHs.
Percent activity was determined by assigning 100% activity to the pH 7.0
3.11 Comparisons of xylose isomerase activity from L. brevis and E. coli in
different concentrations of ethanol....................................46
3.12 Activity of xylose isomerases at different concentration of xylitol. Note that
xylitol concentration in the L. brevis experiments are ten times greater than
in the E. coli experiments......................................46
3.13 Activity of L. brevis xylulokinase at different concentrations of ethanol.
Percent activity is determine by assigning 100% activity to the 0% ethanol
3.14 Activity of xylulokinases from E. coli and L. brevis at different pH.48
3.15 Proposed pathway for xylitol formation in Z. mobilis.................51
3.16 Xylitol production by native Z. mobilis 39676 cell free extract with xylose
and xylulose at 10 gram/liter. Genencor glucose isomerase was added to an
approximate final concentration of 10 units......................52
3.17 Replicative vector pZB1861-GFOR-TC. This vector was designed
incorporating a Zymomonas origin of replication..................55
3.18 Plasmid pZB1861-GFOR-Tc(B). This plasmid contains a Z. mobilis origin
of replication. It also contains a modified glucose-fructose oxidoreductase
gene that is interrupted by the gene for tetracycline resistance. The GFOR
gene has further been altered by the deletion of 300 base pairs resulting in
almost equal sized gene portions on either side of the tetracycline gene
3.19 Suicidal integration vector pYC1865-GFOR-Tc. This vector was designed
lacking a Zymomonas origin of replication.....................57
3.20 Plasmid pYC1865-GFOR-Tc (B). This plasmid does not include a Z.
mobilis origin of replication. It contains a modified glucose-fructose
oxidoreductase gene that is interrupted by the gene for tetracycline
resistance. The GFOR gene has further been altered by the deletion of 300
bp resulting in almost equal sized gene portions on either side of the
tetracycline gene interruption................................58
1.1 Levels of cellulose, hemicellulose, lignin in biomass...................5
1.2 Reported characteristics for xylose isomerase from different species. Metal
cofactors are the cofactors necessary for enzymatic activity. Temperature
and pH are for optimum activity........................................22
"Nature took half a billion years to create the world's oil, but observers
agree that humankind will consume it all in a two-century binge of
profligate energy use."
Science, August 21st, 1998
The need for an environmentally friendly, domestic renewable energy source
has been recognized for some time. In particular there is the need for renewable
transportation fuel, since currently 98% of the fuel consumed in this sector is
petroleum based. There are many reasons to develop such a resource. The three
most obvious reasons are: 1.) Petroleum reserves are limited; 2.) The potential for
national economic instability that dependence on foreign oil creates; 3.) The adverse
effects of petroleum use both on human and global health. Discovery and
development of such a resource would not only address these three concerns, but
would also create a new world fuel paradigm in which any nation that could produce
biomass would be energy self-sufficient.
Experts disagree regarding when global petroleum production will peak.
Actual reserves are particularly difficult to calculate, since it is in many of the oil
producing nations best interest to inflate reserve estimates. This over-estimation
allows these nations to obtain loans more easily, as well as to have increased global
influence. Although there is disagreement among experts exactly when oil
production will begin to decrease, most believe that it will be between 2005 and
2040 at current usage levels (1).
Those most knowledgeable regarding petroleum reserves predict that nations
will begin to experience oil shortages in between 5-45 years if usage remains
constant. However, it is extremely unlikely that petroleum consumption will remain
constant. It is predicted that an increase in urbanized populations in China and India
alone will increase these countries energy consumption by 45 percent. This figure
does not take into account any increase in industrial usage. It is further estimated
that if China and India become as industrialized as South Korea, their combined
energy demand will equal 119 million barrels of oil a day (2). This is twice the
current global demand for petroleum. This increased consumption will surely
deplete reserves more quickly that anticipated. This leads to the conclusion that
dramatic declines in petroleum will be a significant problem early in the next
The United States is increasingly dependent on foreign oil. The danger of
such dependence was dramatically demonstrated in 1973 during the OPEC oil
embargo. This event had a major impact on the U.S. economy, resulting in a 10%
reduction in gross national product (3). In flagrant disregard to the risk of
dependence on foreign oil demonstrated by this crisis, the United States now
imports more oil today than in 1973 (4). Although not as dramatic as the economic
distress caused by the 1973 oil embargo, all sharp increases in the price of oil have
led to economic recession (2). Since oil is a commodity, increased demand by
developing countries will lead to increased cost. Decreased supplies of petroleum
will exacerbate this increase in cost. Since the outcome of higher oil prices has
historically had a negative impact on the U.S. economy, it is imperative that the
nation develops a strategy to deal with the end of cheap oil. It is probable that
national economies that will remain strong in this century will be those that are able
to transition to energy sources other than conventional petroleum.
In addition to the threat posed to the economy through dependence on
foreign oil, there is the very real trade deficit that this dependence produces. It is
estimated that petroleum imports account for more than 50 percent of the U.S. trade
deficit, and that this percentage will increase to between 60 and 70 percent in the
next 10 years (5). This departure of dollars from the U.S. economy represents a loss
of 3.5 million American jobs (3). These jobs could be reclaimed if the nation began
to focus on producing fuel from domestic biomass.
The combustion of oil-based transportation fuels has a significant effect on
the health of the global environment. Although there is still controversy
surrounding the correlation between the combustion of fossil fuels and global
climate change, more and more data continue to be generated that support this
hypothesis (6). Furthermore, prudence would dictate that in dealing with an issue
that could potentially impact the entire planet, great care should be taken. Ethanol
from biomass allows for a sustainable global carbon balance, since the carbon
dioxide produced by combustion is in turn fixed to produce more plant matter.
If one is unconvinced by the research claiming that fossil fuels are damaging
to the environment, there is still the human health impact of conventional fuels. The
U.S. Environmental Protection Agency ranks gasoline as the number one source of
toxic emission (6). Ethanol based fuels have significantly reduced pollutants
compared to traditional fossil fuels. Recently a study of 42 automobiles, 21 fueled
with traditional gasoline, and 21 with E85 (85% ethanol: 15% gasoline) was
conducted (7). In this study it was found that the E85 fleet produced 21% less
hydrocarbon, 28.5% less nitric oxide and 18% less carbon monoxide emissions than
the gasoline fleet. Furthermore, this study determined that E85 vehicles produce
25% less ozone than do traditional automobiles. These data show that impressive
reductions in air pollution can be attained using ethanol-gasoline mixtures; with this
reduction in pollution will come a decrease in pollution related health issues.
Progress is being made towards an ethanol-based fuel future. Automobiles
that can use E85 are now available at no extra charge from both Ford and General
Motors. A major drawback that now needs to be addressed is the cost of ethanol-
based fuel. Ethanol currently costs between $1.10 to $1.25 per gallon to produce,
compared to $0.60 per gallon for gasoline. A significant portion of the cost to
produce ethanol is the feedstock used. Currently most fuel ethanol in the United
States is produced from com. Since com has significant value as either food or
feed, it adds 84% to the price of ethanol (Mark Ruth, National Renewable Energy
Laboratory (NREL), personal communication). It is currently uneconomical to
produce fuel ethanol from costly commodity feedstocks such as grains.
The cost of producing ethanol from grain has led researchers to begin to focus on
using biomass to produce ethanol.
Biomass, particularly agricultural residues, represents a very good source for
renewable energy. Biomass is made up of simple sugars (Table 1) linked together
to form polysaccharides.
Component Percent Dry Weight
Table 1.1 Typical levels of cellulose, hemicellulose, and lignin in biomass (Ref. 8)
Cellulose is made primarily from glucose, a hexose. Hemicellulose is made up
primarily of xylose and arabinose, both pentose sugars. Xylose is the major
component of hemicellulose, but ratios of xylose to arabinose vary in different
plants. Lignin is a poly-phenolic compound that has value as a by-product of
fermentation, but is not converted to ethanol. A good example of biomass source for
the production of ethanol is comstover, the cobs, husks, and stocks from com
production that are currently unused. It is estimated that the conversion of the
sugars in comstover to ethanol would produce over 10 billion gallons of ethanol
(Mike Himmel National Renewable Energy Laboratory, personal communication).
There are two difficulties in converting biomass sugars to ethanol. The first
problem is the breakdown of the cellulose and hemicellulose to monosaccharides.
Efforts that fall beyond the scope of this thesis are currently underway at NREL to
address this problem. The second hurdle facing researchers is the fact that while it
is relatively easy to convert six-carbon sugars to ethanol, no common ethanologenic
organism has the ability to produce ethanol naturally from pentose sugars (9). Since
pentose sugars constitute a significant portion of total carbohydrates in plant
material, it is essential that they be converted to ethanol. The rapid and complete
conversion of the great majority of sugars in biomass to ethanol is key to
inexpensive renewable fuel ethanol.
Currently most research is focused on three biological platforms for
development as ethanologens. These are Saccharomyces cerevisiae, Escherichia
coli and Zymomonas mobilis (9). There are, of course, other organisms that are
being studied for feasibility as ethanologens. Genes for pyruvate decarboxylase and
alcohol dehydrogenase from Zymomonas mobilis have recently been cloned into the
cyanobacterium Synechococcus strain PCC 7942. This produced an organism that is
capable of fixing atmospheric carbon dioxide to produce ethanol (10). This
organism was only able to produce trace amounts of ethanol, but this demonstrates
the exciting prospects for the future of biological ethanol production. Still most
research into biological ethanol production is focused on the three organisms
Two distinct approaches are being taken to produce an efficient recombinant
ethanologen. Both Zymomonas and Saccharomyces are homofermentative; ethanol
is the major product of fermentation. Unfortunately, these organisms lack the
metabolic pathways that allow for pentose metabolism. Research on these
organisms has focused on the introduction of pentose sugar metabolism pathways.
E. coli naturally possesses pathways to metabolize a broad range of sugars including
xylose and arabinose, but this bacterium is heterofermentative, yielding a range of
organic chemicals as products of fermentation (11). Genetic engineering in E. coli
has concentrated on the modification of metabolic pathways to increase ethanol
production. Major strides have been made in all three organisms; however, much
work still needs to be done before commercial ethanol production becomes a reality.
Saccharomyces cerevisiae has a very long history as an ethanologen, and is
currently used to produce ethanol from traditional feedstocks, e.g. com, where it
produces near theoretical yields from hexose sugars (12). While Saccharomyces
cerevisiae is unable to metabolize xylose, it can metabolize xylulose, an isomer of
xylose. This fact has been exploited by researchers who have introduced the gene
for xylose isomerase into yeast, which produces xylulose from xylose. Xylulose is
then able to enter into endogenous metabolic pathways and be converted to ethanol
(13). A problem with this approach is that it appears that the xylose isomerase from
bacterial sources used in these experiments does not fold correctly when expressed
in yeast (14). This has led to a very low efficiency of xylose utilization.
A second strategy that has been explored to create a xylose fermenting
Saccharomyces cerevisiae is to introduce genes for xylose metabolism from Pichia
stipitis. The two enzymes that start xylose metabolism in Pichia are xylose
reductase, which produces xylitol from xylose, and xylitol dehydrogenase that
converts xylitol into xylulose. Xylulose produced this way could then move into the
endogenous xylulose metabolic pathways in S. cerevisiae. This strategy has
produced disappointing ethanol yields. It has been suggested that this is due to either
a redox imbalance, or due to improper ratios of xylose reductase and xylitol
dehydrogenase; however, experiments to address both of these concerns have not
significantly increased ethanol production (15).
Nancy Ho and coworkers at Purdue University have been successful in
generating a xylose fermenting S. cerevisiae by using a third strategy. These
researchers have introduced xylose genes from P. stipidis as well as extra copies of
the S. cerevisiae xylulokinase gene on plasmids. This has resulted in strain 1400, an
ethanol tolerant S. cerevisiae strain that is able to ferment both glucose and xylose
well under laboratory conditions (16). Although this strain shows promise, there
have been no reports of feasibility studies done in an industrial setting, which may
in part be due to plasmid instability.
The ethanol production industry has a significant interest in the development
of recombinant yeast. There is a great deal known about yeast in a commercial
setting and ethanol producers are comfortable with this organism. Unfortunately,
the development of yeast that is able to ferment xylose has been slow and difficult,
and there does not appear to be a good strain to accomplish this even today.
Additionally, there are no reports of a S. cerevisiae that is able to ferment arabinose.
It appears that it may be some time before all metabolic problems regarding pentose
metabolism are solved for yeast.
Escherichia coli has shown great promise as an ethanologen. This organism
is able to metabolize all of the major sugar components of biomass. Additionally E.
coli is arguably the most studied organism on the planet, so there is a plethora of
information regarding genetic manipulation of E. coli. These qualities prompted
Lonnie Ingram and coworkers at the University of Florida to develop E. coli as an
ethanologen. This group started by engineering in the Zymomonas genes for alcohol
dehydrogenase II and pyruvate decarboxlyase under a strong E. coli promoter. This
resulted in an E. coli that produced ethanol as the principal fermentation product
using glucose as a feedstock (17). Next the recombinant organism was tested on
sugars present in cellulose and hemicellulose. In this research, it was determined
that recombinant E. coli was able to ferment to glucose, xylose, arabinose, galactose
and mannose to ethanol with high efficiency (18). This organism was next
improved by the integration of Zymomonas genes for pyruvate decarboxylase and
alcohol dehydrogenase into the chromosome, thus increasing the stability of the
recombinant strain. The stability was further increased by a mutation of the recA
gene to block homologous recombination (19).
Further studies were conducted to increase ethanol tolerance in E. coli. To
accomplish this, the Zymomonas stress proteins, groES and groEL, were expressed
in E. coli. These proteins are produced by Zymomonas in response to high levels of
ethanol. The expression of these genes in E. coli unfortunately did not appear to
have any effect on increasing either ethanol or temperature tolerance (20). This led
the researchers to conclude that the mechanism for ethanol tolerance in Zymomonas
must require other factors in addition to the groES and grEL proteins. Work is
currently underway to better understand the exact expression of genes in E. coli
during stress (Tyrrell Conway Ohio State University, personal communication).
This work should yield insight into both the types of genes expressed during stress
as well as the ratios at which these genes are expressed.
Escherichia coli continues to be improved as an ethanologen. In 1997,
Ingram and coworkers introduced the casAB gene from Klebsiella oxytoca. This
gene codes for cellobiase, an enzyme that is able to break down cellobiose, a
product of the degradation of cellulose, into two molecules of glucose. Expression
of this gene allows E. coli not only to act as an ethanologen, but also to play a role
in the saccharification of biomass. This organism, E. coli KOI 1 was able to ferment
mixed-waste office paper to ethanol (21). Another series of improvements to E. coli
has been the isolation of ethanol-tolerant mutants. Selection for mutants was carried
out by screening for both ethanol tolerance and ethanol production. This yielded a
mutant strain, LY01 that was able not only to produce high levels of ethanol, but
also had a greater tolerance to ethanol than either native Zymomonas or
Saccharomyces (22). All of these improvements have culminated in E. coli being
chosen to be the ethanologen used by BC international in their waste-to-ethanol
plant (23). This represents a huge step towards the realization of fuel ethanol
produced from biomass.
Although E. coli is progressing as an ethanologen, it is plagued by negative
public perception. The perception that E. coli always behaves as a pathogen, though
unfounded, has and will continue to significantly hinder its wide-scale use by
industry, particularly when ethanol production is closely associated with commodity
food items such as com. This fact, in addition to some very good ethanol-producing
attributes, led researchers at NREL to choose Zymomonas mobilis for development
as an ethanologen.
Zymomonas mobilis is a Gram-negative bacterium that is an obligate
ethanologen. It uses a single metabolic pathway to produce cellular energy, the
product of which is ethanol (24). Zymomonas has a long history as a fermentative
organism; it is used throughout tropical regions to produce alcoholic beverages, for
example pulque in Mexico. The organism uses the Enter-Doudoroff pathway
(shown in Figure 1.1) to produce ATP. This pathway yields only one ATP for each
molecule of glucose; this mandates the very rapid consumption and conversion of
glucose in order to maintain constant physiological energy levels (25). It is due to
this low energy yield that Zymomonas is able to produce ethanol from glucose so
efficiently, from three to five times more quickly than yeast is able to. In addition,
Zymomonas is able to uncouple growth from metabolism, so it continues to produce
ethanol without increasing cell mass. This leads to a higher overall yield of ethanol
since less biomass is being formed (26). Zymomonas can also tolerate high levels of
sugars, up to 400 g/1, and ethanol concentrations as high as 12% (27). Moreover,
Zymomonas mobilis has GRAS status; it is Generally Recognized As Safe. GRAS
status is important to industry since it indicates that Zymomonas is not pathogenic
Pentose Metabolism Pathways
D-j lylose L-Arabinose
v Acetaldehyde + CO/
Figure 1.1 The Entner-Doudoroff pathway and the pentose phosphate pathway in
recombinant Zymomonas mobilis.
under normal circumstances. All of these factors make Zymomonas a very good
candidate for development as an ethanologen.
A major drawback in the use of native Zymomonas as an ethanologen is its
limited substrate range; this organism can only metabolize fructose, glucose and
sucrose. Min Zhang and coworkers at NREL have addressed this limitation (Fig
1.1). This group first introduced the pentose phosphate and the xylose assimilation
pathways into Zymomonas. To accomplish this, four genes, xylose isomerase
(xylA), xylulokinase (xylB), transketolase (tktA) and transaldolase (tal), were cloned
from E. coli. The xylA gene and the xylB gene were placed under the control of the
glyceraldehyde-3-phosphate promoter (Pgap), and the tal tktA gene placed under the
control of the enolase promoter. Finally all genes and promoters were ligated into
the plasmid. The resulting strain, CP4 (pZB5) is capable of growth on xylose (25
g/I) as the only carbon source, and was able to produce ethanol at 86% of theoretical
This team next developed a strain that is capable of growth on L-arabinose.
This was accomplished by the introduction of the arabinose assimilation genes L-
arabinose isomerase (araA), L-ribulokinase (araB) and L-ribulose-5-phosphate-4-
epimerase (araD), in conjunction with the pentose phosphate pathway genes
described above. The resultant strain, Zymomonas mobilis 39676 (pZ206) is
capable of growth using arabinose (25 g/1) as the only carbon source and producing
98% theoretical yield of ethanol based on arabinose consumed. Total yield of 86%
has since been accomplished (29).
Zymomonas mobilis has been further improved by the introduction of genes
for both arabinose and xylose utilization. This strain, Zymomonas mobilis 39676
(pZB301), is able to ferment a mixture of 3% glucose, 3% xylose and 2% arabinose
to ethanol at 87% of theoretical yield (30). Work is currently underway to increase
the genetic stability of recombinant Zymomonas through the integration of the
pentose phosphate genes into the genome of the organism (Min Zhang, National
Renewable Energy Laboratory, personal communication).
In addition to the advancements made at NREL toward the development of
Zymomonas, other researchers have also increased substrate range in this organism.
Mannose assimilation has been engineered into Zymomonas by the introduction of
the phosphomannose isomerase gene from E. coli. This recombinant organism is
able to use D-mannose as its sole carbon source, but no data on ethanol production
were provided in this report (31).
Two groups have tried to increase the substrate range of Zymomonas to
include galactose (32, 33). These attempts have enjoyed only marginal success;
most of the galactose in the media was converted to galactonic acid, with only trace
amounts being converted to ethanol. There may be a pathway that, if removed,
would allow for the efficient conversion of galactose to ethanol; however, no
research into this area has been reported.
Although the recombinant Zymomonas strain 39676 (pZB301) is able to
convert both arabinose and xylose to ethanol, the process is not completed as
quickly as the conversion of glucose. Additionally, lagging, or the incomplete
utilization of the sugars, is observed (Fig. 1.2). This phenomenon was thought to be
due to either a lack of pentose sugar transport, sub-optimal recombinant enzyme
activity or a combination of both. It is known that Zymomonas has a single
facilitated diffusion system to transport glucose; there is no other sugar transport
mechanism known for this organism (34). If this transport cannot recognize pentose
sugars, then this may be the limiting step in pentose conversion. Dr. Tyrrell
Conw'ay at Ohio State University is currently investigating sugar transport in
Figure 1.2 In vivo sugar utilization by Z. mobilis 206C/pZB301 grown on pure
glucose:xylose:arabinose (30:30:20 g/1) at pH = 5.5, T = 31.5C. 10 mg/ml total
NREL has also investigated whether recombinant enzymes within Z. mobilis
strains are performing. It proved difficult to determine the in vivo activity of the
recombinant enzymes in Zymomonas. This is reasonable since it is problematic to
separate which effects on sugar metabolism are due to enzyme activity and which
are due to transport issues. This led to the development of a cell free system to
understand enzyme activity better without the complication of sugar transport
(35,36). Once developed, this in vitro system proved to be of key importance for the
study of pentose sugar metabolism in Zymomonas.
Objectives of Research
The primary objectives of this research were to understand the mechanisms
by which pentose sugar metabolism inhibition occurs and to design rational methods
to address these problems. If the nature of sub-optimal pentose utilization could be
determined, then improvements could be made using molecular biological tools in a
systematic fashion. These tools include the modification or destruction of
competing pathways, using directed evolution to modify an enzyme or a suite of
enzymes and screening of enzymes obtained from other organisms that may have
characteristics that would be ideal for ethanol production in Zymomonas. By using
any one or a combination of these methods, it should be possible to create a better
Modification of enzymatic pathways in order to favor the production of a
single product or to abolish the production of an unwanted product is a commonly
used practice. Lonnie Ingram was able to create a strain of E. coli that is nearly
homofermentive, with ethanol being the major product (17). A second example of
using this strategy comes from Zymomonas itself. Wecker and Zall have reported
the production of acetaldehyde in Zymomonas by selecting for mutants with
decreased activity of the alcohol dehydrogenase gene (37). Using this system, this
group was able to select a Zymomonas that was able to produce 4.08 g/liter or 40%
of theoretical yield when grown on 4% (w/w) glucose. This would indicate that
Zymomonas is amenable to diversions in its metabolic pathway.
Xylitol is a by-product formed during the fermentation of xylose by
recombinant Zymomonas. Feldmann and coworkers have attributed the production
of xylitol to a novel xylitol NADPH-dependent aldose reductase (38). Oblation of
this enzyme may have several beneficial effects on ethanol production in
Zymomonas. First, less carbon being converted to xylitol would allow more ethanol
to be formed. Second, xylitol is a potent inhibitor of xylose isomerase (39);
therefore, xylitol production by the recombinant strains may act to inhibit further
xylose conversion to xylulose.
There is the possibility that another pathway exists for the production of
xylitol. Zachariou and Scopes have described an enzyme that appears to be unique
to Zymomonas mobilis: glucose-fructose oxidoreductase (40). They determined
that this enzyme functions to produce gluconolactone and sorbitol from glucose and
fructose, respectively. They further characterized the ability of this enzyme to
convert other sugars, as long as one sugar is a ketose and one is an aldose. This
enzyme shows low activity catalyzing xylose (8%) and xylulose (7%) compared to
glucose and fructose, but it does have activity. It therefore is likely that this enzyme
may be in part responsible for the production of xylitol in recombinant Zymomonas,
and that disruption of its function may increase pentose sugar conversion.
Another recently developed molecular biological tool is the rapid in vitro
evolution of a protein using PCR. Using this technique, Stemmer was able to create
a P-lactamase gene that conferred resistance to the antibiotic cefotaxime at 640
ug/ml. This represents a 32,000-fold increase in resistance to this antibiotic (41). A
diagrammatic representation of this method is shown in Figure 3.1. This technique
cannot only be used to recombine mutated sequences from a single organism, but
also to combine homologous sequences from diverse species (42). It is thought that
this technique may allow for the in vitro evolution of pentose sugar utilization
proteins that will be able to withstand the stresses of a high ethanol, low pH
environment. It has been reported that internal pH in Zymomonas decreases from
6.5 to as low as 5.0 as cells move from logarithmic growth to stationary growth
(43). In addition, ethanol moves out through the cell membrane by passive
diffusion so that cytoplasmic levels of ethanol are as high or higher than in
surrounding media (44). Therefore, if ethanol and pH tolerant characteristics could
be combined, it is likely that this would result in an enzyme that would have a high
level of activity in Zymomonas.
A third strategy that could be utilized to find enzymes that work better in
recombinant Zymomonas is to search the literature for enzymes that have desirable
characteristics, or organisms that live in environments that mimic those of
Zymomonas and screen these bacteria for appropriate enzymes. There is a great deal
of information regarding xylose isomerase (Table 2). Other enzymes in the pathway
are not as well described; however, it seems reasonable that if an organism
possesses a highly ethanol and pH tolerant xylose isomerase, it is likely that the
other enzymes in the metabolic pathway have similar traits. Using the above
methods it is reasoned that it would be possible to determine the cause of sub-
optimal enzyme performance and to increase the efficiency of pentose sugar
conversion to ethanol in Zymomonas. These strategies will form the major focus of
the following experiments
Microorganism metal cofactor Temp.
pH Kmglu Vmaxglu
Vmaxxyl Ki mM
Streptomyces (PLC) Co2+, Mg2+ Mn2+
Streptomyces Thermonitrificans Co2+,Mg2+
Thermotoga maritima Co2+,Mg2+
Neurospora crassa Co2+,Mg2+
Bacillus siearuir ten lupi m us Co2+,Mg2+
Streptomyces rubiginosus* Co2+, Mg2+ Mn2+
70 7 400
85 7 250 191.7
110 6.5-7.5 118 16.2
80 8 220 5.96
80 8 250 5.33
53-90 6.5 400 5.9
90 6.1-7.8 85.5 21 U/mg
7mM 63.8 3.9mM 53
74mM 68.4 46
100mM 44.5 55
33mM 28.26 55
35 mM 12.8 1.5mM 49
16mM 46 U/mg 46
Table 1.2(Cont.). Reported characteristics for xylose isomerase from different species. Metal cofactors are the
cofactors necessary for enzymatic activity. Temperature and pH are for optimum activity. ^Denotes that the gene has
Microorganism metal cofactor
Actinoplanes Missouriensis Co2+,Mg2+
Actinoplanes Missouriensis (mutant) Co2+,Mg2+
Thermus fiavus* Co2+, Mg2+ Mn2+
Thermoanaerobacterium * Mn2+, Mg2+
Thermoamaerobacter ethanolicus * Co2+, Mg2+ Mn2+
Temp. pH 35 xyl 7.0-7.5 60 glu KmGlu mM 290 VmaxGlu
35 xyl 60 glu
85-95 7 106
60 6.4 130 6U/mg
KmXyl I Ki mM Ref.
19mM 16u/mg 38
Escherichia coli* Co2+, Mg2+ Mn2+ 30 7 500 10mM Km 670mM for ribose
Lactobacillus brevis* Mg+2 35 6.0-7.0 920 5mM
Table 1.2. Reported characteristics for xylose isomerase from different species. Metal cofactors are the cofactors
necessary for enzymatic activity. Temperature and pH are for optimum activity. ^Denotes that the gene has been
MATERIALS AND METHODS
Preparation of Cell Free Extracts
Zymomonas mobilis 206C/pZB 301 cells were grown to early stationary phase
at approximately OD6oo 1-78 in rich media (10g/l yeast extract 2g/l K2HP04)
containing 2% glucose, and 10 pg /ml tetracycline at 30 C. Cells used for
experiments investigating the conversion of sugars to ethanol experiments were
washed twice in sonication buffer (10 mM Tris, 10 mM MgCl2, 1 mM dithiothreitol),
and then resuspended in the same buffer with 5 mg/ml lysozyme (Roche,
Indianapolis, IN). Lysozyme treatment was found to increase ethanol production
significantly in cell free experiments. These suspensions were incubated for 30
minutes at room temperature, and then were sonicated. Sonication was conducted
using a Branson Sonifer 450 (Branson Ultrasonics, Danbury, CT) at an output setting
of 2, duty cycle setting of 70% for 30 seconds. Cell free extract was obtained by
centrifugation (1 l,950xg for 45minutes). All cell extracts were kept on ice prior to
Strains L. brevis ATCC 8287 and Lactobacillus Mont 4 were used57. Single
colonies were picked from MRS agar (Difco, Detroit, MI) plate with appropriate
selection factors and used to inoculate 500 ml of MRS media with 2%
xylose (10 p.g /ml erythromycin added into the Mont 4 cultures). Cultures were
grown at 37 C for 48 hours. Cells were harvested by centrifugation (3800Xg for 15
minutes). Cell were next resuspended in sonication buffer and lysed by sonication
followed by centrifugation as above.
Strain E. coli BL21 489 was generously provided by Dr. Juergan Wiegel of
the University of Georgia47. This strain contains the xylA sequence from
Thermoanaerobacterium Strain JW/SL-YS 489. Escherichia coli BL21 strain 489
was grown by picking a single colony from LB plates containing 100 jig /ml
ampicillin and inoculating LB media containing 50 fig/ml ampicillin. Expression of
this gene was induced by the addition of 0.2 mM isopropyl-(3-D-
thiogalactopyranoside (IPTG). Cell free extract was prepared as above.
Acidotkermus cellulolyticus was isolated by NREL researchers from acidic hot
springs in Yellowstone national park58. Cell free extract was prepared by
resuspending frozen cell pellets in sonication buffer followed by sonication.
Acidothermus required five rounds of sonication due to a resilient cell wall. In
experiments requiring a glucose isomerase spike, glucose isomerase from
Streptomyces rubiginosus was obtained from Genencor International (Rochester,
Xylose isomerase assays were conducted according to the protocol
developed by Feldmann et al.59, with higher temperatures used for determination of
thermophilic enzyme activity. Xylulokinase assays were conducted according to the
protocol described by Feldmann et al.37. Spectrophometeric readings were
conducted using a Beckman DU-480 (Beckman Instruments, Fullerton,CA).
In Vitro Sugar Conversion by Cell Free Extracts
Cell-free extract was added to the reaction mixture, resulting in the
following final concentrations: 7 mM magnesium acetate, 5mM ATP, 2mM NAD+,
0.1 mM 2,3- bisphosphoglycerate, and a final protein concentration of 8.33 mg/ml.
Thirty microliters of acetaldehyde was added for each milliliter of reaction mixture.
This addition was later determined to be unnecessary and its use was discontinued.
Sodium arsenate was added to a final concentration of 4mM in order to interrupt the
production of ATP36. Sugars were added to a final concentration of 2% to begin the
reaction. In order to avoid ethanol-induced protein precipitation, sugar-ethanol
mixtures were added to reactions simultaneously. Samples were heat denatured at
97C for ten minutes then filtered through a 0.22-p.M filter, and analyzed by high
performance liquid chromatography (HPLC) using a Bio-Rad Aminex HPX-87H
column (BioRad Laboratories Hercules, CA). All protein concentrations were
determined using the Bradford assay60.
Escherichia coli strains HB101 and DH5a obtained from Life Technologies
were used to determine feasibility of selection based on xylose metabolism on
MacConkey agar. Further these strains were used to determine levels of xylitol
necessary to inhibit xylose consumption. Activity of xylulokinase was assayed in
these strains using the methods described above. All chemicals were purchased
from Sigma Chemicals, St. Louis, MO. All media, unless otherwise noted, were
purchased from Difco Laboratories, Detroit, MI.
RESULTS AND DISCUSSION
The development of a cell-free system to test the function of the recombinant
enzymes in Zymomonas was mandated by the difficulty in determining which
aspects of sub-optimal sugar utilization were caused by sugar transport and which
were caused by enzyme activity. The in vitro system allowed for the study of
isolated enzyme function. Additionally, this system permitted the manipulation of
pH and ethanol concentrations without the potential confusion caused by membrane
based pumps that may alter intracellular conditions.
Algar and Scopes reported a system for cell-free glucose conversion in
Zymomonas that required the use of an ATPase35. ATPase functions to degrade the
ATP produced during glycolysis. This removal of ATP is essential due to the
feedback inhibition that ATP has on the system. Without the removal of ATP,
glycolysis quickly stops. This system proved difficult to recreate since ATPase is
both expensive and problematic to supply in the exact concentrations necessary to
balance ATP production. A later paper from this same laboratory recommended the
use of sodium arsenate at 4mM36. The substitution of sodium arsenate for ATPase
allowed for the rapid conversion of sugars through glycosis. Additionally,
treatment with lysozyme significantly increased sugar conversion, perhaps by
releasing some membrane bound proteins into the cell-free fraction.
In cell-free extract experiments without ethanol challenge, sugar consumption had
the following order: glucose, arabinose and xylose (Fig. 3.1). Not only is the order
of arabinose and xylose reversed compared to in vivo sugar consumption as seen in
Figure 1.2, but also the rate of arabinose conversion approaches that of glucose.
These results indicate that arabinose utilization is most limited by transport, and
xylose utilization could be increased by increased enzyme function. With the cell-
free system working, experiments could be performed to determine the effect of
putative inhibitors; foremost among these was ethanol. Escherichia coli normally
produces very little ethanol; therefore, it is feasible that the recombinant enzymes
from E. coli are unable to withstand the ethanol conditions that exist in Zymomonas.
Because native Zymomonas is able to produce ethanol concentrations of 12% it
was decided to test levels of ethanol around this concentration. When the in vitro
system was challenged with concentration of ethanol at 0, 1, 2, 4, 8 and 16% (v/v),
arabinose was converted more quickly than xylose. In addition, the consumption of
xylose appeared to be more inhibited by ethanol than was arabinose consumption
(Figs. 3.2 and 3.3). Consumption of both sugars was most inhibited at
concentrations of ethanol at 16% (v/v). However, residual arabinose levels
approach 2 g/1 while xylose levels remain at greater than 6 g/1. Although glucose
consumption is only very slightly affected by these ethanol levels, there was some
decrease in sugar conversion at 16% ethanol (Fig. 3.4) Ethanol concentrations of 16
% or greater may represent levels that are not well tolerated by the native enzymes
in the in vitro system. It is however, important to note that the glucose experiment
was done with a much lower protein loading than the previous experiments, 1
mg/ml compared to 10 mg/ml. The lower protein loading allowed for more accurate
measurements to be made due to a slower rate of glucose consumption. One
difficulty faced when working with low levels of proteins is decreased enzyme
stability due to a lack of non-specific and specific protein interactions, e.g.
chaperonins. However, since Zymomonas does not produce concentration of
ethanol this high it seems reasonable that concentrations of ethanol at 16 % may be
Figure 3.1 In vitro sugar conversion by Z. mobilis 206C/pZB301 CFE at 30 C.
10 mg/ml total protein concentration. All sugars were initially 20 grams/liter, but
there is some conversion before the 0 minute sample can be heat denatured.
-A- 2% EtOH
Figure 3.2 In vitro arabinose conversion by Z mobilis 206C/pZB301 at different
concentrations of ethanol at 30 C and 10 mg/ml total protein concentration.
Figure 3.3 Xylose conversion by Z. mobilis 206C/pZB301 CFE
at different concentrations of ethanol at 30C and 10 mg/ml total protein
Figure 3.4 In vitro conversion of glucose by Z. mobilis CFE at 30 C at different
levels of ethanol (v/v). Note that protein loading in this figure is lmg/ml. 1:10 of
normal. This low protein loading was done to facilitate measurement of sugar
The results from the cell-free work suggest the following conclusions. First,
the sub-optimal conversion of arabinose is due largely to transport issues and the
recombinant arabinose specific enzymes appear to be functioning well in
Zymomonas. In the experiments that investigated ethanol effect, it appeared that
arabinose conversion had a profile that was veiy similar to that of glucose. This
would indicate that all of the three arabinose specific enzymes, L-arabinose
isomerase, L-ribulokinase and L-ribulose5-P 4-epimerase, are able to withstand
ethanol at a level similar to endogenous Zymomonas enzymes. Furthermore, these
results show that the two enzymes common to the xylose and arabinose pathway,
transketolase and transaldolase can function well in high concentrations of ethanol.
Therefore, xylose conversion appears to be limited due to enzyme activity of the
xylose specific proteins.
The results of the cell-free studies suggest that the enzymes involved in the
conversion of xylose to D-xylose-5-phosphate, xylose isomerase and xylulokinase.
are limiting. These enzymes do not appear to be as active as the subsequent
enzymes in the pentose phosphate pathway. A series of investigations were
conducted to better understand the exact nature of inhibition of the xylose specific
enzymes. There is a large body of information regarding xylose isomerase. This
enzyme is of significant industrial importance due to its ability to convert glucose to
the sweeter isomer fructose43. From the literature, it was determined that specific
factors could impact xylose isomerase performance in Zymomonas. Among these
factors were pH and xylitol. The impact of ethanol on xylose isomerase has not
been reported. Since ethanol is such a prominent component of the Zymomonas
environment, it was decided to test the effect of ethanol on xylose isomerase, as well
as the effects of xylitol and pH conditions.
Ethanol had a significant impact on the activity of xylose isomerase. In
Figure 3.5 the effect of ethanol on xylose isomerase is shown. This result indicates
that even low levels of ethanol decrease enzyme activity. Ethanol had a much
greater impact in these studies than was seen in the cell free extract results. Again,
this difference may be due to the fact that the protein loading in the xylose
isomerase assay is much lower than in the cell free studies. High protein
concentrations may be protective due to chaperonins or due to nonspecific
stabilization effects. It was necessary to limit the amount of protein loaded in this
series of experiments in order to get accurate readings leading to the above result.
Unfortunately, no other methods were available that would overcome this limitation.
Still these results clearly show that ethanol, even at low levels, decreases xylose
Xylitol is known to be very inhibitory' to xylose isomerase function .
Recombinant Zymomonas can produce levels of xylitol as high as 6.56 grams per
liter (Kent Evans, National Renewable Energy Laboratory, personal
communication). A series of experiments were conducted to determine levels at
which xylitol became inhibitory. Results from these experiments (Fig. 3.6)
demonstrate that xylitol has dramatic effects at levels as low a 1 -gram per liter.
These results indicate that fermentation of xylose by recombinant Zymomonas can
produce levels of xylitol that will inhibit xylose isomerase.
The next inhibitor)' condition that was tested was pH (Fig. 3.7). This is a
particularly important parameter since the internal pH of Zymomonas decreases
from 6.5 during logarithmic growth to as low as 5.0 during stationary' phase24. As
can be seen in Figure 3.7, xylose isomerase activity diminishes greatly as the pH
decreases from 7.0. Since the majority of xylose consumption occurs during late
log or early stationary phase, it is reasonable to assume that xylose isomerase is not
very active at the point in growth when it needs to be most active.
Figure 3.5 Effect of ethanol on E. coli xylose isomerase activity. Percent activity
determined by assigning 100% activity to the 0% ethanol sample.
Figure 3.6 Effect of xylitol on E. coli xylose isomerase activity. Percent activity
determined by assigning 100% activity to the 0% xylitol sample
Figure 3.7 Effect of pH on E. coli xylose isomerase activity. Percent activity
determined by assigning 100% activity to the pH 7.0 sample.
Xylulokinase was also assayed to determine if pH or ethanol inhibited activity.
In Figure 3.8, it is clear that ethanol does not have an effect on the activity of
xylulokinase at levels up to 8% (v/v). However, pH does have a severe impact on
the function of this enzyme (Fig. 3.9). Again, pH is of critical importance due the
intracellular acidification of Zymomonas during culture growth. The inhibitory
effect of xylitol was not tested. It was felt that the chemical structure of the
substrates for xylulokinase, xylulose and xylulose-5-P, were so different from
xylitol that it was unlikely that xylitol would act as an inhibitor.
Figure 3.8 Effect of ethanol on E. coli xylulokinase. Percent activity determined by
assigning 100% activity to the 0% ethanol sample.
Figure 3.9 Effect of pH on E. coli xylulokinase. Percent activity determined by
assigning 100% activity to the pH 7.0 sample
As a result of this research it was decided that the activity of xylose
isomerase and xylulokinase needed to be improved in order to improve xylose
utilization. There has not been a great deal of research done on xylulokinase,
probably because it has little commercial or therapeutic importance. In contrast,
there is a great deal known about xylose isomerase. Xylose isomerase is arguably
the single most important industrial enzyme currently produced in the world.
Xylose isomerases from many organisms have been thoroughly studied for their
kinetics, cofactor requirements and optimal conditions for activity, additionally a
large number of these enzymes have been cloned and sequenced. Table 1.2
represents a compilation of many different xylose isomerase papers. Using this
information, it was possible to quickly identify enzymes that might have
characteristics that would allow them to function well in Zymomonas. Due to the
plethora of information available on xylose isomerase, it was decided to focus on
improving the activity of this enzyme first, and then address the optimization of
A traditional method to identify enzymes with characteristics that would be
resistant to conditions that exist in Zymomonas was undertaken. This method
included the identification of either organisms or enzymes that could function in
conditions similar to those present in Zymomonas during fermentation. The
following criteria were deemed important: ability to withstand low pH, high ethanol
concentrations, and high levels of xylitol. A xylose isomerase that demonstrated all
or some of these qualities would be a very good candidate for production of ethanol
from xylose in Zymomonas. Additionally, it was hoped that identification of a
xylose isomerase with these attributes would allow for the isolation of an equally
Xylose Isomerase Screening
It has been reported that resistance to ethanol corresponds to resistance to
heat Due to this correlation it was decided to begin the search for an ethanol
resistant xylose isomerase by examining thermotolerant xylose isomerases.
Thermotolerant organisms that were able to exist in low pH environments were
particularly scrutinized. The ideal enzyme would be ethanol tolerant at a low pH,
but be able to function at 35 C, which is the optimal temperature for Zymomonas
Initial screens on thermotolerant organisms did not yield promising results
since the proteins retain little or no activity at 35 C. Thermoanaerobacterium
strain JW/SL-YS489 had very high activity at pH 6.5 at 65 C, but had no detectable
activity at 35 C. Acidothennus cellulolyticus was also tested, and had good activity
at 65 C, but it did not show any activity at 35 C either. Although both the xylose
isomerase from Thermoanaerobacterium and A. cellulolyticus show significant acid
tolerance (Table 2), their lack of activity at lower temperatures rendered them
unusable in Zymomonas. A third enzyme from Streptomyces rubiginosus
demonstrated activity over a broad range of temperatures, but lost 54% of activity at
40C. In addition Streptomyces xylose isomerase has a pH optimum of 7.0-7.5.
Due to these facts, it seems unlikely that the xylose isomerase from this organism
would be more effective in our system than the xylose isomerase from E. coli.
These results indicated that it was necessary to identify a xylose isomerase that had
a temperature optimum that was near 35C. This mandated that the focus on
screening organisms be placed on pH characteristics, not on thermotolerance.
In 1968, Yamanaka reported that xylose isomerase from Lactobacillus brevis had a
pH optimum between 6.0-7.050. This is the lowest reported pH optimum for a non-
thermotolerant organism. Additionally, a second species of Lactobacillus had been
isolated by NREL researchers and determined to be resistant to ethanol,
Lactobacillus Mont 4 (Min Zhang, National Renewable Energy Laboratory,
Screening of Lactobacillus proved to be very successful. Two species of
Lactobacillus were tested for xylose isomerase activity, L. brevis and L. Mont 4. An
assay to determine activity at different pH demonstrated that xylose isomerases
from Lactobacillus had higher activities at lower pH than did E. coli or 5.
rubiginosus (Fig. 3.10).
Since there is a great deal of sequence information available for this
organism, experiments to further characterize xylose isomerase focused on L. brevis
instead of L. mont4. Experiments determined that L. brevis xylose isomerase is
more resistant to ethanol than is the enzyme produced by E. coli (Fig. 3.11).
Additionally the L. brevis enzyme shows a greater much greater tolerance to xylitol
than does the E. coli enzyme (Fig. 3.12). The concentrations of xylitol necessary to
inhibit L. brevis xylose isomerase are much greater than are produced during
fermentations. A typical fermentation will produce less than 0.065% xylitol; this
would result in very little inhibition. Even at one percent xylitol, L. brevis xylose
isomerase had greater than 70% activity. These results indicate that the xylose
isomerase from L. brevis may have characteristics that would be ideal for ethanol
production in Zymomonas.
Figure 3.10 Activity of xylose isomerase from different organisms at different pHs.
Percent activity was determined by assigning 100% activity to the pH 7.0 sample.
% ethanol (w/w)
Figure 3.11 Comparisons of xylose isomerase activity from L. brevis and E. coli in
different concentrations of ethanol.
vO O' vO 0s xP 0s- vO O'- xP 0s xP 0s >P 0s- xP O' xP 0s xP O'
O O LO T CM CO ^4* - CM CO
(0 o * o d o (0 CO (0 <0
O > o > > > >
o o o o o o CD 0) cd o
UJ k_ X) o LU o o o V J3 n .Q n
Lii Lii LU Lii _j _i
Figure 3.12 Activity of xylose isomerases at different concentration of xylitol. Note
that xylitol concentrations in the L. brevis experiments are ten times greater than in
the E. coli experiments (n=4).
Xylulokinase from L. brevis was assayed to determine if it would have the
same beneficial characteristics that xylose isomerase from this organism has.
Assays to determine ability to withstand ethanol provided favorable results (Fig.
3.13). This result was expected since xylulokinase from E. coli is very resistant to
ethanol as well. The ability of the enzyme to withstand pH challenge was also
investigated. In Figure 3.14, activity of xylulokinases from both E. coli and L.
brevis are compared. Although the activity of these two enzymes is only slightly
different L. brevis xylulokinase has a slightly lower pH optimum. This may allow
this enzyme to better tolerate conditions that exits in Z mobilis during fermentation
than E. coli xylulokinase.
It is difficult to determine the effect that the expression of the L. brevis
enzymes will have in Zymomonas. These enzymes have biochemical characteristics
that should be well suited to the intracellular environment, but there are other factors
to consider. One of these factors is whether these enzymes will be expressed in
Zymomonas. Both E. coli and Zymomonas are Gram-negative organisms;
Figure 3.13 Activity of L. brevis xylulokinase at different concentrations of ethanol.
Percent activity is determine by assigning 100% activity to the 0%
while Lactobacillus is Gram-positive. Levels of expression are also important.
Lactobacillus may use codons that are unrecognized or under-recognized in
Zymomonas, leading to either no or low expression of these proteins. It may also be
that other protein components in Lactobacillus such as chaperonins enhance ethanol
tolerance. If this is the case, then it is unlikely that the L. brevis enzymes would
have a significant advantage over the E. coli enzymes.
In contrast, it may be that these enzymes are not only expressed, but they
have high activity. This would allow for the rapid conversion of xylose and
subsequent production of ATP. It is suggested that the decreased production of
ATP that results from slow xylose conversion may result in intracellular
acidification. Since most bacteria use active proton pumps to control intracellular
pH, a decrease in available ATP would result in a decrease in intracellular pH in a
fermenting culture55. In recombinant Zymomonas, this may lead to a cycle in which
a decrease in ATP leads to a decrease in pH. It is feasible, then, that increased
utilization of xylose resulting in increased available ATP would allow the cell to
better control its internal pH. This would result in a healthier cell that has a much
greater ability to convert xylose to ethanol.
It is very difficult to know how the expression of recombinant proteins will
affect the host cell without actually cloning and expressing the proteins. It is
reasonable to expect that xylose isomerase and xylulokinase will retain at least some
of their favorable characteristics in Zymomonas.
Xylitol Production in Recombinant Zymomonas
The production of xylitol in Zymomonas mobilis represents a problem for
two reasons. As stated above, xylitol is a strong inhibitor of xylose isomerase. Also
the production of xylitol decreases the production of ethanol by redirecting xylose
into xylitol instead of ethanol. Feldmann and coworkers have attributed the
production of xylitol to a novel xylitol NADPH-dependent aldose reductase41.
Oblation of this enzyme would have the beneficial effects of removing an inhibitor
as well as forcing more xylose to be converted to ethanol. Although there does
appear to be an aldose reductase enzyme, the sequence for it has yet to be
determined. Because the sequence of this gene is unknown, it will be some time
before it is possible to interrupt this gene in a specific way.
There is the possibility that another pathway exists for the production of
xylitol. Zachariou and Scopes have described an enzyme that appears to be unique
to Zymomonas mobilis, glucose-fructose oxidoreductase56. They determined that
this enzyme functions to produce gluconolactone and sorbitol from glucose and
fructose, respectively. They further characterized the ability of this enzyme to
convert other sugars, as long as one sugar is a ketose and one is an aldose. This
enzyme shows low activity catalyzing xylose (8%) and xylulose (7%) compared to
glucose and fructose respectively, but it does show activity with these two sugars. It
therefore is likely that this enzyme may be in part responsible for the production of
xylitol in recombinant Zymomonas, and that disruption of its function may increase
pentose sugar conversion. A schematic of this pathway is presented in Figure 3.15.
Glucose --------y---------------------->r--- Glucolactone
Figure 3.15 Proposed pathway for xylitol production in recombinant Zymomonas
To determine if GFOR had a role in xylitol production, native cell-free
extract was supplemented with only xylose or both xylose and xylulose at 10
grams/liter. Native cell-free extract lacks xylose isomerase. Therefore, if xylitol
production is the sole result of an aldose reductase, equal amounts of xylitol will be
seen in both the native cell-free extract and the native cell-free extract spiked with
Genencor glucose isomerase. Glucose isomerase will produce xylulose from xylose
resulting in the presence of both of these sugars which, in turn, makes available both
an aldose and a ketose to GFOR. If this enzyme functions to produce xylitol under
these conditions then the glucose isomerase spiked sample should have greater
levels of xylitol production than the native cell-free extract only sample. In Figure
3.16, it is clear that the spiked sample produced more xylitol than does the non-
spiked only sample.
0 1 2 3 4 5 24
Figure 3.16 Xylitol production by native Z. mobilis 39676 cell free extract with
xylose and xylulose at 10 gram/liter. Genencor glucose isomerase was added to an
approximate final concentration of 10 units.
This result suggests that in recombinant Zymomonas the main pathway for xylitol
formation occurs through the action of GFOR. There is, however, some xylitol
production jn the sample that had no xylose isomerase added to it. It is unlikely that
this represents the activity of an endogenous xylose isomerase acting in concert with
GFOR. Rather, this xylitol is probably produced through the action of an aldol
reductase as suggested by Feldmann and coworkers. Clearly, the aldol reductase
pathway contributes minimally to overall xylitol production in recombinant
Zvmomonas, but it is likely the predominate pathway in the native organism.
Since the production of xylitol creates not only an inhibitory compound but
also limits ethanol production, it was decided to attempt to interrupt the GFOR
gene. Results of attempts to integrate a non-functional GFOR gene into the
chromosome were disappointing. After two rounds of transformation using plasmid
pZB1861-GFOR-Tc (Fig. 3.17) and growth for 100 generations no colonies were
detected that had lost the chloroamphenicol resistance gene while retaining
tetracycline resistance. All colonies screened were resistant to both antibiotics.
Experiments using the more symmetrical plasmid pZB1861-GFOR-TC (B) (Fig.
3.18) were equally unproductive. After transformation and twelve transfers (-120
generations), no intregrants were detected (Yat-Chen Chou, National Renewable
Energy Laboratory, personal communication).
Transformations with both linear and circular suicide plasmid pYC1865-
GFOR-Tc (Fig. 3.19) were used to transform cells. This vector, if not integrated,
cannot be replicated; therefore any cells that have a tetracycline resistant phenotype
must be intregrants. Although controls transformed with replicative plasmids had a
high transformation rate, showing that the cells were competent, no colonies were
detected from the pYC1865-GFOR-Tc transformations. Experiments using
pYC1865-GFOR-Tc (B) (Fig. 3.20) also yielded nothing. In this series of
experiments, 32 different experiments were conducted using 8 different
concentrations of both linear and circular plasmid. No integrants were obtained for
any of these conditions (Yat-Chen Chou, National Renewable Energy Laboratory,
personal communication). These results indicate that GFOR function is very
difficult to knockout. This may be due to regions of homology that are too short to
allow for efficient recombination. However, since homologous recombination has
been exploited to interrupt the lactate dehydrogenase gene in Zymomonas, this is
unlikely. This interruption was accomplished using constructs very similar to the
ones used for GFOR. A second possibility is that there is an unknown, essential
function of GFOR or one of its products. In this case, interruption of this enzyme
would lead to a lethal mutation.
Whatever the reason for the lack of homologous recombination the fact
remains that the action of GFOR is still a problem for ethanol production from
xylose. A method to address this problem may be to increase the activity of
xylulokinase to such a level that there are extremely low levels of xylulose present
in the cell. Without this aldose present, it is unlikely that any significant levels of
xylitol would be produced.
Z. DNA 2.53 kb
Figure 3.17 Replicative vector pZB1861-GFOR-TC. This vector was designed
incorporating a Zymomonas origin of replication
SphI 42241 A
BsePl/BssHIl 4019 7
Figure 3.18 Plasmid pZB 1861 -GFOR-Tc(B). This plasmid contains a Z. mobilis
origin of replication. It also contains a modified glucose-fructose oxidoreductase
gene that is interrupted by the gene for tetracycline resistance. The GFOR gene has
further been altered by the deletion of 300 bp resulting in almost equal sized gene
portions on either side of the tetracycline gene interruption.
m 77,m wlal 2359
hoi 2718 Hindlll 2364
Figure 3.19 Suicidal integration vector pYC1865-GFOR-Tc. This vector was
designed lacking a Zymomonas origin of replication.
BsePl/BssHII 4019\ l Z. DNA resid.
Figure 3.20 Plasmid pYC1865-GFOR-Tc (B). This plasmid does not include a Z.
mobilis origin of replication. It contains a modified glucose-fructose oxidoreductase
gene that is interrupted by the gene for tetracycline resistance. The GFOR gene has
further been altered by the deletion of 300 bp resulting in almost equal sized gene
portions on either side of the tetracycline gene interruption.
There are two causes of incomplete pentose sugar utilization. Arabinose
utilization is primarily affected by the lack of sugar transport. Sub-optimal xylose
utilization results from E. coli enzymes that are unable to work effectively in
Zymomonas. Xylose conversion may be substantially increased in recombinant
Zymomonas by the expression of xylose isomerase and xylulokinase derived from L.
Sugar to ethanol yield may further be increased by the removal of several
genes that remove carbon from glycolysis. The interruption of both GFOR and the
aldol reductase gene, if possible, would be beneficial in allowing more carbon to be
converted to ethanol. Additionally this would offer more sites for gene integration.
Work is currently in progress to clone the L. brevis xylA and xylB genes into
a Zymomonas expression vector. This work has been helped considerably by the
fact that these genes have recently been cloned57. It is hoped that these proteins can
be expressed in Zymomonas and that they will retain their ability to withstand the
stresses that exist during fermentation. A second exciting technology that has been
recently developed is PCR shuffling This is a technique that allows for rapid in
vitro evolution of proteins through recombination. Using this technique, Stemmer
was able to produce a B-lactamase that had 32,000-fold greater activity than native a
B-lactamase This technique may be applicable to increasing the resistance of
certain enzymes to inhibitors. It might also be useful to increase the substrate range
of the Zymomonas glucose transport protein to include pentose sugars, in particular
Although arabinose assimilation enzymes appeared to be more efficient at
metabolism than did the xylose enzymes, it is highly likely that these enzymes could
also be improved. This research could be carried out in a fashion similar to that
described above for the xylose enzymes. In addition it may be beneficial to
determine if L. brevis is able to metabolize arabinose, or if there is an organism
similar to L. brevis that is capable of metabolizing this sugar. It is hoped through
the application of both traditional and cutting-edge technologies an organism can be
developed that is capable of making inexpensive renewable fuel ethanol a reality.
1. ) 1997 Fuel Ethanol Fact Book: For the Recore, United States of America Fuel
Ethanol Program. Prepared by the Clean Fuels Development Coalition. Arlington,
2. ) Campbell, C.J., Laherrere, J. H., (1997) The End of Cheap Oil. Scientific
American, March 1997, 78-83
3) Romm, J.J., Curtis, C.B., (1996) Mideast Oil Forever? The Atlantic Monthly,
April 1996, 57-74
4. ) Annual Energy Outlook with Projections to 2015. U.S. Department of Energy,
Energy Information Administration. DOE/EIA-0383(96), Washington, D.C.
5. ) Petroleum and Ethanol: A Trade Deficit Analysis, Energetics, June 1994, 1-21.
6. ) Schneider, S. H., (1989) The Greenhouse Effect: Science and Policy. Science,
Vol. 243, 771-781
7. ) Kelly, K.J., Bailey, B.K., Cobum, T.C., Clark. W Lissiuk, P. (1996) Federal
Test Procedure Emissions Test Results from Ethanol Variable-Fuel Vehicle
Chevrolet Luminas. Society for Automotive Engineers International Spring
Fuels and Lubricants Meeting, Dearborn, MI, May 6-8 1996.
8. ) Humphrey, A.E., (1979) The Hydrolysis of Cellulosic Materials to Useful
Products. In Hydrolysis of Cellulose: Mechanisms of Enzymatic and Acid
Catalysis. Advances in Chemistry Series No. 181, American Chemical Society,
Washington D.C. 25-52
9. ) Dumsday, G.J., Jones, K., Stanley, G.A., Pamment, N.B. (1997) Recombinant
Organisms for Ethanol Production from Hemicellulosic Hydrolyzates-A Review
of Recent Progress. Australian Biotechnology, Vol. 7, 285-292
10. ) Deng, M., Coleman, J.R., (1999) Ethanol Synthesis by Genetic Engineering in
Cyanobacteria. Applied and Environmental Microbiology, Vol. 65 No.2, 523-
11. ) Postma, P.W. (1987) "Phosphotransferase Systems for Glucose and Other
Sugar," Chapter 11. Neidhardt, P.C., ed. Escherichia coli and Salmonella
typhimurium. Vol. 1, Washington D.C.: American Society for Microbiology,
12. ) Han-Hagerdal, B., Jeppsson, H., Skoog, K., Prior, B.A., (1994) Biochemistry
and Physiology of Xylose Fermentation by Yeasts. Enzyme Microbial
Technology, Vol. 16, 933-943
13. )Chiang, L.C., Gong, C.S., Chen, L.F., Tsao, G.T., (1981) D-xylulose
Fermentation to Ethanol by Saccharomyces cerevisiae. Applied and
Environmental Microbiology. Vol. 42, 284-289
14. ) Moes, C.J., Pretoruis, I. S., VanZyl W.H., (1996) Cloning and Expression of
the Clostridium thermosulfurogenes D-xylose Isomerase Gene (Xyl A) in
Saccharomyces cerevisiae. Biotechnology Letters, Vol. 18, 269-274
15. ) Walfridsson, M., Hallbom, J., Penttila, M., Keranen, S., Hah-Hagerdal, B.,
(1995) Xylose-metabolizing Saccharomyces cerevisiae Strains Overexpressing
the TKL1 and TALI Genes Encoding the Pentose Phosphate Pathway Enzymes
Transketolase and Transaldolase. Applied and Environmental Microbiology,
16. ) Krishnan M.S., Ho, N.Y.W., Tsao, G.T., (1999) Fermentation Kinetics of
Ethanol Production from Glucose and Xylose by Recombinant Saccharomyces
1400(pLNH33). Applied Biochemistry and Biotechnology, Vol. 77-79, 373-
17. ) Ingram L.O., Conway, T., Clark, D.P., Sewell, G.W., Preston, J.F., (1987)
Genetic Engineering of Ethanol Production in Escherichia coli. Applied
Environmental Microbiology, Vol. 53 No. 10, 2420-2425
18. ) Alterthum, F., Ingram, L.O., (1989) Efficient Ethanol Production from
Glucose, Lactose, and Xylose by Recombinant Escherichia coli. Applied
Environmental Microbiology, Vol. 55 No. 8, 1943-1948
19. ) Ohta, K., Beall, D.S., Mejia, J.P., Shanmugam, K.T., Ingram, L.O., (1991)
Genetic Improvement of Escherichia coli for Ethanol Production: Chromosomal
Integration of Zymomonas mobilis Genes Encoding Pyruvate Decarboxylase and
Alcohol Dehydrogenase II. Applied Environmental Microbiology, Vol. 57 No.
20. ) Barbosa, M.F.S., Yomano, L.P., Ingram, L.O., (1994) Cloning, Sequencing and
Expression of Stress Genes from the Ethanol-Producing Bacterium Zymomonas
mobilis: the groESL Operon. Gene, Vol. 128, 52-57
21. ) Moniruzzaman, M., Lai, X., York, S.W., Ingram, L.O., (1997) Isolation and
Molecular Characterization of High-performance Cellobiose-fermenting
Spontaneous Mutants of Ethanologenic Escherichia coli KOI 1 Containing the
Klebsiella oxvtoca casAB Operon. Applied and Environmental Microbiology,
Vol. 63 No. 12, 4633-4637
22. ) Yomano, Y.P., York, S.W., Ingram, L.O., (1998) Isolation and Characterization
of Ethanol-tolerant Mutants of Escherichia coli KOI 1 for Fuel Ethanol
Production. Journal of Industrial Biotechnology, Vol. 20 No. 2, 132-138
23. ) Energy Department Welcomes New Era in Conversion of Waste to Energy.
Golden Field Office of the Department of Energy News Release. November 11,
24. ) Swings, J., De Ley, J. (1977) The Biology of Zymomonas. Bacteriological
Reviews, Vol. 41 No.l, 1-46.
25. )Conway, T., (1992) The Entner-Duodoroff Pathway: History, Physiology and
Molecular Biology. FEMS Microbiology Reviews, Vol. 103, 1-28
26. )Rodgers P. L., Lee, K. J., Skotnicki, M. L., Tribe, D. E. (1982) Ethanol
Production in Zymomonas mobilis. Advances in Biochemical Engineering, Vol.
27.) Sprenger, G. A., (1996) Carbohydrate Metabolism in Zymomonas mobilis: A
Catabolic Highway with some Scenic Routes. FEMS Microbiology Letters,
Vol. 145, 301-307
28. ) Zhang, M., Eddy, C., Deanda, K., Finkelstein, M., Picataggio, S., (1995)
Metabolic Engineering of a Pentose Metabolism Pathway in Ethanologenic
Zymomonas Mobilis. Science, Vol. 267, 240-243
29. ) Deanda, K., Zhang, M., Eddy, C., Picataggio, S., (1996) Development of an
Arabinose-Fermenting Zymomonas mobilis Strain by Metabolic Pathway
Engineering. Applied and Environmental Microbiology, Vol. 62 No. 12, 4465-
30. ) Chou, Y.C., Zhang, M., Mohagheghi, A., Evans, K., Finkelstein, M., (1997)
Construction and Evaluation of a Xylose/Arabinose Fermenting Strain of
Zymomonas mobilis. Abstracts 19th Symposium on Biotechnology for Fuels and
Chemicals, Colorado Springs CO, May 4-8 1997.
31. ) Weisser, P., Kramer, R., Sprenger, G. A., (1996) Expression of the Escherichia
coli pmi Gene, Encoding for Phosphomannose-isomerase in Zymomonas
mobilis Leads to Utilization of Mannose as a Novel Growth Substrate Which
can be used as a Selective Marker. Applied and Environmental Microbiology,
Vol. 62 No. 11,4155-4161
32. ) Buchholtz, S.E., Eveleigh, D.E., (1990) Genetic Modification of Zymomonas
mobilis. Biotech Advances, Vol. 8, 547-581
33. ) Yamase, H., Masuda, M., Tamaki, T., Tonomura, K., (1990) Expression of the
Escherichia coli Alpha-Galactosidase and Lactose Permase Genes in
Zymomonas mobilis and its Raffinose Fermentation. Journal of Fermentation
Bioengineering, Vol. 70, 1-6
34. ) DiMarco, A.A., Romano, A.H., (1985) D-Glucose Transport System of
Zymomonas mobilis. Applied and Environmental Microbiology Vol. 49, 151-
35. )Algar, E.M. Scopes, R.K., (1985) Studies on Cell Free Extract Metabolism:
Ethanol Production by Extracts of Zymomonas mobilis. Journal of
Biotechnology, Vol. 2, 275-280
36. ) Welch, P., Scopes, R.K., (1985) Studies on Cell-Free Metabolism: Ethanol
Production by a Yeast Glycolytic System Reconstituted from Purified Enzymes.
Journal of Biotechnology, Vol.2, 257-273
37. ) Barre, P., (1978) Identification of Thermobacteria and Homofermentative,
Thermophilic Pentose-Utilizing Lactobacilli from High Temperature
Fermenting Grape Musts. Journal of Applied Bacteriology, Vol. 44, 125-129
38. ) Liu, S., Wiegel, J., Gherardini, F.C., 1996 Purification and Cloning of a
Thermostable Xylose Isomerase with an Acidic pH Optimum from
Themionanerobacterium Strain JW/SL-YS 489. Journal of Bacteriology, Vol.
178 No. 20, 5938-5945
39. )Mohagheghi, A., Grohmann, K., Himmel, M., Leighton, L., Updegraff, D.M.,
(1986) Isolation and Characterization of Acidothermus cellulolyticus gen. nov.,
sp. Nov., a New Genus of Thermophilic, Acidophilic, Cellulolytic Bacteria.
International Journal of Systematic Bacteriology, Vol. 36 No. 3., 435-443
40. ) Feldmann, S.D., Sahm, H., Sprenger, G.A., (1992) Cloning and Expression of
the Genes for Xylose Isomerase and Xylulokinase from Klebsiella pneumoniae
in Escherichia coli K12. Mol. Gen. Genet., Vol. 243 No. 2, 201-210
41. ) Feldmann, S.D., Sahm, H., Sprenger, G.A., (1992) Pentose Metabolism in
Zymomonas mobilis Wild Type and Recombinant Strains. Applied Microbial
Biotechnology, Vol. 38, 354-361
42. ) Bradford, M.M., (1976) A Rapid and Sensitive Method for the Quantification
of Microgram Quantities of Protein Utilizing the Principle of Protein-Dye
Binding. Annals of Biochemsitry, Vol. 72, 248-254
43. ) Antrim, R.L., Colilla, W., Schnyder, B.J., (1979) Glucose Isomerase
Production of High-Fructose Syrups. Applied Biochemistry and
Bioengineering, Vol. 2, 97-155
44. ) Van Tilbeurgh, H., Jenkins, J., Chiadmi, M., Janin, J., Wodak, S.J., Mrabet,
N.T., Limbeir, A., 1992 Protein Engineering of Xylose (Glucose) Isomerase
from Actinoplanes missouriensis 3. Changing Metal Specificity and the pH
Profile by Site-Directed Mutagenesis. Biochemistry, Vol.31, 5467-5471
45. ) Brown, S.H., Sjoholm, C., Kelly, R.M., (1992) Purification and
Characterization of a Highly Thermostable Glucose Isomerase Produced by the
Extremely Thermophilic Eubacterium, Thermotoga Maritima. Biotechnology
and Bioengineering, Vol. 41, 878-886
46. ) Vieille, C., Hess, M., Kelly, R.M., Zeikus, G., (1995) xylA Cloning and
Sequencing and Biochemical Characterization of Xylose Isomerase from
Thennotoga neapolitana. Applied and Environmental Microbiology, Vol. 61
No. 5, 1867-1875
47. ) Weigel, J.G., Ljungdahl, L.G., (1981) Thermoanaerobacter ethanolicus, A new
Extreme Thermophilic, Anaerobic Bacterium. Archives of Microbiology, Vol.
48. ) Schellenberg G. D., Sarthy, A., Larson, A.E., Backer, M.P., Crabb, J.W.,
Lidstrom, M., Hall, B.D., Furlong, C.F., (1984) Xylose Isomerase from
Escherichia coli, Characterization of the Protein and the Structural Gene.
Journal of Biological Chemistry, Vol. 259 No. 11, 6826-6832
49. ) Technical insert for Genecor Spezyme Genecore Glucose Isomerase
Concentrate. Genecor International, Rochester NY 14618
50. ) Yamanaka, K., (1968) Purification, Crystallization and Properties of the D-
Xylose Isomerase from Lactobacillus brevis. Biochimica Et Biophysica Acta,
51. ) Inyang C.U., Gebhart, U., Obi S.K.C., Bisswanger, H., (1995) Isolation and
Characterization of a D-Glucose/Xylose Isomerase from a New Thermophilic
Strain Streptomyces sp. (PLC). Applied Microbial Biotechnology, Vol. 43, 632-
52. ) Deshmukh, S.S., Shankar, V., (1996) Glucose Isomerase from Thermophilic
Streptomyces thermonitriftcans: Purification and Characterization.
Biotechnology and Applied Biochemistry, Vol. 24, 65-72
53. ) Rawat, U., Phadtare, S., Deshpande, V., Rao, M., (1996) A Novel Xylose
Isomerase from Neurospora crassa. Biotechnology Letters, Vol. 18 No. 11,
54. ) Suekane, M., Tamura, M., Tomimura, C., (1978) Physico-chemical and
Enzymatic Properties of Purified Glucose Isomerase from Streptomyces
olivochromogenes and Bacillus stearothermophilus. Agricultural Biological
Chemistry, Vol. 42 No. 5, 909-917
55. ) Booth, I.R., (1985) Regulation of Cytoplasmic pH in Bacteria. Microbiological
Reviews, Vol. 49 No. 4, 359-378
56. ) Zachariou, M., Scopes, R.K., (1986) Glucose-Fructose Oxidoreductase, a New
Enzyme Isolated from Zymomonas mobilis That is Responsible for Sorbitol
Production. Journal of Bacteriology, Vol. 167 No. 3, 863-869
57. )Chaillou, S., Bor, Y.C., Batt, C.A., Postma, P.W., Pouwels, P.H., (1998)
Molecular Cloning and Functional Expression in Lactobacillus plantarum 80 of
xylT, Encoding the D-Xylose-H+ Symporter of Lactobacillus brevis. Applied
and Environmental Microbiology, Vol. 64 No. 12, 4720-4728
58. ) Stemmer, W.P.C., (1994) Rapid Evolution of a Protein in vitro by DNA
shuffling. Nature, Vol. 370, 389-391