Simultaneous cellulase production using domestic wastewater and bioprocess effluent treatment – A biorefinery approach

Accepted Manuscript Simultaneous cellulase production using domestic wastewater and bioprocess effluent treatment – A biorefinery approach Nelson Libardi Junior, Carlos Ricardo Soccol, Júlio César de Carvalho, Luciana Porto de Souza Vandenberghe PII: DOI: Reference:

S0960-8524(18)31759-0 https://doi.org/10.1016/j.biortech.2018.12.088 BITE 20845

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

30 October 2018 20 December 2018 23 December 2018

Please cite this article as: Junior, N.L., Soccol, C.R., César de Carvalho, J., de Souza Vandenberghe, L.P., Simultaneous cellulase production using domestic wastewater and bioprocess effluent treatment – A biorefinery approach, Bioresource Technology (2018), doi: https://doi.org/10.1016/j.biortech.2018.12.088

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Simultaneous cellulase production using domestic

2

wastewater and bioprocess effluent treatment – A

3

biorefinery approach

4

Nelson Libardi Juniora, Carlos Ricardo Soccola, Júlio César de Carvalhoa, Luciana Porto

5

de Souza Vandenberghea*

6 7

a

8

(UFPR), Centro Politécnico, C.P. 19011, 81-531-980, Curitiba, PR, Brazil

9

*corresponding author: [email protected]

Bioprocess Engineering and Biotechnology Department, Federal University of Paraná

10

1

11 12

Abstract The production of cellulases using domestic wastewater as an alternative culture

13

medium and reducing the pollutant charge of the resultant effluents were assessed for

14

the first time in this study. Cellulase production was carried out in a bubble column,

15

column-packed bed and stirred tank reactors by Trichoderma harzianum. Maximum

16

cellulase activity and productivity of 31 UFP/mL and 645 UFP/mL.h, respectively were

17

achieved in the bubble column bioreactor system without immobilization. The

18

fermented broth was microfiltrated and ultrafiltrated, leading to a cellulase recovery of

19

73.5% using a 30 kDa membrane and resulting in a 4.23-fold activity concentration.

20

Chemical oxygen demand and nitrogen concentration were reduced 81.37% and 52.9%,

21

respectively, showing great promise in producing cellulases using domestic wastewater

22

with concomitant development of a medium- to-high added-value process and reduced

23

environmental impact. These results contribute to the development of sustainable

24

bioprocesses approaching a biorefinery concept.

25 26

Keywords: domestic wastewater, FPase, membrane separation, stirred tank reactor

27

(STR), bubble column bioreactor (BCR)

28

2

29 30

1. Introduction Around 80% of all wastewater is discharged into the world’s lakes and oceans,

31

where it creates health, environmental and climate-related hazards. This discharge of

32

untreated effluent in water bodies does not only lead to eutrophication and human health

33

risks, but also contributes significantly to Greenhouse Gas (GHG) emissions in the form

34

of nitrous oxide and methane (IWA, 2018). Urbanization is the primary culprit behind

35

increasing wastewater generation. Increasing attention has been paid to the development

36

of sustainable new industrial processes with reduced environmental impact; the

37

recovery of water, energy, nutrients and other precious materials from wastewater

38

treatment plants (WWTP) is a key opportunity for the production of biomolecules such

39

as enzymes, considering the huge amounts of carbon, nitrogen and phosphorous in its

40

composition.

41

Cellulase production using alternative substrates is an emerging field because it

42

is an interesting way to reduce enzyme production costs, a major bottleneck for

43

applications like second-generation biofuel. The composition of the culture medium

44

significantly affects product concentration, yield and volumetric productivities, which

45

are crucial for economically viable production. A wide range of cellulase production

46

reports is available regarding the use of agro-industrial residues – mainly from

47

lignocellulosic nature, such as sorghum bran, corn cob, sugarcane bagasse, wheat straw

48

and elephant grass – resulting in cellulase activities of 0.1 - 23.0 filter paper units per

49

milliliter (UFP/mL) (Esterbauer et al., 1991; Singhania et al., 2017). Other cellulosic and

50

non-cellulosic carbon sources or cellulase inducers have been also described, including

51

solka floc, microcrystalline cellulose, lactose, sucrose and sophorose (Morikawa et al.,

52

1995; Singhania et al., 2017). Lactose is regarded as one of the most effective carbon

53

sources for fungal cellulases production, particularly in tandem with cellulosic residues. 3

54

Around 15,000 tons of toilet paper is generated per year in WWTP in Europe. In

55

the last decade, many authors have discussed the use of WWTP residues, including

56

sludge for cellulase production (Alam et al., 2008). Libardi et al. (2017) reported the

57

feasibility of using sanitary wastewater for cellulase production by the fungus

58

Trichoderma harzianum. Wastewater, with toilet paper as a major component, consists

59

of cellulose, which is employed as a part of the culture medium for cellulases

60

production, allowing simultaneous production of an added-value product and reduction

61

of the effluent’s pollution charge through its consumption as a source of carbon and

62

nutrients. Other authors (Alam et al., 2003) evaluated the reduction of the Chemical

63

Oxygen Demand (COD) as an indicator of organic matter consumption during

64

fermentation processes, using wastes like cheese whey, starch processing wastes, apple

65

distillery waste and olive mill wastewater. However, the use of such residues brings a

66

new challenge: the culture medium presents many insoluble substances, unwanted

67

metals and other components. The use of wastewater therefore adds further substances

68

to the complex fermented broth, which already contains proteins, biomass and other

69

biomolecules. An efficient separation and purification process must be developed for

70

enzyme recovery.

71

Membrane technologies are adequate for the downstream processing of

72

biological molecules because they operate at low temperatures and pressures and

73

involve no phase changes or chemical additives, thereby minimizing the products’

74

denaturation, deactivation and degradation. A common application of ultrafiltration

75

(UF) in downstream processing is in product concomitant separation and concentration

76

(Charcosset, 2006). The use of UF for cellulase separation and concentration was

77

described using membranes with a molecular weight cutoff (MWCO) varying from 5 to

78

100 kDa (Qi et al., 2012) for cellulases with molecular weights ranging from 60 to 90 4

79

kDa (Mores et al., 2001). The UF allows the concentration of cellulase activity between

80

50 and 75 UFP/mL (Rodrigues et al., 2014). Before the UF step, microfiltration (MF)

81

can be used to clarify the culture broth and to retain suspended solids > 0.2 μm (Mores

82

et al., 2001). Rodrigues et al. (2014) reported 38% recovery of a commercial cellulase

83

obtained from wheat straw hydrolysate. Qi et al. (2012) reached a protein recovery of

84

89.4%, 73.9% and 79.7% using MWCO of 5, 10 and 30 kDa, respectively. Although the

85

use of UF has been described as an interesting way to ensure cellulase separation and

86

concentration, there are some operational problems related to the fouling and

87

concentration polarization, which lead to the reduction of the permeate flux and

88

possibly membrane damage. These issues may increase cleaning costs, lowering the

89

competitiveness of this process (Shi et al., 2014).

90

UF processes have also been described as suitable for wastewater purification,

91

along with MF, nanofiltration and reverse osmosis (RO) (Ryan et al., 2009). The

92

efficiency of these processes could be measured in terms of reducing COD as well as

93

total nitrogen, as the COD has been employed to determine the total amount of organic

94

matter in domestic wastewater (Wan et al., 2016). Different authors have reported

95

limitations in using UF regarding the fouling effects of filtrating the secondary effluent

96

of domestic WWTPs, which implies increased associated process costs that may restrict

97

its application (Verstraete et al., 2009). However, separation and concentration of high

98

added-value products, such as cellulases, could amplify feasibility of using UF

99

processes. This approach is in accordance with wastewater biorefinery concepts, taking

100

advantage of product maximization in combination with reduced effluent concentration

101

(Kleerebezem and Loosdrecht, 2007). In a UF process, where the product of interest is

102

in the retentate fraction, the permeate fraction could be seen as the effluent of the

103

separation process, requiring adequate treatment and relocation.

5

104

This work aims to demonstrate the production of cellulases using domestic

105

wastewater as substrate in bubble column (BCR) and stirred tank (STR) reactors,

106

followed by its recovery and concentration from the culture broth using MF and UF.

107

Simultaneously, reduction of the effluent’s pollution charge was evaluated by

108

monitoring the COD and nitrogen as environmental parameters in a biorefinery concept.

109 110

2. Material and methods

111

2.1 Production of cellulases using domestic wastewater

112

Cellulase production was performed using Trichoderma harzianum (TRIC03)

113

from the Federal University of Paraná Culture Collection, Bioprocess Engineering and

114

Biotechnology Department (LPBII-UFPR). The culture medium was composed of

115

lactose (11.9 g/L), peptone (5.0 g/L), KH2PO4 (2.5 g/L) and MgSO4.7H2O (0.3 g/L),

116

which were added to the raw domestic wastewater. The raw domestic wastewater was

117

collected at the Atuba Municipal Wastewater Treatment Plant from the Sanitation

118

Company of Paraná (SANEPAR), located in Curitiba, Brazil. A fungal spore suspension

119

of approximately 107 spores/mL was used as inoculum.

120

Cellulase production was tested in 4 bioreactors systems: (R1) 1-L bubble

121

column reactor (BCR); (R2) 1-L column packed bed reactor CPBR; (R3) 4-L stirred

122

tank reactor (STR); and (R4) 3-L BCR. The R1 and R4 systems consisted of glass-

123

column bioreactors with porous plates for aeration. The R2 system had 80% of its

124

volume filled with K1 Kaldness support (Veolia Water Technology, France), which is

125

generally used for wastewater treatment purposes by promoting biological biofilm

126

formation over its surface, with a surface area of 500 m2/m3. The STR had a total

127

volume of 5 L, equipped with flat-blade impellers and temperature control, with

128

agitation of 120 rpm. All bioreactors were maintained at 28ºC with an aeration rate of 6

129 130

0.3 vvm (1L of filtered air per minute) during 120 h. Samples of 15 mL were withdrawn every 24 h and were centrifuged at 3000 rpm

131

(2146 g) for 10 min. The supernatant was used for cellulase activity and protein

132

concentration determination, reducing sugars, COD and nitrogen and elementary

133

analysis. The separated solids’ ratios were employed for dry weight biomass

134

determination.

135 136 137

2.2 Cellulase separation and purification procedures Cellulase separation and purification by MF and UF were performed using a

138

VivaFlow 200 (Sartorius Stedim, Germany) crossflow filtration system. The system was

139

equipped with a peristaltic pump connected to a flow restrictor and the pressure

140

indicator to polyethersulfone (PES) membrane cassettes and reservoirs. The membranes

141

had a surface area of 200 cm2. The MF membrane had a pore size of 0.2 μm, and the UF

142

membranes’ molecular weight cut-off (MWCO) were 5 kDa, 10 kDa and 30 kDa.

143

Before experiments, membranes were rinsed by pumping approximately 400 mL

144

of ultrapure water (reverse osmosis, deionization, membrane and ultraviolet processes)

145

(Permution, Brazil); this was because they are stored with a solution of 10% ethanol,

146

according to the manufacturer. After use, membranes were cleaned by recirculating 250

147

mL of 0.5 mM NaOCl: 0.5 M NaOH for 30-40 minutes, according to the

148

manufacturer’s instructions. After fermentation, the cellulase-containing crude broth

149

produced in the BCR was subjected to a vacuum filtration using a Whatman nº. 1 filter

150

paper. The filtrate then passed through MF and UF. The experiments were performed at

151

room temperature, and the samples in the feed tank, retentate and permeate reservoirs

152

were maintained and cooled in an ice bath. A sample volume of 250 mL was used for

153

each experiment. All experiments were carried out with a flow rate of 220 mL/min and 7

154

a transmembrane pressure (TMP) of 1 bar. Graduated reservoirs were used in which it

155

was possible to monitor the volumes of permeate and retentate during the process.

156

As an alternative process, the cellulase crude extract was treated with powdered

157

activated charcoal (AC) (Carbomafra, Brazil), with a surface area between 500 to 1,200

158

m2/g, at a concentration of 20 g/L (Poletto et al., 2015) in order to evaluate its influence

159

on the MF and UF processes. The pretreatment was performed in 500-mL Erlenmeyer

160

flasks with 250 mL of crude extract and 5 g of activated charcoal. Flasks were

161

incubated in a rotatory shaker at 120 rpm for 30 min at 25ºC. Samples were centrifuged

162

at 3,034 g (6,000 rpm) for 10 minutes, and the collected supernatant was filtered

163

through Whatman paper no 1. Samples were then microfiltrated (0.2 μm) and

164

ultrafiltrated (30 kDa membrane).

165

The permeate flux J (L/m2h) was determined and is defined as the flow of a

166

sample volume V (L) through the membrane surface area A (m2) during a time interval t

167

(h), according to Eq. 1 (Rashid et al., 2013).

168 169

(Eq. 1)

170 171

The hydraulic permeability was determined by pumping pure water through the

172

system before and after the samples’ processing. The TMP and the volumes of the

173

reservoirs were monitored and the permeate flux calculated. The water permeability of

174

the membrane Lp (L/m2hbar) was expressed as the permeate flux of deionized water

175

(L/m2h) at a specific TMP (bar), according to Eq. 2 (Qi et al., 2012).

176

8

177

(Eq. 2)

178 179

Irreversible fouling (IF), which is defined as the decrease of the membrane’s

180

pure water permeability after filtration (Lpa) divided by the pure water permeability

181

before filtration (Lpb), was determined according to Eq. 3 (Qi et al., 2012).

182 183

(Eq. 3)

184 185 186

The volume concentration ratio (VCR) is the ratio between the feed volume Vf (L) and the concentrate volume Vc (L) (Qi et al., 2012) and was expressed by Eq. 4.

187 188

(Eq. 4)

189 190

The recovery yield (%) represents the performance of the UF process, relating

191

the enzyme activity of the feed (Af) and the concentrate (Ac), both in U/mL, and the

192

volumes of feed (Vf) and concentrate (Vc) streams, which was determined by Eq. 5

193

(Poletto et al., 2015).

194 195

(Eq. 5)

196 197

The fold purification (Eq. 6) is another way to represent the efficiency of the UF

198

process. This is the ratio between the concentrate-specific activity and the crude broth-

199

(initial feed) specific activity (Poletto et al., 2015).

200

9

201 202 203

(Eq. 6) The fold activity concentration is simply the relation between Ac and Af, according to Eq. 7 (Poletto et al., 2015).

204 205

(Eq. 7)

206 207 208

2.3 Analytical methods The crude and fermented broth, as well as the retentate and permeate of each

209

step of MF and UF, were evaluated for cellulase activity, protein concentration, reduced

210

sugars concentration, biomass, COD and nitrogen concentration. Total cellulase activity

211

was measured according to the filter paper assay (FPase) standard method described by

212

Mandels and Reese (1957) and Ghose (1987), adapted for deep-well microplates

213

according to Camassola and Dillon (2012). Filter paper stripes (0.6 x 1.0 cm) were

214

employed as the substrate, and 50 μL of samples were appropriately diluted in sodium

215

citrate buffer (50 mM; pH 4.8) and incubated for 60 min at 50ºC. After the enzymatic

216

reaction, reduced sugars were measured by the dinitrosalicylic acid (DNS) method

217

(Miller, 1959). One unit of cellulase activity was defined as the amount of enzyme

218

producing 1 μmol of reducing sugars per minute. Protein concentration was determined

219

by the Bradford method, using bovine serum albumin as a standard (Bradford, 1976)

220

and the reagents of the QuickStart Bradford kit (Biorad – USA). COD was analyzed

221

through spectrophotometry, and the total organic nitrogen was evaluated using the

222

Kjeldahl method according to the standard method of the American Public Health

223

Association (APHA, 2012). The biomass was gravimetrically determined and the total

224

organic carbon (TOC) and Total Nitrogen were evaluated by elemental analysis

10

225

performed in a CHNS Vario Macro analyzer (Elementar, Germany).

226 227

3 Results and Discussion

228

3.1 Cellulase production strategies using domestic wastewater

229

Maximum cellulase activity was achieved in the 3L-BCR (R4), with 31.0

230

UFP/mL after 48 h, as shown in Fig. 1-d. The results in terms of substrate to enzyme

231

activity yield and productivity (Table 1) were also higher (4206.2 UFP/g and 645.4

232

UFP/Lh, respectively) for this reactor. The immobilization of the biomass in the R2

233

reactor (Fig. 1-b) led to a lower enzyme activity (5.79 UFP/mL) when compared to R1

234

(Fig. 1-a) (10.27 UFP/mL), and all measured kinetic parameters (Table 1) were also

235

lower, whereas the biomass concentration in both reactors were similar (10.2 g/L and

236

12.5 g/L, respectively). On the other hand, the R3 (STR) reactor (Fig. 1-c) highly

237

favored cellulase activity (17.0 UFP/mL), with a conversion of substrate to biomass

238

(YEA/S) of 897.6 UFP/g and maximum productivity (Qpmax) of 236.5 UFP/Lh, but with

239

lower values in terms of biomass concentration (5.53 g/L), probably associated with the

240

impellers’ brakeage of the fungal hyphae. The YEA/COD values revealed that R3 (44.3

241

UFP/gCOD) better converted the chemical oxidable matter for the formation of cellulase;

242

mechanical agitation would improve absorption of the dissolved organic matter. Alam

243

et al. (2008) achieved a YEA/COD of 21.42 UFP/gCOD, a parameter especially important

244

when developing a biorefinery-based process to convert waste in valuable products.

245

In all the tested reactor systems, the cellulase activity achieved a maximum

246

value followed by a decrease, probably due to proteases attack or the denaturation

247

processes. Alam et al. (2008) also verified that the maximum activity (10.2 UFP/mL) in

248

the fermentation process was achieved after 72 h, followed by a reduction to 2.6

11

249

UFP/mL at 168 h, using wastewater sludge as substrate for Trichoderma harzianum

250

fermentation.

251

The maximum enzyme productivity achieved in the 3-L BCR (R4) (645.4 U/Lh)

252

was about 63 times superior to that obtained by Libardi et al. (2017) (10.2 U/Lh). This

253

is likely related to the differences between the reactors regarding the length-to-diameter

254

ratio (L/D) and superficial gas velocities (vs). The R4 has a L/D of 4.16, while the R1

255

presents a L/D of 3.75. The superficial gas velocity of R4 and R1 were 0.44 cm/s and

256

0.99 cm/s, respectively. The R4 reactor presented a homogeneous (bubbly flow) regime,

257

with uniform small-size bubbles and rise velocities, resulting in a gentle mixing over the

258

column’s entire cross-sectional area. The R1 presented a more heterogeneous regime

259

(churn-turbulent), with a turbulent motion of the larger bubbles and liquid circulation.

260

According to Kantarci et al. (2005), the gas-liquid mass transfer coefficient is lower at a

261

heterogeneous regime, compared to homogeneous flow.

262

The morphology of the fungal biomass during cellulase production in different

263

types of bioreactors was observed (data not shown). In the 1L-BCR (R1), it was

264

possible to see a certain aggregation and sporulation of fungal biomass, which was

265

promoted by the heterogeneous air flow. In the 3L-BCR (R4), the fungal mycelia was

266

very well formed and distributed throughout the bioreactor proving that the

267

homogeneity is crucial for efficient gas-liquid mass transfers in BCRs and the

268

maintaining of an appropriate physiological state for cellulase production by T.

269

harzianum. Contrarily to the pneumatic aeration of BCRs, the mechanical agitation,

270

coupled with forced aeration in R3 (STR), led to the acceleration of fungal growth and a

271

rapid sporulation. This fact culminated in lower enzyme production. Undesirable fungal

272

morphology can lead to poor mass transfer and product secretion during fermentation.

273

In the case of R2, with cell immobilization, a visible biofilm was formed with high 12

274

biomass concentration, probably due to the reduction of shear stress caused by aeration

275

and/or agitation. However, it had a negative impact on cellulase production, as it was

276

not expected and previously reported (Abdella et al., 2016). The relationship between

277

fungal morphology and product formation is difficult to investigate, as there are many

278

interrelated factors (viscosity, agitation and aeration) in a fermentation system that

279

affect both morphology and protein production (Ahmed and Vermette, 2009).

280

Besides bioreactor dimensions and characteristics, the addition of lactose to the

281

culture medium certainly boosted enzyme production: it is a carbon source and cellulase

282

inducer, replacing microcrystalline cellulose that was previously proposed (Libardi et

283

al., 2017). Lactose has been described as an interesting cellulase inducer because it

284

prevents catabolic repression that may be provoked by easily metabolized carbon

285

sources such as glucose, glycerol or fructose (Morikawa et al., 1995). High cellulases

286

activities have been already described with the use of lactose as a carbon source,

287

ranging from 2 to 10%, possibly combined with cellulosic materials and achieving

288

cellulase activities up to 30 UFP/mL (Esterbauer et al., 1991). Pourquie et al. (1988)

289

reached a productivity of 243 UFP/Lh, which was similar to the value found in this work

290

with the R1 system (214 UFP/Lh) and 2.6 times lower than R4 (645.4 UFP/Lh), using a

291

3,000-L fermenter and a Trichoderma strain. These results show the great potential of

292

scaling up this process for even better productivities. Higher results have been related to

293

the use of synthetic carbon sources such as lactose, with or without combination with

294

cellulosic residues. The use of lignocellulosic residues for the production of

295

Trichoderma cellulases resulted in enzyme activities and productivities of 5.25 UFP/mL

296

and 54.68 UFP/Lh, respectively, using corn cob residue at 4% (Liming and Xueliang,

297

2004) and 13.7 UFP/mL, and 91.33 UFP/Lh using wheat bran at 2% (Warzywoda et al.,

298

1983). Ivanova et al. (2013) discovered that lactose permeases (transporters) are present 13

299

in Trichoderma strains and are regulated by the presence of lactose in the culture

300

medium, which explains why lactose is the only carbon source that induces the

301

production of cellulases at an industrial level without inhibition.

302

Other WWTP residues than the raw wastewater have been used for cellulase

303

production, such as sludge. Normally, a pre-treatment of the sludge is necessary to

304

hydrolyze the recalcitrant compounds and render them more accessible to

305

microorganisms; Drani et al. (2011) tested WWTP sludge for cellulase production,

306

achieving maximum productivities of 3.2 U/L.h with Trichoderma strains in submerged

307

fermentation.

308 309 310

3.2 COD and nitrogen removal during cellulase production The maximum reduction of COD content (98%) and nitrogen (78%) was

311

achieved in the 3-L BCR (R4) (Fig. 2), showing that lactose and peptone were

312

efficiently used for the production of fungal biomass and metabolites. According to Fig.

313

1, the reduced COD concentration is accompanied by reduced sugars concentrations for

314

all types of bioreactors, indicating that reducing sugars content is related to COD.

315

Alam et al. (2008) also evaluated COD consumption during Trichoderma

316

harzianum fermentation with sanitary wastewater sludge for cellulases production. The

317

COD concentration was reduced in 96% after 7 days of fermentation, achieving a

318

cellulase activity of 10.2 UFP/mL. Alam et al. (2003) obtained 98% of COD removal for

319

domestic wastewater sludge, fermented with a mixed culture of Aspergillus niger and

320

Penicillium corylophilum. COD reduction has also been evaluated in fermentation

321

processes using various wastes, such as cheese whey, starch processing wastes, apple

322

distillery waste and olive mill wastewater. Ghaly and Kamal (2004) achieved 90.6%

323

COD content reduction for protein production using cheese whey as a medium for 14

324

Kluyveromyces fragilis fermentation. Bo et al. (1999) achieved 95% COD reduction

325

during the production of fungal biomass protein using starch processing wastewater as a

326

culture medium for Aspergillus and Rhizopus strains.

327

The total organic carbon (TOC) and nitrogen percentages during the

328

fermentation process in 1L-BCR (R1) were also evaluated. TOC was reduced from 0.32

329

to 0.2% (68.7%) after 72 hours of fermentation. The same reduction was also observed

330

for COD and reducing sugars concentration, as can be seen in Fig. 1-a. This fact

331

demonstrates that the COD can be used as an indirect measurement of the soluble

332

organic matter in the culture medium. Its reduction is therefore related to the

333

consumption of wastewater’s organic matter, but mostly to the consumption of lactose

334

by the fungi. The total nitrogen increased from 0.04 to 0.08% during the first 24 h of

335

fermentation process, and this value was maintained until 72 h. The increase and/or

336

maintenance of the total nitrogen would be related to the release of proteins produced

337

by the fungal strain in the fermented broth, including the enzymes directly associated

338

with the release of organic nitrogen. The same phenomena was observed by Ghaly and

339

Kamal (2004) in cheese whey fermentation for single-cell protein production using the

340

yeast Kluyveromyces fragilis.

341

The C:N ratio of 8, which was measured in the beginning of the fermentation

342

process, is in accordance with the elementary composition of fungal biomass in terms of

343

dry cell weight (Baltz et al., 2010). Liming and Xueliang (2004) tested the C:N ratios of

344

6, 7, 8 and 9 using corn cob residue (40 g/L) as a carbon source with other compounds,

345

according to the medium proposed by Mandels et al (1981), obtaining the best results in

346

terms of enzymatic activity (5.25 UFP/mL) using a C:N of 8. The C:N ratio decreased to

347

2 after 24 hours of processing, indicating a higher consumption of the carbon source

348

compared to nitrogen. 15

349 350

3.3 Cellulase recovery and concentration in MF and UF

351

Table 2 summarizes the main results of each step of MF and UF. First, cellulase

352

crude extract was microfiltrated to remove particles and cell debris. After MF, cellulase

353

activity was reduced from 5.92 to 3.71 U/mL, resulting in a recovery of 58%, with a

354

permeate flux of 172 L/m2h and efficient clarification of the enzymatic extract. It was

355

possible to observe the behavior of enzymatic activity and protein concentration after

356

each step of UF. The enzymatic activity in the retentate fraction increased with the

357

membrane MWCO; however, the proteins in the permeate fraction increased with

358

higher-MWCO membranes, indicating that there are smaller proteins than the

359

membrane’s pores. According to Fig. 3, higher-MWCO membranes led to higher

360

cellulase-recovery percentages. The highest recovery percentage (73.5%) was obtained

361

with the 30 kDa MWCO membrane. This means that 26% of cellulase activity was

362

detected in the permeate fraction and 0.5% of losses were verified, probably due to

363

other causes, such as fouling or enzyme denaturation. The 10 kDa membrane promoted

364

a cellulase recovery of 63.6%, and the loss of cellulase activity (11.8%) was higher than

365

that of the 30 kDa membrane, while the cellulase recovery in the permeate fraction was

366

24.6%. After UF through a 5 kDa membrane, a highest loss of cellulase activity (31.6%)

367

was reached, with lower enzyme recovery in the retentate (48.8%) and in the permeate

368

(20%) fractions.

369

Regarding protein recovery (Fig. 3), smaller-MWCO membranes led to higher

370

protein-recovery percentages in the retentate fraction. The 10 kDa membrane promoted

371

a protein recovery of 20.1% in retentate. It is clear from Table 2 that the highest

372

increase in activity concentration (4.23-fold) was achieved in the 30 kDa membrane

373

compared to the 10 kDa and 5 kDa membranes, which resulted in 3.66- and 2.78-fold 16

374

increases, respectively. In cellulase concentration experiments, Rashid et al. (2013)

375

achieved a total protein percentage of 34.67% in the retentate fraction using a 10 kDa

376

membrane, with 43.9% of the total protein in the permeate and 21.44% in losses. The

377

same authors achieved a 9.09-fold cellulase activity concentration increase using a 10

378

kDa MWCO membrane.

379

The treatment with activated charcoal (AC) negatively affected cellulase activity

380

in the retentate, which was reduced by 71.23%, compared to the UF with 30 kDa

381

membrane without pretreatment. Total enzyme activity of the crude broth was 1,350 U,

382

reduced to 347.1 U after AC pretreatment (Table 2). This means a reduction in total

383

cellulase activity by 3.8 times without a significant reduction of volume (150 to 130

384

mL). This result indicates that cellulase was probably adsorbed by AC, and that the MF

385

membrane retained fragments of AC. A similar behavior was observed for protein

386

concentration, with a reduction from 1.89 to 1.24 mg/mL in the retentate obtained

387

without AC pretreatment for the 30 kDa MCWO membrane, indicating that the total

388

protein content was also influenced by the use of AC.

389

Higher-MWCO membranes promoted higher cellulase activity recovery and

390

lower losses. Rodrigues et al. (2014) recovered a commercial cellulase (Celluclast 1.5

391

FG L – Novozymes) from a wheat straw hydrolysate using a 10 kDa plate membrane in

392

a cross-flow UF system. The authors optimized the process conditions and improved

393

cellulase recovery values from 32% to 67%. Poletto et al. (2015) concentrated

394

pectinases using a cross-flow (hollow fiber) UF system and achieved a protein recovery

395

of 80.4%, almost the same value as Rodríguez-Fernández et al. (2013), who also

396

observed a pectinase recovery rate of 86.1%. According to Powell and Timperman

397

(2005), the main cause of protein losses in membrane separation processes is pore size

398

distribution. Shear forces also contribute by producing smaller fragments. Some

17

399

proteins are close to the MCWO and could pass through the membrane into the filtrate;

400

Rodrigues et al. (2014) observed losses of around 20% in UF of commercial cellulase

401

using a 10 kDa MWCO membrane. The authors observed the same losses when small

402

volumes of samples were processed at laboratory scale, which influences the recovery

403

percentages. This fact would not be as significant at the industrial scale, where the

404

volumes are higher and small losses do not have the same impact.

405

The use of MF followed by the 30 kDa UF membrane resulted in a global

406

recovery of 42.3% when compared with the crude broth from the fermentation process.

407

The results should be compared to those of other authors with caution, as there is a

408

diversity of UF systems, such as hollow fiber, flat-sheet cassettes, spiral-wound

409

cartridges, tubular modules and other enhanced mass transfer devices, such as charged

410

membranes and electric or ultrasonic fields (Saxena et al., 2009). In addition, very few

411

publications specifically describe the UF of non-commercial cellulases, as in this work.

412

Commercial cellulases that were employed by some authors in MF and UF processes

413

have been previously formulated to ensure their stability during storage. According to

414

Esterbauer et al. (1991), sugars such as glucose and lactose (as well as glycerol or

415

polyalcohol) could be added to prevent the loss of enzyme activity. The interference of

416

these compounds in the culture broth would contribute to differences in cellulase

417

recovery percentages when comparing commercial and non-commercial cellulases. UF

418

processes have become more attractive because they offer the advantages of lower costs

419

and ease of scale-up for commercial production when compared to chromatographic

420

methods, e.g. size-exclusion chromatography for protein concentration (Saxena et al.,

421

2009). Since the aim of this work was to produce cellulases with a low to medium

422

degree of purification, the proposed process could certainly be adopted. Additionally,

18

423

the cellulases were produced with domestic wastewater as substrate to reduce costs,

424

from enzyme production to separation and recovery.

425 426 427

3.4 Analysis of the permeate flux and membrane fouling For all tested membranes, two different stages of the permeate flux were

428

observed. See Fig. 4: first, an expressive drop (around 24 L/m2h for the 5 and 10 kDa

429

membrane and 18 L/m2h for the 30 kDa membrane) within the first minute of process,

430

and afterwards, a low variation. The 30 kDa membrane showed the highest permeate

431

flux (109 L/m2h), followed by the 10 kDa (94 L/m2h) and 5 kDa (14.5 L/m2h). Among

432

the UF membranes, the best anti-fouling performance was observed in the 10 kDa

433

membrane (14.5%), followed by the 30 kDa membrane (19.81%). Conversely, the

434

highest fouling effects were observed in the 5 kDa and MF membranes. The pores of

435

the 5 kDa membrane were possibly blocked by smaller particles of the crude broth that

436

passed through the MF membrane. It is also important to point out that the use of AC

437

promoted increased fouling. In this case, there was a 28% reduction in Lp. The AC

438

released some particles in the culture broth that were retained by the membrane,

439

resulting in its dark appearance after the UF process.

440

The UF process’s operational costs are inversely proportional to the permeate

441

flux. Results from several authors presented in Table 3 show that it is not simple to

442

make a direct comparison, as the permeate flux behavior is dependent on the membrane

443

surface area and porosity, as well as the time of the process.

444

MF has been described as an important step to increase the efficiency and

445

permeate flux of the subsequent UF. Poletto et al. (2015) incremented the pectinase

446

recovery after UF from 59% to 81.4% by adding a prior MF step in a hollow-fiber

447

filtration system. The MF step halved the fouling in the UF step, increasing the 19

448

permeate flux from 11 to 22 L/m2h. Rashid et al. (2013) tested MF membrane pore sizes

449

of 0.2 and 0.45 μm in a hollow-fiber system for cellulase filtration. The authors

450

concluded that, although the 0.2 μm membrane suffered a higher fouling (19.37%)

451

compared to the 0.45 μm membrane (3.8%), the recovered cellulase activity in the

452

permeate was 16% higher, indicating that fouling could also be favorable for the

453

process because it rejects suspended matter and high molecules that could interfere in

454

the target product.

455

Anti-fouling performance is one of the most important factors to consider when

456

choosing a suitable membrane. The water permeability (Lp) values of the 10 and 30

457

kDa membranes obtained in this work were 3 and 2.4 times higher than that obtained by

458

Qi et al. (2012). These authors used 10 kDa (43.7%) and 30 kDa (48.7%)

459

polyethersulfone (PES) membranes, respectively, for cellulase concentration from an

460

enzymatic hydrolysate of steam-exploded wheat straw, using a hollow-fiber filtration

461

system. When testing the UF process (10 kDa MWCO) for pectinases purification with

462

and without a previous MF of the crude broth, Poletto et al. (2015) observed a decrease

463

of 50.6% in the Lp with previous MF, and 72.5% without MF. They achieved a 42.5%

464

reduction in the Lp using an AC pretreatment (5 g/L) and a 50.6% without AC

465

pretreatment. In fact, the authors produced the enzyme by solid-state fermentation,

466

which could have released more particles into the culture broth than the submerged

467

fermentation presented in this work, which efficiently removed the particles through AC

468

pretreatment.

469 470

20

471

3.5 Analysis of the COD and nitrogen concentration of different membrane fractions

472

The evaluation of generated effluents during a UF procedure is crucial in

473

developing a sustainable biotechnological process. Since the retentate fraction contains

474

the product of interest (the concentrated cellulase), the permeate is the effluent of this

475

process and should therefore be evaluated for correct management and disposal.

476

Although the objective of the UF process from an industrial biotechnology point of

477

view was to maximize cellulase recovery, the central task from an environmental

478

perspective was to minimize effluent concentration. The COD of the retentate and the

479

permeate fractions of the MF and UF steps were determined according to the latter

480

reasoning.

481

The MF process promoted a reduction in COD concentration from 36 to 32

482

mg/L (11.36%). Therefore, MF alone was not sufficient to remove the chemically

483

oxidable matter in the fermented broth. According to Fig. 5, a completely different

484

behavior was observed when UF membranes were used, with a very good separation

485

and concentration of chemically oxidable matter (COD) detected in the retentate

486

fractions. Lower concentrations of COD were then present in the permeate fractions.

487

The efficiencies of COD removal by each UF membrane were very similar; a removal

488

of COD from 68 to 9 mg/L (86%) was observed when using the 5 kDa membrane, from

489

64 to 8 mg/L (86%) with the 10 kDa membrane and from 72 to 13 mg/L (81%) with the

490

30 kDa membrane. The efficiency of COD removal obtained using the 10 kDa

491

membrane (86%) for cellulase concentration from a culture broth containing palm oil

492

mill effluent as part of the culture medium was higher than the highest efficiency

493

obtained by Rashid et al. (2013) when using a 10 kDa membrane (30.78%). Cassini et

494

al. (2010) recovered proteins from the wastewater of an isolated soy protein production

495

process; the use of UF with 5 kDa membranes retained 52% of the proteins and reduced 21

496

the COD concentration of the final effluent by 30%, showing the viability of coupling a

497

resource-recovery process and a wastewater treatment technology. Alonso et al. (2001)

498

achieved a COD reduction of 47% with a 50 kDa UF spiral membrane and a 46%

499

reduction using a 0.2 μm MF membrane, respectively, when filtering sanitary

500

wastewater. The higher efficiencies obtained in this work indicate that the majority of

501

the macromolecules of the fermented broth are larger than 30 kDa, and that the tested

502

membranes are efficient for retaining the COD.

503

Another important process parameter was the total organic nitrogen

504

concentration of the permeate and retentate fractions determined, as shown in Fig. 5.

505

The 5 kDa membrane showed the highest efficiency of nitrogen removal from the

506

permeate compared to the retentate fraction (83.9%), followed by the 10 kDa (64.7%)

507

and 30 kDa (52.9%) membranes. The efficiency of nitrogen removal by the MF

508

membrane (51%) was similar to that presented by the 30 kDa membrane, which was

509

also observed in the protein recovery concentration. Alonso et al. (2001) achieved

510

nitrogen removal of 12% using a 50 kDa membrane in a spiral membrane system to

511

filter pre-filtered domestic wastewater.

512

The similarity in nitrogen removal efficiency when comparing the UF (30 kDa) to

513

the MF membranes was not observed for COD removal. According to Alonso et al.

514

(2001), this is related to the N-association with organic macromolecular compounds,

515

which are found in colloidal suspension and retained by the UF and MF membranes.

516

This was not observed for COD, which is probably associated with dissolved matter; the

517

COD concentrations in the permeate fractions of all tested UF membranes were very

518

low. Considering the product of interest is in the retentate fraction, the permeate was the

519

effluent of this concentration process. In this way, the UF process resulted in an effluent

520

with COD concentrations below the acceptable limit of 125 mg/L and total nitrogen

22

521

removal efficiencies close to 70-80% for sanitary wastewater disposal; this is in

522

accordance with European Union Wastewater Directive 98/15/EC, which defines urban

523

and industrial wastewater quality required for disposal. The reduction of COD and

524

nitrogen concentration achieved in this work suggests that the resulting effluent from

525

this production process only requires simple treatment steps to be disposed in nature.

526

Other parameters should be evaluated for this assumption; however, the low COD

527

concentration and high nitrogen removal already indicate reduced costs for treating the

528

resulting wastewater from this enzyme production process. Another fact that increases

529

the environmental importance of evaluating the COD content of the UF processes’s

530

permeate fraction is the large effluent volume, which must be correctly disposed. In this

531

process, the permeate fraction corresponds to 84% of the volume of fermented broth

532

passed through the MF and UF membranes.

533 534 535

4 Conclusions A very efficient process of value-added biomolecule production and recovery

536

was developed using sanitary wastewater as source of nutrients, mainly cellulose. High

537

cellulase productivities were reached with simultaneous COD and nitrogen removal

538

efficiencies of 98% and 78%, respectively, during the fermentation process. Membrane

539

separation with consecutive MF and UF resulted in cellulase recovery of above 70%

540

with COD and nitrogen removal of 81% and 52.9%, respectively. Final effluent

541

environmental characteristics adhered to European Union legislation standards. This

542

work proposed the insertion of a traditional biotechnology process in a biorefinery

543

concept, coupling wastewater treatment/reuse and biomolecules production in a

544

sustainable way.

545 23

546

Acknowledgements

547

The authors would like to thank the Coordination for the Improvement of Higher

548

Education Personnel (CAPES) and the National Council of Technological and Scientific

549

Development (CNPq), Brazil for their financial support.

550 551

References

552

1. Abdella, A., Mazeed, T. E., El-Baz, A.F., Yang, ST. 2016. Production of B-glucosidase from

553

wheat bran and glycerol by Aspergillus niger in stirred tank and rotatin fibrous bed bioreactors.

554

Proc Biochem. 51, 1331-1337. http://dx.doi.org/10.1016/j.procbio.2016.07.004

555

2. Ahamed, A., Vermette, P. Effect of culture médium composition on Trichoderma reesei’s

556

morphopolgy and cellulase production. 2009. Biores. Technol. 100, 5979-5987.

557

3. Alam, M.Z., Fakhru’l-Razi, A., Molla, A.H., 2003. Biosolids accumulation and

558

biodegradation of domestic wastewater treatment plant sludge by developed liquid state

559

bioconversion

560

https://doi.org/10.1016/S0043-1354(03)00260-4

561

4. Alam, M.Z., Muyibi, S.A., Wahid, R., 2008. Statistical optimization of process conditions for

562

cellulase production by liquid state bioconversion of domestic wastewater sludge. Bioresour.

563

Technol. 99, 4709–4716. https://doi.org/10.1016/j.biortech.2007.09.072

process

using

a

batch

fermenter.

Water

Res.

37,

3569–3578.

564

5. Alonso, E., Santos, A., Solis, G.J., Riesco, P., 2001. On the feasibility of urban wastewater

565

tertiary treatment by membranes: A comparative assessment. Desalination 141, 39–51.

566

https://doi.org/10.1016/S0011-9164(01)00387-3

567 568 569 570

6. APHA, 2012. Standard methods for the examination of water and wastewater. twenty second edition, American Public Health Association, Washington. 7. Baltz, R.H., Demain, L., Davies, E., 2010. Manual of industrial microbiology and biotechnology. ASM Press, Washington.

571

8. Bo, J., Hans, L., Q, Y., B, P., 1999. Screening and selection of microfungi for microbial

572

biomass protein production andwater reclamation from starch processing wastewater. J.

573

Chem. Technol. Biotechnol. 74, 106–110.

574

9. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram

575

quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–

24

576

254. https://doi.org/10.1016/0003-2697(76)90527-3

577

10. Camassola, M., Dillon, A.J.P., 2012. Cellulase determination: modifications to make the

578

filter paper assay easy, fast, practical and efficient. J Anal and Bioanal Tech, v.1, p.1–4,

579

2012. https://doi.org/10.4172/scientificreports.458

580

11. Cassini, A.S., Tessaro, I.C., Marczak, L.D.F., Pertile, C., 2010. Ultrafiltration of wastewater

581

from isolated soy protein production: A comparison of three UF membranes. J. Clean. Prod.

582

18, 260–265. https://doi.org/10.1016/j.jclepro.2009.10.016

583 584

12. Charcosset, C., 2006. Membrane processes in biotechnology: An overview. Biotechnol. Adv. 24, 482–492. https://doi.org/10.1016/j.biotechadv.2006.03.002

585

13. Drani, A., Alam, M., Muyibi, S., 2011. Screening of potential strains and co-substrate for

586

production of cellulase enzyme using sewage treatment plant sludge as major source.

587

Biotechnology 1–6. https://doi.org/10.3923/biotech.2011.197.202

588

14. Esterbauer, H., W, S., Labudova, I., Hermann, a., Hayn, M., Steiner, W., Labudova, I.,

589

Hermann, a., Hayn, M., 1991. Production of Trichoderma cellulase in laboratory and pilot

590

scale. Bioresour. Technol. 36, 51–65. https://doi.org/10.1016/0960-8524(91)90099-6

591

15. Ghaly, A.E., Kamal, M.A., 2004. Submerged yeast fermentation of acid cheese whey for

592

protein production and pollution potential reduction. Water Res. 38, 631–644.

593

https://doi.org/10.1016/j.watres.2003.10.019

594 595

16. Ghose, T.K., 1987. Measurement of Cellulases. Pure and Applied Chemistry, v. 59, n. 2, p. 257–268. https://doi.org/10.1351/pac198759020257

596

17. Ivanova, C., Bååth, J.A., Seiboth, B., Kubicek, C.P., 2013. Systems analysis of lactose

597

metabolism in Trichoderma reesei Identifies a lactose permease that is essential for cellulase

598

induction. PLoS One 8, 1–10. https://doi.org/10.1371/journal.pone.0062631

599

18. International Water Association (IWA), 2018. Waste water report 2018: The use

600

opportunity. http://www.iwa-network.org/publications/the-reuse-opportunity/ Acessed in

601

30th October 2018

602 603 604 605

19. Kantarci, N., Borak, F., Ulgen, K.O., 2005. Bubble column reactors. Process Biochem. 40, 2263–2283. https://doi.org/10.1016/j.procbio.2004.10.004 20. Kleerebezem, R., Loosdrecht, M.C.M. Van, 2007. Mixed culture biotechnology for bioenergy production 207–212. https://doi.org/10.1016/j.copbio.2007.05.001

606

21. Libardi, N., Soccol, C.R., Góes-Neto, A., Oliveira, J. de, Vandenberghe, L.P. de S., 2017.

607

Domestic wastewater as substrate for cellulase production by Trichoderma harzianum.

608

Process Biochem. 57, 190–199. https://doi.org/10.1016/j.procbio.2017.03.006

25

609

22. Liming, X., Xueliang, S., 2004. High-yield cellulase production by Trichoderma reesei ZU-

610

02 on corn cob residue. Bioresour. Technol. 91, 259–262. https://doi.org/10.1016/S0960-

611

8524(03)00195-0

612

23. Mandels, M., Medeiros, J.E., Andreotti, R.E., Bissett, F.H., 1981. Enzymatic Hydrolysis of

613

Cellulose: Evaluation of cellulose culture filtrates under use conditions. Biotechnol. Bioeng.

614

XXIII, 2009–2016.

615

24. Mandels, M., Reese, E.T., 1957. Induction of cellulase in Trichoderma viride as influenced

616

by

617

https://doi.org/10.1002/path.1700730133

618 619

carbon

sources

and

metals.

J.

Bacteriol.

73,

269–278.

25. Miller, G.L., 1959. Use of Dinitrosalicylic Acid Reagent for Determination of Reducing Sugar. Anal. Chem. 31, 426–428. https://doi.org/10.1021/ac60147a030

620

26. Mores, W.D., Knutsen, J.S., Davis, R.H., 2001. Cellulase recovery via membrane filtration.

621

Appl. Biochem. Biotechnol. - Part A Enzym. Eng. Biotechnol. 91–93, 297–309.

622

https://doi.org/10.1385/ABAB:91-93:1-9:297

623 624

27. Morikawa, Y., Ohashi, T., Mantani, O., Okada, H., 1995. Cellulase induction by lactose in Trichoderma reesei PC-3-7. Appl Microbiol and Biotechnol, v.44, p. 1992–1993.

625

28. Poletto, P., Da Rocha Renosto, D., Baldasso, C., Zeni, M., Da Silveira, M.M., 2015.

626

Activated charcoal and microfiltration as pretreatment before ultrafiltration of pectinases

627

produced by Aspergillus niger in solid-state cultivation. Sep. Purif. Technol. 151, 102–107.

628

https://doi.org/10.1016/j.seppur.2015.07.029

629

29. Pourquie, J., Warzywoda, M., Chevron, F., Thery, M., Lonchamp, D., Vandercasteele, J.P.,

630

1988. Scale up of cellulase’s production and utilization. Biochemistry and Genetics of

631

Cellulose Degradation. pp. 71–86.

632

30. Powell, M.J., Timperman, A.T., 2005. Quantitative analysis of protein recovery from dilute,

633

large volume samples by tangential flow ultrafiltration. J. Memb. Sci. 252, 227–236.

634

https://doi.org/10.1016/j.memsci.2004.12.024

635

31. Qi, B., Luo, J., Chen, G., Chen, X., Wan, Y., 2012. Application of ultrafiltration and

636

nanofiltration for recycling cellulase and concentrating glucose from enzymatic hydrolyzate

637

of

638

https://doi.org/10.1016/j.biortech.2011.10.049

639 640 641

steam

exploded

wheat

straw.

Bioresour.

Technol.

104,

466–472.

32. Rashid, S.S., Alam, M.Z., Fazli, M.B.F.A., 2013. Separation of Cellulase Enzyme from Fermentation Broth of Palm Oil Mill Effluent by Ultrafiltration Process 1, 501–506. 33. Rodrigues, A.C., Felby, C., Gama, M., 2014. Cellulase stability, adsorption/desorption

26

642

profiles and recycling during successive cycles of hydrolysis and fermentation of wheat

643

straw. Bioresour. Technol. 156, 163–169. https://doi.org/10.1016/j.biortech.2014.01.019

644

34. Rodríguez-Fernández, D.E., Parada, J.L., Medeiros, A.B.P., De Carvalho, J.C., Lacerda,

645

L.G., Rodríguez-León, J.A., Soccol, C.R., 2013. Concentration by ultrafiltration and

646

stabilization of phytase produced by solid-state fermentation. Process Biochem. 48, 374–

647

379. https://doi.org/10.1016/j.procbio.2012.12.021

648

35. Ryan, D., Gadd, A., Kavanagh, J., Barton, G.W., 2009. Integrated biorefinery wastewater

649

design. Chem. Eng. Res. Des. 87, 1261–1268. https://doi.org/10.1016/j.cherd.2009.04.016

650

36. Saxena, A., Tripathi, B.P., Kumar, M., Shahi, V.K., 2009. Membrane-based techniques for

651

the separation and purification of proteins: An overview. Adv. Colloid Interface Sci. 145, 1–

652

22. https://doi.org/10.1016/j.cis.2008.07.004

653

37. Shi, X., Tal, G., Hankins, N.P., Gitis, V., 2014. Fouling and cleaning of ultrafiltration

654

membranes:

655

https://doi.org/10.1016/j.jwpe.2014.04.003

A

review.

J.

Water

Process

Eng.

1,

121–138.

656

38. Singhania, R., Adsul, M., Pandey, A., Patel, A., 2017. Cellulases, in: A. Pandey; S. Negi; C.

657

Soccol (Eds.); Current Developments in Biotechnology and Bioengineering - Production,

658

Isolation and Purification of Industrial Products. Elsevier, Amsterdam, pp.73–97.

659

39. Verstraete, W., Van de Caveye, P., Diamantis, V., 2009. Maximum use of resources present

660

in

661

https://doi.org/10.1016/j.biortech.2009.05.047

662

domestic

“used

water.”

Bioresour.

Technol.

100,

5537–5545.

40. Wan, J., Gu, J., Zhao, Q., Liu, Y., 2016. COD capture: A feasible option towards energy

663

self-sufficient

664

https://doi.org/10.1038/srep25054

domestic

wastewater

treatment.

Sci.

Rep.

6,

1–9.

665

41. Warzywoda, M., Vandercasteele, J. R. & Pourquic, J., 1983. A comparison of genetically

666

improved strains of the cellulolytic fungus Trichoderma reesei. Biotechnol. Lett. 5, 243–246.

667

27

668 669

Figure Captions

670

Fig. 1. Kinetics of cellulase (FPase), biomass and protein production and consumption

671

of reducing sugars, COD and nitrogen in a 1L-BCR (2a), 1-L CPBR (2b), 4-L STR (2c)

672

and 3L-BCR (2d).

673

Fig. 2. Efficiency of COD and nitrogen removal for the 1-L BCR (R1), 1-L 1-L CPBR

674

(R2), 4-L STR (R3) and the 3-L BCR (R4).

675

Fig. 3. (a) Cellulase recovery and (b) protein recovery percentages in the UF permeate

676

(white bars) and retentate (dark grey bars) fractions, and the losses (light grey bars)

677

during the processes.

678

Fig. 4. (a) Permeate flux in 5 (closed diamond), 10 (open square) and 30 (open triangle)

679

kDa ultrafiltration membranes of crude cellulase extract pretreated by microfiltration;

680

(b) permeability of pure water before (dark grey bars) and after (light grey bars) the

681

ultrafiltration, and the fouling (%) of the membranes of different MWCO.

682

Fig. 5. (a) COD and (b) nitrogen concentration and percentage of reduction of the

683

retentate (dark grey bars) and permeate (light grey bars) fractions for different MF and

684

UF membranes (5, 10 and 30 kDa).

28

Table 1. Kinetic parameters of cellulase production in different bioreactor systems Parameters

R1

R2

R3

R4

1-L BCR

1-L CPBR

4-L STR

3-L BCR

EAmax (UFP/mL)

10.27

5.79

17.02

31.0

Biomass (g/L)

12.5

10.2

5.53

13.52

Qpmax (UFP/Lh)

214.0

80.4

236.5

645.4

Qp (UFP/Lh)

60.3

38.3

36.9

186.0

YEA/S (UFP/g)

342.3

196.1

897.0

4206.2

YEA/COD (UFP/gCOD)

20.2

17.9

44.3

34.0

CODremoval (%)

97.7

89.6

85.7

98.1

Nitrogenremoval (%)

60.3

71.9

70.6

77.7

Table 2. Separation, purification and concentration of cellulases by MF and UF using different membranes

Step

Activity (UFP/mL)

Protein (mg/mL)

Specific

Total

Total

Cellulase

Protein

activity

activity

Protein

Recovery

Recovery

(UFP/mg)

(UFP)

(mg)

(%)

(%)

250

5.92

1.41

4.2

1480

352.5

100%

100%

b

MF – 0.2 μm

230

3.71

1.52

2.4

853.3

349.6

58%

99%

UF - 30 kDa

40

5.68

1.89

8.3

627.2

75.6

73.5%a

14.8%

UF - 30 kDa - dAC

30

4.51

1.24

3.6

135.3

37.2

39%a

11%

UF - 10 kDa

40

3.57

1.83

7.4

542.8

73.2

63.6%a

20.1%

UF - 5 kDa

40

0.32

1.79

5.8

412.8

71.6

48.4%a

23%

The presented recovery values are related to the comparison with the MF.

b

c

(mL)

Enzyme

Crude Broth

c

a

Volume

Microfiltration

Ultrafiltration

d

Activated carbon

29

30

31

32

33



First time wastewater re-used as a great opportunity for cellulase production



High cellulase productivity achieved in a 3L-bubble column reactor (645 UFP/L.h)



Enzyme recovery (73.5%) using micro and ultrafiltration 30 kDa membrane



Cellulase production with significant COD (81.37%) and nitrogen (52.9%) reduction

34