Effects of cultural conditions

on denitrification by
Pseudomonas stutzeri
measured by Membrane Inlet
Mass Spectrometry
J.R. Firth* and C. Edwards
School o f B iological S ciences, U niversity o f L iverpool,
L iverpool, U K 7056/02/99: received 1February 1999 and accepted 27
April 1999
J .R.F I RTH AN D C.E DWARDS.1 999. Denitrification is a
globally important process leading to loss of fertiliser
efficiency and the production of the greenhouse gas
nitrous oxide and nitric oxide, an ozone depleter.
Membrane inlet mass spectrometry (MIMS) was
employed to study the effect of different variables on
the process of denitrification by P seudomonas stutzeri
in a defined salts medium. MIMS was used for
concomitant measurements of nitrous oxide, nitrogen
and oxygen and showed that denitrification occurred
in the presence of dissolved oxygen. A nitrate
concentration of 15 mmol l#1 and a nitrite
concentration of 5 mmol l#1 were found to be
optimum for complete denitrification of nitrate or
nitrite to nitrogen and varying these concentrations
had a marked effect on the ratio of gaseous products
released. Denitrification products were also dependant
on pH with neutral or alkaline conditions being best
for production of gaseous end products. Our results
suggest that under nutrient rich conditions the most
important factor in the regulation of denitrification by
Ps. stutzeri is the amount of nitrite generated at the
first enzymatic stage of the process. This appears to
cause inhibition of the denitrification pathway above 5
mmol l#1 and at high enough concentrations
(15 mmol l#1) restricts growth.
INTROD UCTION
Denitrification is an important step in the nitrogen cycle in
which nitrate (NO3$), is converted via nitrite (NO2$), and
nitric oxide (NO), to nitrous oxide (N2O), or dinitrogen gas
(N2). It is the only biological mechanism by which N2 is
returned to the atmosphere (Ferguson 1994). The process is
carried out by a variety of b acterial species, which are found
throughout natural environments. NO3# is used in place of
oxygen (O2), as a terminal electron acceptor during respiration
(Stouthamer 1991). Because this is an energetically unfavour-
able option, it has often been suggested that denitrification is
a strictly anaerobic process (Tiedje etal. 1982). The beneficial
aspects of denitrification include control or bioremediation of
NO3# contaminated waters, which can cause eutrophication
(Ku$ cera et al. 1986). However, it has also been linked to
stomach cancer (O%Donnell and Edwards 1992), and b lue
baby syndrome (Mirvish 1985). Denitrifiers may play an
important part in the b reakdown of various hydrocarbon
Correspondence to: J .R. Firth, School of Biological Sciences, University of
Liverpool, Life Sciences Building, Crown Street, Liverpool, L69 7ZB
(e-mail: jfirth@liverpool.ac.uk).
compounds (Evans et al. 1991; Altenschmidt and Fuchs
1992). In agricultural areas denitrification leads to a loss of
fertiliser efficiency (Goulding et al. 1993), and incomplete
denitrification can lead to the release of NO and/or N2O,
both of which are known to be involved in the formation of
acid rain, ozone depletion and global warming (Knowles 1982;
Mosier et al. 1991; Culotta and Koshland 1992).
The method used here to study denitrification was mem-
brane inlet mass spectrometry (MIMS). This technique has
been used to study various microbial processes, such as
methanogenesis, nitrification, and the rumen ecosystem, as
well as denitrification (Lloyd and Scott 1983). The principles
behind MIMS have been described elsewhere (Lloyd et al.
1986), and it provides continuous, sensitive and virtually
noninvasive measurements of gases and volatiles in the gas or
liquid phase.
MIMS was employed here to assess the effect of different
parameters on denitrification by Ps. s tutzeri, a known deni-
trifier, which is widely disseminated in many environments
(e.g. van Rijn et al. 1996). It possesses the full complement
of denitrifying enzymes and is therefore capable of producing
inert nitrogen as the end product.
MATE RIALS AND METHO DS
Bacteria and growth conditions
Pseudomonas stutzeri NCIMB 11056 was maintained on nutri-
ent agar (Amersham, Bury, UK) at 4 °C and subcultured
every three weeks. Seed cultures were grown from individual
colonies inoculated into 50 ml of minimal salts medium in
static 250 ml conical flasks at 30 °C. The salts medium con-
tained (g per litre): sodium succinate, 27; KNO3, 1; K2HPO4,
1; (NH4)2SO4, 1; Na2SO4, 0·054; CaCl2, 0·05; MgCl2, 0·025;
and 1ml trace element solution containing (g per litre):
MnCl2, 1; FeSO4, 1; ZnSO4, 1(BDH). The final pH was 7·0.
Cells for inocula were collected by centrifugation and washed
three times in equal volumes of 25% (w/v) Ringer#s solution
to prevent carry over of nutrients.
Salts medium experiments
Pseudomonas stutzeri was grown in a 2-L fermentor (LH
Fermentor Products, Reading, UK), with temperature, pH
and stirrer control. The salts medium was amended with
different nitrate or nitrite concentrations, and pH varied as
required. Washed cell suspensions were inoculated to give an
initial OD550 of ¼ 0· 1.Temperature was maintained at 30 °C
with a stirring rate of 200 r.p.m. to keep the cells in suspen-
sion. A reservoir of 5 mol#1 HCl was used to control pH
when necessary. No oxygen other than that diffusing from
the atmosphere was added.
Analytical methods
Growth was monitored by measuring OD550 of culture sam-
ples and cell free supernatant fluids were used to analyse
nitrite concentration by the hydrazine reduction method
(Hall 1984). Gas analysis was carried out using a quadrupole
membrane inlet mass spectrometer (Hiden Analytical, War-
rington, England), that had a 4-probe manifold capable of
sequential sampling. The probes were covered with a silicon
rubber membrane and had an external diameter of 0·7 mm.
Gases were drawn into the probe b y a vacuum and ionized
by an electron beam. The ions produced were then sorted
according to their mass/charge ratio (m/z). Gases producing
predominant ions of the same m/z ratio (e.g. CO2 and N2O
give ions which both have an m/z-value of 44), were then
distinguished by studying additional m/z channels which
measure minor ion peaks produced from these gases. To
distinguish between CO2 and N2O gases, CO2 was measured
at m/z channel 12 where it causes a peak corresponding to
6% of its contribution at channel 44. By multiplying up this
value to give a 100% value and subtracting this amount from
channel 44, the contribution made by N2O could then be
calculated. These data were obtained from a mass spectral
cracking pattern (Lloyd and Scott 1983). The data were
continuously stored on computer disk and analysed ina spre-
adsheet at the end of the experiment using Excel 4·0 (Micro-
soft).
RESU LTS
Denitrification during growth in the defined medium.
Denitrification was followed during batch growth of Ps. s tut-
zeri with 10 micromoles l#1 nitrate and pH controlled to a
value of 7 (Fig. 1). Under these growth conditions the oxygen
concentration in the media remained at a level of approxi-
mately 25$50 mmol l#1 throughout the experiment. Deni-
trification occurred during the exponential growth phase and
was monitored as nitrogen production. As the culture entered
exponential phase, nitrite levels rose to a peak concentration
of approximately 7 mmol l#1 after 9 h; subsequently, they fell
to undetectable levels by 13 h. Nitrogen production continued
Fig. 1 Denitrification by a 14h-batch culture of P seudomonas
stutzeri in a defined salts medium measured using Membrane Inlet
mass Spectrometry. The gas traces illustrated are nitrogen ##,
nitrous oxide $ $$$, carbon dioxide ······ and oxygen $ - $ -$. Also
illustrated are nitrite concentrations detected in the medium
#E#, and growth as measured as OD550 #R#.
to rise during this period, along with nitrous oxide and carbon
dioxide, but no nitric oxide could b e measured. When nitrite
levels fell, nitrogen production continued to a peak value of
approximately 7 mmol l%1 at 11h. At the end of exponential
growth, maximum levels of carbon dioxide and nitrous oxide
were detected. Nitrous oxide could still be measured even
when net nitrogen production had finished. When the experi-
ment was repeated without pH control so that the pH rose
as growth occurred, less nitrous oxide was released and higher
concentrations of nitrogen were detected (results not shown).
Effect of nitrate and nitrite concentration
Conditions for denitrification were optimized by varying
some of the parameters for the experiment described in Fig.
1. Peak nitrogen and nitrous oxide concentrations, used as a
measure of efficiency of nitrogen production, revealed that a
nitrate concentration of approximately 15 mmol l#1 was opti-
mal for maximum nitrogen production; above this value
amounts of nitrogen produced fell (Fig. 2a). Nitrous oxide
Fig. 2 The effect of nitrite and nitrate concentrations on the
production of nitrogen # E#, and nitrous oxide #$# by
Pseudomonas. stutzeri during denitrification in a defined salts
medium. Data are representative of typical experiments.
production was also maximal at 15 mmol l%1 nitrate; however,
unlike nitrogen, the amount of nitrous oxide measured only
fell slightly to a concentration of approximately 420 micro-
moles l%1 even when the nitrate supplied was at 100 mmol
l%1. In contrast, an optimum of 5 mmol l%1 nitrite was found
for maximum nitrogen and nitrous oxide production. This
value corresponded well with the peak nitrite concentration
(7 mmol l%1) measured in the medium during denitrification
at the optimum nitrate concentration of 15 mmol l%1 (Fig.
2b). Nitrite accumulation in the medium over the range of
nitrate concentrations examined increased to a maximum of
about 20 mmol l#1 incultures grown with 30 mmol l#1 nitrate
(Fig. 3). Above 30 mmol l#1 nitrate, the levels ofnitrite which
could be detected in the medium fell steadily to a value of ¼
5 mmol l#1 when the nitrate concentration supplied was 250
mmol l#1.
Although nitrogen production was inhibited at nitrate con-
centrations above 15 mmol l#1 this was not true for growth.
The final stationary phase biomass, measured as OD550, of
the culture after 24 h incubation was highest at 30 mmols l#1
nitrate and only above this concentration was growth yield
reduced. Reduced biomass values after 24 h incubation
resulted in higher residual oxygen concentrations (Fig. 4 a).
During denitrification using nitrite, final OD550 values
remained fairly constant up to 15 millimoles l#1 and then fell.
In general, cultures grown on nitrite as terminal electron
acceptor produced lower cell concentrations compared with
ones grown on nitrate. Residual oxygen concentrations rec-
orded during the experiments remained at approximately 50
Fig. 3 Maximum nitrite concentrations detected during
denitrification by P seudomonas stutzeri in a defined salts
medium containing different initial nitrate concentrations. Data
are taken from representative experiments.
Fig. 4 The effect of nitrite and nitrate concentration on
concentrations of residual oxygen # $#, and growth as
measured as OD550 #R# of Pseudomonas stutzeri during
denitrification in a defined salts medium. Data are taken from
representative experiments.
micromoles l%1 until growth yield became reduced, at which
point its concentration rose noticeably (Fig. 4b).
Effect of pH
An optimum pHof 7 was found for both nitrogen and nitrous
oxide production (Fig. 5a), although the process proceeded
more efficiently when no pH constraint was applied and the
medium allowed to b ecome more alkaline throughout the
experiment, as a consequence of denitrification. Cor-
responding growth values measured as OD550 also increased
up to pH7 at which point they remained fairly constant (Fig.
5b). A more acidic medium caused a more severe reduction
in nitrogen production than nitrous oxide production.
DISCUSSION
Nitrite production and accumulation in the medium together
with nitrous oxide and nitrogen were always detected in
Fig. 5 The effect of pH on (a) the production of nitrogen
#E#, and nitrous oxide #?# b y and (b) growth of
Pseudomonas stutzeri during denitrification ina defined salts medium.
Data are taken from representative experiments.
denitrifying cultures of Ps. s tutzeri. Nitrite rose during
exponential growth inthe presence ofnitrate, but disappeared
before stationary phase was reached. Other work has shown
that Ps. stutzeri released nitrogen only, from nitrate without
any detectable nitrous oxide, and that its production occurred
before any nitrite could b e detected (Carlson and Ingraham
1983). Xu and Enfors (1996) found that nitrate availability
over the 24 h prior to inoculation had a marked effect on the
transient accumulation of nitrite due to a need to synthesize
new nitrite reductase. Different volatile fatty acids used as
carbon sources have also been shown to influence nitrite
accumulation in Ps. stutzeri (van Rijn etal. 1996), and pH
influences the process in Paracoccus denitrificans (Ku# cera
1986). K o¨rner and Zumft (1989) found that nitrate inhibition
of nitrite reductase could lead to a nitrite build up in Ps.
stutzeri. Nitrate reduction occurred more rapidly than nitrite
reduction and this did seem to have an inhibitory effect on
nitrous oxide reductase. This probably explains the shoulder
seen in the nitrogen production curve in Fig. 1at around 4$
8 h where the rate of nitrogen production was temporarily
reduced until the nitrite levels in the medium began to fall.
Gonz ´alez et al. (1994) have previously demonstrated the inac-
tivation of nitrate reductase by nitrite in the yeast Hansenula
anomala. Although the starter cultures used inthis study were
all grown aerobically overnight, denitrification commenced
almost immediately indicating that the enzymes required for
denitrification were already present or synthesized very rap-
idly as oxygen levels fell. This is contrary to the work of
Blaszczyk (1992), who showed that nitrite was utilized b efore
the production of nitrogen and without any nitrous oxide
being detected. In Ps. s tutzeri, nitrous oxide and nitrogen
production occurred simultaneously with concentrations of
nitrogen peaking around 2 h before nitrous oxide reached a
maximum. The nitrogen peak corresponded to the maximum
amounts of carbon dioxide detected, but began to fall at the
end of exponential growth. If complete conversion of each of
the intermediates of the denitrification pathway had occurred
before utilization of the next, peak nitrogen levels would have
been expected after that of its precursor nitrous oxide. The
timing ofmaximum amounts ofnitrous oxide could be explai-
ned by the observation that all nitrite had disappeared from
the medium at this point.
Nitrate concentration affected denitrification, both directly
and indirectly. Although denitrification was inhibited at
nitrate concentrations higher than 15 mmol l# 1, higher cell
concentrations were measured above this concentration,
despite denitrification being a growth-related process. T his
can be explained by different rates of nitrogen production
over the range of nitrate concentrations. Nitrite accumulation
may be the most critical step during denitrification in this
bacterium. Reduction in growth yield at nitrate con-
centrations in excess of 30 mmol l#1 is probably due to cells
reducing nitrate to nitrite concentrations sufficient to inhibit
growth and thus reducing denitrification further. Nitrite is a
well-documented bacterial inhibitor and is even used as a
preservative.
The optimum level of nitrate in this study (15 mmol l# 1),
was similar to the lower optimum of approximately 20 mmol
l#1 found by Thomas et al. (1994) for denitrification by Ps.
aeruginosa and Ps. f luorescens, but m uch lower than the value
of 100 mmol l#1 found for denitrifying Ps. aeruginosa reported
by Williams et al. (1978). Our failure to detect nitric oxide in
the medium is not a limitation of MIMS as a technique,
because we have been able to distinguish nitric oxide from
nitrous oxide and nitrogen during denitrification by an
unidentified denitrifier isolated from aquatic environments
(Firth and Edwards, Unpublished Data). As no nitric oxide
was detected it appears that the nitric oxide reductase is
extremely efficient in Ps. s tutzeri, as reported by Goretski
et al. (1990), producing nitrous oxide at approximately the
same time as nitrite is produced. Similarly, there was no
inhibition of nitrite reductase which would have led to a
reduction in nitrous oxide concentrations. One possibility is
nitrite inhibition of the nitrous oxide reductase leading to a
build-up of nitrous oxide. At elevated concentrations this
nitrous oxide may then exert a negative feedback effect upon
the nitrite or nitrate reductases, switching off the pathway.
Schulthess etal. (1995) found that high nitrite concentrations
affect all denitrification enzymes leading to a build up of
nitric and nitrous oxides in a denitrifying activated sludge.
The most preferred situation, where the highest possible
concentration of nitrate is removed from the medium with
the least amount of nitrous oxide released to the atmosphere,
is one in which the nitrate concentration is at 10 mmol l#1.
Since nitrous oxide concentrations produced did not fall as
quickly as nitrogen concentrations at nitrate concentrations
above optimum, a higher percentage of the nitrate was
released into the atmosphere as this harmful gas. Nitrite
concentration was not quite as critical in these terms, since
little difference was seen in the ratio of nitrogen to nitrous
oxide production when the culture was grown in the presence
of between 5 and 15 mmol l#1 nitrite.
Although it has been suggested that denitrification is a
strictly anaerobic process, this work clearly showed that Ps.
stutzeri is capable of carrying out the process in the presence
of oxygen. The use of a stirred fermentor vessel reduced
the possibility of anaerobic microniches occurring. Oxygen
concentrations remained approximately constant throughout
our experiments as an equilibrium was reached between dif-
fusion into solution from the headspace and consumption b y
the respiring cells. The culture as a whole, and perhaps even
individual cells, appeared capable of utilizing oxygen and
nitrate as electron acceptors simultaneously, despite the latter
being less energetically favourable.
Membrane inlet mass spectrometry is proving to b e a
valuable tool for studying elemental cycling. The ability to
provide real time data provides new evidence in support of
existing ideas on how denitrification is regulated within the
intact cell and these studies show the complexity of inter-
actions between the denitrification enzymes, their products
and the surrounding environment.
ACKNOWLE DG EM ENTS
This work was supported by a studentship from the Natural
Environment Research Council, UK.
REFER ENCES
Altenschmidt, U. and Fuchs, G. (1992) Anaerobic toluene oxidation
to benzyl alcohol and benzaldehyde in a denitrifying P seudomonas
strain. Journal of Bacteriology 174, 4860#4862.
Blaszczyk, M. (1992) Comparison of denitrification by P aracoccus
denitrificans, P seudomonas stutzeri and Pseudomonas aeruginosa.
Acta M icrobiological Pologica 41, 203#210.
Carlson, C.A. and Ingraham, J .M. (1983) Comparison of deni-
trification b y Pseudomonas s tutzeri, Pseudomonas aeruginosa and
Paracoccus denitrificans. Applied Environmental M icrobiology 45,
1247#1253.
Culotta, E. and Koshland, D.E. (1992) NO news is good news.
Science 258, 1862#1865.
Evans, P.J., Mang, D.T. and Young, L.Y. (1991) Degradation of
toluene and m-xylene and transformation ofo-xylene by denitrify-
ing enrichment cultures. Applied and Environmental M icrobiology
57, 450#454.
Ferguson, S.J. (1994) Denitrification and its control. Antonie van
Leeuwenhoek 66, 89#1 10.
Firth and Edwards Unpublished data.
Gonz ´alez, C., Gonz ´alez, G., Avila, J . et al. (1994) Nitrite causes
reversible inactivation of nitrate reductase in the yeast Hansenula
anomala. M icrobiology 140, 2633#2637.
Goretski, J., Zafiriou, O.C. and Hollocher, T.C. (1990) Steady-
state nitric oxide concentrations during denitrification. Journal of
Biological Chemistry 265, 11535#11538.
Goulding, K.W.T., Webster, C.P., Powlson, D.S. and Poulton,
P.R. (1993) Denitrification losses of nitrogen fertiliser applied to
winter wheat following ley and arable rotations as estimated by
acetylene inhibition and 15N balance. Journal of Soil Science 44,
63#72.
Hall, G.H. (1984) Measurement of nitrification rates in lake sedi-
ments. Comparison of the nitrification inhibitors nitrapyrin and
allylthiourea. M icrobial Ecology 10, 25#36.
Knowles, R. (1982) Denitrification. M icrobiological Reviews 46, 43#
70.
K o¨rner, H. and Zumft, W.G. (1989) Expression of denitrification
enzymes in response to the dissolved oxygen level and respiratory
substrate in continuous culture of P seudomonas s tutzeri. Applied
and E nvironmental M icrobiology 55, 1670#1676.
Kucera, I., Maty ´asek, R. and Dad´ ak, V. (1986) The influence of pH
on the kinetics of dissimilatory nitrite reduction in Paracoccus
denitrificans. Biochimica et Biophysica A cta 848, 1#7.
Lloyd, D., Davies, K.J.P. and Boddy, L. (1986) Mass spectrometry
as an ecological tool for in s itu measurement of dissolved gases in
sediment systems. FEMS M icrobiology Ecology 38, 11#17.
Lloyd, D. and Scott, R.I. (1983) Direct measurement of dissolved
gases in microbial systems using inlet mass spectrometry. Journal
of Microbiological Methods 127, 313#328.
Mirvish, S.S. (1985) Gastric-cancer and salivary nitrate and nitrite.
Nature 315, 461#462.
Mosier, A., Schimel, D., Valentine, D. et al. (1991) Methane and
nitrous oxide fluxes in native, fertilised and cultivated grasslands.
Nature 350, 330#332.
O$Donnell, C.M. and C., Edwards. (1992) Nitrosating activity in
Escherichia coli. FEMS M icrobiology Letters 95, 95#98.
van R ijn, J., Tal, Y. and Barak, Y . (1996) Influence of fatty acids on
nitrate accumulation by a Pseudomonas stutzeri strain isolated from
a denitrifying fluidised bed reactor. Applied and Environmental
Microbiology 62, 2615#2620.
Schulthess, R.V., K u¨hni, M. and Gujer, W. (1995) Release of nitric
and nitrous oxides from denitrifying activated sludge. Water
Research 29, 215#226.
Stouthamer, A.H. (1991) Metabolic regulation including anaerobic
metabolism in Paracoccus denitrificans. Journal of Bioenergetics
Biomembranes 23, 163#185.
Thomas, K.L., Lloyd, D. and Boddy, L. (1994) Effects of oxygen,
pH and nitrate concentration on denitrification by P seudomonas
species. FEMS M icrobiology Letters 118, 181#186.
Tiedje, J .M., Sexstone, A.J., Myrold, D.D. and Robinson, J .A.
(1982) Denitrification: Ecological niches, competition and
survival. Antonie van Leeuwenhoek 48, 569#583.
Williams, D.R., R owe, J .J., Romero, P. and Eagon, R.G. (1978)
Denitrifying P seudomonas aeruginosa: some parameters of growth
and active transport. Applied and Environmental M icrobiology 36,
257#263.
Xu, B. and Enfors, S.-O. (1996) Influence of nitrate starvation
on nitrite accumulation during denitrification by P seudomonas
stutzeri. Applied M icrobiology and Biotechnology 45, 229#235.