University of Groningen

Growth of bacteria at low oxygen concentrationsGerritse, Jan

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:1993

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Gerritse, J. (1993). Growth of bacteria at low oxygen concentrations. Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 15-06-2020

Chapter 3

Modelling of mixed chemostat cultures of an aerobic bacterium,

Comamonas testosteroni, and an anaerobic bacterium,

Veillonella alcaleseens: Comparison with experimental data

Jan Gerritse, Frits Schut and Jan C. Gottschal

In: Applied and Environmental Microbiology (1992) 58:1466-1476

Chapter 3

Modelling of Mixed Chemostat Cultures of an Aerobic Bacterium, Comamonas testosteroni, and an Anaerobic Bacterium, VeillonelIa

alcaleseens: Comparison with Experimental Data JAN GERRITSE,' FRITS SCHUT, AN D JAN C. GOTTSCHAL

Departmellt oJ Microbiology, University oJ Gronillgen, Kerklaan 30, 9751 NN Haren, Th e Netherlands

Received 14 November 1991/Acceptcd 7 Fcbruary 1992

A malhemalical model of mixed chemoslal cultures of Ihe obligalely aerobic baclerium Comamonas teslosteroni and the anaerobic bacterium Veillonella alcaleseens grown under dual limitation of L- lactate and oxygen was conslrucled. The model was based on MichaeIis·Menlen·type kinelics for Ihe consumplion of subsirales, wilh noncompelilive inhibition of V. a/ea/eseens by 02' The growth characleristics of Ihe aerobic and 3naerobic organisms were determined experimentally wUb pure cultures of the individual species in (oxygen·limiled) chemostals. Using Ihese pure·cullure dala in Ihe model of Ihe mixed culiure resulled in a good descriplion of Ihe aclual mixed cullures of Ihe two bacleria. In Ihe aclual mixed·culture experimenls, coexistence of tbe two species occurred only when the cultures were oxygen limited. With increasing oxygen supply (Ihe aclual oxygen concenlralion in Ihe culiure remaining al < 0.2 ILM), Ihe biomass of C. testosteroni increased, whereas Ihal of V. a/ea/eseens decreased. Apparently, C. testosteron; prolecled V. a/ea/eseens from inhibition by oxygen by maintaining sufficiently low oxygen concentrations. The model calculations indicated Ihal compelilion between Ihe aerobic and Ihe anaerobic baclerium for common subsirales (L· laclale and oxygen) occurred and Ihallhe anaerobe was Ihe beller compelilor. Analysis of Ihe culiure Huid indicaled Ihal C. testosteron; grew primarily allhe expense oflhe fermenlalion produels of V. a/ea/eseens, i.e., propionale and acelale. The model furlher indicaled Ihal wilh differenl values of severa l growth para meiers (e.g., subsirale affinity and/or inhibition consianis), Ihe affinily of Ihe aerobic organism for oxygen and Ihe sensilivily of Ihe anaerobic organism for oxygen were the most important properties determining the coexistence of tbese MO physiologically differenl Iypes of bacleria.

Microbial ccosystcms with aerobic and anacrobic zones are found in many environments. Examples of this arc sediments (16, 31, 32), microbial mats (33), soils (39), (strat· ified) lakes and se.as (1, 22, 35, 40), biofilms (25), periodontal pockets (20), the rumen (37), and baclerial colonies (42). At the interface of aerobic and anacrobic zones, the concentra­lion of dissolved oxygen is of ten very low but slill sufficienl for the growth of ae robes (9) and not high enough for tOlal inhibilion of anaerobes. Recently, wc showed that strictly aerobic and anaerobic bacteria can grow side by side under microaerobic conditions in chemoslat cultures provided the supply of oxygen does nol exceed the maximum 0 , con· sumption ra te of the bacteria (8). Under such oxygen-limiled conditions~ fermentativc, sulfate-reducing, and even very oxygen·sensitive melhanogenic bacteria have been shown to coexist with obligately aerobic bacteria (8, 11, 13, 26). The anaerobic bacteria werc apparenlly protected against 0 , damage thanks to the high oxygen uplake affinity of the (facultalively) aerobic bacleria ensuring a very low dis· solved·oxygen concenlralion (3, 4, 34). The ae robes con· sumed the end products of the metabolism of Ihe anaerobic bacleria.

To funher investigate Ihe environmental conditions and growth charactcristics which may influcncc the cocx.istencc of ae robes and anaerobes, wc sludied mixed cultures of the anacrobic fcrmentative bacterium Veillollella a lcaleseens and the aerobic bacterium Comamona.'l testo.',teroni as an experimental model system. In preliminary experiments, these two bacteria were shown 10 eoe xi st over a consider-

.. Corrc~ponding aulhor.

24

able range of oxygen supply rates in chemostat cultures grown under dual limitation of L·lactale and oxygen (8).

Under anaerobic condilions , L·laclale is fermenled by V. alcaleseens to propionate, acelate, H" and CO, . Although oxygen can be reduced by V. alcalescem, it also inhibits the growth of this anacrobe. The oxygen mctabolism of V. alcalescens in pure culture has been studied extensively by De Vries el al. (7). It was demonsiraled Ihal Ihis baclerium can grow in the presence of oxygen when the concentralion of 0 , remains Icss than approximalely 0.8 fLM . This sludy further revealed th at in the presence of oxygen, V. alca· leseens is able to consurne L-Iactate via a membrane-bound lactale oxidase syslem, resulting in the produclion of super· oxide anion radicals and hydrogen peroxide. Although cal· alase activity was delected in V. alcalescens, inhibilion of growlh of Ih is organism by oxygen was correlated with the accumulalion of toxic oxygcn melabolites in the medium and a rapid (reversibie) inaclivalion of lactale dehydrogenase.

The strictly aerob ic strain of C. testosteron; used in Ih c present study was enriched selectively in a microaerobic chemoslal (pania l 0 , pressure of < 1 fLM) with L· lactate as Ihe carbon and energy source (8). In add ition to L· laclale, Ihe two major fermenlation producls of V. alcaleseem (propionate and acelate) can be uscd by C. testosterolli. Hydrogen could nOl be metabolizcd.

The main objective of Ihis sludy was to cxplore in more detail thc cocxistence of V. alcalescefls and C. testo~·((,'-()lli at various degrces of nUlrient availability. Ta Ihis cnd. " mathematical model of thi s mixed culture was construclcJ la oblain abelter understanding of the changes in Ihc f1uxc, of substrates and fcrmentation produels in response lO a changing supp ly of oxygcl1 and growlh subslrates. With thc

CXCI

pan min vidl

o 31ld

wer Thc min acc' pre' and pol< obt. thc rcfe rate

A org cell 0,,= sur, out me; nin Ihe sial

[

for cor del mo Oh ver sur to Cel per teil blo ma cel obi int. sal sir: qUl lim kin ade usc an lirr Ih, s ul COl

bel esl pa les de ex pH

exception of thc oxygcn inhibition constant, all growth parameters used in this mathematical model were de ter­mincd cxpcrimentally in oxygen-limited cultures of the indi­vidual organisms.

MATERIALS AND METHODS

Organisms and growth conditions. V. a/ca/escellS NS. L49 <lnd C. testosteron; (previously Pseudomonas testosteroni) wcrc isolated in our laboratory as described before (8, 17). The bacteria we re grown at 30°C in a phosphate·buffered minimal medium (pH 7.0) with L·lactate, propionate, or acelale as a carbon souree in chemostats as described previously (8). Oxygen concentrations in the culture Iiquid and in the headspace we re measured continuously with polarographic oxygen probes (Jngold) . Redox readings we re obtained continuously with a platinum electrode and with the Ag·AgCi reference electrode of the pH electrode as the reference cell (8). Thc chemostat was operated at a dilution rate of 0.1 h- I.

Analytical procedures. Gasses (H2 , CO2, and O2), total organic carbon, and maximum oxygen consumption rates of cell suspensions were determined as described before (8) . Oxygen consumption rales in chemostat cultures were mea­sured continuously by oxygen gas analysis of the in· and the outtlowing gas (air·N2 mixtures) (8) . Organic acids were measured by gas chromatography using the method of Nan· ninga and Gottschal (23). Protein was quantified according to the method of Lowry et al. with bovine serum albumin as the standard (21).

Determination of kinetie parameters. Kinetic parameters for L-Iactate, acetate, propionate oxidation, and oxygen consumption by C. testosteron; and V. a/ca/escens were determined respirometrically at 30°C in a biological-oxygen monitor (Yellow Springs Instruments Co., Yellow Springs, Ohio) equipped with a polarographic oxygen electrode. At very low oxygen concentrations, the electrical current mea­sured at the output of the oxygen electrode was amplified 10 to 100 times with a Keithley model 485 picoamperometer. Cells were taken from the chemostat, washed , and sus­pended in a phosphate buffer (50 to 250 mg of cell r.ro­tein · liter- I) containing chloramphenicol (30 mg · liter - ) to block protein synthesis during the incubations (8) . Initial maximum oxygen consumption rates of the resuspended cells did not differ much «15% difference) from those obtained with cells transferred directly ffom the chemostat into the biological-oxygen monitor. To determine the half saturation constants for the consumption of growth sub­strates, cell suspensions were saturated with air and subse­quently supplied with L-Iactate, propionate, or acetate in limiting concentrations (5 to 50 ILM). To determine the kinetics for oxygen consumption, these substrates were added at saturating concentrations (2 mM) . The bacteria used in these incubations were pregrown on L-Iactate or on a mixture of L-Iactatc, propionate, and acetale in an oxygen­limited chemosta!. O2 consumption rates we re calculated as the tangent of O2 concentration versus time plots at different substrate concentrations. Apparent Km and Qm~ (rate of consumption at saturating concentration of substrate; sec below) va lues for L-Iactate, ace ta te, and propionate we re estimated by the direct linear-plot method (27). )(jnetic parameters for anaerobic L-Iactate consumption by V. a/ca­/escellS we re determined by directly measuring L-Iactate depletion in an anaerobic batch culture. In all oxygen uptake experiments, dilute cell suspensions were used (5 to 50 mg of protein . liter - I) to minimize interference of membrane O2

Modelling of mixed chemostat cultures

l-lactate

?-o,~ \I V

proplonllle

lICIIIIIIe

FIG. 1. Schematic present at ion of a mixed culture of V. a/ca· leseens and C. teslOsteroni growing on L-lactate under mcygen limitation. L-Lactate is fermented by V. a/calescens 10 propionate, acetate, H2• and CO2- L-Lactate, propionate, and acetate can be used by C. testosteron;. Oxygen is used by C. testosteron; and 10 some extent by V. alcaleseens. but at higher oxygen concentrations, (he growth of this anaerobe is inhibited.

diffusion characteristics with O2 consumption. This allowed accurate measurement of oxygen concentrations of 2:0.01 ILM. It was assumed that air-saturated buffer contained 250 ILM dissolved O2 ,

Mathematical model. The model is based on the assump­tion that substrate-dependent growth of bacteria in mixed culture can be described according to the same mathematica I formulations as are used for the description of growth in pure cultures (10, 12, 18, 19, 28, 38, 41; sec below also) . The model was designed according to the following three basic steps.

(ij Step I. Determine the specific rates of consumption (Q) of the various substrates by the baeteria. This was done according to Michaelis-Menten-type kinetics.

(1)

In th is equation, Qm.. represents the value of Q at a saturating concentration of s (s j!> Km)' S is the substrate concentration, and Km is the half saturation constant.

ii. Step 11. Calculate the corresponding specific growth rates (IL) of the individual organisms. The following equation was used.

IL=QxY (2)

In this equation, Y is the amount of biomass forrned per substrate consumed (yield coefficient).

(iij) Step 111. Determine the changes in the concentrations of L-Iactate, oxygen, propionate, acetate, and biomass in the chemostat. This was done according to standard differential equations describing growth in continuous culture (sec be­low). For explanation of the symbols and their default values, see Table 1.

Growth of C. testosteron; in the mixed culture is intlu­enced by the availability of three potential carbon sub­strates, i.e., acetate, propionate, and L-Iactate, and by the presence of oxygen (Fig. 1). V. a/ca/escens is also capab le of reducing oxygen. lts specific growth rate is to some extent inhibited by O2 , depending on the actual concentration of O2 , Growth in the presence of very low dissolved-oxygen concentrations results in a change in the relativc amounts of

25

Chapter 3

TABLE 1. Explanation of symbols used in the model and their default values

Symbol

D L, 0 2'n 0, L P A QL Ven

QLOXm v.;,

Q02.max~eil

QO:z,mBJt2Vcil

K .Q Veil

iN·.l X

Vcil

QL Com QP::Com

Q0 2,max2 Com

YLCom

ypCom YACom

x Com

MOLCom

MopCom

Description

Dilution ra Ie L-Lactate eonen in reservoir medium Oxygen supply in chemostat liquid Dissolved-oxygen eonen in chemostat L-Lactate eonen in chemostat Propionate eonen in chernostat Acetale eonen in chemostat Maximum specific L-Iactate consumption Tale of V. a/calescens Maximum specitic L-Iactate oxidation fale of V. alcaleseens Maximum specifk oxygen consumption rale of high-affinity

system of V. alcaJescens Maximum specific oxygen consumption rale of low-affinity

system of V. a/ca/escens Half saturatio" constant for L-Iactate of V. a/calescens Half saturation constant for oxygen of high-affinity syslem of

V. alcaleseens Half saturation constant for oxygen of low-affinity system of V.

alcaLescens lnhibition constant of V. aLcaLescens by oxygen Growth yield of V. alcaleseens on L-Iactate Biomass of V. alcaleseens in chemostat Maximum specific L-Iactate consumption rate of C. testosteroni Maximum specific propionate consumption rale of C.

testasterani Maximum specific acetate consumption rate of C. testosteroni Maximum specific oxygen consumption rate of high-affinity

system of C. testosteron i Maximum specitic oxygen consumption rate of low-affinity

system of C. testosteroni Half saturation constant for L-Iactate of C. testosteroni Half saturation constant for propionate of C. testosteroni Half saturation constant for acetale of C. testosteroni Half saturation constant for oxygen of high-affinity system of

C. testosteroni Half saturation constant for oxygen of low-affinity system of C.

testasterani Growth yield of C. testosteron; on L-Iactate Growth yield of C. testosteroni on propionale Growth yield of C. testosteroni on acetale Biomass of C. testosteroni in chemostat L-Iactate-oxygen consumption stoichiomelry of C. testosteroni Propionate-oxygen consumption stoichiometry of C.

testosteron i Acetate-oxygen consumption stoichiometry of C. testosteron;

h- ' mM mM . h- t

mM mM mM mM

Unit

!-Lmol · mg of cell earbon- t . h- 1

!-Lmol . mg of eell carbon - I . h- I

JLmol . mg of ce)) carbon- t . h- t

!-Lmol . mg of cell carbon- I. h- J

JLM JLM

JLM

JL~ gaf cell carbon · mol- t

mg of cell carbon . liter - ~I JLmol . mg of ce)) carbon . h- ' !-Lmol · mg of ceU carbon - t . h- t

!-Lmol . mg of cell earbon - 1 . h- I

).Lmol · mg of cell earbon - 1 . h- t

JLmol . mg of ce)) carbon - t . h- t

g of eell carbon . mol- t

g of eell carbon · mol- I g of eell carbon . mol - 1

mg of cell carbon. liter- t

mol . mol - 1

mol · mol - I

mol · mol - I

Value

0.1 17.9

a...3.6 Q-j).250 Varia bie Variabie Variabie 155

6.9 8.4

13.2

5.1 0.11

350

10 Variabie Variabie

17.5 16.0

4.2 22.4

29.3

5.2 6.0 4.3 0.35

400

1l.7 15.3 8.7

Variabie 0.5 0.5

0.8

produets formed during L-Iactate ferment at ion and in the molar growth yield (7; th is study).

The respiration rates of C. testosteron; (Q02 Com) and V alcalescens (QO, Ve;') are composed of the combined activi­ties of two enzyme systems, one with high affinity and one with much lower affinity (see Results). Thus, for C. tes­tosteron;, the specific oxygen consumption rate was de­scribed as follows .

The specific substrate consumption rates were used to caJculate the specific growth rates (",) of the bacteri a in the culture by using the following procedure.

QO,Com = {Q0 2,maxt Com X [0,1/([0,1 + Kmt02Com)} +

{Q02.max2 Com X [021/([0 21 + K, .. 20 2 Com)} (3)

A slrictly analogous de script ion is used for oxygen con­sumption by Valealeseens .

The specific L-Iactate consumption ra te of C. testosteron; is expressed as follows .

QLCom = QLmaxCom x U(L + KmLCom) (4)

The specific propionate and acetate consumption rates of C. testosteron; and the L-Iactate consumption rate of V. alcaleseens determined either anaerobically or aerobically were formulated analogously.

26

(i) Determine the actual oxygen consumption rates by the populations of C. testosteron; and Valealeseens in the culture.

To this end, the specific substrate consumption rates were multiplied by the respective biomass concentrations and a n oxygen consumption factor (F) representing the fraction of the oxygen supplied to the culture that was consumcd by the individu al species C. testosteron; and Valealeseens . This fraction was determined by dividing the sum of the specific substrate consumption rates of onc of the populations in the culture by the sum of these rates for the total culture . For example, oxygen consumption by the population of C. testosteron; in the culture becomes

vO,Com = Q0 2Com X xCom x pO>m (5)

For oxygen consumption by the population of V a/ca­leseens, a strictly analogous equation was used.

----- ---

v g

(ii) Determine the actual consumption rates of carbon sources by the C. testosteron; and V. a/ca/escens popula­tions in the culture.

L-Lactate, propionate, and acetate consumption by C. testosteron; is proportional to the amount of O~onsumed by this bacterium. L-Lactate consumption (vL m) by C. testosteron; in the culture was calculated by multiplying vO, Com by the molar amount of L-Iactate used per mole of oxygen consumed (MOLCom) and by the fraction of the Oz consumed by C. testosteron; (FL) for the oxidation of this substrate. This 0 z-L-Iactate consumption fraction was de­termined by dividing the specific consumption rate of L-Iac­tate by the sum of the spccific consumption rates of L-Iac­late, propionate, and acetatc.

vLCom = vOzCom X MaLCom x FL (6)

Rates of propionate and acetate consumption by C. test­osteron; were calculated analogously.

L-Lactate consumption by the population of V. a/ca­/escellS in the culture was assumed to be independent of the availability of oxygen and was calculated accQrding to the following equation.

vL Veil = QL Veil X X Vcil (7)

(iii) Calculate the speeitic growth rates of C. testosteron; and V. alcalescellS at the expense of the individual sub­strates.

The specific growth rales were calculated according to equation 2. For example, for growth of C. testosteron; on L-Iactate, the following equation was used.

~LCom = vLCom x YLCom (8)

At growth-limiting substrate concentrations, it is reason­able to assume that the spccific growlh rate of an organism corresponds to Ihe sum of the spccific growth rates at the expcnse of the individual substrates (10, 12). Thus, the spccific growlh rate of C. testosteron; becomes

fLCom = fLLCom + fLPCom + fLACom (9)

where fL Com is :50.66 h - I, the maximum specific growth rate of C. testosteron; in an aerobic batch culiure on a mixture of L-Iactate, propionate, and acetate.

Inhibition of Ihe growth of V. a/ca/escellS by a, was assumed to be noncompclitive. Thus, Ihe specific growth rate of V. alca/escellS was expressed asrollows.

fL Vc;' = (vL Vc;' x y"'dl)/{l + ([Ozl/KPz VC;')} (10)

where (vL Vc;' x y"'c;') is :50.3 h- I, the maximum specific growth rale of V. alca/escellS on L-Iactate in an anaerobic

Modelling of mixed chemostat cultures

batch culture. In equation 10, KP2 ve;', the inhibition con­stant, rcpresents the oxygen concentration at which Jl. Veil

equals 0.5 X ~max Veil.

Changes in the densities of V. alca/escellS and C. test­osteron; in the chemostat were calculated on Ihe basis of the following general differential equation.

dx/dt = x x (fL - D) (11)

The changes in substrate and product concentrations in the culture were calculated as follows.

ds/dt = supply to the culture - outflow by dilution - con­sumption by V. alcalescellS - consumption by C. tes­

tosteron;

The primary growth substrate, L-Iactate, was supplied to the chemostat from the reservoir medium (i.e., D xL,).

Supply of propionate and acetate was through the fermen­tation of L-Iactate by V. alcalescellS (vL Vd'). Because the stoichiometries of the formation of propionate and acetate from L-Iactate by V. a/ca/escellS depended on the amount of Oz the organism consumed, these were determined experi­mentally in pure cultures of V. a/calescellS at different rates of aeration (this study). The Hz and biomass product ion rates of V. a/calescellS in the mixed culture were calculated on the basis of the same experiments.

Oxygen supply rates in the range of 0.0 to 3.6 mM . h- I

were used in the various mixed·culture simulations. The differential equations were solved numerically by

using the method described by Van der Hoeven and Gottschal (41). This provided automatic adjustment of the time intervals (dt) to avoid neg.tive substrate concentra­tions. The computer program was written in Turbo Pascal version 4.0 and can be executed on IBM-compatible com­puters containing the 8087 mathematical coprocessor. The model can be adapted relatively easily to describe mixed cultures of other aerobic and anaerobic bacteria growing under oxygen limitation.

RESULTS Experiments with pure cultures of V. a/ca/escens and C.

testosteron; were perforrned to obtain information on the essen ti al growth parameters of these bacteria under oxygen­limiting conditions.

Growtb of C. testosteron; on L-Iaetate, propionate, and acetate io batcb aod O,-limited contiouous cultures. During growth of C. testosteron; under oxygen limitation in the ehemostat, the actual oxygen concentration in the culture remained below the detection level of the electrode «0.2 fLM). The molar growth y ield on L-Iactate under these

TABLE 2. Analysis of chemOslat cultures of C. testosteron; grown at various oxygen supply rates on various substratesU

Steady Redox Rate (mM . h- I ) orb:

Carbon state reading Consumption Production recovery YicldC

(mV) 0 , Lact Prop Acet CO, Py, Cell C

(%)

34 1.43 0.91 1.30 0.06 1.17 97 15.4 14 1.24 0.27 0.40 0.13 - d 0.12 1.13 17.0 19 2.15 0.50 0.54 0.28 1.62 0.16 1.33 93 12.1 12 2.57 0.69 0.63 0.53 2.43 0.00 2.05 89 13.3

• Substrates were L·laClale (sleady state 1; S. = 20.0 mM) and a mixture or L·lactate. propionale. and acetale (stcady statcs 2. 3. and 4; Sf = 7.0 mM each) as the carbon and energy source (D = 0.1 h- I

).

h Abbreviations: Laci . L-Iactale; Acet, aCClate; Prop, propionale; Pyr = pyruvate; Cclt C. cen carbon. C Expressed as grams of eeU carbon produced per mole of carbon substrate consumed. ti _, COl analysis nol re liable .

27

Chapter 3

TABLE 3. Chemostat cultures of V. a/calescens in steady state at (hree different oxygen supply (ales in an L-Iactate-limited chemOslatU

Steady Redox Rate (mM · h I) of h:

C ubon stale reading Consumption Production recovery Yiel&'"

(mV) 0, Lact CO, H, Prop Cell C

(%) Acet

-73 0.00 1.79 0.42 0.53 0.83 0.93 0.42 99 2.8 -56 0.35 1.95 0.90 0.31 0.97 0.80 0.58 99 3.6 -56 0.70 1.74 1.33 0.15 1.00 0.57 0.63 109 4.3

" D = 0.1 h- t; Sr = 19.8 mM L-Iactate. The actual oxygen concentrat ion in the culture liquid remained below the detection level of the oxygen electrodes used

«0.2 ~M). b Abbreviations: Lact, L-Iactate: Acet, acetate; Prop, propionate; eell C. ce l! carbon. ~ Expresscd as grams of eell carbon produced pcr mole of l...Iactate consumed.

conditions was 15.4 g of cell carbon per mol of L-Iactate consumed (Tabie 2, steady state 1), which is very similar to that obtained in an oxygen-sufficient batch culture (15.1 g . mol - I). From aerobic batch cultures grown on prop ion­ate and acetate, yield va lues of 18.8 and 10.6 g . mol - I, respectively, were obtained. Growth yields of C. testoster· oni in statically incubated (02-limited) batch cultures were significantly lower than those in well-aerated cultures. Val­ues of 11.7,15.3, and 8.7 glmol of L-Iactate, propionate, and acetate, respectively, were obtained. On the basis of these yield values, the amounts of substrate used per oxygen consumed appeared to be 0.5 (L-Iactate), 0.5 (propionate), and 0.8 (acetate) mol/mol.

The maximum specific growth rate of C. testosteron i in a shake culture on L-Iactate or ~ropionate was 0.54 h - t. The I'-m~ on acctate was 0.43 h- . On a mixture of L-Iactate, propionate, and acetate, the organism grew at a slightly higher rate of 0.66 h- l Wh en grown in spent supernatant of an L-Iactate-grown culture of V. alcaleseens, C. testosteron; had a I'-max of 0.43 h- ' at the expense of propionate and acetate formed during growth of V. alcaleseens .

Hydrogen was not oxidized by C. testosteron i in batch cultures incubatcd under an air-H2 headspace in the pres­ence or the absence of L-Iactate.

Oxygen-limited chemostat cultures of C. testosteron i grown on a mixture of L-Iactate, propionate, and acetate (7 mM each in the reservoir medium) at different oxygen supply rates used all three substrates simultaneously. However, L-Iactate and propionate were consumed to a larger extent than acetate (Tabie 2, steady states 2, 3, and 4). Similar results were obtained with 02-limited and 0 2-sufficient batch cultures grown on mixtures of these substrates (data not shown). Small amounts of pyruvate we re excreted by C. testosteron; grown under oxygen limitation in continuous culture (Tabie 2).

Growtb of V. alcaJeseens in batch and aerated cODtiDUOUS cultures. V. alcaleseens NS.lA9 was grown in an L-Iactate­Iimited chemostat (D = 0.1 h- ' ) at different rates of oxygen supply. The anaerobe consumed the O2 that entered the culture liquid, and steady states were obtained at O2 con­sumption rates of 0.00, 0.35, and 0.70 mM . h- t (Tabie 3). Aeration of the culture resulted in an increase of the con­centrations of relatively oxidized fermentation products (C02 and acetate) and a decrease of more-reduced products (H2 and propionate). The molar growth yield increased significantly with the higher O2 supply rates. On the basis of these steady-state data, the formation of fermentation prod­ucts by V. alcaleseens was described as a function of the amount of oxygen consumed (Fig. 2). Second-order polyno­mials were used to fit data on the fermentation paltern and growth yield of V. alcaleseens to curves describing the response to increased oxygen consumption. The equations

28

of these curves were used subsequently to approximate the changes anticipated in the oxygen-limited mixed cultures.

The maximum specific growth rate of V. alcaleseens on L-Iactate in an anaerobic batch culture was 0.30 h - I, which increased slightly (to 0.34 h- 1

) if measured in spent super­natant of a culture of C. testosteroni af ter readdition of 20 mM L-Iactate.

Growth of V. alcaleseens on pyruvate was analyzed because C. testosteroni produced pyruvate from L-Iactate when this organism was grown under oxygen limitation. In an anaerobic batch culture with pyruvate as the carbon and energy source, V. alcaleseens grew at a rate of 0.12 h- ' . Compared with the results of growth on L-Iactate, the products were more oxidized. One mole of pyruvate was fermented to 0.14 mol of propionate, 0.68 mol of acetate, 0.39 mol of H2' 0.68 mol of CO2, and 0.32 mol of cell carbon. The growth yield was 3.9 g of cell carbon· mol of pyru­vate - I.

EstimatioD of substrate consumption kinetics. Kinetics for the consumption of oxygen and the oxidation of L-Iactate,

0.8 ,..--------------,

, ... u .!!!

g 0.6

] § 0.4

~ .E

~ 0.2

à: 0.0 L-__ -'--__ -'-__ -'-__ -'

0.0 0.1 0.2 0.3 0.4

O. -col'1SllT"Ption (mol. mollacC')

FIG. 2. Changes in the produets of L-Iactate fermentation in steady-state cultures of V. a/calescens grown in continuous cuhure~ (D = 0.1 h- 1, Sr = 19.8 mM L-Iactate) with different aeration fales. Symbols: /'" , propionate; \7, acetate; 0, H , ; 0 , CO,; and 0 , cell carbon. Second-order polynomials were used 10 fit the fermentation pattern.

la

h!"

8 6 3

Je

Jn

:h :r-20

,d te In ld I

Ie iS

e, n. J.

Jr

~ ~ c o

.~

~

300r-----------------------, A

200

§ 100 IN o

OL---~--~-----'----'------'

o 2 4 6 8 10

Time (min)

4.-----------------,

5 10 15 25 30

Acetate concentration (tJ,M)

Modelling of mixed chemostat cultures

30 B

\ 20 ~

10 ~ ~ ~~

0 y~-v-

0 2 4 P 8 10

Time (min)

5r-------------------,---~~

o

4

3

2

o ~~-~4_~~_+_4~~~-~ -30 -20 -10 o 10

Acetate concentration (/.LMl FIG. 3. Respirometric detennination of thc kinetic parameters fOT acetate consumption. (A) Thc change in oxygen concentration upon the

addition of 25 IJ.M acetate was traeed until thc respiratio" fale had returned to the endogenous Tale. (8) Thc change in the concentratio" of acetate during the oxygen uplake experiment was calculated from the tracing in panel A. (C) Thc rale of acetale consumptio" was plotted as a functio" of the acetate concentrations as obtained by measuring thc tangent of the substrate depletion plot (panel B) at different times. (D) A direct linear plot was constructed from the rate-versus-concentration data obtained from panel C. In this case of acetate consumption, a Km of 4.3 ~M and a Qmax of 4.2 ~mol mg of cell carbon - I . h- 1 we re obtained.

acetate, and propionate were determined respirometrically with resting-cell suspensions of C. testosteron; and V. alca· lescens . An example of the determination of the kinetic parameters for acetate consumption by C. testosteron; grown on a mixture of L·lactate, propionate, and acetate in an 0 2-limited chemostat is shown in Fig. 3. An oxygen tracing (Fig. 3A) from which an acetate depletion curve is obtained (Fig. 3B) is subsequently converted into a graph representing the acetate consumption rates at different ace­tate concentrations (Fig. 3C). From such graphs , the maxi-

mum specific consumption ra te (QAma. Com) and the half saturation constant (KmA Cam) for acetate were estimated by using a direct linear plot (Fig. 3D).

The half saturation constants for the consumption of L·lactate, propionate, and acetate by C. testosteron; and L·lactate by V. alcalescens were very similar, ranging from 4.3 to 6.0 ILM (Tab Ie 1). The half saturation constant of V. alcalescens for L-Iactate oxidation, as determined in the biological·oxygen monitor (5.1 ILM), was essentially the same as that measured directly as L-Iactate depletion in an

29

Chapter 3

15.-------------------------,

A

10

5

oL-----~----~----~--~

o 50 100 150 200

02-concentration <.u.Ml

FIG. 4. Plot of oxygen consumpt io" rale versus oxygen concentratio" in the prese nce of excess L-Iactate in a washed eeU suspensio" of V. a/calescens pregrown on L-Iactate in an aeraled chemostat. (A) The plot reveals the apparent presence of high- and low-affini ty systems for oxygen consumption in Ihis organism. (8) Response at very low oxygen concentrations.

anaerobic cell suspension (5.1 IJoM). However, lhe maximum specific L-Iaclate consumption rates of V. alcaleseens mea­sured under aerobic conditions were much lower than those measured under anaerobic conditions (6.9 and 155 IJomol · mgofcell carbon - I . h- I, respectively). Acetale and propionate (and pyruvate) were not oxidized significantly by V. alcaleseens «O.llJomol. mg of cell carbon - I . h- I). The maximum specitk consumption rales of C. testosteron; for L-Iactate (QLmox Cam) and propionate (QP m;"f Cum) were 17.5 and 16.0 IJomol . mg of cell carbon - I. h- , respectively . These rates were approximately four times higher than the maximum specific rate of acetate consumftion (QA,.ax Com), which was 4.2 IJomol . mg of cell carbon - . h- I.

Plots of the respiration ra te versus dissolved-oxygen con­centration of C. testosteroni and V. alcaleseens yielded biphasic cUlVes. In Fig. 4, the rate of oxygen consumption by V. alcaleseens is shown as a function of the oxygen concentration as obtained from oxygen depletion plots. Two distinct Km and Qmax values, apparently representing a high-affinity system (Km = 0.11 IJoM ; Qmax = 8.4 IJomol of O2 . mg of cell carbon - I . h - I) and a low-affinity system (Km = 350 IJoM ; Qmax = 13.2 IJomol of O2 , mg of cell carbon - I . h- I), were estimated. The parameters of the low-affinity system are not very precise, because O2 con­sumption rates at concentrations above air saturation (>250 IJoM) were not deterrnined. The kinetic parameters for oxy­gen uptake by C. testosteroni were in the same range as lhose obtained for V. alcaleseens (Tabie 1). The O2 con­sumption rale of C. testosteron; with a mixture of L-Iactate, propionate, and acetate was within 87 to 98% of the sum of the rates obselVed on these substrates separately (data not shown) . Therefore, in the model calculations, it was as­sumed that the act ua I O2 consumption rate of C. testosteroni was the sum of the rates obtained with L-Iactate, propionate, and acetate.

Estimation of the oxygen inhibition constant of V. alea· leseens. Simulation of a pure culture of V. alealeseens

30

y ielded an accurate description of the experimentally ob­selVed growth pattern (i.e., steady-state fermentation prod­uct and biomass concentrations), provided that the O2 sup­ply rate remained between 0.0 and 0.5 mM . h- I and the oxygen inhibition constant for growth of V. alcaleseens (KP2Ve

;l) was >6 IJoM. At O2 supplies of > 0.5 mM . h- I or with O2 inhibition constants of <6 IJoM , V. alcaleseens could not establish a steady state in the chemostat. Fig. 6 shows the relationships among the specific growth rate of V. alcaleseens , the inhibition constant for oxygen, and the actual oxygen concentration in the culture. In the mixed­culture simulations, a value of 10 IJoM was used .

Modelling of mixed cultures of V. alcaJeseens and C. tes­tosteroni and comparison with experimental observations. A mathematical model of the mixed culture of V. alcalescens and C. testosteron; th at used the experimentally determined growth parameters (Tabie 1) predicted steady states in which lhe !wo organisms coexisted as long as oxygen remained limiting, i.e. , insufficient for C. testosteron; to oxidize all available carbon sou rees in lhe culture. The predicted steady-slatc biomass of C. testosteroni and V. alcalescens in the chemostat (Fig. SA) as weil as the product ion rales of propionate and acelate (Fig. SB) calculated by the model were quite similar to those obselVed experimentally. How­ever, the simulated H 2 production rales we re somewhat higher than those obselVed in the chemostat studies. Al­though the dissolved-oxygen concentration remained below the detection limit of the oxygen probe «0.2 IJoM) in the experimentally obtained steady states, lhe model predicted concentrations of 2:0 .5 IJoM in these mixed cultures.

The model did predict coexistence of both species at higher oxygen supply rales than were obtained in the pub­lished experiments on which the model was based. The biomass of C. testosteron; increased linearly with the amount of oxygen supplied to the culture. lncreased L-Iac­tate concentrations in the reselVoir medium resulted in an increase in V. alcalescens both in the chemostat experiments

5

of ms

lb­ld­Ip­he ons or

lid ws V. he :d-

.s­A

ns ed ch ed all ed in of lel oN­

lat \1-,w he od

at b-]e

1e c­l n

.ts

Modelling of mixed chemostat cultures

20 20 1.0 A

.p, B

1 ~ 0.8 t5

/ 15 ~ V,

I j 1 " 5 0.6

~ I ~ 8 ---<>.

tO 10 ~ ~ lil

0

8 ~ 1 0.4 äi 0

~ 0 I

5 5 0" 0 cl. 0 .2

0 0 0 .0 0 2 3 4 0 2 3 4

O.-SLWly UrM. hO') O.-SLWly (!TM. hO')

2.0 4.0 C 0

'.:

I 1.5 L: 3.0

8 i

t 8 1.0 i

2.0

0 c .! 0

0

! 0.5 I 1.0 Ö

0 .0 2 3 4 0 2 3 4

O.-SLWly UrM. hO') O.-SLWly UrM. hO')

FIG. 5. Model predictions and experimental resuJts of mixed cultures of V. alca/escens and C. testosteroni at various rales of oxygen supply. (A) Steady-state concentrations of biomass of V. alcaleseens and C. testosteron; and dissolved oxygen in an L-lactate·Jimited chemostat (D = 0.1 h- I

• Sr = 19.1 mM L-Iactate). Open symbols (with error bars representing standard deviations) represent the measured cell densities (expressed as eell carbon) of V. a/calescer.s (0 ) and C. testosteron; (0 ). Closed symbols represent the concentrations predicted by the model, using the experimentally determined default parameters given in Table 1. Oxygen concentrations predicted by the model are also shown (4). (B) Measured (open symbols) and predicled (closed symbols) formation rates of propionale (6 ), acelate (\7), and H, (0 ) by the mixed-chemostat cuhure of V. a/ca/escens and C. testosteron; grown on L-Iactate and subjected to several oxygen supply rates. (C) Model calculations of the carbon substrate consumption rates in steady states of mixed cultures of C. testosteron; and V. a/ca/escens. L-Lactate consumption by V. a/ca/escens (+) and C. testosteroni (e). propionate consumption by C. testosteroni (A), and acetale consumption by C. testosteron; (Y) are shown. (0) Model calculations for oxygen consumption rates by the individual bacteria V. a/calescens (+) and C. testosteron; (e) in the mixed-culture steady states .

and in the simulations (not shown). This clearly demon­strated that the mixed cultures were growing under a dual limitation of L-Iactate and oxygen.

For completely anaerobic conditions, the model predicted a complete washout of C. testosteron i and the establishment of a monoeulture of V. a/ca/escens (Fig. SA) . Also, at very low oxygen supply rates «0.025 mM · h - 1

), the model predicted that V. a/ca/escens would complctely outcompete C. testosteroni. Under such conditions, V. alcalescells con-

sumed all L-Iactate and oxygen supplied 10 the culture. However, at higher rates of oxygen supply, most of the oxygen was consumed by C. testosteroni (Fig. 5D) and most of the L-Iactate was consumed by V. a/ca/escens (Fig. SC), resulting in stabie coexistence. Increasing oxygen supply rates resulted in increasing oxygen. L-Iactate. propionale, and acetate consumption by C. testosteroni and decreasing L-Iactate consumption by V. a/ca/escens. Only after most of the propionate and acetate had been used did C. testosteron i

31

Chapter 3

0.3 =:-------,---------------,

1(, . 100 .......

0.2 K, ,. '0 uM

K, = l JAM

0. ' - - - - - - - - - - - - - - - - - - - - - - - -- -

1<.,·0, 1 uM

O .OL-------~~L-______ ~ __ L_~ ______ ~

0 . 1 10 100

02- c onc entrat ,on Cu M)

FIG. 6. Calcu lated specific growth rale of V. alcaleseens at different oxygen concentrations as influenced by different values of the in hibition constant (K,02v e ll

) for oxygen . In the model , a value of 10 ,.M was used (solid line). The dashed line indicates the specific growth rale (0.1 h- 1

) predicted at a dissoJved-oxygen concentration of 20 ,.M.

consurne a substantial part of the L-Iactate supplied to the culture. Together with the increasing level of oxygen (Fig. 5A), this resulted in a decreasing steady-state biomass concentration of V. alcaleseens . At oxygen supply rates above 3.6 mM . h - I

, the anaerobe washed out of the culture and a s teady-state pure culture of C. testosteron i with L-Iactate as the limiting substrate, oxygen in excess, and a biomass concentration of 17.5 mM cell carbon was obtained.

ED'ecls of cbanges in parameter va lues of the model. To s tudy the importanee of the various growth paramete rs for coexistence of V. alcalescells and C. testosteroni, the o ut­come of the model with the measured parameter values (TabIe 1) at an O2 supply rate of 2 mM . h - 1 was compared with the predicted outcome following a change of one of these parameter values. With a 10-fold increase or decrease in the carbon substrate consumption parameters (Qmax or Km values) or yield coefficients (Y) of V. alcaleseens or C. testosteroni , the model still predicted coexistence of the two bacteria but with altered steady-state substrate and biomass concentrations. However, when the O2 consumption C3-

pacity of C. testosteron; was reduced by decreasing the Q02.max Cem from 22.4 to < 15 ,.mol · mg of cell carbon- ! . h - 1 or by increasing the Km,02c om from 0.35 to >5 ,.M, coexistence was no longer predicted. With these paramete r values, the caJculated dissolved-oxygen concen­tration increased to values (> 20 ,.M) which repressed the specitic growth rate of V. alcalescens to a value below the dil ut ion ra te «0.1 h- I

), which can be seen in Fig. 6. Obviously, coexistence also depended on the oxygen inhibi­tion constant of V. a/ca/escens. When the KiOz V e il was reduced from 10 to <0.5 ,.M, V. alcaleseens could no longer maintain itself in the culture.

32

DISCUSSION

The a im of this s tudy was to investigate the interactions bClwccn thc strictly anaerobic V. alca/escens and the obli­gately aerobic C. testosteron i , which in a prelim inary study (8) we re shown to coexist in chemostats operated under dual \imitation of oxygen and L-Iacta te . To this end , a mathemat­ical model was constructed 10 describe the experimenta l mixed cultures in more detail, th us helping us understand the observed pattern of growth in response to various degrees of oxygen supply. The current model is a further elaboration of earlier mode Is that is based on Monod-type growth kinetics and is used 10 describe both pure and mixed continuous cultures (10, 12, 19, 38, 41).

A basic assumption common to most of these models is that the specific growth rate (,.) can be described as a saturating function of the growth substrate concentration (S) with a constant maximum specific growth ra te ("m,.) and a half saturation constant (K.): the fami liar Monod-equation, ~ = "max X S/(Ks + SJ.

Significant modifications of this equation are necessary if more than one substrate simultaneously limits growth of an organism. One way of doing th is is by summation of the specific growth rates on the individual substrates by using the Monod equation (10). However, an important assump­tion to be made is that each individual substrate does not affect the kinetics for growth on the other compounds. This assumption could not be made for the C. testosteroni strain used in the present mixed-culture experiments, as it ap­peared that under oxygen-limiting conditions, L-Iactate and propionate we re used preferentially and very little aceta te was consumed (see also below). Therefore, the present model was based o n the measured values of the kinetic parameters for substrate consumption, which together deter­mine the specific growth rate. These parameters could be obtained in pure cultures of V. alcalescens and C. testoster­oni grown in oxygen-limited chemostats in the presence of the potential substrates in the mixed cultu re. It was then assumed that these subst rate consumptio n kinetics were not s ignificantly different from those in the mixed cultu res. The observation that the maximum specific growth rates of C. testosteroni in spen t supernata nt of V. alcalescells and of V. a/calescens in spenl supernatant of C. testosteron i wcre close to the rates in freshly prepared media provided strong support for the validity of this assumption.

Pure·culture observatioos. To determine the substrate con­sumptio n kinetics of V. alcaleseens and C. testosteroni , pure cultures of these bacteria were grown at a dilution rate of 0.1 h - I under oxygen limitation with either L-Iactate a lone or a mixture of L-Iactale, propionate, and acetate as growth substrate.

Steady states of L-lactate-Iimited pure cultu res of V. alcalescens were obtained at several rates of O2 supply with the O2 concentration in the cu lture liquid remaining < 0.2 ,.M. The consumpt ion of oxyge n by V. alcalescens caused a change in the fermentation pattern from that o f strictly anaerobic cultures , resulting in the format ion of more rela­tively oxidized produets (acetale and CO2 ) and less rela ­tive ly reduced produets (propionate and H2). The increased acetate formation was parallelled by a significant increase in the molar growt h yield on L-Iacta te that presumably was due to increased ATP formation in the metabolic route via acetyl-p hosphate to acctate. Several examples of organisms responding in a similar way have been reported for the genera Lactococcus , Selenomonas, Propionibacterium , Bacteroides, and Actinomyces (2, 5, 6, 30, 36).

ions ,bli­udy ju al nal­nlal Ihe sof n of tics ous

s is s a (S)

,d a on,

y if an

the ing np· not 'his ain ap­Ind ate ent tt ic er­be er· of

len lot 'he C. V. !re ng

,",

lre

1.1 • a 1h

V. Ih

1.2 I a Iy a· a· ~d in Je ia 1S

le 'I,

Pyruvale, a major end producl found by De Vries el al. (7) in cullures of V. a/ca/escens cxposcd la oxygen concenlra­lions higher Ihan Ihose used in Ihis sludy (0.8 10 250 fLM ), was not produced in the steady-state cultures of our strain. The V. a/ca/escens slrain used in Ihe presem sludy appeared la use pyruvale as a growlh subsIrale.

When grown on a mixture of L-Iaclale, propionale, and acetate undcr oxygcn limitation, C. testosteron; consumed all of Ihese subsIrales simullaneously. bul acclale was used at a s ignificantly lower rate. Pyruvate was no l detected in the mixed-cuhure sleady slales, indicaling Ihal V. a/ca/escens may have consumed il. However, pyruvale production by C. testosteroni and subscquent consumption by V. alcalescens wc re not included in the mathemalical model o f such mixed cultures, since it would have caused only minor changes in the predictions of the model.

Measuring Ihe potentia l substrale-oxidizing capacities by respirometry proved useful for estimating kinetic parameters for the consumption of substrates and oxygen (18). The oxidation of all three carbon s ubstratcs (L-Iactale, propion­ate, and acetate) obeyed Michaelis-Menten kinetics, and apparent Km values were in the micromolar range , as ob­served for other bacteria grown on organic acids (27). Moreover, the oxidation kinetics of bath V. a/ca/escens and C. testosteron; could best be described by assuming Ihe presence of more than one Michaelis-Menten type of oxida­tion system with different Km and Qma,.;. values. Such results indicale the presence of a branched electron transport chain and Ihe concurrent presence of (at least) two oxidases (3, 4, 14, 15, 24, 29, 34).

Modelling of mixed cultures of V. alcakscens and C. tes· tosteroni. Using parameter values determined in pure cul­tures, numerical solution of Ihe model fitted the experimen­tally observed changes in mixed chemostal cuhures of V. a/ca/escens and C. testosteron; grown at different oxygen supply rates quite weil. Coexistence was predicted as long as the oxygen supplied to the culture did not exceed the amount needed by C. testosteron; to oxidize all available carbon sources in the culture. The high affinity of C. testosteron; for oxygen ensured that under these conditions, the dissolved­oxygen concentration remained weil below 20 fLM , a value which would bring the specific growth rale of V. a/ca/escens to below the dilution rate. The oxygen concentrations cal· culated in the model (0.5 to 12 fLM) were above the detection limit (0.2 fLM ) ofthe oxygen probe used in the culture . Vet in the mixed cultures, oxygen concentrations always remained below this detection level. Maybe this discrepancy is ex­plained by the fact that the true oxygen affinity of C. testosteron; is even better than was measured with the respirometric met had used. This would result in significamly lower steady·state oxygen concentrations. Although the increased sensitivity of this polarographic method allowed delection of oxygen concentrations as low as 0.01 fLM , the use of (Ieg)hemoglobin as an ahernative met had has indi· cated that oxygen uptake systems with even lower K~ values may be characterized (3, 4).

At oxygen supplies resuhing in consumption rates of more than 3.6 mM . h - I, washout of V. a/ca/escens occurred and L-Iactate became the sole limiting substrate of the cuhure. At very low oxygen supply rates «0.025 mM O2 , h - I

), estab­lishment of a pure cuhure of V. a/ca/escens was predicted. The complete exclusion of the aerobic organism is most likely explained by the fact that the biomass formation of the aerobe does and th at of the anaerobe does not dcpend on the consumption of 02' Bclow a su pply of 0.025 mM . h- I

, this resulls in a culture in which the cell density and hence the 0 ,

Modelling of mixed chemostat cultures

consumplio n of V. alcalescens arc much higher than those of C. testosteron;. This res ults in a concentration of oxygen below the valuc necdcd by C. testosteron i la sustain a spccifk grawlh rale equal 10 the dilution rate . Thus. the paradaxica l si tuation occurs in which an anaerobe outcam· petes an aerobe for oxygen. Experimental data from mixed cuhures are not available to substantiate this theoretical prediction, since oxygen supply rates below 0.025 mM . h - I

wc re not testcd. At oxygen supply rates for which coexistence was pre·

dicted, it appeared that most of the L-Iactate was used by V. a/ca/escens and that C. testosteron; grew predominamly at the expense of the fermentation products of V. a/ca/escens, propionate and acetate. So, V. a/ca/escens appeared the better campetito r for L-Iactate, even under canditions of microaerobic growth. These considerations explain why the biomass of V. a/ca/escens in the mixed·culture steady states did not vary much , whereas that of C. testosteron; increased Iinearly with increasing oxygen supply.

Changing the growth parameters of the rwo species used in the model revealed which parameters were most significani in determining coexistcnce of the two organisms. It appeared that coexistence is ruled predominantly by the inhibilion of V. a/ca/escefls by oxygen in combination with the high O2

consumption capacity of C. testosteron i , a situatian vcry similar ta that encountered in mixed cultures of Actinomyces viscosus and Streptococcus mutans grown under oxygen­limiting conditions (4 1). Coexistence of these organisms under a dual limitation of oxygen and glucose depended on consumption of oxygen by A. viscosus and inhibition o f the growth of S. mutans by oxygen. In general, continuous growth of an anaerobe in an aerated chemostat is possible only if the consumption of oxygen by bacteria present in the culture is sufficient to ensure a dissolved·oxygen concentra­tion low enough to a llow for aspecific growth rate exceeding the dilution rate of the culture. Of course, the oxygen­consuming bacteria may be the (oxygen·tolerant) anaerobic organisms themselves , or they may be accompanying (fac· ultative ly) aerobic bacteria.

The results of th is s tudy demonstrate that the combined use of chemostat experiments and a malhematical model can help us understand interactions between aerobic and anaer· ob ic bacteria and better de fine the conditions allowing such organisms to grow as stabie mixed cultures.

Further investigations of species with other characteristics (e.g., anaerobes more sensitive to oxygen than V. a/ca· lescens) are essential for obtaining full confidence in the use of such mode Is and for evaluating the importance of these results for the growth of various aerobic and anaerobic bactcria in the environment.

ACKNOWLEDGMENTS

This work was supported by the Netherlands Integrated Soil Resea rch Programme.

We thank G. Berrelkamp-Lahpor ror technical assistance.

REFERENCES L Alldredge, A. L., and Y. Cohen. 1987. Can microscale chemica I

patches persist in Ihe sea? Microelectrode study of marine snow, faecal pelIeis. Science 23S:689-<i91.

2. Bentzen, G., and H. Larsen. 1989. Oxygen activation and defence against oxygen toxicity in a psychrophilic bacle­roidaceae. Arch. Microbiol. 151:95- 100.

3. Bergersen, F. J., and G. L. Turner. 1980. Properties of terminal oxidase systems of bacteroids from root nodules of soybean and cawpea and of Nz·fixing bacleria grown in conlinuous culture. J . Gen. Microbiol. 118:235- 252.

33

Cbapter 3

4. Bergerse n ~ F. J ., and G . L. Turner. 1988. Glulamate as a carbon SQurce for N2- fixing bactcroids prepared from soybea n root nodules. J. Gen. Microbiol. 134:2441- 2448.

5. Condon, S. 1987. Responses of lactic acid bacteria to oxygen. FEMS Microbiol. Rev . 46:269-280.

6. Oe Jong, M. H., J . S. Van der Hoeven, C. W. A. Van den Kieboom, and P. J . M. Camp. 1988. Effec ts ol oxygen on the growlh aod metabolism of Acr jnomyces lliscosus. FEMS Micro­biol. Ecol. 53:45- 52 .

7. De Vries, W., C. Donkers, M. Boellaard, aod A. H. Stouthamer. 1978. Oxygen me tabolism by the anaerobic bacterium Vei/· lonel/a alcaleseens. Arch. Microbiol. 119:167- 174.

8. Gerritse, J. , F. Schut, and J . C. Gottschal. 1990. Mixed chemo­SIal cultures of obligately aerobic aod fermentative or methano­genie bacteria grown under oxygen limi ting condit ions. FEMS Microbiol. Lett . 66:87-94.

9. Gottschal, J. C. 1985. Occurrence aod func lioning of anaerobic bacteria in oxidized environments , p. 263-273. In V. Jensen, A. Kjö ller, and L. H. S0rensen (ed.), Microbial communities in soil. Thirty-third symposium of the Federation of European Microbiologica l Societies. Elsevier Science Publishing Co., London.

10. Gottschal, J. C. 1986. Mixed substrate ut ilizat ion by mixed cu ltures, p. 261- 292. In J. S. Poindexter and E. R. Leadbetter (ed.), Bacteria in nature, vol. 2. Plenum Publishing Corp., New Vo rk.

11 . Gottschal, J . C., aod R. Szewzyk. 1985. Growth of a facultative anaerobe under oxygen-limit ing conditions in pure cul ture wi th a sulfate-reducing bac terium. FEMS Microbiol. Let t. 31:159-170.

12. Gottschal, J . C. , and T. F. Thingstad. 1982. Mathematical description of competi tion between two and th ree bacteria l species under dua l substrate limi tation in the chemosta t: a comparison with experimental data . Biotechnol. Bioeng. 24: 1403- 1418.

13. Häggstrom, M. H., and M. Dostàlek. 1981. Regulatio n of a mixed culture of Streptococcus lactis and Saccharomycopsis fibuliger . EuT. J . Appl. Biotechnol. 12:216-219.

14. Harrison, D. E. F. , D. G. MeLennan, and S. J . Pirt. 1970. Responses of bacte ria to dissolved oxygen ten sion, p. 117- 144. In D. Perlman (ed.), Fermentation advances. Academic Press, loc. , New Vork.

15. Henry, M. F., and P. M. Vignais. 1978. Induction of cytochrome din the plasma membrane of Paracoccus denitrifica ns. FEBS Lett. 100:41-46.

16. Jones, J. G. 1982. Activities of aerobic and anaerobic bacteria in lake sediments and their effect on the water column, p. 107-145. In D. B. Nedwe ll and C. M. Brown (ed .), Sediment mic robiol· ogy. Academic Press, Inc ., London.

17. Laanbroek, H. L. , H. J. Geerlings, A. A. C. M. Peynenburg, and J . Siesling. 1983. Competition for L- Iacta te between De­sulfovibrio , Veillonella, and Acetobacterium species isolated lrom intertidal sediments. Microb . Ecol. 9:341- 354.

18. Legan, J . 0., and J. D. Owens. 1987. Determination of growth parameters of methy lamine-using bacteria . J . Gen. Microbiol. 133:1075-1080.

19. Legan, J. D., J. D. Owens, and G. A. Chilvers. 1987. Compet i­tion between specialist and generalist methylotrophic bacteria for an intermitte nt supply of methylamine . J. Gen. Microbio l. 133:1061- 1073 .

20. Loesche, W. J ., F. Gusberti, G. Mettraux, T. Higgins, and S. Syed. 1983. Relationship be tween oxygen tension and subgingi­val bacterial flora in untreated human periodontal pockets . Inlect. Immun. 42:659-667.

21. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265- 275 .

22. Murray, J. W. , H. W. Jannasch, S. HOl\Îo, R. F. Anderson, W. S. Reeburgh, Z. Top, G. E. Friederich, L. A. Codispoti, and E. lzdar. 1989. Unexpected changes in the oxic/anoxic interface

34

in the Black Sea. Natu re (London) 338:4 11-413. 23. Nanninga. H. J ., and J . C. Gotlschal. 1985. Amino acid fe rmen·

tat io n and hydroge n transfe r in mixed cu ltu res. FEMS Micro­biol. Ecol. 31:261- 269.

24. Niekus, H. G. D. 1980. Oxygen consumption by Campylobacler spulOfum subspecies bubulus with formate as subst rate . Ph .D. thesis. University of Amsterdam, Amsterdam. The Nether· lands.

25. Nielsen, L. P., P. B. Christensen, N. P. Revsbech, and J. Sorensen. 1990. Denitrifica tion and oxygen respiratian in bio­films studied with a microsensor fa r nitrous oxide and oxyge n. Microb. Ecol. 19:63- 72.

26. Ohla, H. , J. C. Gottscbal, K. Fukui, and K. Kato. 1990. In terre lat ionships between Wolinella recta and Srreptococcus sanguis in mixed continuous cu ltures. Microb. Ecol. Health Dis. 3:237-244.

27. Owens, J. 0., and J. D. Legan. 1987. Determination of Monod substrate sa turation constant fo r microbial growth . FE MS Mi­crobiol. Rev. 46:419-432.

28. Pirt , S. J. 1975. Princ iples of microbe and ce ll c ultivation. Blackwell Scientific Publications, Ltd ., Londa n.

29. Poole, R. K. 1983. Bacterial cytochrome oxidases. A structur­ally and fun ct ionally diverse group of electron-transfer prote ins. Biochem . Biophys. Acta 726:205- 243.

30. Pritchanl , G. G., J . W. T. Wimpenny, H. A. Morris, M. W. A. Lewis, and O. E. Hughes. 1977. Effects of oxygen on Propioni. bacrerium shennanü grown in continuous culture. J . Gen. Microbiol. 102:223- 233.

31. Revsbech, N. P. , and B. B. Jorgensen. 1986. Microelectrodes: thei r use in microbial ecology. Adv. Microb. Eco!' 9:293- 352.

32. Revsbech, N. P., B. B. Jergenseo, aod T. H. 8lackbum. 1980. Oxygen in the sea bottom measured with a microelectrode. Science 207:1355- 1356.

33. Revsbech, N. P., 8 . 8. JtJrgensen, T. H. Blackburn, and Y. Cohen. 1983. Mic roelectrode s tudies of the photosynthesis and 02' H2S and pH profiles of a microbial mat. Limno!. Oceangr. 28:1062- 1074.

34. Rice, C. W., and W. P. Hempfting. 1978. Oxygen· limited con· tinuous c ulture and respiratory energy conservation in Esche­richia coli . J . Bacterio l. 134:115- 124.

35. Rudd, J. W. M., and R. D. Hamilton. 1978. Methane cycl ing in a eut rophic shie ld lakt and its effects on whole lake metabolism. Limnol. Oceangr. 23:337-348.

36. Samah, O. A., and J. W. T. Wimpenny. 1982. Some effects ol oxygen on the physiology of Selenomonas ruminantium WPL15111 grown in continuous c ultu re. J . Gen. Microbiol. 128:355-360.

37. Scoll, R. I. , N. YarleU, K. Hillman, and D. Lloyd. 1983 . Presence of axygen in rumen liquor and its effecl upon met ha­nogenesis. J . Appl. Bacterio l. 55:143- 149.

38. Sinclair~ C. G., aod D. N. Ryder. 1975. Models fo r the cont in­uous culture of microorganisms under both oxygen and ca rbon limit ing condi tions . Bio technol. Bioeng. 27:375- 398.

39. Smith, K. A., and J . R. M. Anh. 1985. Anaerobic micro­environments in soil and the occurrence of anaerabic bacteria , p. 247- 261. In V. Jensen, A. Kjölle r, and L. H. S0rensen (ed .), Microbial communities in soi l. Thi rty-thi rd symposium of the Federation of European Microbiological Societies . Elsevier Science Publishing Co., Londo n.

40. Sweerts, J.·P. R. A. 1990. Oxygen consumption processes, mineralization and nitrogen cycling at the sediment-wate r inter· face of north tempe rate lakes. Ph .D. thesis. University of Groningen, Groningen, The Netherlands.

41. Van der Hoeven, H. S., and J. C. Gollschai. 1989. Growth ol mixed cultures of Act;nom;ces viscosus and S treptococcus mUlans under a dual limitation of glucose and oxygen. FEMS Microbiol. Ecol. 62:275- 284.

42. Wimpe:nny, J. W. T. 1981. Spatial orde r in microbial ecosys­tems. Bio l. Rev. 56:295- 342.