Subscriber access provided by SELCUK UNIV

Article

Modeling Nitrous Oxide Production and Reduction in Soil Through Explicit Representation of Denitrification Enzyme Kinetics Jianqiu Zheng, and Paul V. Doskey Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/es504513v • Publication Date (Web): 14 Jan 2015 Downloaded from http://pubs.acs.org on January 20, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

1 2 3 4 5 6 7 8 9

Environmental Science & Technology

Modeling Nitrous Oxide Production and Reduction in Soil Through Explicit Representation of Denitrification Enzyme Kinetics Jianqiu Zheng† and Paul V. Doskey*†,‡,§ †

Atmospheric Sciences Program, ‡Department of Civil and Environmental Engineering, §School of Forest Resources and Environmental Science, Michigan Technological University, Houghton, Michigan 499311295, United States

10

ABSTRACT: An enzyme-explicit denitrification model with representations for pre- and de

11

novo synthesized enzymes was developed to improve predictions of nitrous oxide (N2O)

12

accumulations in soil and emissions from the surface. The metabolic model of denitrification is

13

based on dual substrate utilization and Monod growth kinetics. Enzyme synthesis/activation was

14

incorporated into each sequential reduction step of denitrification to regulate dynamics of the

15

denitrifier population and the active enzyme pool, which controlled the rate function.

16

Parameterizations were developed from observations of the dynamics of N2O production and

17

reduction in soil incubation experiments. The model successfully reproduced the dynamics of

18

N2O and N2 accumulation in the incubations and revealed an important regulatory effect of

19

denitrification enzyme kinetics on the accumulation of denitrification products. Pre-synthesized

20

denitrification enzymes contributed 20, 13, 43, and 62% of the N2O that accumulated in 48-hr

21

incubations of soil collected from depths of 0-5, 5-10, 10-15, and 15-25 cm, respectively. An

22

enzyme activity function (E) was defined to estimate the relative concentration of active

23

enzymes and variation in response to environmental conditions. The value of E allows activities

24

of pre-synthesized denitrification enzymes to be differentiated from de novo synthesized

25

enzymes. Incorporating explicit representations of denitrification enzyme kinetics into

26

biogeochemical models is a promising approach for accurately simulating dynamics of the

27

production and reduction of N2O in soils.

ACS Paragon Plus Environment

Environmental Science & Technology

28 29

INTRODUCTION Nitrous oxide (N2O) is a long-lived greenhouse gas and the principal biogenic source of

30

stratospheric N2O contributing to the destruction of the ozone layer.1,2 Global N2O emissions

31

from cultivated soils have been estimated at 4.2 Tg yr-1, which accounts for 50% of global

32

anthropogenic N2O sources.3 Application of synthetic nitrogen fertilizers has substantially

33

stimulated emissions of N2O from agroecosystems.4-6 Sequential reduction of nitrate (NO3-),

34

nitrite (NO2-), and nitric oxide (NO) during denitrification produces N2O.7 Low levels of N2O

35

relative to NO3- and NO2-, which are the principal products of nitrification, were observed in

36

tracer studies of nitrification and suggest denitrification is a more significant source of N2O than

37

nitrification.8,9 Nitrous oxide reductase (N2OR) is the only known enzyme that mediates the

38

reduction of N2O to molecular nitrogen (N2), and thus, denitrification is also an important

39

biological sink for N2O.

40

The sequential reduction of nitrogen oxides during denitrification is typically modeled with

41

Michaelis-Menten kinetics using enzyme-specific values of the maximum rate of the reaction

42

(Vmax) and the Michaelis or substrate affinity constant (Km), which is the substrate concentration

43

at which the reaction rate is 0.5 Vmax.10,11 Reaction intermediates accumulate if rates of

44

consumption are slower than rates of production. Dual substrate, Michaelis-Menten kinetic

45

models that use Vmax and two Km values for each denitrification step to account for affinities of

46

the electron donor and acceptor (i.e. reduced carbon and nitrogen oxides, respectively), have also

47

been developed to explain observations.12,13 Pan et al. decoupled reduced carbon oxidation and

48

nitrogen oxide reduction in a denitrification model by introducing reduced and oxidized electron

49

carriers in the Michaelis-Menten kinetic expression and by using different substrate affinity

50

constants to explain competition for electrons.14 Dynamics of the denitrifier population are 2

ACS Paragon Plus Environment

Page 2 of 32

Page 3 of 32

Environmental Science & Technology

51

usually simulated in denitrification models with the Monod equation that relates microbial

52

growth rates to the concentration of a limiting substrate. Relative rates of the sequential

53

denitrification reactions are parameterized using unique values of Vmax or are simulated by

54

assuming reductases are unique to the various denitrifying populations.15 Thus, sequential

55

induction of denitrification enzymes is represented by differences in growth rates of the various

56

reductase populations.

57

Accumulation of denitrification intermediates result from unbalanced rates of the sequential

58

denitrification reactions, which can be ascribed solely to enzyme kinetics,10,12 or together with

59

sequential gene expression.16 Denitrification is essentially regulated at the enzymatic level rather

60

than solely at the population level. Enzyme synthesis and the dynamics of enzyme activities are

61

relatively fast compared to population dynamics. Transcriptional analysis showed that expression

62

of genes encoding all denitrification reductases peaked within 2-3 hours after switching from

63

oxic to anoxic conditions in cultured Pseudomonas fluorescens.16 De novo syntheses of NO3-

64

and NO2- reductases (NAR and NIR, respectively) were observed within 2-3 h and 4-12 h,

65

respectively, following incubations under anaerobic conditions.11,17 Rapid accumulation of N2O

66

in incubations of soil-extracted bacteria under oxic conditions immediately following incubations

67

under anoxic conditions could not be ascribed to growth of the denitrifying communities.18 The

68

primary denitrification product in incubations under anoxic conditions was N2; however, a shift

69

to N2O was observed when the incubations were re-exposed to molecular oxygen (O2), which

70

deactivated N2OR. The activity of N2OR recovered when O2 was depleted after the O2-exposure

71

period.18 The results suggest an important role of the pre-synthesized denitrification proteome in

72

the accumulation of N2O. Thus, denitrification models based solely on population dynamics

73

might have difficulty in simulating transient N2O accumulations in soil under rapidly fluctuating 3

ACS Paragon Plus Environment

Environmental Science & Technology

74

environmental conditions that regulate O2 tensions like the infiltration of water during

75

precipitation.

76

Enzyme dynamics can be incorporated into a conventional denitrification model by

77

representing the rate of each denitrification step as a function of both population dynamics and

78

the active enzyme pool.19 Here we formulate a mathematical model for denitrification with

79

explicit representation of enzyme dynamics. Enzyme synthesis and activation are regulated by

80

enzyme substrate saturation and inhibitor concentrations. An enzyme activity function is defined

81

to estimate the relative concentration of active enzymes and allows activities of pre-synthesized

82

denitrification enzymes to be differentiated from de novo synthesized enzymes. Parameters

83

describing enzyme dynamics are estimated and validated with experimental data obtained from

84

soil incubations. The primary objective of the work is to improve model forecasts of N2O

85

accumulations in soil and emissions by developing a model framework that incorporates

86

representations of explicit enzyme dynamics into denitrification models.

87 88

EXPERIMENTAL SECTION

89

Soil collection and analysis. Soils were collected 19-26 July 2012 as part of a rainfall

90

simulation study20 at the Bondville, Illinois AmeriFlux site (40°00′N, 88°18′W). No-till

91

agriculture has been practiced at the site for more than 20 a, and soybeans and corn have been

92

rotated annually since 2000.21 The soil type is silt loam, with an average porosity of 45%

93

between 0-50 cm and an inorganic fraction composed of 25% clay, 70% silt and 5% sand.21,22

94

Total precipitation was 2 mm during the previous 10 days before the field experiment.

95 96

Soil samples were collected in-the-row of soybean down to a depth of 25 cm using a 1.27-cm o.d. stainless steel sampler (AMS, Inc. American Falls, ID) and sectioned into 4 depth 4

ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32

Environmental Science & Technology

97

increments (i.e., 0-5, 5-10, 10-15 and 15-25 cm). Soil core sections were stored in 15-mL sterile

98

plastic tubes (Fisher Scientific, Pittsburgh, PA), flash-frozen in the field in liquid N2, and

99

transported to the laboratory in a liquid N2 dewar (PrincetonCryo, Flemington, NJ). Details of

100

the analytic techniques and the complete results can be found in Zheng.20 Briefly, subsamples

101

from soil cores sections were sieved (4 mm) prior to analyses of soil pH, NO3-, extractable

102

organic carbon (i.e., dissolved organic carbon; DOC), and microbial biomass carbon (SMBC).

103

Soil pH was determined in a soil suspension using the 1:1 slurry method. The DOC and soil

104

soluble N were extracted with potassium sulfate and analyzed with a TOC Analyzer (Sievers

105

900, GE Analytical Instruments, CO) and Rapid Flow Analyzer (Perstorp Analytical Inc., Silver

106

Spring, MD), respectively. The SMBC was determined through a correlation with the

107

phospholipid fatty acid (PLFA) content of soil.23,24 Lipids were extracted from freeze-dried soils

108

with chloroform-methanol25 and the methylated PLFAs were quantified by high-resolution gas

109

chromatography with flame ionization detection (FID; HP6890; Agilent, Palo Alto, CA).

110

Calculation of SMBC was based on the following correlation:20

111 112

SMBC=4.5 PLFAT +33 (R2=0.85)

113 114 115

where SMBC and total PLFAs (PLFAT) are expressed as µg C g-1 and nmol g-1, respectively. Soil Incubations. Subsamples of soil core sections (3 g) were incubated in 40 mL amber vials

116

containing 5 mL of synthetic rainwater and sealed with mininert valves (Sigma Aldrich, MO).

117

The average volumetric air content was between 40-50% before the incubation. Levels of

118

chemical constituents in the synthetic rainwater were determined from the average

119

concentrations in annual precipitation.26 Air was evacuated from the vial headspace for 30 min 5

ACS Paragon Plus Environment

Environmental Science & Technology

120

and replaced by helium (He) for a total of 3 times to reduce headspace O2 levels below1% (v/v).

121

Treatments with chloramphenicol (CHL; 2.5 g L-1) were used to inhibit protein synthesis and

122

estimate the activity of pre-synthesized denitrification enzymes.27 To inhibit the activity of

123

N2OR, which reduces N2O to N2, 3.5 mL of He was removed from the headspace and replaced

124

with acetylene (C2H2) to make the headspace concentration 10% v/v. The activity of pre-

125

synthesized N2OR was estimated by adding both CHL and C2H2. Vials with the various

126

treatments were prepared in triplicate, incubated at 25°C, and gently mixed on a rotary shaker

127

(250 rpm) during the 48-h experiments.

128

The headspace of each vial was sampled at 0, 3, 6, 12, 24, 36, and 48 h to match the sampling

129

schedule of the rainfall simulation study. Samples of headspace were injected into a 1-mL

130

stainless steel sample loop connected to a 2-position, 6-port valve (VICI, Houston, TX) upstream

131

of a high-resolution gas chromatograph with electron capture detector (ECD; HP5890; Hewlett

132

Packard, Palo Alto, CA). The N2O was separated from other electron capturing species with a

133

30-m × 0.530-mm fused silica capillary coated with a 3.00-µm carbon film (GS-CarbonPlot;

134

Agilent). The carrier and ECD makeup gases were He and N2, respectively. The C2H2

135

diminished sensitivity and impeded recovery of the ECD, and thus, was removed from the

136

column effluent by redirecting the column flow through a 2-position, 4-port valve (VICI) to an

137

FID after N2O eluted from the column. The precision for N2O quantitation was better than 2%

138

and the detection limit was less than 5 ppbv.

139

Model development. The model represents the enzyme dynamics of each sequential

140

denitrification reaction by including an enzyme activity function, which simulates changes in

141

size of the active enzyme pool with fluctuating environmental signals, i.e. substrate level and O2

6

ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32

Environmental Science & Technology

142

concentration. The rate of enzyme synthesis/activation is assumed to obey Michaelis-Menten

143

kinetics as follows (Table 1):

144

145

K I ,R Ci dE R = Vmax, R ⋅ ⋅ (1 − E R ) K E ,i + Ci K I ,R +C O2 dt

146 147

where ER is the enzyme activity function (dimensionless), Vmax,R is the maximum specific

148

synthesis/activation rate of the corresponding reductase reaction (h-1), KE,i is the half-saturation

149

constant for enzyme induction (M), KI,R is the O2 inhibition coefficient for the corresponding

150

reductase (M), Ci is the aqueous-phase concentration of the substrate (M), R is the reductase (i.e.,

151

NAR, NIR, NOR, and N2OR), and i is the corresponding substrate (i.e., NO3-, NO2-, NO, and

152

N2O, respectively). Values of E range from 0-1, where 0 represents no active enzymes and 1

153

represents full activation of the total enzyme pool. Estimates of KE,i and KI,R are based on

154

properties of denitrification enzymes and their sensitivity to the O2 level.7,28 A low KI value was

155

assigned to the O2 inhibition coefficient for N2OR to compensate for the strong inhibitory effect

156

of O2.28,29 The value of E is 1 under conditions of anoxia and ideal substrate concentration.

157

Under suboptimal conditions, the initial estimate of E and the temporal variation of enzyme

158

synthesis are the parameters that control enzyme activity.

159

Denitrification reactions are simulated based on dual substrate utilization and Monod growth

160

kinetics.12,13 Microbial oxidations of C using O2, NO3-, NO2-, NO, and N2O as electron acceptors

161

are considered and stoichiometric relationships are obtained through electron balance between

162

the C source (electron donor) and electron acceptors. Microbial mediated transformations are

163

assumed to occur in the aqueous phase with equilibrium for O2, NO, N2O and N2 between the 7

ACS Paragon Plus Environment

Environmental Science & Technology

164

gas and aqueous phases following Henry’s law. All chemical species follow a time-dependent

165

mass balance in the gas and liquid phase. The specific reaction rate depends upon the maximum

166

utilization rate of the substrate (q), microbial biomass (B), and the term describing dual substrate

167

utilization. All denitrifiers are assumed to be equally saturated with different denitrification

168

enzymes, and thus, a linear dependency of E is applied in the rate expressions to approximate the

169

relative active enzyme concentrations. The net variation in Ci and CO2 depends on the rate of

170

production and consumption by the corresponding biomass (Bi), which is calculated from the C

171

utilized as substrate and the growth efficiency via microbial respiration using O2 and nitrogen

172

oxides.

173

Kinetics and stoichiometry of the transformations are summarized in Table 2. Respiration is

174

blocked by NO through binding to cytochrome oxidase and nM levels of NO can cause

175

substantial inhibition of respiration.30 Competitive inhibition from NO increased the apparent

176

value of KNO, which is determined by KO2, CNO, and KI,NO,O2 in the rate expression for O2

177

respiration.31,32 Two molecules of NO are bound to NOR during reduction of NO and substrate

178

inhibition was observed to occur at µM levels.33 Thus, the kinetics of NO reduction follows the

179

classic Haldane formula for substrate inhibition.34 However, levels of NO in the soil incubations

180

are unlikely to reach µM levels due to the lower levels of initial substrate concentrations. Kinetic

181

reaction parameters previously estimated and validated by laboratory studies and process-scale

182

soil denitrification models35-37 are listed in Table 2S.

183

Soil slurries were sufficiently buffered and remained constant at about pH 7 over 48 hr, and

184

thus, inhibition of N2OR activity at suboptimal pH (6.0) is not considered in the model.38-40 In

185

the absence of inhibitory effects, denitrification rates are related to availability of electron

8

ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32

Environmental Science & Technology

186

acceptors and donors and active enzymes mediate the reactions. The rate of volatilization of

187

gaseous chemical species from the aqueous phase is calculated as follows:41 Rtr,i = K L (

188

Ci ,g Hi

− Ci ,aq )

189

where Rtr,i is the transfer rate of the chemical species (M h-1), Ci,g and Ci,aq are gas and liquid

190

phase concentrations (M), Hi is Henry’s law constant expressed as LH2O Lair-1 and KL is the

191

overall liquid-phase mass transfer coefficient (h-1). The value of KL depends on the

192

physicochemical properties of the chemical species and the depth of liquid in the soil slurry.

193

Estimates of KL for O2, NO, N2O and N2 were calculated based on Henry’s law constants and

194

reported values of individual gas- and liquid-phase mass transfer coefficients for H2O and CO241

195

and varied from 16.1-19.3 h-1. Estimates of KL for the incubation system are at the low end of

196

experimentally determined values (19.4-20.2 h-1) from a robotic incubation system.42

197

The system of differential equations (Tables 1 and 2) is solved numerically using Matlab (The

198

Mathworks, Inc., Natick, MA, USA) with ODE solvers. The average time step is about 0.003 h.

199

Initial conditions are assigned according to levels measured in the incubations,20 including

200

concentrations of O2, NO3-, DOC, SMBC, and the measured activity of pre-synthesized enzymes

201

prior to incubation (Table 3S). Parameters developed in the model were optimized by the

202

Levenberg-Marquardt method.43 Model fitness was evaluated by calculating the coefficient of

203

determination as follows: R

204

2

∑ (C = 1− ∑ (C

exp

− C mod el ) 2

exp

− Cexp ) 2

205

where Cexp and Cmodel are experimentally determined and model simulated concentrations (M),

206

respectively.

207 9

ACS Paragon Plus Environment

Environmental Science & Technology

208

RESULTS

209

Model Evaluation. The dynamics of N2O in the headspace of all soil incubations were similar

210

(Figure 1S). Levels of N2O increased sharply within the first 12-24 h in the headspace of the soil

211

slurries in synthetic rainwater (CTR) and then ceased when N2O was likely being reduced to N2.

212

Production of N2O in slurries treated with C2H2 to block N2OR activity exhibited a continuous

213

increase in headspace N2O and indicates the activity of N2OR is represented by the difference in

214

headspace N2O between the CTR and C2H2 treatment. Accumulation of N2O from the

215

CHL+C2H2 treatment represents the activity of pre-synthesized denitrification enzymes and

216

differences in the accumulation of N2O between the CHL and CHL+C2H2 treatments represents

217

pre-synthesized N2OR activity. Negligible pre-synthesized N2OR activity was observed;

218

however, a substantial amount of N2O accumulation was attributed to the activities of pre-

219

synthesized NAR, NIR and NOR.

220 221 222

The initial value of E was estimated from the measured denitrification rates with and without CHL treatment as follows: E = RCHL / RCTR

223

where RCHL and RCTR are derived from the initial, linear portions of the N2O production curves

224

from the CHL and CTR treatments, respectively (Figure 1S). Activities of NAR, NIR, and NOR

225

can not be distinguished unless products of each of the sequential reductions are measured. The

226

rate of N2O production is limited by the slowest reduction step that is catalyzed by NAR, NIR, or

227

NOR. A lower activity of NAR or NIR would not produce sufficiently high concentrations of

228

NO2- and NO to produce the observed levels of N2O. Thus, initial values of E for NAR, NIR, and

229

NOR are assumed to be equal and are estimated directly from the experimental data. The

10

ACS Paragon Plus Environment

Page 10 of 32

Page 11 of 32

Environmental Science & Technology

230

assumption also limits the number of unknown variables and avoids introducing unnecessary

231

uncertainties into the model.

232

The initial biomass of denitrifiers, which carry reductases for the various nitrogen oxides and

233

switch from O2 respiration to NO3- respiration, is assumed to be a fraction of the measured total

234

microbial biomass as follows:

235 236

BDen = δ ⋅ BTotal

237 238

where BDen is the initial biomass (on a mol C basis) of the total denitrifiers in the incubation

239

solution (mol C L-1), δ is the fraction of the total microbial biomass represented by the

240

denitrifiers (dimensionless), and BTotal is SMBC (on a mol C basis) in the incubation solution

241

(mol C L-1). Initial biomasses of denitrifiers carrying NAR (BNO3-), NIR (BNO2-), NOR (BNO), and

242

N2OR (BN2O) in the rate expressions for Monod growth kinetics (Table 2) are assumed to be

243

equal to BDen. The initial value of BO2, which is the microbial biomass mediating respiratory

244

metabolism using O2 as the electron acceptor, is assumed to be equal to BTotal. Values of δ are

245

estimated from the soil incubations by using data derived from the CHL treatment and gradually

246

increased with depth in the soil (Figure 2S). The average value of δ was 0.0461 ± 0.0013.

247

Values of Vmax,R are estimated from data obtained from incubation of the 0-5 cm soil core

248

section. The increased accumulation of N2O in the C2H2 treatment relative to the CHL+C2H2

249

treatment is attributed to de novo activity of NAR, NIR, and NOR that produces N2O; however,

250

in the CTR relative to the CHL treatment, N2O accumulation is due to the net effect of N2O

251

production by the combined de novo activities of NAR, NIR, and NOR and reduction by de novo

252

N2OR activity. The experiment does not differentiate rates of de novo synthesis for NAR, NIR, 11

ACS Paragon Plus Environment

Environmental Science & Technology

253

and NOR, and thus, values of Vmax,NAR, Vmax,NIR, and Vmax,NOR are assumed to be equal and

254

designated as Vmax,1. Initial estimates of Vmax,1 and Vmax,N2OR (Vmax,2) are made by examining the

255

temporal sequence in de novo enzyme synthesis. De novo synthesis of NAR, NIR, and NOR

256

begins within the first 3 h with synthesis of N2OR beginning about 12 h later. Values of Vmax,1

257

and Vmax,2 were optimized with the CTR and the C2H2 treatment data by minimizing the sum of

258

squares of the deviations between the measured and predicted N2O and N2 mixing ratios,

259

respectively. The correlation between Vmax,1 and Vmax,2 is low (Figure 4aS). The range in estimates

260

of Vmax,1 are wider than the range in estimates of Vmax,2, which is likely due to the assumption that

261

Vmax for NAR, NIR, and NOR are equal.

262

Model Sensitivity Analysis. A sensitivity analysis was performed on several key model

263

parameters (Figure 2). Variations of ±5, ±10, ±15, and ±20% were applied to parameters and

264

resulting changes in the cumulative concentrations of NO3-, NO2-, NO, N2O, and N2 were

265

normalized to the corresponding reference simulation (Figure 2). Activities of NAR, NIR, and

266

NOR are regulated by Vmax,1, which determines the sequential flux of N substrates.

267

Concentrations of NO3-, NO2-, and NO are sensitive to changes in Vmax,1, which creates an

268

imbalance between production and reduction rates. Cumulative concentrations of N2O and N2 are

269

slightly influenced by variations in Vmax,1. The low sensitivity of the model performance to Vmax,1

270

is likely related to the wide range in estimated values of Vmax,1. Changes in Vmax,2 have a more

271

direct effect on the accumulation of N2O and N2 through regulation of EN2OR. Variations in the

272

parameter enlarge the imbalance between activities of NAR, NIR, NOR, and N2OR, resulting in

273

a greater accumulation of N2O. The parameters KE,N2O and KI,N2OR are key regulators of N2OR

274

activity and variations demonstrate a strong impact on accumulation of N2O and N2 and minimal

275

impact on accumulation of NO3-, NO2-, and NO (Figure 2). Other important parameters include 12

ACS Paragon Plus Environment

Page 12 of 32

Page 13 of 32

Environmental Science & Technology

276

KO2, which regulates O2 consumption, and KL, which controls volatilization of NO, N2O and N2

277

from the aqueous phase. Accumulation of N2O and N2 are both highly influenced by KO2, which

278

is mainly due to the sensitivity of N2OR to O2 concentration. Model sensitivity to variations in

279

KL was low. The first gaseous intermediate product of denitrification is NO, which shows the

280

strongest response.

281

Additional sensitivity analyses were performed with parameters derived from the literature.

282

Twelve parameters were grouped into three categories: half-saturation constants for nitrogen

283

oxide uptake, Michaelis constants for denitrification enzyme induction, and substrate utilization

284

rates. Variations in half-saturation constants for nitrogen oxide uptake and Michaelis constants

285

for enzyme induction cause limited changes in the accumulation of nitrogen oxides (< 5%).

286

However, variations in the substrate utilization rates (q) show significant influence on the

287

accumulation of the corresponding substrate (Figure 3S). In particular, changes in the utilization

288

rate of NO3- (qNO3-) strongly influenced accumulations of N2O and N2, which were expected due

289

to a forced minimization of the accumulations of NO2- and NO by the model.

290

Variation in Production and Reduction of N2O with Depth. Transformation rates of N2O

291

are regulated by denitrification enzyme activity, which is parameterized by E. Values of E for

292

each enzyme are dependent on the initial enzyme activity (E0) and Vmax,1 and Vmax,2. Measured

293

and model predicted temporal variations of N2O and N2 during the incubations are presented in

294

Figure 1. Estimated values of Vmax,1 and Vmax,2 for different soil core sections are summarized in

295

Table 3. The value of Vmax,1/Vmax,2 is significantly greater than 1, which implies a lag in the

296

synthesis of N2OR. Measurements of reaction rates derived from the CHL and CHL+C2H2

297

treatments of incubations of the 5-10 cm soil core section were substantially different, which is

298

due to the activity of pre-synthesized N2OR (Table 3S). The estimated value of Vmax,1/Vmax,2 is 13

ACS Paragon Plus Environment

Environmental Science & Technology

299

significantly lower than the other three enzymes (Table 3). The lag in N2OR synthesis causes an

300

extremely high value of N2O/(N2O+N2) during the first few hours of the incubation (Figure 3).

301

Production of N2 from pre-synthesized N2OR in the 5-10 cm soil core section diminishes the

302

early plateau in the magnitude of the ratio.

303

Role of pre-synthesized denitrification proteome. Contributions from pre-synthesized

304

denitrification enzymes are evaluated with the model by setting Vmax,1 to zero to suppress de novo

305

synthesis of NAR, NIR and NOR, which ascribes N2O accumulation solely to the activity of pre-

306

synthesized enzymes. Pre-synthesized enzymes contributed 63, 38, 48, and 48% of the total

307

cumulative N2O flux during incubations of the 0-5, 5-10, 10-15, and 15-25 cm soil core sections,

308

respectively (Figure 4). Contributions of pre-synthesized enzymes normalized to SMBC

309

increased with soil depth and were 20, 13, 43, and 62%. Due to the activity of pre-synthesized

310

N2OR (Figure 1S), the contribution of pre-synthesized enzymes to N2O accumulation in the 5-10

311

cm soil core section is relatively low.

312 313 314

DISCUSSION Representations of N2O production in soils in current denitrification models are based on

315

growth of the microbial biomass.35,36 An enzyme-explicit model of the biological regulation of

316

denitrification that incorporates dynamics of pre- and de novo synthesized denitrification

317

enzymes was developed in the subject study to improve forecasts of the production of N2O in

318

soils. Incubation studies have demonstrated persistence of denitrification enzymes in soils

319

subjected to aerobic conditions.17,27,44 Denitrification activity and product gases observed 1-3 h

320

after the onset of anaerobiosis during the incubations were ascribed to the activity of pre-

321

synthesized enzymes.17 After anaerobiosis was imposed, de novo synthesis of NAR, NIR, and 14

ACS Paragon Plus Environment

Page 14 of 32

Page 15 of 32

Environmental Science & Technology

322

N2OR were found to occur within 2-3, 4-12, and 24-42 hr, respectively.11 A similar dynamic was

323

observed in the subject study. Rapid accumulation of N2O was attributed to the activity of pre-

324

synthesized NAR, NIR, and NOR. Pre-synthesized levels of N2OR were negligible and de novo

325

synthesis of N2OR lagged behind the other 3 denitrification enzymes, which contributed to the

326

rapid accumulation of N2O (Figure 1S). Incorporating values of E for the denitrification enzymes

327

in the model allowed quantification of pre-synthesized enzymes and parameterization of the

328

delay in N2OR synthesis.

329

The sensitivity analysis demonstrated the important role of Vmax,2 in regulating accumulation of

330

N2O. Differences between Vmax,1 and Vmax,2 lead to imbalances in E (Figure 5S) and result in

331

transient accumulation of N2O. Previous modeling work identified a critical role of qN2O when

332

estimating the accumulation of N2O.36 The model developed in the subject study also showed a

333

high sensitivity of N2O accumulation to the value of qN2O. However, in the enzyme-explicit

334

model, qN2O regulates the magnitude of the N2O maxima and Vmax,2 regulates the width of the

335

N2O peak (Figure 5).

336

Estimates of cumulative N2O emissions by conventional denitrification models are in general

337

agreement with low-temporal resolution measurements of the N2O flux; however, the dynamics

338

of the emissions are poorly represented.45,46 The temporal sequence of denitrification enzyme

339

synthesis likely explains differences between model simulations of the timing of peak N2O

340

emissions and field observations of maxima in N2O fluxes.45-47 High temporal resolution

341

measurements reveal that pulse emissions of N2O related to precipitation events are a significant

342

contribution to the annual emission inventory.48 The enzyme-explicit denitrification model with

343

representations for pre- and de novo synthesized enzymes (Figure 5S) is able to capture rapid

15

ACS Paragon Plus Environment

Environmental Science & Technology

344

changes in N2O production and reduction and should improve forecasts of the timing and

345

magnitude of pulse emissions.49

346

Development of the parameter, E, in the subject study enabled parameterization of pre- and de

347

novo synthesized enzyme activity in the enzyme-explicit denitrification model. The parameter

348

allows quantification of the “denitrification population” that mediates the various sequences of

349

denitrification and integrates variations in denitrification enzyme activity in response to

350

environmental conditions. The enzyme-explicit denitrification module could be coupled with

351

other denitrification or biogeochemical models like the electron competition model of

352

denitrification14, DNDC45 or ecosys46, to improve model predictions of N2O emissions from soil.

353 354

ASSOCIATED CONTENT

355

Supporting Information

356

Additional tables with references and figures are provided.

357 358

AUTHOR INFORMATION

359

Corresponding Author

360

*E-mail: [email protected]. Phone: 906-487-2745. Fax: 906-487-2943.

361

Notes

362

The authors declare no competing financial interest.

363 364

ACKNOWLEDGEMENTS

365

The authors acknowledge start-up funding, which supported Jianqiu Zheng, and an equipment

366

loan to Paul V. Doskey through Michigan Technological University and Argonne National 16

ACS Paragon Plus Environment

Page 16 of 32

Page 17 of 32

Environmental Science & Technology

367

Laboratory, respectively. Partial support for Jianqiu Zheng through the Atmospheric Sciences

368

Program is also greatly appreciated.

369 370 371 372 373 374 375

REFERENCES (1) Seinfeld, J. H.; Pandis, S. N., Atmospheric Chemistry and Physics, John Wiley and Sons, New York, 1998. (2) Ravishankara, A. R.; Daniel, J. S.; Portmann, R. W., Nitrous oxide (N2O): The dominant ozone-depleting substance emitted in the 21st Century. Science 2009, 326, 123-125. (3) IPCC, Climate Change 2001: The Scientific Basis. Contribution of Working Group 1 to

376

the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge

377

University Press, Cambridge, United Kingdom and New York, NY, USA. 2001.

378 379 380

(4) Kroeze, C.; Mosier, A.; Bouwman, L., Closing the global N2O budget: A retrospective analysis 1500–1994. Global Biogeochem. Cy. 1999, 13, 1-8. (5) Sowers, T.; Rodebaugh, A.; Yoshida, N.; Toyoda, S., Extending records of the isotopic

381

composition of atmospheric N2O back to 1800 A.D. from air trapped in snow at the South Pole

382

and the Greenland Ice Sheet Project II ice core. Global Biogeochem. Cy. 2002, 16, 1129,

383

doi:1110.1029/2002GB001911.

384 385 386 387 388 389

(6) USEPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2007, http://epa.gov/climatechange/emissions/usinventoryreport.html, 2009. (7) Zumft, W. G., Cell biology and molecular basis of denitrification. Microbiol. Mol. Biol. R. 1997, 61, 533-616. (8) Mørkved, P. T.; Dörsch, P.; Bakken, L. R., The N2O product ratio of nitrification and its dependence on long-term changes in soil pH. Soil Biol. Biochem. 2007, 39, 2048-2057. 17

ACS Paragon Plus Environment

Environmental Science & Technology

390

(9) Mørkved, P. T.; Dörsch, P.; Henriksen, T. M.; Bakken, L. R., N2O emissions and product

391

ratios of nitrification and denitrification as affected by freezing and thawing. Soil Biol. Biochem.

392

2006, 38, 3411-3420.

393

(10) Betlach, M. R.; Tiedje, J. M., Kinetic explanation for accumulation of nitrite, nitric

394

oxide, and nitrous oxide during bacterial denitrification. Appl. Environ. Microbiol. 1981, 42,

395

1074-1084.

396

(11) Dendooven, L.; Anderson, J. M., Use of a “least square” optimization procedure to

397

estimate enzyme characteristics and substrate affinities in the denitrification reactions in soil.

398

Soil Biol. Biochem. 1995, 27, 1261-1270.

399 400

(12) Almeida, J. S.; Reis, M. A. M.; Carrondo, M. J. T., A unifying kinetic model of denitrification. J. Theor. Biol. 1997, 186, 241-249.

401

(13) Thomsen, J. K.; Geest, T.; Cox, R. P., Mass spectrometric studies of the effect of pH on

402

the accumulation of intermediates in denitrification by Paracoccus denitrificans. Appl. Environ.

403

Microbiol. 1994, 60, 536-541.

404

(14) Pan, Y.; Ni, B.-J.; Yuan, Z., Modeling electron competition among nitrogen oxides

405

reduction and N2O accumulation in denitrification. Environ. Sci. Technol. 2013, 47, 11,083-

406

11,091.

407 408

(15) Leffelaar, P. A.; Wessel, W. W., Denitrification in a homogeous, closed system: Experiment and simulation. Soil Sci. 1988, 146, 335-349.

409

(16) Philippot, L.; Mirleau, P.; Mazurier, S.; Siblot, S.; Hartmann, A.; Lemanceau, P.;

410

Germon, J. C., Characterization and transcriptional analysis of Pseudomonas fluorescens

411

denitrifying clusters containing the nar, nir, nor and nos genes. Biochim. Biophys. Acta 2001, 16,

412

436-440. 18

ACS Paragon Plus Environment

Page 18 of 32

Page 19 of 32

413 414

Environmental Science & Technology

(17) Smith, M. S.; Tiedje, J. M., Phases of denitrification following oxygen depletion in soil. Soil Biol. Biochem. 1979, 11, 261-267.

415

(18) Morley, N.; Baggs, E. M.; Dorsch, P.; Bakken, L., Production of NO, N2O and N2 by

416

extracted soil bacteria, regulation by NO2- and O2 concentrations. FEMS Microbiol. Ecol. 2008,

417

65, 102-112.

418 419 420 421 422 423 424 425 426

(19) Wallenstein, M. D.; Weintraub, M. N., Emerging tools for measuring and modeling the in situ activity of soil extracellular enzymes. Soil Biol. Biochem. 2008, 40, 2098-2106. (20) Zheng, J., Denitrification in soils: From genes to environmental outcomes. Ph.D. Dissertation, Michigan Technological University, Houghton, MI, 2014. (21) Bondville AmeriFlux Site, AmeriFlux US-Bo1 sponsed by NOAA/GEWEX http://ameriflux.ornl.gov/fullsiteinfo.php?sid=44. 2012. (22) Illinois Climate Network, sponsed by Illinois State Water Survey http://www.isws.illinois.edu/warm/datatype.asp. 2012. (23) Leckie, S. E.; Prescott, C. E.; Grayston, S. J.; Neufeld, J. D.; Mohn, W. W., Comparison

427

of chloroform fumigation-extraction, phospholipid fatty acid, and DNA methods to determine

428

microbial biomass in forest humus. Soil Biol. Biochem. 2004, 36, 529-532.

429

(24) Bailey, V. L.; Peacock, A. D.; Smith, J. L.; Bolton Jr., H., Relationships between soil

430

microbial biomass determined by chloroform fumigation–extraction, substrate-induced

431

respiration, and phospholipid fatty acid analysis. Soil Biol. Biochem. 2002, 34, 1385-1389.

432 433 434

(25) Bligh, E. G.; Dyer, W. J., A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 1959, 37, 911-917. (26) NADP, National Atmospheric Depostion Program http://nadp.isws.illinois.edu/. 2012.

19

ACS Paragon Plus Environment

Environmental Science & Technology

435 436

(27) Dendooven, L.; Anderson, J. M., Dynamics of reduction enzymes involved in the denitrification process in pasture soil. Soil Biol. Biochem. 1994, 26, 1501-1506.

437

(28) Korner, H.; Zumft, W. G., Expression of denitrification enzymes in response to the

438

dissolved oxygen level and respiratory substrate in continuous culture of Pseudomonas stutzeri.

439

Appl. Environ. Microbiol. 1989, 55, 1670-1676.

440

(29) Wild, D.; Von Schulthess, R.; Gujer, W., Synthesis of denitrification enzymes in

441

activated sludge: Modelling with structured biomass. Water Sci. Technol. 1994, 30, 113-122.

442

(30) Giuffre, A.; Borisov, V. B.; Mastronicola, D.; Sarti, P.; Forte, E., Cytochrome bd

443

oxidase and nitric oxide: From reaction mechanisms to bacterial physiology. FEBS Lett. 2012,

444

586, 622-629.

445 446 447 448 449 450 451 452 453 454

(31) Koivisto, A.; Matthias, A.; Bronnikov, G.; Nedergaard, J., Kinetics of the inhibition of mitochondrial respiration by NO. FEBS Lett. 1997, 417, 75-80. (32) Copeland, R. A., Time-dependent inhibition. In Enzymes, John Wiley & Sons, Inc.:

2002; pp 318-349. (33) Girsch, P.; de Vries, S., Purification and initial kinetic and spectroscopic characterization of NO reductase from Paracoccus denitrificans. Biochim. Biophys. Acta 1997, 16, 1-2. (34) Copeland, R. A., Kinetics of single-substrate enzyme reactions. In Enzymes, John Wiley & Sons, Inc.: 2002; pp 109-145. (35) Gu, C.; Riley, W. J., Combined effects of short term rainfall patterns and soil texture on soil nitrogen cycling — A modeling analysis. J. Contam. Hydrol. 2010, 112, 141-154.

455

(36) Maggi, F.; Gu, C.; Riley, W. J.; Hornberger, G. M.; Venterea, R. T.; Xu, T.; Spycher,

456

N.; Steefel, C.; Miller, N. L.; Oldenburg, C. M., A mechanistic treatment of the dominant soil

20

ACS Paragon Plus Environment

Page 20 of 32

Page 21 of 32

Environmental Science & Technology

457

nitrogen cycling processes: Model development, testing, and application. J. Geophys.Res. 2008,

458

113, G02016, doi:10.1029/2007JG000578.

459

(37) Blagodatsky, S.; Grote, R.; Kiese, R.; Werner, C.; Butterbach-Bahl, K., Modelling of

460

microbial carbon and nitrogen turnover in soil with special emphasis on N-trace gases emission.

461

Plant Soil 2011, 346, 297-330.

462

(38) Bergaust, L.; Mao, Y.; Bakken, L. R.; Frostegård, Å., Denitrification response patterns

463

during the transition to anoxic respiration and posttranscriptional effects of suboptimal pH on

464

nitrogen oxide reductase in Paracoccus denitrificans. Appl. Environ. Microbiol. 2010, 76, 6387-

465

6396.

466

(39) Bakken, L. R.; Bergaust, L.; Liu, B.; Frostegård, Å., Regulation of denitrification at the

467

cellular level: A clue to the understanding of N2O emissions from soils. Philos. T. Roy. Soc. B.

468

2012, 367, 1226-1234.

469

(40) Liu, B.; Morkved, P. T.; Frostegård, Å.; Bakken, L. R., Denitrification gene pools,

470

transcription and kinetics of NO, N2O and N2 production as affected by soil pH. FEMS

471

Microbiol. Ecol. 2010, 72, 407-417.

472 473 474 475

(41) Thibodeaux, L. J., Environmental Chemodynamics. John Wiley and Sons, New York,

1996. (42) Molstad, L.; Dorsch, P.; Bakken, L. R., Robotized incubation system for monitoring gases (O2, NO, N2O, N2) in denitrifying cultures. J. Microbio. Meth. 2007, 71, 202-211.

476

(43) Ashyraliyev, M.; Jaeger, J.; Blom, J., Parameter estimation and determinability analysis

477

applied to Drosophila gap gene circuits. BMC Syst. Biol. 2008, 2, 83, doi:10.1186/1752-0509-2-

478

83.

21

ACS Paragon Plus Environment

Environmental Science & Technology

479 480 481

(44) Firestone, M. K.; Tiedje, J. M., Temporal change in nitrous oxide and dinitrogen from denitrification following onset of anaerobiosis. Appl. Environ. Microbiol. 1979, 38, 673-679. (45) Li, C.; Frolking, S.; Frolking, T. A., A model of nitrous oxide evolution from soil driven

482

by rainfall events: 1. Model structure and sensitivity. J. Geophys. Res. Atmos. 1992, 97, 9759-

483

9776.

484

(46) Metivier, K. A.; Pattey, E.; Grant, R. F., Using the ecosys mathematical model to

485

simulate temporal variability of nitrous oxide emissions from a fertilized agricultural soil. Soil

486

Biol. Biochem. 2009, 41, 2370-2386.

487

(47) Rabot, E.; Cousin, I.; Hénault, C., A modeling approach of the relationship between

488

nitrous oxide fluxes from soils and the water-filled pore space. Biogeochemistry 2014, 1-14.

489

(48) Groffman, P.; Butterbach-Bahl, K.; Fulweiler, R.; Gold, A.; Morse, J.; Stander, E.;

490

Tague, C.; Tonitto, C.; Vidon, P., Challenges to incorporating spatially and temporally explicit

491

phenomena (hotspots and hot moments) in denitrification models. Biogeochemistry 2009, 93, 49-

492

77.

493

(49) Butterbach-Bahl, K.; Baggs, E. M.; Dannenmann, M.; Kiese, R.; Zechmeister-

494

Boltenstern, S., Nitrous oxide emissions from soils: How well do we understand the processes

495

and their controls? Philos. T. Roy. Soc. B. 2013, 368, doi:10.1098/rstb.2013.0122.

496 497 498

22

ACS Paragon Plus Environment

Page 22 of 32

Page 23 of 32

499 500

Environmental Science & Technology

Table 1. Enzyme synthesis/activation kineticsa Enzyme

Rate expression R E NAR = Vmax, NAR ⋅

Nitrate Reductase (NAR)

RE NIR = Vmax, NIR ⋅

⋅ (1 − E NAR )

3

C NO −

K I , NIR



2

K E , NO − + C NO − K I , NIR + CO2

RENOR = Vmax, NOR ⋅

Nitrous Oxide Reductase (N2OR) a

K E , NO − + C NO − K I , NAR + CO2

2

Nitric Oxide Reductase (NOR)

K I , NAR



3

3

Nitrite Reductase (NIR)

501 502 503

C NO −

⋅ (1 − E NIR )

2

K I , NOR C NO ⋅ ⋅ (1 − E NOR ) K E , NO + C NO K I , NOR + CO2

R E N 2OR = Vmax, N 2OR ⋅

C N 2O



K I , N 2OR

K E , N 2O + C N 2O K I , N 2OR + CO2

Parameters are listed in Table 1S.

23

ACS Paragon Plus Environment

⋅ (1 − E N 2OR )

Environmental Science & Technology

504

Page 24 of 32

Table 2. Rate expressions for Monod growth kineticsa Biological Reactions

Rate Expressions

CO2 COC ⋅ C NO ) + CO2 K OC + COC K O2 (1 + K I , NO ,O2

CH 2O + O2 ( aq) → CO2 ( aq) + H 2O

RO2 = qO2 ⋅ BO2 ⋅

2 NO3− + CH 2O → 2 NO2− + CO2 ( aq) + H 2O

RNO− = qNO− ⋅ BNO− ⋅ 3

3

3

CNO− 3

KNO− + CNO− 3

− 2

R NO − = q NO − ⋅ B NO − ⋅

+

4 NO + CH 2O + 4H → 4NO(aq) + CO2 (aq) + 3H 2O

2

2

2

3



COC ⋅ ENAR KOC + COC

C NO − K NO − + C NO − 2

CNO

8 NO ( aq ) + 2CH 2 O → 4 N 2 O ( aq ) + 2CO 2 ( aq ) + 2 H 2O

RNO = qNO ⋅ BNO ⋅

4 N 2O ( aq ) + 2CH 2O → 4 N 2 ( aq ) + 2CO2 ( aq ) + 2 H 2 O

R N 2O = q N 2O ⋅ B N 2O ⋅

a

2

COC ⋅ E NIR K OC + COC

2

 CNO  ) K NO + CNO ⋅ (1 + K I ,NO  

C N 2O K N 2O + C N 2O

505 506 507



2

Variables and parameters are listed in Table 2S.

24

ACS Paragon Plus Environment



2



COC ⋅ ENOR KOC + COC

COC ⋅ E N 2OR K OC + COC

Page 25 of 32

508

Environmental Science & Technology

Table 3. Model parameter estimations from soil core incubations

509 510

0-5 cm

5-10 cm

10-15 cm

15-25 cm

Vmax,1 (h)

0.4455

0.5008

3.7982

1.4011

Vmax,2 (h)

0.0512

0.1401

0.2712

0.0695

9

4

14

20

Vmax,1/Vmax,2 511 512 513 514

25

ACS Paragon Plus Environment

Environmental Science & Technology

515

Figure Captions

516 517

Abstract Art

518 519

Figure 1. Measured (Open Circles) and modeled (Line) temporal variation of N2O concentrations

520

in the He headspace of soil core sections (0-5 cm, 5-10 cm, 10-15 cm and 15-25 cm from top to

521

bottom) incubated in synthetic rainwater (CHL+ C2H2, synthetic rainwater containing

522

chloramphenicol with He headspace containing acetylene; CTR, synthetic rainwater; C2H2, He

523

headspace containing acetylene).

524 525

Figure 2. Simulated variations in the accumulation of NO3-, NO2-, NO, N2O, and N2 when

526

variations of ±5, ±10, ±15, and ±20% were applied to model parameters.

527 528

Figure 3. Simulations of the temporal variation of the N2O/(N2O+N2) product ratio in the He

529

headspace of 4 soil core sections incubated in synthetic rainwater.

530 531

Figure 4. Simulations of the temporal variation of N2O in the He headspace of 4 soil core

532

sections incubated in synthetic rainwater with both pre- and de novo synthesized enzymes

533

(PE+DE) and with pre-synthesized enzymes (PE) only.

534 535

Figure 5. Simulations of the temporal variations of N2O with variations of ±10 and ±20% applied

536

to (a) qNO3- and (b) Vmax,2 .

537 538

26

ACS Paragon Plus Environment

Page 26 of 32

Page 27 of 32

539

Environmental Science & Technology

Abstract Art

540 541 542

27

ACS Paragon Plus Environment

Environmental Science & Technology

543

Figure 1.

544

545 546

28

ACS Paragon Plus Environment

Page 28 of 32

Page 29 of 32

547

Environmental Science & Technology

Figure 2.

548

549

29

ACS Paragon Plus Environment

Environmental Science & Technology

550

Figure 3.

551 552

30

ACS Paragon Plus Environment

Page 30 of 32

Page 31 of 32

553

Environmental Science & Technology

Figure 4.

554 555

31

ACS Paragon Plus Environment

Environmental Science & Technology

556

Figure 5.

557 558

32

ACS Paragon Plus Environment

Page 32 of 32

Modeling nitrous oxide production and reduction in soil through explicit representation of denitrification enzyme kinetics.

An enzyme-explicit denitrification model with representations for pre- and de novo synthesized enzymes was developed to improve predictions of nitrous...
3MB Sizes 0 Downloads 22 Views