Activity and Stability of a Recombinant
Plasmid-Borne TCE Degradative Pathway
in Suspended Cultures
Robert R. Sharp,1 James D. Bryers,2 Warren G. Jones,2 Malcolm S. Shields3
1Department of Environmental Engineering, Manhattan College, Riverdale,
New York 10475; telephone: (718) 862-7169; fax: (718) 862-8018; e-mail:
RSHARP@Manhattan.edu
2Center for Biofilm Engineering, Montana State University,
Bozeman, Montana
3Department of Biology, University of West Florida, Pensacola, Florida
Received 20 September 1996; accepted 8 July 1997
Abstract: The retention and expression of the plasmid-
borne, TCE degradative toluene-ortho-monooxygenase
(TOM) pathway in suspended continuous cultures of
transconjugant Burkholderia cepacia 17616 (TOM31c)
were studied. Acetate growth and TCE degradation kinet-
ics for the transconjugant host are described and utilized
in a plasmid loss model. Plasmid maintenance did not
have a significant effect on the growth rate of the trans-
conjugant. Both plasmid-bearing and plasmid-free
strains followed Andrews inhibition growth kinetics
when grown on acetate and had maximum growth rates
of 0.22 h−1. The transconjugant was capable of degrading
TCE at a maximum rate of 9.7 nmol TCE/min z mg pro-
tein, which is comparable to the rates found for the origi-
nal plasmid host, Burkholderia cepacia PR131 (TOM31c).
The specific activity of the TOM pathway was found to be
a linear function of growth rate. Plasmid maintenance
was studied at three different growth rates: 0.17/h, 0.1/h,
and 0.065/h. Plasmid maintenance was found to be a
function of growth rate, with the probability of loss rang-
ing from 0.027 at a growth rate of 0.065/h to 0.034 at a
growth rate 0.17/h. © 1998 John Wiley & Sons, Inc. Biotech-
nol Bioeng 57: 287–296, 1998.
Keywords: expression; plasmid; stability; TCE; continu-
ous culture; activity
INTRODUCTION
Remediation of soil and ground water contaminated by or-
ganic pollutants has been the focus of much research in
recent years. Some of the most ubiquitous contaminants in
America’s ground water aquifers are volatile organics, a
large number of which are chlorinated aliphatic compounds,
including the EPA priority pollutant, trichloroethylene
(TCE). Many biological methods have been developed to
detoxify TCE-laden soil and ground water, including an-
aerobic dechlorination (Bouwer and McCarty, 1983; Ewers
et al., 1990; Fogel et al., 1986; Freedman and Gossett, 1989;
Kleopfer et al., 1985; Vogel and McCarty, 1985) and oxy-
genase-driven TCE degradation (Alvarez-Cohen and Mc-
Carty, 1991a,b; Arciero et al., 1989; Finette et al., 1984;
Kaphammer et al., 1990; Nelson et al., 1986; Oldenhuis et
al., 1989; Shields and Reagin, 1992; Tsien et al., 1989;
Vandenbergh and Kunka, 1988; Wackett and Gibson, 1988;
Wackett et al., 1989; Winter et al., 1989). One of the most
notable aerobic, cometabolic TCE degrading bacterial spe-
cies is the environmental isolate Burkholderia cepacia G4
(formerly Pseudomonas cepacia G4) (Nelson et al., 1986).
B. cepacia G4 degrades TCE via a cometabolic pathway
that must be induced by a cosubstrate, either phenol, tolu-
ene, o-cresol, m-cresol, or catechol (Nelson et al., 1987,
1988).
In laboratory studies performed by Shields et al. (1992),
Tn5 transposon mutants of B. cepacia G4 were developed
that are capable of constitutive mineralization of TCE. The
genesis of these transposon mutants is depicted in Figure 1.
The transposon mutation resulted in two phenol revertants
(B. cepacia PR131 [TOM31c] and B. cepacia PR123
[TOM23c]), both which were capable of constitutive degra-
dation of TCE. One of these transposon mutants, B. cepacia
PR131 (TOM31c), carries the complete Tn5 transposon, in-
cluding kanamycin (Km) resistance, on the toluene/TCE
degradative plasmid TOM31c.
The utility of TOM31c for TCE degradation in either open
ecosystems or closed bioreactor bioremediation systems
rests on the ability to find an appropriate host for the plas-
mid. A suitable host for TOM31c must have a resistance to
waste-related cell injury and toxicity, an ability to retain and
express the desired plasmid-borne phenotype during long-
term operation, and an ability to persist in a desired eco-
system or bioreactor environment.
Microorganisms carrying plasmids are susceptible to
plasmid instability which can lead to the loss of a desired
plasmid-borne phenotype or complete loss or inactivation of
a desired plasmid-borne genotype. Plasmid instability can
occur in two ways: (1) segregational instability; and (2)
Correspondence to: R. R. Sharp
Contract grant sponsor: National Science Foundation
Contract grant number: EC8907039
Contract grant sponsor: Center for Biofilm Engineering at Montana State
University
Contract grant sponsor: Center for Environmental Diagnostics and Bio-
remediation at the University of West Florida
© 1998 John Wiley & Sons, Inc. CCC 0006-3592/98/030287-10
structural instability. Segregational instability is the conse-
quence of random and irregular partitioning of plasmids to
both daughter cells during cell division, and can lead to new
generations of daughter cells that do not contain the plas-
mid. Structural instability of a plasmid involves the actual
change or recombination (deletion, insertion, and/or rear-
rangement) of a single gene or several genes on the plasmid.
Structural instability can also occur when a portion of the
plasmid DNA is incorporated into the chromosomal DNA.
In addition, structural instability may involve the loss of a
certain plasmid-borne phenotype, but not the loss of other
phenotypes associated with the plasmid. For instance, struc-
tural instability of plasmid TOM31c could result in the loss
of the TCE degrading phenotype, but not the loss of the
plasmid-borne, Tn5-encoded, kanamycin resistance.
Much published research on the stability of plasmids ex-
ists. Table I lists some of the factors affecting plasmid re-
tention and stability. A great deal of this research has fo-
cused on suspended cell cultures and the effects of continu-
ous culture parameters on plasmid stability (Dwilvedi et al.,
1982; Dykhuizen and Hartl, 1983; Ensley, 1986; Jones et
al., 1980; Kadam et al., 1987; Kumar et al., 1990, 1991;
Noack et al., 1982; Ollis, 1982; Roth et al., 1980; Roth and
Noack, 1982; Wood et al., 1990). It is important to under-
stand plasmid/host interactions to utilize recombinant plas-
mids in environmental remediation systems. In addition, the
use of plasmid recombinant microorganisms, such as B.
cepacia 17616 (TOM31c), in open ecosystems mandates re-
search into the fundamental plasmid/host processes that in-
fluence plasmid retention, stability, expression, and transfer.
Transconjugant B. cepacia 17616 (TOM31c) was used in
all experiments to evaluate plasmid maintenance, expres-
sion, and stability in suspended cultures. Plasmid TOM31c
was chosen for these studies for three reasons: (1) TOM31c
is capable of constitutively degrading TCE and thus has
obvious applications to the bioremediation of TCE-laden
air, soil, and ground water; (2) TOM31c originates from an
environmental isolate and is not considered an industrial or
genetically designed plasmid; and (3) TOM31c is easily se-
lected for, using either kanamycin (resistance on plasmid-
borne Tn5 insertion, see Fig. 1) or various specific hydro-
carbons as growth substrates. 17616 was chosen to host
TOM31c because it has been well characterized (Cheng and
Lessie, 1994) and it is relatively easy to work with and
maintain.
This article reports on the laboratory evaluation of the
retention and expression of the TCE degrading plasmid
TOM31c in the transconjugant host B. cepacia 17616 in
suspended cultures. A subsequent article examines the re-
tention and expression of the recombinant TCE degrading
plasmid in B. cepacia biofilm cultures (Sharp et al., 1997).
MATERIALS AND METHODS
Plasmid/Host System
Plasmid-bearing strains of B. cepacia were obtained by
solid surface conjugation between PR131 (TOM31c) and
17616. Cultures of the original plasmid host, PR131
(TOM31c), were used in a number of studies as either con-
trols or for comparison with 17616 (TOM31c) cultures.
Glycerol/peptone frozen cultures (−70°C) of plasmid-free
and plasmid-bearing cultures were maintained and used for
Figure 1. Genesis of plasmid TOM31c via the Tn5 transposon mutagen-
esis of Burkholderia cepacia G4 to produce Burkholderia cepacia PR131
(TOM31c) (Shields and Reagin, 1992).
Table I. Factors affecting plasmid stability and retention.
Environmental and
physiological factors References
Growth rate. Increases plasmid loss
with increased growth rate.
Stewart and Carlson (1986); de
Taxis du Poet et al. (1987);
Seo and Bailey (1985)
Plasmid copynumber. Decreased
plasmid loss rate with increased
plasmid copynumber. Plasmid
copynumbers can range from 1 to
700.
Huang et al. (1993); Sayadi et
al. (1989); Uhlin and Nord-
strom (1978); Peretti and Bai-
ley (1987)
Carbon-to-nitrogen ratios. Increased
nitrogen growth conditions can
increase plasmid stability.
Huang et al. (1994); Sayadi et
al. (1989)
Selection. Selection of plasmid us-
ing a selective carbon source or
antibiotic resistance markers can
increase plasmid retention in a
given population.
Lauffenburger (1987); Tiedje et
al. (1989); Wood et al. (1990)
Nutrient limitations. Nitrogen, phos-
phorus, potassium, magnesium,
and carbon limitations may either
increase or decrease plasmid sta-
bility.
Godwin and Slater (1979); Jones
and Melling (1984); Noack et
al. (1982)
Immobilization and attachment.
Plasmid-bearing populations that
are immobilized or in a biofilm
culture may display either in-
creased or decreased plasmid sta-
bility.
Huang et al. (1993); Kumar and
Schugerl (1990); Inloes et al.
(1983); Dykhuizen and Hartl
(1983)
Exposure to injurious or toxic sub-
stances. Injury and toxicity may
lead to increased plasmid loss.
Ridgway (personal communica-
tion, 1994); Sharp (1995)
288 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 57, NO. 3, FEBRUARY 5, 1998
starter cultures in all of the experiments. Selective phenol-
kanamycin agar plates of all plasmid-bearing cultures were
maintained and restreaked every week, as were phenol-
kanamycin slants, which were restreaked from frozen cul-
ture every month. All batch and continuous culture experi-
ments were inoculated with a culture of 17616 (TOM31c)
cells harvested from highly selective starter cultures (phenol
HCMM2 medium with 80 mg/mL of Km). In addition, all
continuous flow experiments began with 80 mg/mL of kana-
mycin in the initial reactor volume to insure each experi-
ment started with a pure plasmid-bearing culture.
Media
Three different media were used in this research: (1) rich
general growth medium—Luria broth glucose (LBG; 10 g/L
tryptone, 5 g/L yeast extract, 5 g/L NaCl, 1 g/L glucose); (2)
nonselective medium, hydrocarbon minimal medium
(HCMM2) amended with the nonselective carbon source,
acetate (0.5 to 40 mM sodium acetate); and (3) selective
medium, HCMM2 amended with phenol (2 mM) and kana-
mycin (50 mg/mL). Acetate was chosen as the nonselective
growth substrate because both 17616 (TOM31c) and PR131
(TOM31c) grow well on acetate, it is easily analyzed using
ion chromatography, and it does not competitively inhibit
TCE co-oxidation. All nonselective agar plates were made
with 15 g/L bacto-agar (Difco), whereas all plasmid-
selective plates using phenol were made with low carbon
Noble agar (Difco) to ensure that phenol was the sole
growth substrate.
TOM Specific Activity Analyses
The m-trifluoromethylphenol (TFMP) assay was used to
indicate the expression and activity of both toluene ortho-
monooxygenase (TOM) and catechol 2,3-dioxygenase
(C230) (Shields et al., 1989), which make up the TOM
pathway. The TFMP assay is a colorimetric assay, which
involves the transformation of TFMP to TFM-catechol
(TFMC) via TOM, followed by the conversion of TFMC to
7,7,7-trifluoro-2-hydroxy-6-oxo-2,4-heptadienoic acid
(TFHA) via C230. TFHA is a bright yellow product (ad-
sorption maximum 4 386 nm) that is quantified using color
spectrometry (A386) with a relatively high extinction coef-
ficient (E386 4 26,900) (Engesser et al., 1988).
The TFMP assay is used to determine the presence of
TOM31c through the activities of two TOM-encoded en-
zymes. The TFMP assay was performed using two methods:
(1) a colony assay used to make a positive/negative deter-
mination of TOM expression in single colonies (Shields et
al., 1991); and (2) a suspended culture assay used to quan-
tify the specific activity of the TOM pathway in suspended
cultures (Sharp, 1995).
Suspended culture TFMP assay results represent the ac-
tivity, or rate of TFHA production, of the TOM pathway
borne on the plasmid TOM31c per unit of bacterial culture.
These results are reported as either specific TOM activity of
the whole culture, referred to as ‘‘total TFMP activity’’
(milligrams TFHA produced/milligram total biomass per
minute), or as specific TOM activity relative to just the
plasmid-bearing fraction of a culture, referred to as ‘‘true
TFMP activity’’ (milligrams TFHA produced/milligram
plasmid-bearing biomass per minute).
TCE Degradation Rate Studies
TCE Disappearance Studies
Plasmid TOM31c activity and expression was determined
using batch reactor TCE disappearance assays. Assays were
carried out as per Folsom et al. (1990), using 25 mL of
17616 (TOM31c) culture harvested from a 2 mM phenol-
HCMM2 continuous culture. TCE degradation rate con-
stants for relevant TCE concentrations were determined for
17616 (TOM31c) by running a series of TCE disappearance
assays at different TCE concentrations (5 to 60 mM in so-
lution). The initial specific TCE degradation rates (nano-
moles TCE utilized per minute per milligram protein) were
determined at several discrete TCE concentrations and plot-
ted as specific TCE degradation rate versus TCE concen-
tration. The data were assumed to follow Michaelis–Menten
kinetics. A Henry’s law constant of 0.38 (± 0.03) [mol TCE
vapor phase]/[mol TCE solution phase] was determined and
used for the 17616 (TOM31c) batch cultures. Appropriate
cell-free and inactivated cell controls, and TCE calibration
curves were used for each experiment and in determining
the dimensionless Henry’s law constant.
TCE Analysis
TCE analysis was performed using a Shimadzu Gas Chro-
matograph (Model GC-9a) equipped with a Shimadzu elec-
tron capture detector (ECD) and a 30-m, 0.53-mm-i.d. Vo-
col megabore column (Supelco, Inc). The analytical method
was isothermal (110°C oven temperature, 200°C detector
and injector temperature) with a carrier gas flow of 5 mL/
min and a make-up gas flow of 50 mL/min. Ultrahigh-purity
nitrogen was used as both the carrier and make-up gas.
Vapor phase TCE samples were injected directly onto the
column and liquid samples were extracted with pentane and
then analyzed.
Acetate Growth Kinetics
The metabolic demand of maintaining the plasmid TOM31c
was determined by comparing acetate growth kinetics for
both plasmid-free and plasmid-bearing 17616 cultures. Ac-
etate growth kinetics (biomass and acetate concentrations
over time) were determined from batch reactor data ana-
lyzed using an initial rate method (D’Adamo et al., 1984),
where the initial acetate concentrations ranged from 0.5 to
50 mM diluted in HCMM2 medium. Each acetate batch
reactor study was run in triplicate with appropriate cell-free
SHARP ET AL.: PLASMID ENCODED ACTIVITY AND MAINTENANCE IN SUSPENDED CULTURES 289
and inactive-cell controls. Resultant rates versus initial ac-
etate concentrations were correlated using Andrews inhibi-
tion kinetics (Andrews et al., 1968). Kinetic constants from
this correlation were used both to evaluate plasmid meta-
bolic demand on the growth rate of transconjugant 17616
(TOM31c) and in the implementation of a plasmid loss
model. Acetate concentrations were determined with a
Dionex Ion Chromatagraph (Model AI-450; Dionex Co.,
San Francisco, CA) equipped with a pulse electrochemical
detector (Model DX300) using a 4-mm Ionpac AS10 col-
umn. Appropriate calibration curves were determined for
each set of acetate samples.
Plasmid Stability and Expression Studies in
Suspended Culture
Batch Reactor Studies
Experiments using the original plasmid host B. cepacia
PR131 (TOM31c) and the transconjugant B. cepacia 17616
(TOM31c) growing on nonselective acetate-HCMM2 me-
dium were carried out to compare levels of TOM expres-
sion, acetate growth characteristics, and TOM31c stability
between the two strains during batch growth. The batch
reactors used were custom made from 1000-mL graduated
glass beakers to provide a 500-mL total liquid volume. Re-
actors were continuously stirred at a rate of approximately
350 rpm and were aerated with 0.2 m filtered air at a rate of
approximately 300 mL/min. The pH was monitored during
each experiment and was found to range from 7.0 and 7.4.
A series of batch growth studies using acetate concentra-
tions ranging from 2 to 20 mM were carried out. Acetate
growth characteristics were determined by monitoring ac-
etate and biomass concentrations over the duration of com-
plete batch growth. Plasmid TOM31c expression and activity
were determined periodically during each batch experiment
using the TFMP-suspended culture assay and 24-h over-
night batch TCE disappearance assays.
Total, plasmid-free, and plasmid-bearing cell concentra-
tions were determined periodically throughout each batch
experiment to study the loss of plasmid TOM31c in the batch
17616 (TOM31c) cultures. Total cell numbers were deter-
mined by dilution plating on LBG agar plates. Ratios of
plasmid-free to total cell counts and plasmid-bearing to total
cell counts were determined using the plasmid TOM31c-
selective direct-colony transfer method (PSDCT method)
described in what follows.
Continuous Culture Studies
All continuous culture experiments used custom-made che-
mostats with 500-mL working liquid volume. The chemo-
stats were continuously stirred at a rate of approximately
350 rpm and were aerated with 0.2 mm of filtered air at a
rate of approximately 300 mL/min. Chemostats were
changed every 7 days to reduce the effects of wall growth.
Changes in pH during each continuous culture experiment
were found to be insignificant with an average pH of 7.1.
Acetate-HCMM2-fed chemostat studies were carried out to
determine the stability and activity of TOM31c in host 17616
during continuous culture. Chemostats were run at dilution
rates ranging from 0.05 to 0.19 h−1. Steady-state TOM spe-
cific enzyme activities were determined using the sus-
pended culture TFMP assay to determine if TOM expres-
sion was a function of growth rate. Plasmid-bearing, plas-
mid-free, and total cell counts were periodically determined
at three of the different dilution rates using the PSDCT
method to monitor plasmid loss during continuous growth
and to determine if plasmid loss was a function of growth
rate.
A series of TOM-selective, phenol-fed chemostat studies
were performed to determine if selective phenol growth
could be used to either stabilize or select TOM31c-bearing
cells in continuous culture. Phenol concentrations were de-
termined using a modified colorimetric phenol assay pre-
sented by Folsom et al. (1990).
Analytical Methods and Protocols
Protein Assay
Protein content of all strains used was determined using the
enhanced BA protein assay (Pierce Co.) using a Milton-Roy
Spectronic 601 Photospectrometer. Cell number versus pro-
tein content, as well as A600 versus protein content calibra-
tion curves were determined for each specific microbial
strain harvested from each growth medium. The protein
assay was used in all of the TFMP assays, TCE disappear-
ance assays, and suspended culture studies as a measure of
biomass concentration.
TOM31c-Selective Direct Colony Transfer
Method (PSDCT)
The PSDCT method was used to determine total cell counts
and the fraction (0.01 to 1.0) of plasmid-bearing and plas-
mid-free cells in a given culture (Sharp, 1995). The method
involved the transfer of colonies grown on nonselective me-
dium (LBG) to selective medium (phenol and phenol Km).
The transferred colonies were then checked for both growth
on the selective medium and TFMP activity (TOM activity).
The PSDCT method determined the presence of TOM by
indicating both the presence of the plasmid-borne kanamy-
cin resistance (Tn5 transposon) and the expression of TOM
and C230 (TFMP colony assay). In addition, the cells’ abil-
ity to grow on phenol is a direct result of the presence of
plasmid TOM31c, giving a third indication of the presence
and activity of TOM31c.
MATHEMATICAL MODELS
Single substrate Michaelis–Menten kinetics obtained from
sets of batch disappearance assays were used to model TCE
290 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 57, NO. 3, FEBRUARY 5, 1998
biodegradation by B. cepacia 17616 (TOM31c). Andrews
inhibition kinetics [Andrews et al., 1968; see eq. (1)] were
used to model acetate growth of both the plasmid-bearing
and plasmid-free strains of 17616. Acetate studies were car-
ried out to determine if the TOM31c caused a detectable
metabolic demand affecting the growth of 17616:
m =
mmax S
SKs + S + S2KiD
(1)
A model for plasmid loss in continuous growth cultures
was presented by Bailey and Ollis (1986). A mass balance
on a plasmid-bearing population in a chemostat accounts for
the rate of plasmid-bearing cell growth (mX+), the rate of
segregational plasmid loss (pmX+), and the rate of cell mass
leaving the continuous flow reactor via the effluent (−DX+).
A similar complementary balance on the plasmid-free cell
population (X−) can also be written. Both balances can be
applied to a system where plasmid maintenance does not
affect cell growth rate (m+ 4 m− 4 m). In this system, the
total biomass concentration is the sum of the plasmid-
bearing and plasmid-free populations in the system (X 4 X−
+ X+). For this system, Bailey and Ollis illustrate that the
total biomass concentration (X) will have a nonzero steady-
state value when the dilution rate is equal to the growth rate,
D 4 m. In a steady-state chemostat system, the value for
plasmid-bearing cell concentration (X+) at time t is repre-
sented by eq. (2):
lnX+(t) 4 lnX+(0) − pmt (2)
Each data set was correlated as per eq. (2), where the slope
of a linear regression on lnX+ versus time would yield
(p z m). With the assumption that growth rate is equal to
dilution rate (m 4 D) at steady state, an estimate of the
probability of plasmid loss (p) was made.
RESULTS
Growth and TCE Disappearance Studies
Overnight (24-h) TCE disappearance and TFMP specific
activity assays showed that TOM31c expression in the trans-
conjugant is equivalent to that of the TOM31c in its original
host (Fig. 2). In addition, the plasmid-free strains demon-
strated no TCE degradative ability nor any TFMP activity.
Results from a series of batch TCE disappearance assays
show that specific TCE degrading activity of transconjugant
17616 (TOM31c) follows simple single-substrate Michaelis–
Menten saturation kinetics. Figure 3 shows that the satura-
tion model correlated well with the kinetic data collected.
From this correlation, the maximum TCE specific activity
(Vmax) was found to be 9.7 nmol TCE/min z mg protein and
the corresponding half-saturation (Km) constant was 5.4 mM
TCE.
Batch acetate growth studies show that both the plasmid-
bearing and plasmid-free 17616 strains follow Andrews in-
hibition kinetics when grown on acetate (Fig. 4). Figure 4
illustrates that there is no significant difference between the
growth kinetics of 17616 and 17616 (TOM31). Resulting
acetate growth kinetic model parameters for each strain
were mmax 4 0.49 h−1, Ks 4 3.5 mM acetate, and Ki 4 9.4
mM acetate. Using these parameters, the highest obtainable
inhibited growth rates for both the plasmid-bearing and
plasmid-free cultures was found to be 0.22 h−1.
Suspended Culture Plasmid Stability and
Expression Results
Batch
Batch reactor plasmid loss studies showed that there was no
measurable plasmid loss during batch growth of either the
Figure 2. TCE removal by TOM31c-free and TOM31c-bearing 17616 and
PR131 strains. Solid bars indicate the initial TCE concentration; empty bars
indicate the TCE concentration after 24 h of shaking/incubation at 25°C.
Figure 3. Monod kinetics for TCE mineralization by 17616 (TOM31c),
where Vmax 4 9.7 nmol TCE/mg protein z min and Km 4 5.4 mM TCE.
Experiment was carried out in aerated HCMM2 medium at a pH of 7.2 and
a temperature of ∼25°C. Nonlinear curve fit (solid line) of R2 4 0.94 and
95% confidence intervals (dashed lines) are shown.
SHARP ET AL.: PLASMID ENCODED ACTIVITY AND MAINTENANCE IN SUSPENDED CULTURES 291
17616 (TOM31c) or PR131 (TOM31c). This could be inter-
preted in two ways: (1) plasmid loss was not occurring in
batch growth; or (2) the PSDCT method was unable to
measure the plasmid loss in the finite number of cell divi-
sions represented in the batch growth studies. These results
suggest that plasmid loss in 17616 (TOM31c) would be best
studied under continuous culture conditions so that signifi-
cant plasmid loss and/or loss of phenotype could be mea-
sured over many generations and under steady-state growth
conditions.
Chemostat
Results from the nonselective continuous culture experi-
ments showed that, after an initial lag phase (∼10 genera-
tions), there was a significant loss of TOM31c in 17616
(TOM31c) cultures at all three growth rates. Figure 5 shows
a plot of plasmid-bearing cell fractions and the total and true
TFMP specific activities of the continuous cultures at three
different dilution rates. The initial lag in measured plasmid
loss was attributed to both the lack of sensitivity in the
PSDCT method and the high kanamycin concentration (80
mg/mL) used to insure a pure plasmid-bearing culture in the
initial chemostat reactor volume. It is believed that the high
kanamycin concentration in the initial reactor volume re-
sulted in the killing or inactivation of plasmid-free cells,
thus not allowing them to persist in the reactor at the initial
stages of each experiment.
Figure 5 shows the different magnitudes of TFMP spe-
cific activity for each of the three growth rates for which
plasmid loss was measured. It should be noted that all of the
colonies that did grow on phenol also proved to be kana-
mycin resistant and TFMP positive. In addition, all of the
colonies that did not grow on phenol tested TFMP negative
and were not resistant to kanamycin. These results indicate
that, when one plasmid-borne phenotype was lost, all of the
tested phenotypes were lost, suggesting segregational plas-
mid instability. Figure 6 shows the linear relationship be-
tween growth rate and the true TFMP specific activity of the
continuous cultures. Plasmid-loss factors (p) were deter-
mined at three different growth rates (m 4 0.065, 0.10,
and 0.17) using eq. (2). These results are summarized in
Table II.
Although a plasmid-loss factor was not determined for
substrates other than acetate, significant plasmid loss was
also noted in both nonselective LBG- and phthalate-fed con-
tinuous cultures. No plasmid loss was measured in selective
phenol continuous cultures of 17616 (TOM31c). In addition,
no measurable plasmid loss was observed in any nonselec-
tive parallel experiments run with the original plasmid host
PR131 (TOM31c).Figure 4. Andrews substrate inhibition growth kinetics for 17616
(TOM31c) growing on nonselective. noncompetitive acetate medium,
where mmax 4 0.49/h, Ks 4 3.5 mM acetate, and Ki 4 9.1 mM acetate.
Andrews model fit (solid line) with R2 4 0.956 and 95% confidence
intervals (dashed lines) shown.
Figure 5. Plasmid-bearing cell fractions, total TFMP activities, and true
TFMP activities versus number of generations for acetate-fed continuous
17616 (TOM31c) cultures, where dilution rates (D) 4 0.17, 0.1, and 0.065/
h. Note the lag in measurable plasmid loss and the relative stability of true
TFMP activities.
292 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 57, NO. 3, FEBRUARY 5, 1998
DISCUSSION
Results from the batch reactor studies indicate that B. ce-
pacia 17616 harbored TOM31c and constitutively expressed
the TOM pathway. This is evident from results of TCE
disappearance assays (Figs. 2 and 3) and TFMP assays. The
TCE mineralization rates for 17616 (TOM31c) shown in
Figure 3 are equivalent to or slightly higher than those
found for the original plasmid host B. cepacia G4 under
similar growth conditions (Folsom et al., 1990). These re-
sults indicate that the 17616 (TOM31c) transconjugant has
successfully incorporated the plasmid and expressed the
plasmid-borne TOM pathway at levels equivalent to those
of the original plasmid host.
Results from acetate growth kinetics for both the plas-
mid-bearing and plasmid-free 17616 cultures (Fig. 4) show
that plasmid maintenance had little or no significant meta-
bolic demand on the host and did not result in a growth rate
advantage for the plasmid-free cells. If maintenance and
expression of the plasmid were to have a significant meta-
bolic demand on 17616 (TOM31c) cells (m+ < m−), the plas-
mid-free cells would have a growth rate advantage and
would eventually overtake a culture, resulting in an inevi-
table loss in the desired TCE degradative phenotype. Such
a loss would lead to inadequate TCE degradation efficiency
and obvious process control problems.
Results from continuous culture experiments show that
expression and subsequent activity of TOM31c-encoded en-
zymes in 17616 is a linear function of growth rate, which
can be described by eq. (3):
True TFMP specific activity (nmol TFHA/mg
protein z min) 4 2.1 + 2382 z [m(min−1)] (3)
Figure 6 demonstrates the linear relationship between
TOM specific activity and growth rate. It can be seen
that, at zero growth rate, resting cells had a residual
specific activity of 2.1 mmol TFHA/mg protein z min. The
residual activity illustrates the continuous expression of
TOM31c under extremely low nutrient conditions. Because
the maximum specific activity of TOM31c changes with
growth rate, so should the maximum TCE degradation
specific activity (Vmax). The effect of growth rate on
specific activity indicates that, to attain an efficient TCE
degradative TOM31c-bearing culture, whether it be in a
reactor or an in situ remediation system, there must be
ample growth substrate and nutrients available. However,
modest rates of TCE degradation can be achieved at
minimal growth, as is illustrated by the residual activity
of the resting cells.
Continuous culture plasmid-loss shown in Figure 5 illus-
trates that plasmid loss is a function of growth rate. The
dichotomy between the effects of growth rate on TOM31c
expression (as measured by enzyme specific activity), and
the effects of growth rate on plasmid loss demonstrates the
need to maintain a B. cepacia 17616–TOM31c culture at a
relatively low growth rate to balance plasmid loss and TCE
degradative activity. In addition, the apparent plasmid loss
in continuous culture indicates that some form of selection
for plasmid-bearing cells must be applied to the culture
periodically to maintain a dominant and active TCE degrad-
ing population. Bias toward the plasmid-bearing population
could be achieved through either antibiotic selection pres-
sures using periodic doses of kanamycin to kill the plasmid-
free population or the intermittent use of phenol or toluene
as growth substrates, which only the plasmid-bearing cells
can utilize.
Use of kanamycin is not economically feasible, but the
intermittent use of selective carbon sources could be effec-
tive, especially in continuous culture reactors and high rate
biofilters. Yet, only periodic use of selective carbon sources
would be feasible, due to interactions with the plasmid-
selective carbon sources and TCE, which result in competi-
tive inhibition of TCE oxidation (Folsom et al., 1990).
Table II. Results from continuous culture plasmid retention and activity experiments: Using 17616
(TOM31c) grown under nonselective conditions (acetate-HCMM2) and the plasmid loss model pro-
posed by Ollis (1982).
Growth rate/
dilution rate
(h−1)
Initial biomass
Xo (cells/mL)
Plasmid loss
factor (p)
True TFMP
(spec. act.)
Model
fit (R2)
0.065 3.15 × 108 0.027 4.1 0.94
0.1 4.55 × 107 0.031 6.9 0.97
0.17 1.02 × 106 0.0335 8.25 0.95
Figure 6. Linear relationship between growth rate and TOM31c specific
enzyme activity, where slope, m 4 −27.8, and y-axis intercept, b 4 2.1
mmol TFHA/mg protein z min (R2 4 0.96).
SHARP ET AL.: PLASMID ENCODED ACTIVITY AND MAINTENANCE IN SUSPENDED CULTURES 293
Results from additional continuous culture experiments
using other nonselective growth substrates (LBG and
phthalate) also exhibited considerable plasmid loss (Sharp,
1995), indicating that plasmid loss is not an artifact of ac-
etate growth but a consequence of either continuous culture,
which has been suggested by Dwilvedi et al. (1982) and
Primrose et al. (1984), or a process inherent in plasmid/host
systems such as the one described here. A method for de-
termining whether or not the plasmid instability described in
this study is an artifact of continuous culture is to perform
the same types of studies on biofilm cultures of the trans-
conjugant 17616 (TOM31c). This subject is addressed in a
subsequent article (Sharp et al., 1997).
Our results show that, even though a plasmid may be
initially incorporated into a desired population, its mainte-
nance and activity may incur instabilities and inconsisten-
cies depending on growth rate, selection parameters, and the
relationship/interactions between the specific plasmid/host
system.
CONCLUSIONS
The incorporation and expression of the TOM pathway in
the transconjugant Burkholderia cepacia 17616 (TOM31c)
was demonstrated. B. cepacia 17616 (TOM31c) was capable
of degrading TCE at rates equivalent to those of the original
TOM host (the phenol induced B. cepacia G4). Enzyme
activity encoded by TOM31c was found to be a linear func-
tion of growth rate, with a basal activity in the resting cells
that results from the continuous, noninduced expression of
TOM31c. Degradative activity was maintained in plasmid-
bearing cells during long-term growth in continuous sus-
pended cultures.
Plasmid loss in 17616 (TOM31c) determined using a con-
tinuous suspended culture mathematical model was consid-
erable, resulting in an order-of-magnitude decrease in plas-
mid-bearing cells over 60 to 120 generations under nonse-
lective growth conditions. This degree of plasmid loss
would result in the critical loss of the TCE degradative
phenotype in the culture, which would have a profoundly
negative effect on the performance of TCE degrading bio-
remediation technology. Results show a need for plasmid
selection and/or process control methods aimed at reducing
plasmid loss while enhancing degradative activity. Such
methods could lead to the long-term maintenance of an
effective TCE degrading population of 17616 (TOM31c) to
be used in bioremediation technologies. Our studies illus-
trate a need to examine plasmid loss in environmentally
relevant plasmid/host systems to determine the plasmid/host
interactions that affect the performance and efficiency of
bioremediation technologies. In addition, the study of these
plasmid/host interactions in biofilm cultures is also sug-
gested, due to the possibility of greater plasmid stability
among biofilm populations and the abundance of biofilm
processes associated with both in situ and reactor-based
bioremediation technologies.
We thank John Newman at the Center for Biofilm Engineering
for his analytical expertise and Jayne Billmayer at the Center for
Biofilm Engineering for her assistance in the laboratory.
NOMENCLATURE
m specific growth rate (time−1)
m+ specific growth rate of plasmid-bearing cells (time−1)
m− specific growth rate of plasmid-free cells (time−1)
Vmax Michaelis–Menten maximum specific activity (time−1)
mmax maximum specific growth rate (time−1)
Ks Monod half-saturation constant (mM or mg/L)
Km Michaelis–Menten half-saturation constant (mM or mg/L)
Ki Andrews substrate inhibition half-saturation constant (mM or mg/
L)
S concentration of growth substrate (mM or mg/L)
X+ plasmid-bearing cell concentration (cells/mL)
X− plasmid-free cell concentration (cells/mL)
t time (h or days)
p plasmid loss factor
D dilution rate (time−1)
References
Alvarez-Cohen, L., McCarty, P. L. 1991a. Product toxicity and cometa-
bolic competitive inhibition modeling of chloroform and trichloroeth-
ylene transformation by methanotropic resting cells. Appl. Environ.
Microbiol. 57: 1031–1037.
Alvarez-Cohen, L., McCarty, P. L. 1991b. Effects of toxicity, aeration, and
reductant supply on trichlorethylene transformation by a mixed metha-
notrophic culture. Appl. Environ. Microbiol. 57:228–235.
Andrews, J. F. 1968. A mathematical model for the continuous culture of
microorganisms utilizing inhibitory substrates. Biotechnol. Bioeng.
10: 707–723.
Arciero, D. T., Vannelli, ., Logan, M., Hooper, A. B. 1989. Degradation
of trichloroethylene by ammonia-oxidizing bacterium Nitrosomonas
europea. Biochem. Biophys. Res. Commun. 159: 640–643.
Bailey, J. E., Ollis, D. E. 1986. Biochemical engineering fundamentals.
2nd edition. McGraw-Hill, New York.
Bouwer, E. J., McCarty, P. L. 1983. Transformations of 1- and 2-carbon
halogenated aliphatic compounds under methanogenic conditions.
Appl. Environ. Microbiol. 45: 1286–1294.
D’Adamo, P. D., Rozich, A. F., Gaudy, A. F., Jr. 1984. Analysis of growth
data with inhibitory carbon sources. Biotechnol. Bioeng. 26: 397–402.
de Taxis du Poe¨t, P., Arcand, Y., Bernier, R., Barbotin, J.-N., Thomas, D.
1987. Plasmid stability in immobilized and free recombinant E. coli
JM105(pKK223-200): importance of oxygen diffusion, growth rate,
and plasmid copy number. Appl. Environ. Microbiol. 53: 1548–1555.
Cheng, H.-P., Lessie, T. G. 1994. Multiple replicons constituting the ge-
nome of Pseudomonas cepacia 17616. J. Bacteriol. 176: 4034–4042.
Dwilvedi, C. P., Imanaka, T., Aiba, S. 1982. Instability of plasmid harbor-
ing strain E. coli in continuous culture. Biotechnol. Bioeng. 24:
1465–1468.
Dykhuizen, D. E., Hartl, D. L. 1983. Selection in chemostats. Microbiol.
Rev. 47: 150–168.
Engesser, K. H., Cain, R. B., Knackmuss, H. J. 1988. Bacterial metabolism
of side chain fluorinated aromatics: cometabolism of 3-trifluoromethyl
(T.M.)-benzoate by Pseudomonas putida (arvilla) mt-2 and Rhodococ-
cus rubropertinctus. N657-Arch. Microbiol. 149:188–197.
Ensley, B. D. 1986. Stability of recombinant plasmids in industrial micro-
organisms. CRC Crit. Rev. Biotechnol. 4: 263–277.
Ewers, J., Frier-Schroder, D., Knackmuss, H. J. 1990. Selection of trichlo-
294 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 57, NO. 3, FEBRUARY 5, 1998
roethylene degrading bacteria that resist inactivation by TCE. Arch.
Microbiol. 154: 410–413.
Finette, B. A., Subramanian, V., Gibson, D. T. 1984. Isolation and char-
acterization of Pseudomonas putida PpF1 mutants defective in the
toluene dioxygenase enzyme system. J. Bacteriol. 160: 1003–1009.
Fogel, M. M., Taddeo, A. R., Fogel, S. 1986. Biodegradation of chlorinated
ethenes by a methane-utilizing mixed culture. Appl. Environ. Micro-
biol. 51: 720–724.
Folsom, B. R., Chapman, P. J., Pritchard, P.H. 1990. Phenol and trichlo-
roethylene degradation by Pseudomonas cepacia G4: kinetics and in-
teractions between substrates. Appl. Environ. Microbiol. 56:
1279–1285.
Freedman, D. L., Gossett, J. M. 1989. Biological reductive dechlorination
of tetrachloroethylene and trichloroethylene to ethylene under metha-
nogenic conditions. Appl. Environ. Microbiol. 55: 2144–2151.
Godwin, D., Slater, J. H. 1979. The influence of growth environment on
the stability of a drug resistant plasmid in E. coli K12. J. Gen. Micro-
biol. 111: 201–210.
Grandi, G., Mottes, M., Sagaramella, V. 1981. Specific pattern of insta-
bility of E. coli HisG gene cloned Bacillus subtilis via the Staphylo-
coccus aureus plasmid pCS194. Plasmid 6: 99–111.
Huang, C. T., Peretti, S. W., Bryers, J. D. 1993. Plasmid retention and gene
expression in suspended and biofilm cultures of recombinant E. coli
DH5a (pMJR1750). Biotechnol. Bioeng. 41: 211–220.
Huang, C. T., Peretti, S. W., Bryers, J. D. 1994. Effects of carbon–nitrogen
ratio on biofilm formation and plasmid retention. Biotechnol. Bioeng.
44: 329–336.
Inloes, D. S., Smith, W. J., Taylor, D. P., Cohen, S. N., Michaels, A. S.,
Robinson, C. R. 1983. Hollow fiber membrane bioreactors using im-
mobilized E. coli for protein synthesis. Biotechnol. Bioeng. 25:
2653–2681.
Jones, S. A., Melling, J. 1984. Persistence of pBR322-related plasmids in
E. coli grown in chemostat cultures. FEMS Microbiol. Lett. 22:
239–243.
Jones, I. M., Primrose, S. B., Ellwood, D. C. 1980. Maintenance of some
ColE1-type plasmid in continuous culture. Mol. Gen. Genet. 180:
579–584.
Kadam, K. L., Wollweber, K. I., Grosch, J. C., Jao, Y.C. 1987. Investiga-
tion of plasmid stability in amylase-producing B. subtilis using con-
tinuous culture. Biotechnol. Bioeng. 29: 859–872.
Kaphammer, B., Kukor, J. J., Olsen, R. H. 1990. Cloning and character-
ization of a novel toluene degradative pathway from Pseudomonas
picketti PKO1 (abstr.K-145), p. 243. In: Abstracts of the 90th annual
meeting of the American Society of Microbiology, 1990. American
Society for Microbiology, Washington, DC.
Kleopfer, R. D., Easley, D. M., Haas, B. B., Jr., Deihl, T. G. 1985. An-
aerobic degradation of trichloroethylene in soil. Environ. Sci. Technol.
19: 277–280.
Kumar, P. K., Maschke, H. E., Friechs, K., Schugerl, K. 1991. Strategies
for improving plasmid stability in genetically modified bacteria in
bioreactors. Trends Biotechnol. 9: 279–284.
Kumar, P. K., Schugerl, K. 1990. Immobilization of genetically engineered
cells: a new strategy for higher plasmid stability. J. Biotechnol. 14:
255–272.
Lauffenburger, D. A. 1987. Bacterioicin production as a method of main-
taining plasmid-bearing cells in continuous culture. Trends Biotech-
nol. 5: 87.
Noack, D., Roth, M., Geuther, R., Muller, G., Undisz, K., Hoffmeier, C.,
Gaspar, S. 1982. Maintenance and genetic stability of vector plasmids
pBR322 and pBR325 in E. coli K12 strains grown in a chemostat. Mol.
Gen. Genet. 184: 121–124.
Nelson, M. J. K., Montgomery S. O., O’Neill, E. J., Pritchard, P. H. 1986.
Aerobic metabolism of TCE by a bacterial isolate. Appl. Environ.
Microbiol. 52: 383–384.
Nelson, M. J. K., Montgomery, S. O., Mahaffey, W. R., Pritchard, P. H.
1987. Biodegradation of trichloroethylene and involvement of an aro-
matic biodegradative pathway. Appl. Environ. Microbiol. 53:
949–954.
Nelson, M. J. K., Montgomery, S. O., Pritchard, P. H. 1988. Trichloroeth-
ylene metabolism by microorganisms that degrade aromatic com-
pounds. Appl. Environ. Microbiol. 54: 604–606.
Oldenhuis, R., Vink, R. L., Janssen, D. B., Witholt, B. 1989. Degradation
of chlorinated aliphatic hydrocarbons by Methylosinus trichosporium
OB3b expressing soluble methane mono-oxygenase. Appl. Environ.
Microbiol. 55: 2819–2826.
Ollis, D. F. 1982. Industrial fermentations with unstable recombinant cul-
tures. Phil. Trans. R. Soc. Lond. B 297: 617–629.
Peretti, S. W., Bailey, J. E. 1987. Simulations of host–plasmid interactions
in E. coli: copy number, promoter strength, and ribosome binding site
strength effects on metabolic activity and plasmid gene expression.
Biotechnol. Bioeng. 29: 316–325.
Primrose, S. B., Derbyshire, P., Jones, I. M., Robinson, A. 1984. The ap-
plication of continuous culture to the study of plasmid stability. In:
A. C. R. Dean, D. C. Ellwood, and C. G. T. Evans (eds.), Continuous
culture 8. Society of Chemical Industry/Ellis Horwood, West Sussex,
UK.
Roth, M., Muller, G., Noack, D. 1980. Loss of the multi-copy resistance
plasmid pBR325 from E. coli GY2354 pBR325 during continuous
cultivation, pp. 143–146. In: S. Mitsuhasi, L. Rosival, and V. Krec-
mery (eds.), Antibiotic resistance: transposition and other mechanisms.
Springer, New York.
Roth, M., Noack, D. 1982. Genetic stability of differentiated functions in
Streptomyces hygroscopicus in relation to conditions of continuous
culture. J. Gen. Microbiol. 128: 107–114.
Sayadi, S., Nasri, M., Barbotin, J. N., Thomas, D. 1989. Effect of envi-
ronmental growth conditions on plasmid stability, plasmid copy num-
ber, and C230 activity in free and immobilized E. coli cells. Biotech-
nol. Bioeng. 33: 801–808.
Seo, J. H., Bailey, J. E. 1985. Effects of recombinant plasmid content on
growth properties and cloned gene product formation in E. coli. Bio-
technol. Bioeng. 27: 1668–1674.
Sharp, R. R. 1995. Stability and expression of a plasmid-borne TCE deg-
radative pathway in suspended and biofilm cultures. Ph.D. dissertation,
Montana State University, Bozeman, MT, USA.
Sharp, R. R., Bryers, J. D., Jones, W. L. 1997. Retention and expression of
a plasmid-borne TCE degradative pathway in biofilm cultures of the
transconjugant Pseudomonas cepacia 17616-TOM31c. Biotechnol.
Bioeng. (in press).
Shields, M. S., Montgomery, S. O., Chapman, P. J., Cuskey, S. M., Pritch-
ard, P. H. 1989. Novel pathway of toluene catabolism in the TCE
degrading bacterium G4. Appl. Environ. Microbiol. 55: 1624–1629.
Shields, M. S., Montgomery, S. O., Cuskey, S. M., Chapman, P. J., Cus-
key, S. M., Pritchard, P. H. 1991. Mutants of Pseudomonas cepacia
G4 defective in catabolism of aromatic compounds and trichloroeth-
ylene. Appl. Environ. Microbiol. 57: 1935–1941.
Shields, M. S., Reagin, M. J. 1992. Selection of a Pseudomonas cepacia
strain constitutive for degradation of trichloroethylene. Appl. Environ.
Microbiol. 58: 3977–3983.
Shields, M. S., Reagin, M. J., Gerger, R., Campbell, R., Sommerville, C.
1995. TOM, a new aromatic degradative plasmid from Burkholderia
(Pseudomonas) cepacia G4. Appl. Environ. Microbiol. 61:
1352–1356.
Stewart, G. J., Carlson, C. A. 1986. The biology of natural transformations.
Annu. Rev. Microbiol. 40: 211–235.
Tiedje, J. M., Colwell, R. K., Grossman, Y. L., Hodson, R. E., Lenski,
R. E., Mack, R. N., Regal, P. J. 1989. The planned introduction of
GEOs: ecological considerations and recommendations. Ecology 70:
298–315.
Tsien, H. C., Brusseau, G. A., Hanson, R. S., Wackett, L. P. 1989. Biodeg-
radation of TCE by Methylosinus trichosporium. Appl. Environ. Mi-
crobiol. 55: 3155–3161.
Uhlin, B. E., Nordstrom, K. 1978. A runaway-replication mutant of plas-
mid R1drd-19: temperature-dependent loss of copy number control.
Mol. Gen. Genet. 165: 167–179.
SHARP ET AL.: PLASMID ENCODED ACTIVITY AND MAINTENANCE IN SUSPENDED CULTURES 295
Vandenbergh, P. A., Kunka, B. S. 1988. Metabolism of volatile chlorinated
aliphatic hydrocarbons by Pseudomonas fluorescens. Appl. Environ.
Microbiol. 54: 2578–2579.
Vogel, T. M., McCarty, P. 1985. Biotransformation of tetrachloroethylene
to TCE, DCE, VC, and carbon dioxide under methanogenic conditions.
Appl. Environ. Microbiol. 49: 1080–1083.
Wackett, L. P., Brusseau, G. A., Householder, S. R., Hanson, R. S. 1989. A
survey of microbial oxygenases: trichlorethylene degradation by pro-
pane-oxidizing bacteria. Appl. Environ. Microbiol. 55: 2960–2964.
Wackett, L. P., Gibson, D. T. 1988. Degradation of TCE by toluene di-
oxygenase in whole-cell studies with Pseudomonas putida F1. Appl.
Environ. Microbiol. 54: 1703–1708.
Winter, R. B., Yen, K. M., Ensley, B. D. 1989. Efficient degradation of
trichloroethylene by a recombinant Escherichia coli. Bio/Technology
7: 282–285.
Wood, T. K., Kuhn, R. H., Peretti, S. W. 1990. Enhanced plasmid stability
through post-segregational killing plasmid-free cells. Biotechnol.
Technic. 4: 39–44.
296 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 57, NO. 3, FEBRUARY 5, 1998