See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7422643

Growth Kinetics of Thiobacillus ferrooxidans Isolated from Arsenic Mine

Drainage

Article  in  Applied and Environmental Microbiology · August 1984

Source: PubMed

CITATIONS READS
52 63

3 authors, including:

Joan Braddock Hoang Van Luong

University of Alaska Fairbanks Vietnam Military Medical University
50 PUBLICATIONS   1,969 CITATIONS    21 PUBLICATIONS   317 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Joan Braddock on 22 May 2014.

The user has requested enhancement of the downloaded file.

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, JUIY 1984, p. 48-55 Vol. 48, No. 1
0099-2240/84/070048-08$02.00/0
Copyright © 1984, American Society for Microbiology

Growth Kinetics of Thiobacillus ferrooxidans Isolated from Arsenic

Mine Drainage
JOAN FORSHAUG BRADDOCK,* HUAN V. LUONG, AND EDWARD J. BROWN
Institute of Water ResourceslEngineering Experiment Station, University of Alaska, Fairbanks, Alaska 99701
Received 16 January 1984/Accepted 2 April 1984

Thiobacillus ferrooxidans is found in many Alaskan and Canadian drainages contaminated by metals
dissolved from placer and lode gold mines. We have examined the iron-limited growth and iron oxidation
kinetics of a T. ferrooxidans isolate, AKI, by using batch and continuous cultures. Strain AKI is an arsenic-
tolerant isolate obtained from placer gold mine drainage containing large amounts of dissolved arsenic. The
steady-state growth kinetics are described with equations modified for threshold ferrous iron concentrations.
The maximal specific growth rate (limax) for isolate AK1 at 22.5°C was 0.070 h-1, and the ferrous iron
concentration at which the half-maximal growth rate occurred (K,) was 0.78 mM. Cell yields varied inversely
with growth rate. The iron oxidation kinetics of this organism were dependent on biomass. We found no
evidence of ferric inhibition of ferrous iron oxidation for ferrous iron concentrations between 9.0 and 23.3 mM.
A supplement to the ferrous medium of 2.67 mM sodium arsenite did not result in an increased steady-state
biomass, nor did it appear to affect the steady-state growth kinetics observed in continuous cultures.
The bacterium ThiobacillusJerrooxidans is responsible for growth and iron oxidation kinetics of isolate AK1 appear to
metal leaching and acid production in water draining from deviate from traditional kinetic models. There was no appar-
sulfide deposits. T. ferrooxidan.s can catalyze the dissolution ent effect on the growth or iron oxidation kinetics of this
of metals either by directly attacking sulfide ores or by isolate when the initial ferrous iron concentration was in-
oxidizing reduced iron leading to indirect dissolution via the creased from 9.0 to 23.3 mM. Additionally, the presence of
ferric ion (3). Sulfides and metals (such as arsenic) are reduced arsenic in the feed ferrous iron did not affect the
commonly associated with lode gold deposits and are some- kinetics of this arsenic-tolerant isolate.
times associated with placer deposits (46). Recently, T.
ferrooxidans has been isolated from streams in Alaska and MATERIALS AND METHODS
northern Canada that are affected by mining. Microorganisms. T. ferrooxidans AK1 was originally iso-
Brown et al. (6) found large numbers of T. Jerrooxidans in lated from Eva Creek, located near Fairbanks, Alaska, and
drainage from 90% of the placer mines that they sampled. is the same isolate as strain TF133 described by Martin et al.
Concentrations of dissolved arsenic >10 parts per billion (10 (34). Eva Creek is heavily affected by placer gold mining and
,uig/liter) were found in 30% of those same streams. T. is known to contain large amounts of dissolved arsenic (6).
ferrooxidans and arsenic were also present in streams affect- Growth medium. The growth medium was modified from
ed by lode mining. In addition, T. Jerrooxidans will leach that described by Tuovinen and Kelly (45) and contained
substantial amounts of arsenic from placer gold mine materi- MgSO4 7H20 (0.4 g liter-1), (NH4)2SO4 (0.4 g * liter-1),
al (H. V. Luong, J. F. Braddock, and E. J. Brown, Geomi- KH,PO4 (0.1 g liter- i), and 10 N H2S04 to bring the
crobiol. J., in press). These results suggest that T. Jerrooxi- medium to pH 1.85 to 1.95. Reduced iron as a sole energy
dans may be directly involved in the leaching of arsenic and source was supplied as FeSO4 7H,O, and concentrations
other metals associated with active and abandoned mines to varied in different experiments from 5 to 30 g * liter-1 for
subarctic streams and groundwaters. However, the mecha- batch cultures and 2.5 to 6.5 g liter-1 for continuous
nisms by which these organisms are involved in the cycling cultures. To eliminate iron precipitate formation, the compo-
of metals such as arsenic are not fully understood. nents of the medium were autoclaved in two separate
Many investigators have studied the microbiologically solutions and were aseptically mixed after the solutions
mediated dissolution kinetics of various sulfide minerals (10, cooled to room temperature. Solution A contained the
23, 26, 32, 42, 44). Although the growth kinetics of various FeSO4 and about 2.5 ml of 10 N H2SO4 per liter of final
thiobacilli when grown on reduced sulfur compounds have medium volume. Solution B contained the remaining salts
been described (2, 18, 25, 31), the iron-limited growth diluted in water. The water used throughout these experi-
kinetics of T. ferrooxidans are not well understood. The ments was American Society for Testing Standards Type I
partial kinetic descriptions that are found in the literature quality.
(21, 27-29, 33, 43) show inconsistent results. This may be In some continuous-culture experiments, 2.67 mM re-
due to the difficulties of growing this organism, particularly duced arsenic (supplied as NaAsO2) was also incorporated
in continuous culture, and to taxonomic uncertainty about into the culture medium. The arsenite was dissolved in water
isolated strains of acidophilic iron oxidizers (13). (pH 2) and was then aseptically filtered into the prepared
In this study, we used batch and continuous cultures to medium.
describe the iron-limited growth and iron oxidation kinetics Batch culture. For batch culture experiments, 5 to 10 ml of
of a T. ferrooxidans isolate (AKI) obtained from a stream exponentially growing cells was inoculated into 500-ml Er-
near Fairbanks, Alaska. This stream is affected by placer lenmeyer flasks containing iron medium. The flasks were
mining and contains large amounts of dissolved arsenic. The placed on a rotary shaker and were maintained at 22.5°C for
the duration of each experiment. Microbiological and chemi-
* Corresponding author. cal parameters were measured periodically in small samples
48
VOL. 48, 1984 GROWTH KINETICS OF T. FERROOXIDANS 49

vessel contents were placed in small test tubes containing

SAMPLING
MAGNETIC
~~~~PORT
STIRRER
PERISTALTIC
PUMP
WASTE
FIG. 1. Two-phase continuous-culture system used for cultiva-
tion of T. ferrooxidans.
aseptically removed from each flask. Triplicate flasks were
used so that a mean and standard error could be calculated
for the measured quantities at each sampling time.
Continuous culture. A two-phase continuous-culture appa-
ratus (Fig. 1) was designed to provide the optimum condi-
tions for iron oxidation by T. ferrooxidans. Culture medium
stored in a 20-liter glass carboy was pumped by a peristaltic
pump directly into a specially designed reactor vessel. The
reactor was adapted from a 1-liter boiling flask by adding an
effluent tube, atn air-sparging tube, and a sampling port. The
effluent port was placed to maintain the volume of liquid in
the flasks at 500 to 600 ml. The exact volume of liquid was
measured for each flask. Compressed air to supply oxygen
and carbon dioxide was added via the air port at a rate of 500
to 600 ml * min- The sampling port was sealed with a
. the graphite furnace atomic absorption method described by
silicone stopper through which a small piece of glass tubing
was inserted. A small valve connected the glass tubing to a
sterile syringe. We could rapidly and efficiently remove
sterile samples with this system.
A magnetic stirring device continually agitated the con-
tents of the reactor flask. The flasks were supported slightly
above the stirrer to avoid any effect on the culture vessel
from the heat generated by the motor of the stirrer. Silicone
stoppers sealed the reactor vessel and feed carboy. Sterile
cotton filters inserted in the feed carboy stopper and reactor
filtered water. Two drops of filtered 0.1% (in water) acridine
orange solution were then added to each tube to stain the
cells. The stained-cell suspension in each tube was then
filtered with a 3-ml syringe and a Swinnex filter apparatus
(Millipore Corp.) onto an irgalan black-stained (22) filter
(0.2-,um pore size, 25-mm diameter; Nuclepore Corp.). Fil-
tered water (three 2-ml portions) was used to rinse each tube
and the filter apparatus. The damp Nuclepore filter was then
placed, with a drop of immersion oil, under a cover slip on a
microscope slide and was held at 4°C in the dark for 30 min
to 1 h, after which 15 to 20 microscope fields were counted
per slide by using a Zeiss Standard microscope with an
epifluorescence light source. Duplicate or triplicate filters
were prepared and counted for each sample.
Ferrous iron. Ferrous iron concentrations in solution
(filtered through a Millipore type HA filter [0.22-Rm pore
size; 25-mm diameter]) and total ferrous iron concentrations
(nonfiltered) were determined by the ortho-phenanthroline
spectrophotometric method of the American Public Health
Association (1). Ferrous iron concentrations were monitored
in continuous-culture feed reservoir medium to assure chem-
ical stability for the duration of an experiment. Ferrous iron
was stable in the acid media for at least 6 months.
Arsenic. Arsenic concentrations were designated as ar-
senic in solution (filtered through a Millipore type HA filter
[0.22-iim pore size; 25-mm diameter]) or as total arsenic
(including arsenic in solution and associated with bacterial
cells).
Total arsenic (micrograms per liter range) was analyzed by
Wilson and Hawkins (46) as modified with a nickel matrix
(19) or (milligrams per liter range) by the flame atomic
absorption method with an Electrodeless Discharge Lamp at
193.7 nm as described for the Perkin-Elmer model 4000
atomic absorption spectrophotometer (Publication 0303-
0152, 1982, The Perkin-Elmer Corp., Norwalk, Conn.).
Ionic species of arsenic were determined in culture fil-
trates by treating 2.0 ml of sample with 0.2 ml of a modified
Murphy and Riley (36) molybdate reagent, containing 4 ml of
ammonium molybdate (30 g liter-'), 5 ml of 10 N H2SO4, 2
-

vessel stopper maintained ambient pressure throughout the ml of potassium antimony tartarate (1.36 g liter-'), and
system. Glass tubing and a minimal amount of silicone 0.108 g of ascorbic acid. The samples were incubated for ca.
tubing connected the feed carboy to the reactor vessel. 6 h, at which time the arsenate fraction was extracted with
The continuous cultures were maintained at 22 to 23°C in a
temperature-controlled room. At the start of an experiment,
5 to 10 ml of exponentially growing cells was added to 200 to TABLE 1. Continuous-culture parameters
300 ml of fresh culture medium. The peristaltic pump was
then started, the vessel was filled, and continuous cultiva- Symbol Parameter Units
tion was begun. Generally, a slow flow was used for several SR Influent limiting nutrient con- mM
days until organism growth was observed, at which time the centration
pump was adjusted to the desired flow rate. The flow rate for S Ambient limiting nutrient con- mM
each culture was measured several times to assure continu- centration (external limiting
ous and constant delivery at a given flow rate by each pump. nutrient concentration)
The dilution rate (D) was calculated as the flow rate divided St Threshold mM
Ycarbon Yield mg of C * mM
by the volume of the culture vessel. The value of D was Ycell Yield cells * mM-'
assumed to be equivalent to the specific growth rate (,u) of Specific growth rate h-'
the organism at steady state. After at least three residence Gross iron oxidation rate mM * h-'
times (R, where R = D-' or 7-1) had passed at a given flow I1max Maximum growth rate h-I'
rate, a steady state was assumed, and samples of the reactor Vmax Maximum iron oxidation rate mM h-'
-

vessel contents were removed for microbiological and chem- K,^ Growth half-saturation constant mM
ical analysis. KFe Iron oxidation half-saturation mM
Enumeration. Numbers of cells in laboratory cultures constant
were determined by epifluorescent direct-count microscopy 1'carbon Net iron oxidation rate mmol clmg of C h
by a method adapted from that described by Hobbie et al. 1'cell Cellular iron oxidation rate fmol cell' h'
as Steady-state affinity liters mg of C` h`
(22). Measured aliquots (0.1 to 0.5 ml) of flask or reactor
50 BRADDOCK, LUONG, AND BROWN APPL. ENVIRON. MICROBIOL.

1.4 0.06r

0.02 - 0.2 0.051-

9.0 0.15 '
2
120 X0.o 0.041-
X,1-. +N-
0 .c *n
0 0 E
o 0.03 1
at
a
7.0
+

_u
40 o o
6.0 E
o
*
5.0
Time, hr
FIG. 2. Ferrous iron in solution (A), log1l number of cells (0),
and carbon yield and cell yield (O) as functions of time for batch-
grown cells.
isoamyl alcohol. The aqueous (arsenite-containing) phase
was analyzed, as described above, by atomic absorption
spectrophotometry (4). Arsenate ion concentrations were
calculated as the differences between total arsenic and
aqueous phase concentrations.
Carbon. Organic carbon was measured by the wet-chemi-
cal method described for the Technicon Auto Analyzer II
(industrial method no. 451-76W/A, 1978, Technicon Instru-
ments Corp., Inc., Tarrytown, N.Y.). This method was
selected for its sensitivity (0.4 to 20.0 mg of C * liter-') and
small sample volume requirement (2 ml).
Kinetic measurements. A list of continuous-culture param-
eters is found in Table 1. The relationships among these
parameters were used to describe the growth and iron
oxidation kinetics of T. ferrooxidans in continuous culture.
Growth rates (pL) were plotted against residual ferrous iron
concentrations with a modified Monod model (8). The modi-
fied Monod relationship can be expressed as
I- = .max(S - St)[K + (S - St)' (1)
and can be linearized by the relationship
(S - St)([u)-. = Kp,(PmaxJ' + (S - St)(LmaxFW (2)
(see Table 1 for definition of symbols). This linearization was
selected rather than the more commonly used double-recip-
rocal plot since it is less biased by deviations at low substrate
concentrations (11) and, for these experiments, yields a
more accurate estimate of the maximum growth rate (P.max).
Z; /4 000
_ 0.02-
o
0.01 F
0.0 0.5 1.0 1.5 2.0 2.5
External ferrous iron, mM
FIG. 3. Steady-state growth rate as a function of external ferrous
iron concentration for T. ferrooxidans. Symbols represent influent
ferrous iron concentrations (mM) as follows: K, 9.0 to 9.5; *, 9.0 +
1.3 to 2.7 mM As3"; A, 17.9; 0, 22.4 to 23.3.
of cells were monitored over time are shown in Fig. 2. The
initial ferrous iron concentration in the culture medium was
108 mM in this experiment. These data can be used to
predict an apparent maximum specific growth rate, a maxi-
mum iron oxidation rate (Vmax), and growth yields for this
isolate.
The plot of logl0 numbers of cells over time can be used to
estimate the apparent P-max. The apparent P-max was calculat-
ed to be 0.089 h- l for strain AK1 in this experiment and in a
duplicate experiment (data not shown). Generation time (tg),
the time required for the population to double during expo-
nential growth, was about 7.7 h. The relationship between
ferrous iron in solution and time was used similarly to
estimate an apparent maximum iron oxidation rate of 3.34
mmol liter-' h-' for this isolate.
The organic-carbon and cell number data were used to
calculate growth yields, Ycarbon and Ycell, for strain AK1.
The Ycarbon ranged from 0.12 to 0.20 mg of carbon per mmol
of Fe2+ oxidized, whereas the Ycell demonstrated an increase
40
30
0
0

Alternatively, P-max was estimated with equations analogous E20

to thqse from the internal stores model (17, 40). 20
e iron oxidation rate (v) was analyzed as a function of the
residual ferrous iron concentration, specific growth rate, and cn
carbon yield (Ycarbon). The relationship between v and the cn K,, = O. 78mA
external ferrous iron concentration (S) was linearized as 10
described for the Monod model. Regression analysis was ltno =W 0. 070hr -/
used to estimate best-fit lines for all linear relationships.
Curvilinear relationships were drawn by hand to indicate
general trends in the data. The significance of the slope of 0
linear relationships at the 95% confidence level was deter- 0.0 0.5 1.0 1.5 2.0
mined with F tests. S-St, mM
RESULTS FIG. 4. Linearization of the Monod growth curve (threshold
corrected) for T. ferrooxidans. The equation for the line at the 95%
Batch culture. The results of a batch culture experiment in confidence level is y = (14.30 + 2.83)x + 11.12 ± 2.78 with r = 0.87.
which ferrous iron in solution, organic carbon, and numbers Symbols are the same as those in Fig. 3.
VOL. 48, 1984 GROWTH KINETICS OF T. FERROOXIDANS 51

1.2
A
. 0.5 I.. 7 1.0 a
x
0
e 6
*~04
.- OA" I- 2E
0
E 0
-

u 0.3 - ca 0.6
0
._.
0aax
la Q4 c
0.2
c
p
0.1

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.5 1.0 1.5 2.0 2.5

Growth rate, hr-1 External ferrous Iron, mM

FIG. 5. The steady-state yield of cellular carbon as a function of
growth rate for T. ferrooxidans. Symbols are the same as those in
Fig. 3.
over time from 0.25 109 to 1.20 109 cells per mmol of Fe2+
oxidized. In a duplicate experiment (data not shown), a
similar final Ycell value of 1.3 109 cells per mmol of Fe'2
oxidized was calculated. These Ycarbon and Ycell values
suggest that the amount of carbon per cell decreased over
time in batch-grown cells. At the stationary phase of growth
(near the end of the experiment), there was ca. 0.12 pg of
carbon per cell.
Continuous culture. (i) Growth kinetics. The Monod rela-
tionship between growth rate and external ferrous iron
concentration for isolate AKi is shown in Fig. 3. This
relationship describes the growth of this isolate at several
different influent ferrous iron concentrations (SR) and in the
presence and absence of reduced arsenic. An apparent
threshold ferrous iron concentration (St; intercept) of
about 0.25 mM was estimated from the results in Fig. 3. This
extrapolated threshold value represents ferrous iron unavail-
able to the organism for growth (8).
The apparent threshold value was applied to a lineariza-
FIG. 7. Steady-state gross ferrous iron oxidation rate as a func-
tion of external ferrous iron concentration at the three different
influent ferrous iron concentrations (A, B, C) as defined by the
symbols in Fig. 3.
tion of the Monod (35) growth curve (Fig. 4). The apparent
P-max was 0.070 h- l, and the threshold-corrected growth half-
saturation constant (K, ) was 0.78 mM (Fig. 4).
Ycarbon, expressed as carbon produced per iron oxidized as
a function of growth rate, is shown in Fig. 5. The steady-
state Ycarbon decreased with the increasing dilution rate for
this isolate. This decrease is apparently related to a decrease
in carbon per cell as a function of growth rate (Fig. 6) for this
isolate. The carbon-to-cell ratio was not constant when
strain AK1 was grown in batch cultures (Fig. 2).
(ii) Iron oxidation kinetics. The Michaelis-Menten relation-
ship between the gross (total) iron oxidation rate and the
external ferrous iron concentration at three different influent
ferrous iron concentrations is shown in Fig. 7. The apparent
x
threshold ferrous iron concentration for gross iron oxidation
at three SR values was ca. 0.25 mM, the same value
c
5

0.5 r 4

0.4 F b-

z
U

C.)
B
0.3 F U 2
1
C D A
0
.0
0

a
U.
0.2 V
0

I-)

0.1 F
ao 0.5 1.0 1.5 2.0
S-St, mM
0.0 L-

0.00 0.01 0.02 0.03 0.04 0.05 FIG. 8. Threshold-corrected linearization of Fig. 7. The equa-
tions for the lines at the 95% confidence level are as follows: A, y =
Growth rate, hr-1 (0.45 + 0.39)x + 0.91 0.48 with r = 0.85; B, y = (0.57 + 0.78)x +
FIG. 6. The steady-state carbon/cell ratio as a function of growth 1.06 + 1.00 with r = 0.47; and C, y = (2.44 + 0.56)x + 0.87 + 0.39
rate. Symbols are the same as those in Fig. 3. with r = 0.91. Symbols are the same as those in Fig. 3.
52 BRADDOCK, LUONG, AND BROWN
-
7
0.5 D>
£
E 0.4 -
E
O/
00
%~~~~~~
0
0~~~~
'R 0.2 0z ,|
0..
0.0 0.5 1.0 1.5 2.0 2.5
External ferrous iron, mM
FIG. 9. Net iron oxidation rate as a function of external ferrous
iron concentration. The equation for the line at the 95% confidence
level is y = (0.22 ± 0.04)x - 0.05 0.04 with r = 0.91. Symbols are
±
the same as those in Fig. 3.
estimated for growth kinetics. This indicates that iron oxida-
tion and growth kinetics are tightly coupled. Threshold-
modified linearization of the gross iron oxidation rate versus
the external ferrous irori concentration is shown in Fig. 8.
This figure can be used to predict a vmmax and a half-saturation
constant (KFe) for different values of SR. The apparent vmax
predicted from the reciprocal slope of the best-fit lines when
SR = 9.0 mM was 0.41 mmol * liter-' h- l. The Vmax for SR =
22.4 to 23.3 mM was 2.22 mmol * liter-' h-. The KFe is
estimated from vmax times the y intercept. Thus, KFe was
0.36 and 2.02 mM for SR = 9.0 and 22.4 to 23.3, respectively.
Insufficient data were available at low dilution rates to yield
a significant linear relationship when SR = 17.9 mM, so no
attempt was made to estimate I'max or KFe for this value of
SR .
The Michaelis-Menten relationship between net iron oxi-
dation rate (icarhon) and external ferrous iron concentration
is shown in Fig. 9. Since saturation was not apparent, we
assumed that the relationship between 1'carhon and S was
linear (first order). A theoretical maximum net iron oxidation
rate must exist; therefore, the rates we observed for AKi
must be significantly below the maximum. The x intercept
(threshold) concentration of the linear relationship (Fig. 9) is
0.24 mM. The slope of the line (Fig. 9) is mathematically
analogous to steady-state affinity (as) as defined by Button
(9). The a, for isolate AKI is ca. 0.22 liters * mg of C'
and is a good parameter to use for comparative measures of
specific (subsaturating) iron oxidation rate at low iron con-
centrations.
(iii) Arsenic. Reduced arsenic (1.33 or 2.67 mM) was added
to the ferrous iron medium in several continuous-culture
experiments. The results indicated that reduced arsenic was
not oxidized in the presence of isolate AK1 growing at four
different growth rates ranging from 0.007 to 0.034 h-l.
Virtually none of the reduced arsenic added was recovered
in the oxidized form.
DISCUSSION
Growth kinetics. We have used batch and dilute (low
biomass) continuous cultures to describe the growth kinetics
APPL. ENVIRON. MICROBIOL.
of an arsenic-tolerant subarctic isolate of T. ferrooxidans.
The growth and iron oxidation kinetics of this isolate deviate
from traditional kinetic models.
The conventional Monod model which was originally
derived from Michaelis-Menten enzyme kinetics is useful for
describing the carbon-limited growth of heterotrophs when
the growth-limiting substrate is incorporated as cellular
material. When T. ferrooxidans is grown on reduced iron,
the energy-limiting substrate is not incorporated as cellular
material. The Monod model depends on a constant yield
which links specific growth rate to substrate removal. The
Monod model has also been modified to accommodate the
possibility that yield may increase with increasing growth
rate (i.e., maintenance energy; see reference 39). The model
is less readily adaptable to the possibility that yields may
decrease with increasing growth rate.
Decreasing yields with increasing growth rate have com-
monly been observed for a variety of microorganisms grown
in chemostats under the limitation of inorganic ions such as
phosphate, magnesium, silicate, ammonium, and nitrate and
of vitamins such as B12 (5, 12, 30, 40). The internal-stores
kinetic model originally proposed by Droop (17) and modi-
fied by Rhee (40) has been used to describe growth rates as a
function of internal rather than external concentrations of
limiting substrates. It is essential in this kinetic model that
yields vary inversely with growth rate since concentrations
of intracellular substrate increase with increasing growth
rate (38, 40). The application of either the Monod or the
internal-stores model to the observed growth of dilute, iron-
limited cultures of isolate AK1 is limited. Despite these
recognized limitations, either model can be used to mathe-
matically describe some growth and iron oxidation functions
for this isolate.
The Monod model was used to estimate iron-limited
growth parameters for strain AKi with respect to external
ferrous iron concentrations. The values we found for PLmax
and K. were somewhat lower than those reported for other
isolates of T. ferrooxidans. Values reported for other isolates
cultured in chemostats include the following: a Pmax of 0.14
h-V and a K of 37 mM (43); a Pmax of 0.161 h-l and a K of
3.8 mM (33); and a -max of 1.25 to 1.78 h-1 and a K of 0.7 to
2.4 mM (27, 28). These apparent discrepancies are probably
a result of growing the organism under different physiologi-
cal regimes (i.e., pH and temperature) and of variations in
the microorganisms classified as T. ferrooxidans. Several
investigators recently demonstrated that variations of taxo-
nomic importance may occur among different isolates of T.
ferrooxidans. Variations have been found in DNA base
composition studies (20, 34), physiological studies (41), and
morphological studies (13). Growth and iron-oxidation kinet-
ics of these iron-oxidizing bacteria may be yet another way
of describing a functional classification for these microorga-
nisms.
Our results also show apparent threshold ferrous iron
concentrations of about 0.25 mM for both iron-limited
growth and iron oxidation. The threshold ferrous iron con-
centration observed here represents a concentration of sub-
strate unavailable to strain AKi for ehergy utilization but is
not analogous to maintenance energy since the same thresh-
old exists for both growth rate and iron oxidation rate (i.e.,
maintenance energy is observed as a positive threshold only
for growth in conventional kinetic models). Thresholds have
not been reported for other T. ferrooxidans strains growing
on reduced iron. However, most other studies have used
much higher iron concentrations, resulting in much more
biomass in the continuous-culture vessel. Eccleston and
VOL. 48, 1984
Kelly (18) did observe a maintenance energy coefficient for a
T. Jerrooxidans strain grown in either tetrathionate- or
thiosulfate-limited chemostat cultures. However, the lowest
dilution rate used in these studies was 0.02 h'. The appar-
ent threshold of 0.25 mM for both growth and iron oxidation
indicates that growth and iron oxidation are tightly coupled
in isolate AKI. The threshold value may be an indication of
inhibition of ferrous oxidation at high ratios of ferric iron to
ferrous iron. However, increasing ferric iron/ferrous iron
ratios do not change specific (biomass-adjusted) ferrous
oxidation rates (7).
The yield of organic carbon for strain AK1 was relatively
constant in batch cultures, ranging from 0.12 to 0.20 mg of
carbon per mmol of iron oxidized. However, steady-state
(continuous-culture) data indicate a decreasing curvilinear
dependence of Ycarbon on the specific growth rate. particular-
ly below the point where p. = 0.03 h- 1 (Fig. 5). If Y,[rbon data
below this point had not been available, the yield might have
been assumed to be constant with increasing dilution rate.
Yce,i (number of cells mmol of iron oxidized-') also de-
creased as a function of p. (data not shown). The exponential
function (Fig. 5) can be approximated by the linearization In
Y = -22.19p. - 0.89 (r = -0.89) which describes the
observed relationship between Y.r,,tn and p.
The steady-state carbon/cell ratio also appears to decrease
with increasing dilution rate for isolate AKI (Fig. 6). We
observed, when counting AK1 cells by epifluorescence
microscopy, that cell size markedly increased at low dilution
rates. Other investigators have reported similar observations
(D. W. Tempest and 0. M. Neijssel, Abstr. Third Int. Symp.
Microbial Ecol., 1983, 112, p. 8). Our data indicate that at
low growth rates strain AK1 is most efficient at utilizing
reduced iron. The highest Ycarbo)n we observed was 0.50 mg
of C per mmol of iron oxidized at p = 0.007 h- l. This yield is
close to the theoretical maximum growth yield of 0.53 mg of
C per mmol of iron for T. ferlooxidans growth at pH 2 (24).
The steady-state growth kinetics of strain AKI apparently
do not depend on the initial ferrous iron concentration (Fig.
3) over the range of SR values used in these experiments.
Additionally, the growth kinetics of this isolate are not
affected by the addition of reduced arsenic to the ferrous
iron-containing feed.
Iron oxidation kinetics. The gross ferrous iron oxidation
rate at steady state depends on the influent ferrous iron
concentration (see Fig. 7 and 8 for strain AK1). The imjx and
the KFe both appear to increase with increasing SR. This
increase in maximum iron oxidation rates with increasing SR
was also observed by Guay et al. (21). Guay et al. examined
the steady-state iron oxidation kinetics of a T. jerrooxidans
strain at influent ferrous iron concentrations of 89 to 161
mM. Their strain did not reach its maximum capacity for
iron oxidation even at an SR of 161 mM.
The observed dependence of 1'niax on influent substrate
concentration is predicted from Michaelis-Menten kinetics
applied to substrate oxidation in a chemostat. The Michaelis-
Menten parameter, i'mj,x. is a relative rather than a specific
value. The apparent maximum rate of the enzyme is attained
when all of the enzyme sites are saturated with substrate.
Therefore, if more enzyme is present at a given substrate
concentration, the apparent V'm.tx is greater. When the influ-
ent ferrous iron concentration is increased, a larger concen-
tration of organic carbon (as a function of ambient substrate
concentration or of growth rate) is produced by strain AKI.
Assuming that cellular carbon represents about 50% of the
dry weight of the AK1 cell under iron-limited, steady-state
conditions, an increase in cell dry weight (biomass) should
GROWTH KINETICS OF T. FERROOXIDANS 53
result in an increased concentration of ferro-oxidase per unit
volume. Therefore, the total activity of the enzyme is greater
at increasing influent ferrous iron concentrations as long as
the cells remain iron limited.
The gross iron oxidation rate can be converted to a
specific rate by defining iron oxidation per unit of carbon
('carhon) or per numbers of cells (OceIl). These specific iron
oxidation rates are analogous to the specific growth rate
constant described in the Monod growth model. The l carbon
as a function of the S-S, is independent of the SR for strain
AKI (Fig. 9). Therefore, enzyme concentration, not ferric
inhibition of ferrous-iron oxidation, appears to be responsi-
ble for the results observed in Fig. 7 and 8, which may also
be used as evidence that ferrous iron remained the limiting
factor in these experiments. Additionally, reduced arsenic
added to the feed does not affect the net iron oxidation rate
of this organism.
Expressed mathematically, 1'carhon p :Ycarbo 1. When
11 Ycarbj l is plotted against Y.,trh- ' by analogy to the
internal-stores growth model, the relationship described by
the equation pL = 0.070 - 0.179Y(r = 0.933) predicts a p-max
of 0.070 h-', the same value predicted by the Monod
equation.
The steady-state 1'carhon as a function of S (Fig. 9) for strain
AKI appears to be a linear relationship. Since a theoretical
maximum net iron oxidation rate must exist, the function
must asymptotically approach a maximal value. I'his indi-
cates that the observed rates are significantly lower than the
maximum rate obtainable by strain AK1. The maximum rate
would presumably occur when much higher ferrous iron
concentrations are added to iron-starved batch cultures (30).
Arsenic. T. fe'rooxidans oxidizes many reduced metals
including cuprous copper (37), elemental selenium (44), and
uranous sulfate (14-16). Furthermore, T. fjrrooxidans may
be able to obtain energy from some of these inorganic
oxidations (15, 16).
Strain AK1 was isolated from placer gold mine drainage
containing dissolved arsenic. Speciation of the dissolved
arsenic in this stream indicated that arsenate species and
other species (including arsenite) were present in about
equal amounts (6). Arsenate was not the only species present
despite the fact that the stream was saturated with oxygen.
In laboratory cultures, strain AK1 was tolerant to at least
900 mg of arsenite or arsenate liter-' when the organism
was grown in a medium containing ferrous iron (J. M.
Forshaug, M.S. thesis, University of Alaska, Fairbanks,
1983).
For these reasons, we hypothesized that T. fe,,ooxidans
isolate AKI can oxidize reduced arsenic and derive energy
from such an oxidation. However, the results of steady-state
experiments to test this hypothesis were inconclusive. When
arsenite was added to the continuous-culture medium con-
taining ferrous iron, we observed no oxidation of the arsenite
ion. The presence of either 1.33 or 2.67 mM arsenite in the
medium did not yield an increase in biomass, nor did it have
any observable effect on growth-kinetic or iron oxidation
kinetic parameters being measured. If arsenite can be used
for energy production by isolate AK1, an observable in-
crease in biomass would be expected.
Summary. The importance of the iron-oxidizing thiobacilli
in catalyzing biogeochemical cycles has only recently been
recognized. With the depletion of high-grade ores. methods
of bioleaching for the recovery of valuable metals are
becoming increasingly important. At some time in the future.
genetic manipulation may be used to produce highly efficient
leaching microorganisms (3). The iron-limited growth kinet-
54 BRADDOCK, LUONG, AND BROWN
ics have not been well described for T. jerooxidans. Pub-
lished kinetics are inconsistent. These inconsistencies prob-
ably result from apparent strain-to-strain differences and
from the difficulties in culturing this organism, especially in
iron-limited continuous cultures. A complete kinetic model
must describe both cell growth and substrate utilization with
some function that relates the specific growth rate to the
substrate utilization rate. The yield constant links the equa-
tions of the Monod model. Alternatively, the Monod model
can be modified to accommodate an increasing yield with
increasing growth rate when a maintenance coefficient is
observed. Other kinetic models have been developed (17. 38.
40) to describe the growth of microorganisms as a function of
internal rather than external concentrations of a limiting
nutrient.
We have described the growth kinetics, ferrous iron
oxidation kinetics, and arsenite oxidation potential of an
arsenic-tolerant T. fjerrovidans isolate. AK1. The growth of
this organism is described in both transient (batch) culture
and steady-state (continuous) culture. We have observed
that the growth and iron oxidation kinetics of this autotro-
phic isolate deviate from Monod kinetics. The kinetics of
this isolate show some similarities to kinetic models derived
to describe substrate-limited growth from internal rather
than external substrate concentrations. These observations
indicate that additional studies are needed, particularly at
low dilution rates, to accurately model the growth and
substrate utilization of T. jerrooxidans and possibly other
autotrophic microorganisms.
We have also described a method for the successful
continuous cultivation of T. Jfrrooxiidans in dilute, iron-
limited continuous cultures. This continuous-culture method
can be used to predict the growth and iron oxidation kinetics
of other strains of T. jerrooxidins and to investigate the
ability of T. jerrooxiclans to oxidize and produce energy from
reduced metals other than iron. Such studies are important
for understanding the biogeochemical cycling of many met-
als and for applying these microorganisms to biohydrometal-
lurgy.
ACKNOWLEDGMENTS
Support for this study provided in part by grants B-024-ALAS
was
and A-070-ALAS from the Office of Water Policy. U.S. Department
of the Interior. and by grant MIN-1-81 from the Alaska Council on
Science and Technology.
LITERATURE CITED
1. American Public Health Association. 1975. Standard methods for
the examination of water and wastewater, 14th ed.. p. 207-213.
American Public Health Association. Inc.. Washington. D.C.
2. Bounds, H. C., and A. R. Colmer. 1972. Comparisoni of the
kinetics of thiosulfate oxidation by three iron-sulfur oxidizers.
Can. J. Microbiol. 18:735-740.
3. Brierley, C. L. 1982. Microbiological mining. Sci. Am. 247:44-
53.
4. Brown, E. J., and D. K. Button. 1979. A simple method of
arsenic speciation. Bull. Environ. Contam. Toxicol. 21:37-42.
5. Brown, E. J., and D. K. Button. 1979. Phosphate-limited growth
kinetics of green alga. J. Phycol. 15:305-311.
a lism and ferrous iron oxidation in suspensions and batch cul-
6. Brown, E. J., H. V. Luong, and J. M. Forshaug. 1982. The
occurrence of 7Thioba(c(illius f'erloo.xitlaoins and arsenic in subarctic
streams affected by gold mine drainage. Arctic 35:417-421.
7. Brown, E. J., H. V. Luong, and j. M. Forshaug. 1983. Geomi-
crobiology of arsenic associated with gold deposits in Alaska. p.
571-580. In G. Rossi and A. E. Torma (ed.). Recent progress in
biohydr-ometallurgy. Associlzione Mineraria Sarda. Iglesias.
Itally.
8. Button, D. K. 19)78. O)n the theory of control of microbiall
Appi- ENVIRON. MICROBIOL.
kinetics by limiting nutrient concentrations. Deep-Sea Res.
25:1163-1177.
9. Button, D. K. 1983. Differences between the kinetics of nutrient
uptake by microorganisms. growth and enzyme kinetics. Trends
Biochem. Sci. 8:121-124.
10. Chang, Y. C., and A. S. Myerson. 1982. Growth models of the
continuous bacterial leaching of iron pyrite by Thiobacillius
ferroo.voidans. Biotechnol. Bioeng. 24:889-902.
11. Cornish-Bowden, A. 1979. Fundamentals of enzyme kinetics, p.
15-33. Butterworth. Woburn. Mass.
12. Cunningham, A., and P. Maas. 1982. The growth dynamics of
unicellular algae. p. 167-188. In M. J. Bazin (ed.). Microbial
population dynamics. vol. 1. CRC Press. Inc.. Boca Raton. Fla.
13. DiSpirito, A. A., M. Silver, L. Voss, and 0. H. Tuovinen. 1982.
Flagella and pili of iron-oxidizing thiobacilli isolated from a
uranium mine in northern Ontario, Canada. AppI. Environ.
Microbiol. 43:1196-1200.
14. DiSpirito, A. A., and 0. H. Tuovinen. 1981. Oxygen uptake
coupled to uranous sulfate oxidation by Thiobacillhs ferrooxi-
(lans and T. acidophilus. Geomicrobiol. J. 2:275-291.
15. DiSpirito, A. A., and 0. H. Tuovinen. 1982. Uranous ion
oxidation and carbon dioxide fixation by Thiobacillius ferrooxi-
dlans. Arch. Microbiol. 133:28-32.
16. DiSpirito, A. A., and 0. H. Tuovinen. 1982. Kinetics of uranous
and ferrous iron oxidation by Tlhiohbc('illius f'('OO-i-oxdn.s. Arch.
Microbiol. 133:33-37.
17. Droop, M. R. 1968. Vitamin B,2 and marine ecology. IV. The
kinetics of uptake. growth and inhibition in Monochr.vsis lhtheri.
J. Mar. Biol. Assoc. U.K. 48:689-733.
18. Eccleston, M., and D. P. Kelly. 1978. Oxidation kinetics and
chemostat growth kinetics of Tliioba(cillius ferroo.idalns on te-
trathionate and thiosulfate. J. Bacteriol. 134:718-727.
19. Ediger, R. D. 1975. Atomic absorption analysis with the graphite
furnatce using matrix modification. At. Absorpt. Newsl. 14:127-
132.
20. Guav, R., M. Silver, and A. E. Torma. 1976. Base composition
of DNA isolaited from 7lhiobacillhs fierrooxidans grown on
different substrates. Rev. Can. Biol. 35:61-67.
21. Guay, R., M. Silver, and A. E. Torma. 1977. Ferrous iron
oxidation and uranium extraction by Tlhioba(illius ferroo.idans.
Biotechnol. Bioeng. 19:727-740.
2 . Hobbie, J. E., R. J. Daley, and S. Jasper. 1977. U se of
Nuclepore filters for counting bacteria by fluorescence micros-
copy. Appl. Environ. Microbiol. 33:1225-1228.
23. Hoffman, M. R., B. C. Faust, F. A. Panda, H. H. Koo, and H. M.
Tsuchiya. 1981. Kinetics of the removal of iron pyrite from coal
by microbial catalysis. Appl. Environ. Microbiol. 42:259-271.
24. Ingledew, W. J. 1982. Tliiob(acillis f e('lrloox.iitdans: the bioenerget-
ics of an acidophilic chemolithotroph. Biochim. Biophys. Acta
683:89-117.
25. Justin, P., and D. P. Kellv. 1978. Growth kinetics of 7liiobacillhs
clenitrificains in anaerobic and aerobic chemostat culture. J.
Gen. Microbiol. 107:49-58.
26. Karaivko, G. 1., S. 1. Kuznetsov, and A. I. Golonizik. 1977. The
bacterial leaching of metals from ores. Technicopy. Ltd.. Stone-
house. U.K.
27. Kelly, D. P., M. Eccieston, and C. A. Jones. 1977. Evaluation of
TIhiohbcillins
continuous chemostat cultivation of fi ,.,.oovidaccns
on ferrous iron or tetrathionate. p. 1-7. In W. Schwartz (ed.),
Gesellschaft fur biotechnologische Forschung Conference-
Bacterial Leaching. Verlag Chemie GmbH. Weinheim. Federal
Republic of Germany.
28. Kelly, D. P., and C. A. Jones. 1978. Factors affecting metabo-
tures of 71hiobacillus fierrooxidans: relevance to ferric iron leach
solution regeneration. p. 19-43. In L. E. Murr. A. E. Torma,
and J. E. Brierley (ed.). Metallurgical applications of bacterial
leaching and related microbiological phenomena. Academic
Press, Inc.. New York.
29. Lacey, D. T., and F. Lawson. 1970. Kinetics of the liquid-phase
oxidation of acid ferrous sulfate by the bacterium Thiobczcillhs
fcnrooxidans. Biotechnol. Bioeng. 12:29-50.
growth 30. Lang, D. S., and E. Jl. Brown. 1981. Phosphorus-limited growth
 VOL. 48, 1984
of a green alga and a blue-green alga. Appl. Environ. Microbiol.
42: 1002-1009.
31. Leefeldt, R. H., and A. Matin. 1980. Growth and physiology of
Thliobacillls notvel/ius under nutrient-limited mixotrophic condi-
tions. J. Bacteriol. 142:645-650.
32. Lundgren, D. G., and M. Silver. 1980. Ore leaching by bacteria.
Annu. Rev. Microbiol. 34:263-283.
33. MacDonald, D. G., and R. H. Clark. 1970. The oxidation of
aqueous ferrous sulphate by Thiobacillusferrooxidans. Can. J.
Chem. Eng. 48:669-676.
34. Martin, P. A., P. R. Dugan, and 0. H. Tuovinen. 1981. Plasmid
DNA in acidophilic, chemoautotrophic thiobacilli. Can. J. Mi-
crobiol. 27:850-853.
35. Monod, J. 1942. Recherches sur la croissance des cultures
bacteriennes. Hermann S.A., Paris.
36. Murphy, J., and J. P. Riley. 1962. A modified single solution
method for the determination of phosphate in natural waters.
Anal. Chim. Acta 27:31-36.
37. Nielsen, A. M., and J. V. Beck. 1972. Chalcocite oxidation and
coupled carbon dioxide fixation by Thiobacillus ferrooxidans.
Science 175:1124-1126.
38. Nyholm, N. 1976. A mathematical model for microbial growth
under limitation by conservative substrates. Biotechnol.
Bioeng. 18:1043-1056.
View publication stats
GROWTH KINETICS OF T. FERROOXIDANS 55
39. Pirt, S. J. 1975. Principles of microbe and cell cultivation, p. 4-
75. John Wiley & Sons, Inc., New York.
40. Rhee, G.-Y. 1980. Continuous culture in phytoplankton ecology.
p. 151-203. In M. R. Droop and H. W. Jannasch (ed.), Ad-
vances in aquatic microbiology, vol. 2. Academic Press. Inc.,
New York.
41. Silver, M. 1978. Metabolic mechanisms of iron oxidizing thioba-
cilli, p. 3-17. In L. E. Murr, A. E. Torma. and J. A. Brierley
(ed.), Metallurgical applications of bacterial leaching and related
microbiological phenomena. Academic Press, Inc., New York.
42. Silver, M., and A. E. Torma. 1974. Oxidation of metal sulfides
by Thiobacillus ferrooxidans grown on different substrates.
Can. J. Microbiol. 20:141-147.
43. Tomizuka, N., M. Yagisawa, J. Someya, and Y. Takahara. 1976.
Continuous leaching of uranium by Thiobacillus ferr ooxidans.
Agric. Biol. Chem. 40:1019-1025.
44. Torma, A. E., and F. Habashi. 1972. Oxidation of copper (11)
selenide by Thiobacillus ferrooxidans. Can. J. Microbiol.
18:1780-1781.
45. Tuovinen, 0. H., and D. P. Kelly. 1974. Use of microorganisms
for recovery of metals. Metall. Rev. 19:21-31.
46. Wilson, F. H., and D. B. Hawkins. 1978. Arsenic in streams.
stream sediments, and ground water, Fairbanks area, Alaska.
Environ. Geol. 2:195-202.