Biological abatement of inhibitors in rice hull ... - PubAg

ww.sciencedirect.com

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 8

Available online at w

ScienceDirect

http: / /www.elsevier .com/locate/biombioe

Biological abatement of inhibitors in rice hullhydrolyzate and fermentation to ethanol usingconventional and engineered microbes

Nancy N. Nichols*, Ronald E. Hector, Badal C. Saha, Sarah E. Frazer,Gregory J. Kennedy

Bioenergy Research Unit, National Center for Agricultural Utilization Research, Agricultural Research Service, U.S.

Department of Agriculture, 1815 N. University St., Peoria, IL 61604, USA1

a r t i c l e i n f o

Article history:

Received 12 September 2013

Received in revised form

25 March 2014

Accepted 18 April 2014

Available online

Keywords:

Rice hulls

Fermentation

Inhibitors

Bioabatement

Hemicellulose

* Corresponding author. Tel.: þ1 309 681 627E-mail address: [email protected].

1 Mention of trade names or commercial pnot imply recommendation or endorsemeemployer.http://dx.doi.org/10.1016/j.biombioe.2014.04.0961-9534/Published by Elsevier Ltd.

a b s t r a c t

Microbial inhibitors arise from lignin, hemicellulose, and degraded sugar during pretreat-

ment of lignocellulosic biomass. The fungus Coniochaeta ligniaria NRRL30616 has native

ability to metabolize a number of these compounds, including furan and aromatic alde-

hydes known to act as inhibitors toward relevant fermenting microbes. In this study, C.

ligniaria was used to metabolize and remove inhibitory compounds from pretreated rice

hulls, which comprise a readily available agricultural residue rich in glucose (0.32e0.33 g

glucan/g hulls) and xylose (0.15e0.19 g xylan/g hulls). Samples were dilute-acid pretreated

and subjected to bioabatement of inhibitors by C. ligniaria. The bioabated rice hull hemi-

cellulose hydrolyzates were then utilized for ethanol fermentations. In bioabated liquors,

glucose was converted to 0.58% (w/v) ethanol by Saccharomyces cerevisiae D5a at 100% of

theoretical yield, while fermentations of unabated hydrolyzates either failed to exit lag

phase or had reduced ethanol yield (80% of theoretical). In fermentations using ethanol-

ogens engineered for conversion of pentoses, bioabatement of hydrolyzates similarly

improved fermentations. Fermentation of xylose and arabinose by ethanologenic Escher-

ichia coli FBR5 yielded 2.25% and 0.05% (w/v) ethanol from bioabated and unabated samples,

respectively. Fermentations using S. cerevisiae YRH400 had decreased fermentation lag

times in bioabated hydrolyzates. However, xylose metabolism in S. cerevisiae YRH400 was

strongly affected by pH and acetate concentration.

Published by Elsevier Ltd.

1. Introduction

Production of rice, the third most abundant grain crop in the

world behind wheat and corn (http://www.irri.org), generates

1.gov (N.N. Nichols).roducts in this article is

nt by the U.S. Departme

026

quantities of rice hulls as a by-product. Global production of

rice hulls is estimated at 139 million tonnes annually [1]. The

hulls, which are harvested with the grain and equal 20% of

harvested rice on a dry weight basis, are traditionally

considered a low-value or waste residue due to high lignin

solely for the purpose of providing specific information and doesnt of Agriculture. USDA is an equal opportunity provider and

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 880

(16%) and ash (20%) content. The material is poorly suited as

animal feed because of its low density and high silica content

and resulting abrasiveness [2]. Relative to other types of

lignocellulosic biomass, rice hulls are lower in hemicellulose

and higher in ash. Rice hulls contain 31e35.6% w/w cellulose

and 12.0% hemicellulose [2,3].

Although rice hulls presently have limited commercial

value, they could be converted by fermentation of cell wall

sugars, primarily glucose and xylose, to an end-product such

as ethanol used for fuel [4e6]. The biomass to ethanol process

requires grinding and pretreatment of the material in order to

deconstruct cell wall polymers. Pretreatment renders cellu-

lose accessible to saccharification by cellulases and de-

constructs hemicellulose to monomers or oligomers,

depending on the pretreatment method. Dilute acid pre-

treatment of rice hulls has previously been used to generate

monomeric sugars from rice hulls [2,6e8].

The fibrous nature of lignocellulosic biomass necessitates

harsh pretreatment conditions, which leads to the presence of

fermentation inhibitors in the resulting sugar mixture

because lignin and sugars decompose during pretreatment.

Inhibitors formed during dilute acid hydrolysis of biomass

include the furan aldehydes furfural and 5-hydrox-

ymethylfurfural (HMF) derived from sugar degradation, ace-

tate from hemicellulose, and a number of aromatic aldehydes

and acids from lignin [9]. Fermenting microbes respond to the

presence of furfural and HMF by reduction to alcohols, which

are less toxic to cells. Fermentation may stall while the

microbe attempts detoxification of inhibitors, and in many

cases, the fermentation reaction does not recover [10]. Along

with the development of resistant microbes engineered to

better resist inhibitors, mitigation strategies are needed to

facilitate fermentation of biomass sugars. Methods of inhibi-

tor abatement include dilution, adsorption, extraction, and

precipitation [11,12]. These methods, however, add expense

and generate additional waste.

Biological abatement of inhibitors is an alternate strategy

for eliminating fermentation inhibitors from biomass hydro-

lyzates [13e15]. Biological inhibitor abatement, in which

inhibitory compounds are removed via microbial metabolism,

has the advantage of being suitable for treating liquidesolid

mixtures, without need for chemical inputs and with no

generation of chemical waste. Bioabatement could also pre-

vent concentration of inhibitors in process water, which may

facilitate water recycling. The bioabatement microbe Con-

iochaeta ligniaria NRRL30616 was enriched and isolated from

furfural-contaminated soil based on its ability to both

metabolize furans and withstand the mixture of inhibitory

compounds present in corn stover dilute acid hydrolyzate [16].

Bioabatement was found to be as effective as overliming for

inhibitor mitigation and reducing fermentation lag times [15].

Biomass-to-ethanol fermentations differ from conven-

tional starch fermentations because the sugar streams

derived from biomass feedstocks comprise a mixture of hex-

oses and pentoses rather than the glucose syrup derived from

starch processing. Therefore, fermentation of the resulting

sugars, obtained from cellulose and hemicellulose, optimally

requires a microbe capable of converting a mixed stream of

hexoses and pentoses quantitatively to ethanol. In addition to

efforts to harness naturally-occurring yeasts for efficient

fermentation of pentoses plus hexoses to ethanol, significant

research efforts have resulted in development of recombinant

pentose fermenting ethanologens [17]. Work has been

directed both at incorporation of genes for ethanol fermen-

tation in native pentose-utilizing bacteria such as Escherichia

coli [18,19] and toward engineering ethanol fermenting yeast

(Saccharomyces cerevisiae) for fermentation of xylose [20e24].

In this study, the inhibitor-metabolizing fungus C. ligniaria

NRRL30616 was examined for its ability to remove fermenta-

tion inhibitors from rice hull dilute acid hemicellulose hy-

drolyzate (RHH). The fermentability of the biologically

conditioned RHH was evaluated in ethanol fermentations

using the conventional yeast S. cerevisiae D5a, E. coli strain

FBR5 engineered for high-yield ethanol production, and S.

cerevisiaeYRH400, a strain engineered for xylose fermentation.

2. Materials and methods

2.1. Microbial strains and culture media

S. cerevisiae D5a is available as ATCC 200062 (American Type

Culture Collection, Manassas, VA, USA). S. cerevisiae YRH400

[21] maintains integrated copies of the xylose reductase,

xylitol dehydrogenase, and xylulokinase genes necessary for

xylose utilization. S. cerevisiae D5a and YRH400 were cultured

at 32 �C in YPD medium (10 g yeast extract, 20 g peptone

(Becton Dickinson and Company, Sparks, MD, USA), and 20 g

glucose per L). E. coli FBR5 [18] carries the Zymomonas mobilis

pdc and adhB genes for fermentation of pyruvate to ethanol. E.

coli FBR5 was grown at 35 �C in LB medium (10 g peptone, 5 g

yeast extract, and 5 g NaCl per L) containing 4 g/L xylose. C.

ligniaria NRRL30616 was subcultured periodically in mineral

medium (12.5 mM each Na2HPO4 and KH2PO4, 0.1% (w/v)

(NH4)2SO4, and 1 ml/L of Hutner’s mineral solution [25])

containing 10 mM furfural as the source of carbon and

energy.

2.2. Preparation of hydrolyzates

Rice hulls were obtained whole from Label Peelers (Kent, OH)

or as a 20/80 grind fromRice Hull Specialty Products (Stuttgart,

AR). Whole rice hulls were ground in a knife mill (Grindomix

GM200, Retsch GmbH, Haan, Germany) at 6000 rpm for 10 s

followed by grinding twice at 10,000 rpm for 30 s. The resulting

ground material passed through a No. 18 (1 mm) screen.

Ground rice hulls had a moisture content of 6.2% w/w and

contained 35.6 � 0.1% cellulose, 12.0 � 0.7% hemicellulose,

15.4 � 0.2% lignin, and 18.7 � 0.0% ash (w/w, dry basis).

The groundmaterial was hydrolyzed in a rotating stainless

steel reactor equipped with infrared heating (Labomat BFA-12

v200, Werner Mathis, Concord, NC). Samples containing 30 g

biomass in 100 ml 1.0% (v/v) H2SO4 were heated (45 min) to

150 �C, held for 20 min, and cooled (35 min) to room temper-

ature. Samples were rotated at 50 rpmduring pretreatment. In

cases where solids were removed (centrifugation for 10 min,

25 �C, 15,000 g), the solids were washed with a 10% volume of

sterile water, which was added back to the supernatant. The

combined supernatant plus wash liquid was adjusted to pH

6.5 with Ca(OH)2 and filter-sterilized prior to inoculation with

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 8 81

the bioabatement strain. Hydrolyzate composition is shown in

Table 1.

2.3. Inhibitor bioabatement and ethanol fermentations

C. ligniaria NRRL30616 was precultured for 20e22 h in RHH

(200 rpm; Innova 4230 incubator; New Brunswick Scientific,

New Brunswick, NJ). A cell pellet equal to a 10% v/v inoculum

was collected by centrifugation and washed with an equal

volume of mineral medium, then added to RHH to initiate

bioabatement. The inoculated hydrolyzates (200e250 ml per

1000 ml Erlenmeyer flask) were incubated at 30 �C with

shaking (225 rpm). An uninoculated flask containing hydro-

lyzate served as an unabated control for each bioabatement

experiment and subsequent fermentation.

Fermentations of RHHwere carried out by inoculation with

either E. coli FBR5, S. cerevisiae D5a, or S. cerevisiae YRH400 as

follows. E. coli FBR5 was cultured in LB containing 20 g/L

xylose, without shaking. A cell pellet, washed with mineral

medium and equal to 5% v/v inoculum, was added to initiate

fermentation. S. cerevisiaeD5awas grown in YPDwith shaking

at 225 rpm, washed with mineral medium, and added as a

5% v/v inoculum. S. cerevisiae YRH400 was cultured in YPD

medium containing 5% w/v glucose, and fermentations were

inoculated to an optical density (A600 nm; 1.0 cm pathlength)

of 2.0; this was approximately equal to a 10% inoculum.

Fermentations using S. cerevisiae D5a and YRH400 were

initiated at pH 4.5 unless otherwise stated. Fermentations

using E. coli FBR5 contained 1X LB medium and 0.05 M MOPS,

pH 7. Theywere carried out in duplicate or triplicate in Ankom

(Macedon, NY) gas production systems and monitored by

measuring CO2 production, a by-product of ethanol fermen-

tation, at 15 min intervals. Fermentation broths (100 ml) were

mixed with magnetic stir bars and sampled for HPLC analysis

of fermentation products at the beginning and end of incu-

bation. Gas production was calculated by converting the

measured pressure accumulation to CO2 using the ideal gas

law. When pH control was required for some YRH400 fer-

mentations, fleaker bioreactors [26] were used and sampled

periodically for HPLC analysis. Temperature was maintained

at 32 �C by a circulating water bath and pH was controlled by

addition of 1 M NaOH. Fleaker fermentations (100 ml) were

carried out in duplicate.

In simultaneous saccharification and fermentation (SSF)

reactions, pretreated RHH containing solids was autoclaved

for 20 min before bioabatement. After bioabatement was

carried out with solids present, an additional 0.1% (w/v)

(NH4)2SO4 along with 50 mM citric acid, pH 4.5, were added.

GC220 cellulase (10 Filter Paper U/g (dw) solids; Genencor,

Beloit, WI), Novo188 beta-glucosidase (26 U/g (dw) solids), and

a washed inoculum of E. coli FBR5, S. cerevisiae D5a, or S. cer-

evisiae YRH400 cells were then added to initiate fermentations

Table 1 e Concentration of furans, sugars, and acetate in RHH

Abatement time (h) Furfural (mM) HMF (mM) Acetate (% w

0 13.6 � 5.2 2.8 � 0.6 0.63 � 0.11

22.2 � 1.4 2.6 � 2.7 1.6 � 0.9 0.60 � 0.11

48.0 � 0.0 2.5 � 1.0 1.4 � 1.0 0.57 � 0.19

in 100 ml total volume. SSF fermentations were carried out

and monitored in gas production systems or fleakers as

described above.

2.4. Analytical methods

Cellulose, hemicellulose, lignin and ash content in rice hulls

were measured according to National Renewable Energy

Laboratory analytical procedures. After two-stage acid hy-

drolysis, structural carbohydrates were determined using

HPLC measurement of sugars and acid-insoluble lignin was

determined gravimetrically [27]. Ash content was determined

using a muffle furnace [28]. Moisture content was determined

by measurement of loss on drying (IR60, Denver Instrument,

Bohemia, NY). Sugar and ethanol concentrations were deter-

mined using HPLC separation (Aminex HPC-87H column, Bio-

Rad, Richmond, CA) and refractive index detection. Samples

were run at 65 �C and eluted (0.6 ml/min) with 5 mM H2SO4.

Furfural and HMF were measured in hydrolyzate using

reverse-phase HPLC with ultraviolet detection at 277 nm as

described by Lopez et al. [16].

3. Results

3.1. Bioabatement of furans and acetate in RHH

RHHwas prepared by dilute acid pretreatment of rice hulls, to

solubilize hemicellulose while largely leaving cellulose as re-

sidual solids. Furan aldehydes (furfural and HMF) averaged

14.8 � 6.3 mM and 3.0 � 0.5 mM, respectively, in RHH. Bio-

abatement, using C. ligniaria NRRL30616, was carried out to

mitigate the fermentation inhibitors present in RHH. In bio-

abatement, C. ligniaria was inoculated into the hydrolyzate

supernatant to remove inhibitors by metabolism. The results

of bioabatement are shown in Table 1. By 22 h, furfural was

depleted 81% on average, while HMFwas reduced by 43% from

its starting concentration. At 22e24 h, 0.04% w/v glucose,

corresponding to 5% of soluble glucose in RHH, had been

consumed by the bioabatementmicrobe, and no pentoses had

been consumed.

Unlike furfural and HMF, little acetate was consumed from

RHH during bioabatement by C. ligniaria, with only 9.5% of the

acetate depleted after 48 h (Table 1). However, in separate

experiments, over a longer time course (not shown), 74% of

the starting acetate concentration was consumed in 96 h of

bioabatement incubation. In those experiments, at the 96 h

time point, 0.45% w/v glucose and 0.02% w/v xylose were also

consumed by C. ligniaria, corresponding to loss during bio-

abatement of 31% of the starting glucose monomers and 0.7%

of xylose monomers. Therefore, to avoid loss of fermentable

sugars that occurred during longer bioabatement incubations,

after bioabatement with C. ligniaria NRRL30616.

/v) Glucose (% w/v) Xylose (% w/v) Arabinose (% w/v)

1.59 � 0.52 3.28 � 0.23 0.40 � 0.07

1.50 � 0.52 3.27 � 0.34 0.40 � 0.07

1.35 � 0.58 3.12 � 0.26 0.40 � 0.07

Fig. 1 e Fermentation of rice hull dilute acid hydrolyzate by

A. E. coli FBR5 B. S. cerevisiae D5a, and C. S. cerevisiae

YRH400. Ethanol fermentations were monitored by

measuring pressure accumulation due to CO2 production.

Each line represents an individual fermentation of

bioabated or unabated RHH.

bioabatedandunabatedRHH

withE.coliFBR5.

Sugarco

nsu

med

duringferm

entation

Sugarremainingatth

eendofferm

entationa

Finalxylitol

(%w/v)

Finaleth

anol

(%w/v)

Eth

anolyield

(g/g)c

Gluco

se(%

w/v)

Xylose

(%w/v)

Arabinose

(%w/v)

Gluco

se(%

w/v)

Xylose

(%w/v)

Arabinose

(%w/v)

0.03�

0.01

0.09�

0.05

0.00�

0.00

1.13�

0.03

3.06�

0.03

0.36�

0.01

0.00�

0.00

0.05�

0.02

0.42

1.01�

0.01

2.92�

0.22

0.29�

0.06

0.05�

0.00

0.19�

0.24

0.14�

0.12

0.00�

0.00

2.25�

0.19

0.53

calculatedasth

etimeto

reach

5%

ofmaxim

um

psi

inferm

entationbottles.

þxylose

þarabinose

consu

med.

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 882

fermentation experiments were carried out after 22e24 h

bioabatement unless otherwise stated.

Table

2e

Ferm

entationof

Bioabatement

Lagtime

(h)b

NN

Y6.3

�1.1

a93h.

bFerm

entationlagtimeswere

cgeth

anolpro

duce

d/g

gluco

se

3.2. Effect of bioabatement on fermentation of rice hullhydrolyzate

Fermentations of RHH that had been subjected to bio-

abatementwere compared to fermentations of unabated RHH.

Bioabatement was carried out on hydrolyzates from which

solids were removed and washed, with the wash liquid added

back to the RHH supernatant. In these fermentations of

hemicellulose hydrolyzate, the solids were removed after

Table 3 e Fermentation of bioabated and unabated RHH with S. cerevisiae D5a.

Bioabatement Lag time(h)b

Sugar consumed during fermentation Sugar remaining at the end of fermentationa Final xylitol(% w/v)

Final ethanol(% w/v)

Ethanol yield(g/g)c

Glucose(% w/v)

Xylose(% w/v)

Arabinose(% w/v)

Glucose(% w/v)

Xylose(% w/v)

Arabinose(% w/v)

N 4.2 � 4.2d 0.64 � 0.48 0.12 � 0.18 0.04 � 0.03 0.48 � 0.43 3.07 � 0.22 0.44 � 0.08 0.11 � 0.01 0.27 � 0.2 0.42

Y 0.8 � 0.5 0.98 � 0.20 0.10 � 0.09 0.04 � 0.04 0.02 � 0.03 3.03 � 0.33 0.43 � 0.05 0.03 � 0.03 0.50 � 0.10 0.51

a 72 h.b Fermentation lag times were calculated as the time to reach 5% of maximum psi in fermentation bottles.c g ethanol produced/g glucose consumed.d Fermentation lag times could be calculated for only the three fermentations (of seven total) that exited lag phase.

Table 4 e Fermentation of bioabated and unabated RHH with S. cerevisiae YRH400.

Bio-abatement pHcontrol

Lag time(h)a

Sugar consumed during fermentation Sugar remaining at the end of fermentation Final xylitol(% w/v)

Final ethanol(% w/v)

Ethanol yield(g/g)b

Glucose(% w/v)

Xylose(% w/v)

Arabinose(% w/v)

Glucose(% w/v)

Xylose(% w/v)

Arabinose(% w/v)

N Nc 1.4 � 1.6 0.79 � 0.76 0.24 � 0.14 0.01 � 0.00 0.52 � 0.72 3.27 � 0.02 0.39 � 0.02 0.01 � 0.01 0.42 � 0.41 0.38

Y Nc 0.6 � 0.2 1.13 � 0.03 0.17 � 0.06 0.01 � 0.00 0.02 � 0.01 3.17 � 0.12 0.39 � 0.02 0.04 � 0.00 0.65 � 0.03 0.50

N Yd NA 1.04 � 0.05 0.85 � 0.29 0.05 � 0.05 0.01 � 0.01 2.62 � 0.29 0.38 � 0.06 0.19 � 0.09 0.68 � 0.01 0.36

Y Yd NA 1.22 � 0.28 1.18 � 0.30 0.06 � 0.05 0.00 � 0.00 2.21 � 0.31 0.34 � 0.05 0.41 � 0.23 0.79 � 0.14 0.33

a Fermentation lag times were calculated as the time to reach 5% of maximum psi in fermentation bottles.b g ethanol produced/g glucose þ xylose consumed.c Fermentations were initiated at pH 4.5, without pH control during fermentations, and monitored for 72 h.d pH was controlled at 5.5 throughout the fermentations, and sampled after 165 h.

bio

mass

and

bio

energy

67

(2014)79e88

83

0

10

20

30

40

50

60

70

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 25 50 75 100

Xylo

se u

lized

(%)

Acet

ate

(% w

/v)

Bioabatement me (h)

Fig. 2 e Extended length of bioabatement using C. ligniaria

NRRL30616 results in decreased acetate concentration in

RHH at the end of bioabatement. Longer bioabatement of

RHH, and decreased acetate concentration, were associated

with increased xylose consumption by the xylose-utilizing

yeast strain YRH400. Samples were bioabated for 24e96 h,

then fermented using YRH400. B acetate concentration at

the start of fermentation after bioabatement for 24e96 h;C

% of xylose consumed by the end of fermentation after

bioabatement for 24e96 h. Best fit line is shown for xylose

consumption.

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 884

pretreatment, prior to bioabatement and fermentation.

Sugars and fermentation products were measured at the

beginning and end of fermentations, and gas (CO2) pressure

detectors were used during fermentations to conveniently

assess fermentation progress and measure fermentation lag

times. In bioabated RHH supernatants, fermentations pro-

ceededwith little delay (Fig. 1aec), and glucosewas consumed

by all three fermenting microorganisms (Tables 2e4). In

contrast, and as expected based on the fermentation profiles

shown in Fig. 1, a significant amount of residual sugars

remained after fermentations of unabated RHH. As shown in

Fig. 1 and Tables 2e4, the unabated RHH was fermented

poorly by S. cerevisiae strains YRH400 and D5a and not at all by

E. coli FBR5.

E. coli FBR5, an engineered ethanologen with native ability

to metabolize xylose and arabinose, consumed most of the

xylose and arabinose (Table 2) present in bioabated RHH. RHH

supernatant that was abated prior to fermentation yielded

2.25%w/v ethanol at a yield essentially equal to the theoretical

maximum of 0.51 g/g.Without bioabatement, little to no sugar

was consumed and fermented to ethanol by E. coli FBR5. A

summary of E. coli FBR5 fermentations is shown in Table 2.

In fermentations using S. cerevisiae D5a, glucose was

completely consumed by the yeast in fermentations of bio-

abated RHH supernatant (Table 3). The ethanol produced in

these fermentations averaged 0.50% w/v, nearly twice that

obtained in unabated hydrolyzate. Without inhibitor abate-

ment, fermentations with strain D5a either failed to exit lag

phase or had reduced ethanol yield, as shown in Table 3. S.

cerevisiaeD5a does not have genes for completemetabolism of

xylose and although low levels of xylitol were occasionally

detected in these experiments due to the presence of an

endogenous aldose reductase [29], strain D5a did not consume

a significant amount of xylose in fermentations of RHH.

Therefore, fermentations were also carried out using a yeast

strain, S. cerevisiae YRH400, engineered with Pichia stipitis XYL1

and XYL2 and S. cerevisiae XKS1 for xylose metabolism [21].

Strain YRH400 completely fermented glucose from bioabated

RHH while without abatement, 0.52% w/v glucose was not

consumed (Table 4).

Under the conditions initially used for fermentation, strain

YRH400 did not metabolize xylose in bioabated RHH (Table 4).

These fermentations using YRH400 were initiated at pH 4.5,

without pH control during the 72 h fermentations. When the

starting pH was increased for YRH400 fermentations,

increased utilization of xylose was observed as follows: 15% of

xylose was consumed when the pH at the start of fermenta-

tion was 6.5, compared to 5% utilization of xylose when

fermentation was initiated at pH 4.5. The pH in the former

reaction drifted from 6.5 to 5.5 by the fermentation end point

(72 h).

3.3. Effect of acetate on metabolism of xylose by YRH400

In the experiments described above, xylose consumption by

the engineered strain YRH400 was impacted by the starting

pH. However, the fermentation apparatus used to monitor

CO2 production did not allow continuous pH control during

the course of fermentation. Therefore, additional fermenta-

tions were carried out utilizing pH control. Constant pH,

ranging from 4.5 to 6.5, was maintained by automated addi-

tion of NaOH. In these fermentations, highest utilization of

xylose by YRH400 occurred with a controlled pH of 5.5 (Table

4), with 35 � 9% of the xylose consumed from bioabated

RHH supernatant. In unabated hydrolyzates, 24 � 8% of the

xylose was metabolized when fermentations were controlled

at pH 5.5.

Minimal utilization of xylose by YRH400 at pH 4.5 (and

improved utilization at pH 5.5) suggested that acetate present

in RHH may limit xylose fermentation. As shown in Table 1,

only 4.8% of the acetate present in RHHwas consumed during

bioabatement for 22 h. Therefore, additional experiments

with extended bioabatement times were carried out to

determine whether more complete removal of acetate resul-

ted in increased xylose consumption by YRH400. After

extended bioabatement, the hydrolyzates were fermented by

strain YRH400. In these experiments (Fig. 2), xylose con-

sumption correlated positively with bioabatement time

(r2 ¼ 0.94) and negatively with acetate concentration in RHH

(r2 ¼ 0.71.) In RHH bioabated for 72 h, 51.2� 3.4% of xylose was

consumed by YRH400, compared to 23 � 6.3% of xylose

consumed in unabated RHH. Bioabatement beyond 72 h did

not result in further increases in xylose utilization. Notably,

although the amount of xylose consumed increased in RHH

bioabated up to 72 h, the amount of ethanol as fermentation

product did not increase correspondingly. Rather, the pro-

duction of xylitol by YRH400 increased in these samples (Table

5).

3.4. Simultaneous saccharification and fermentation

In order to measure conversion of cellulose to ethanol, bio-

abatement was carried out on RHHwithout removal of fibrous

solids, which were then saccharified with cellulase enzymes

and fermented. Simultaneous saccharification and fermen-

tations (SSF) using E. coli FBR5 in unabated samples failed

Table

5e

Ferm

entationofRHH

withS.cerevisiaeYRH400afterextendedbioabatem

enttim

e.a

Bio-abatement

tim

e(h)

Ace

tate

(%w/v)c

Sugarco

nsu

medduringferm

entation

Sugarremainingatth

eendofferm

entationb

Finalxylitol

(%w/v)

Finaleth

anol

(%w/v)

Eth

anolyield

(g/g)d

Gluco

se(%

w/v)

Xylose

(%w/v)

Arabinose

(%w/v)

Gluco

se(%

w/v)

Xylose

(%w/v)

Arabinose

(%w/v)

00.59�

0.01

1.04�

0.05

0.85�

0.29

0.05�

0.05

0.01�

0.01

2.61�

0.29

0.38�

0.06

0.19�

0.09

0.68�

0.01

0.36

24

0.57�

0.03

1.22�

0.28

1.18�

0.30

0.06�

0.05

0.00�

0.00

2.21�

0.31

0.34�

0.05

0.41�

0.23

0.79�

0.14

0.33

48

0.56�

0.05

1.01�

0.16

1.23�

0.30

0.08�

0.07

0.01�

0.00

2.17�

0.34

0.33�

0.08

0.30�

0.02

0.83�

0.08

0.37

72

0.42�

0.09

0.72�

0.15

1.74�

0.07

0.10�

0.00

0.01�

0.00

1.66�

0.13

0.30�

0.01

0.49�

0.06

0.77�

0.09

0.32

96

0.20�

0.01

0.70�

0.37

1.72�

0.53

0.06�

0.05

0.01�

0.00

2.04�

0.61

0.35�

0.07

0.45�

0.25

0.71�

0.15

0.35

apH

wasmaintainedat5.5

byco

ntrolledadditionofNaOH.

b165h.

cAce

tate

conce

ntrationatth

ebeginningofferm

entation.

dgeth

anolpro

duce

d/g

gluco

seþ

xylose

consu

med.

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 8 85

completely (i.e. no ethanol was produced) while SSF of bio-

abated RHH proceeded after a 6.4 h� 0.8 h lag and yielded 65%

of the theoretical maximumbased on hydrolyzed glucose plus

xylose in the SSF (Fig. 3A). Fermentations using the conven-

tional yeast S. cerevisiae D5a are shown in Fig. 3B. Ethanol

yields were 36% and 41% of the theoretical maximum for

unabated and bioabated samples, respectively. The effect of

bioabatement in fermentations using D5a was also apparent

in the reduced lag time (3.0 � 0.7 h) observed for bioabated

samples compared to unabated samples, which averaged

9.1 � 4.0 h lag time. For the xylose-fermenting yeast YRH400

(Fig. 3C), ethanol yields were 49% and 51% of the theoretical

maximum for bioabated and unabated samples, respectively,

with a slightly reduced lag time associated with bioabatement

(2.4 � 0.5 h) compared to unabated samples (4.1 � 1.6 h). Fer-

mentations using YRH400 (Fig. 3C) were complete sooner than

those using D5a (Fig. 3B).

Fig. 3 e SSF of rice hull dilute acid hydrolyzate containing

30% (w/v) solids by A. E. coli FBR5 B. S. cerevisiae D5a, and C.

S. cerevisiae YRH400. Fermentations were monitored by

measuring pressure accumulation due to CO2 production.

Each line represents an individual fermentation of

bioabated or unabated RHH.

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 886

4. Discussion

Rice hulls comprise a feedstock that is centrally collected in

modern harvesting operations, but has low value as animal

feed. Both glucose derived from cellulose and xylose solubi-

lized from dilute-acid pretreatment of hemicellulose are uti-

lized in conversion of lignocellulosic biomass. Therefore, RHH

was targeted for bioabatement and fermentation due to the

glucan and xylan content of rice hulls (0.32e0.33 and

0.15e0.19 g glucan and xylan, respectively, per g hulls) and due

to the large regional availability of the feedstock. We evalu-

ated the utility of C. ligniaria for abatingmicrobial inhibitors in

dilute-acid pretreated RHH. Of particular note are furfural and

HMF, potent inhibitors of microbial fermentations, and ace-

tate, which significantly inhibits xylose metabolism in engi-

neered yeast strains [30e33]. Hydrolyzate conditioning,

therefore, is important for removal of furans and other

fermentation inhibitors.

Biological conditioning of hydrolyzates could avoid some

of the difficulties associated with physicalechemical methods

of inhibitor mitigation [34,35]. C. ligniaria grows in liquid cul-

ture with a yeast-like appearance [36] and is amenable to

growth in a hydrolyzate containing solids. The fungus C. lig-

niaria NRRL30616 has native ability to utilize a number of

inhibitory compounds known to act as fermentation in-

hibitors [37]. Furans as well as aromatic and aliphatic acids

and aldehydes in corn stover hydrolyzate were shown to be

metabolized in corn stover hydrolyzates by C. ligniaria,

resulting in improved fermentability of lignocellulosic hy-

drolyzates [15,37]. In the present study, furfural and HMFwere

also efficiently metabolized during bioabatement of RHH

(Table 1); 80% of furfural and 43% of HMFwere metabolized by

NRRL30616 by 22 h.

Acetate, which arises in pretreated hydrolyzates from

hemicellulose, is also of interest because it is inhibitory to

fermenting microbes and negatively affects utilization of

xylose by engineered ethanologens [30e33]. C. ligniaria

NRRL30616 can grow using acetate as a sole source of carbon

and energy, although in the experiments described here, only

15% of the acetate was depleted after bioabatement for 48 h

(Table 1). This is presumably due to a higher priority assigned

to detoxification of furfural, HMF, and other inhibitory com-

pounds by the bioabatement strain NRRL30616. In longer time

course experiments, 73% of the starting acetate concentration

was consumed after 96 h of bioabatement. However, longer

incubations to consume acetate may not be desirable for

bioabatement using wild-type C. ligniaria because 38% of the

starting glucose and 0.7% of the starting xylose was also

consumed (not shown). Because engineered xylose-utilizing

yeast have heightened sensitivity to the effects of acetate,

enhanced removal of acetate during bioabatement by C. lig-

niaria NRRL30616 may facilitate efficient fermentation of

xylose by engineered yeast.

For ethanol fermentations, RHH was subjected to bio-

abatement using C. ligniaria and then inoculated with either a

conventional strain of S. cerevisiae, an engineered E. coli

ethanologen, or a S. cerevisiae strain engineered for xylose

metabolism. In all cases, bioabatement of RHH was associ-

ated with decreased fermentation lag times and in some

cases, bioabatement was the difference between successful

fermentation of hydrolyzate and fermentations that failed

completely (Figs. 1, 3). In other fermentations, the effect of

bioabatement was seen in the ability of the fermenting

microbe to consistently and completely consume glucose in

RHH. Although there is relatively little solubilized glucose

present in the hemicellulose hydrolyzate (without biomass

solids and saccharifying enzymes), the concentration of

glucose consumed by the fermenting yeast or bacterial strain

serves as a useful marker for the effectiveness of bio-

abatement. E. coli FBR5 failed to exit lag phase and consumed

little to no glucose in unabated hydrolyzates (Fig. 1, Table 1),

while the conventional yeast strain S. cerevisiae D5a per-

formed inconsistently. D5a sometimes failed to consume any

glucose in unabated hydrolyzates (Fig. 1) and other times

proceeded with fermentation after a variable delay (Table 3).

S. cerevisiae YRH400 exited lag phase relatively quickly, but

left some glucose unfermented in the unabated hydrolyzate

(Table 4).

Conversion of xylose, arising from hemicellulose, is an

important consideration in utilization of lignocellulosic

biomass. Here, fermentations with two engineered microbes

were used to examine conversion of xylose in dilute acid

hemicellulose hydrolyzates prepared from rice hulls. E. coli

FBR5 is an engineered ethanologen with natural ability to

ferment xylose and arabinose [18]. In fermenting bioabated

RHH, E. coli FBR5 converted xylose at essentially theoretical

yield to ethanol, producing 2.3% w/v ethanol (Table 2).

Without bioabatement, essentially no glucose, xylose, or

arabinose was consumed by E. coli FBR5.

S. cerevisiae YRH400 ferments xylose via stably integrated

Pichia stipitis xylose reductase (XYL1) and xylitol dehydroge-

nase (XYL2) genes and the S. cerevisiae xylulokinase (XKS1)

gene. In culture medium containing xylose, YRH400 yielded

0.24 g ethanol/g xylose, while in fermentations of alkaline

pretreated switchgrass, the strain produced 14%more ethanol

than the xylose-nonfermenting parental strain [21]. In the

present study, initial fermentations of RHH using strain

YRH400 showed little xylose consumption in both bioabated

RHH and unabated controls, possibly due to acetate present in

RHH. As stated previously, little acetate was consumed during

the standard bioabatement timeframe (Table 1) and the initial

fermentations in which S. cerevisiae YRH400 consumed little

xylose were initiated at pH 4.5 and conducted without pH

control (Table 4). The inhibitory effect of acetate is more

pronounced at lower pH because the concentration of the

undissociated (acetic acid) form increaseswith decreasing pH.

Undissociated acetic acid diffuses across cell membranes and

thus is more inhibitory than the dissociated (acetate) form

[30].

Two approaches were used to mitigate the potential

negative effects of acetate on xylose utilization by the yeast

strain YRH400. In the first, fermentation of RHH at varied pH

resulted in improved xylose utilization by YRH400, as shown

in Table 4 for fermentations controlled at pH 5.5. The greater

consumption of xylose at pH 5.5 did not, however, result in

increased ethanol production, because the strain produced

additional xylitol rather than ethanol. As noted by Hector et al.

[21], activities of pentose phosphate pathway enzymes and

xylose transporters affect xylose fermentation in strain

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 8 87

YRH400, which may require further genetic modification to

decrease xylitol production and increase ethanol yield.

We also sought to determine whether increased time of

bioabatement would improve xylose fermentation by YRH400.

Some improvement may be expected if the increased abate-

ment incubation time resulted in increased consumption of

acetate and other inhibitors affecting xylose utilization. In a

series of fermentations, a time course of inhibitor abatement

ranging from 0 h (unabated) to 96 h of bioabatement was fol-

lowed by YRH400 fermentation. Xylose consumption

increased up to 56%, with a final ethanol concentration of

0.83% w/v (Table 5). However, the increased metabolism of

xylose in these experiments cannot be tied definitively to

removal solely of acetate, since C. ligniaria also metabolizes

numerous additional inhibitory compounds present in

biomass dilute acid hydrolyzates [36]. As in earlier fermenta-

tions, a significant portion of xylose was converted to xylitol

rather than ethanol.

In experiments containing rice hull solids, bioabatement

resulted in SSF ethanol yields ranging from 41 to 65% of the

theoretical maximum based on saccharified sugars whereas

SSF of unabated RHH using E. coli FBR5 yielded no ethanol.

However, SSF ethanol yields for both yeast strains in un-

abated RHH were close to those obtained in bioabated RHH.

For the D5a and YRH400 SSF reactions, the impact of bio-

abatement on SSF can be seen in reduced fermentation lag

times (Fig. 3) although the effect for YRH400 is modest due to

relatively short lag times observed in unabated RHH. Xylose

reductase from Pichia stipitis has been shown to reduce fur-

aldehyde compounds in lignocellulosic hydrolyzates [38].

Xylose reductase from Zymomonas mobilis was also recently

shown to exhibit activity toward furfural and HMF [39]. The

improved ability of YRH400 to ferment the unabated RHH is

most likely due to increased expression of xylose reductase

engineered into this strain, which required less time to

complete fermentations than D5a (Fig. 3). These data also

indicate that it may be possible to use harsher pretreatment

conditions and/or higher solids loading (i.e., increased in-

hibitor concentrations) when using bioabatement and

YRH400.

5. Conclusion

This work demonstrates that biological abatement facili-

tates fermentation of sugars obtained from dilute acid pre-

treatment of rice hulls. A specially selected microbe, C.

ligniaria NRRL30616, was used to metabolize undesirable

compounds which, if left untreated, can complicate micro-

bial conversion of sugars to the desired end product. In

ethanol fermentations, conversion of glucose in bioabated

hemicellulose hydrolyzates proceeded to completion, while

fermentations in unabated samples either stalled or failed to

completely utilize glucose. Increased utilization of xylose by

ethanologenic E. coli FBR5 and the xylose-utilizing yeast S.

cerevisiae YRH400 was also observed. Bioabatement was

associated with enhanced ethanol production in fermenta-

tions of dilute acid-pretreated rice hulls and may facilitate

utilization of this feedstock, which presently has little

commercial use.

Acknowledgements

The authors thank Patricia O’Bryan for excellent technical

assistance and Rice Hull Specialty Products (Stuttgart, AR) for

the gift of ground rice hulls.

r e f e r e n c e s

[1] Lopez Y, Gullon B, Puls J, Parajo JC, Martın C. Dilute acidpretreatment of starch-containing rice hulls for ethanolproduction. Holzforsch 2011;65:467e73.

[2] Saha BC, Iten LB, Cotta MA, Wu YV. Dilute acid pretreatment,enzymatic saccharification, and fermentation of rice hulls toethanol. Biotechnol Prog 2005;21:816e22.

[3] Juliano BO. Rice hulls and straw. In: Juliano BO, editor. Rice:chemistry and technology. 2nd ed. St. Paul, MN: AmericanAssociation of Cereal Chemists; 1985. pp. 685e743.

[4] Hickert LR, Souza-Cruz PBD, Rosa CA, Ayub MAZ.Simultaneous saccharification and co-fermentation of un-detoxified rice hull hydrolysate by Saccharomyces cerevisiaeICV D254 and Spathaspora arborariae NRRL Y-48658 for theproduction of ethanol and xylitol. Bioresour Technol2013;143:112e6.

[5] Binod P, Sindhu R, Singhania RR, Vikram S, Devi L,Nagalakshmi S, et al. Bioethanol production from rice straw:an overview. Bioresour Technol 2010;101:4767e74.

[6] Moniruzzaman M, Ingram LO. Ethanol production fromdilute acid hydrolysate of rice hulls using geneticallyengineered Escherichia coli. Biotechnol Lett 1998;20:943e7.

[7] Martin C, Alriksson B, Sjode A, Nilvebrant N-O, Jonsson LJ.Dilute sulfuric acid pretreatment of agricultural and agro-industrial residues for ethanol production. Appl BiochemBiotechnol 2007;136-140:339e52.

[8] Wei G-Y, Lee Y-J, Kim Y-J, Jin I-H, Lee J-H, Chung C-H, et al.Kinetic study on the pretreatment and enzymaticsaccharification of rice hull for the production of fermentablesugars. Appl Biochem Biotechnol 2010;162:1471e82.

[9] Palmqvist E, Hahn-Hagerdal B, Galbe M, Zacchi G. The effectof water-soluble inhibitors from steam-pretreated willow onenzymatic hydrolysis and ethanol fermentation. EnzymeMicrob Technol 1996;19:470e6.

[10] Palmqvist E, Hahn-Hagerdal B. Fermentation oflignocellulosic hydrolysates. I: inhibition and detoxification.Bioresour Technol 2000;74:17e24.

[11] Mussatto S, Roberto I. Alternatives for detoxification ofdiluted-acid lignocellulosic hydrolyzates for use infermentative processes: a review. Bioresour Technol2004;93:1e10.

[12] Jonsson LJ, Alriksson B, Nilvebrant N-O. Bioconversion oflignocellulose: inhibitors and detoxification. BiotechnolBiofuels 2013;6:16.

[13] Jonsson LJ, Palmqvist E, Nilvebrant N-O. Detoxification ofwood hydrolysates with laccase and peroxidase from thewhite-rot fungus Trametes versicolor. Appl Microbiol Technol1998;49:691e7.

[14] Dong H, Bao J. Metabolism: biofuel via biodetoxification. NatChem Biol 2010;6:316e8.

[15] Nichols NN, Dien BS, Cotta MA. Fermentation of bioenergycrops into ethanol using biological abatement for removal ofinhibitors. Bioresour Technol 2010;101:7545e50.

[16] Lopez MJ, Nichols NN, Dien BS, Moreno J, Bothast RJ. Isolationof microorganisms for biological detoxification oflignocellulosic hydrolysates. Appl Microbiol Biotechnol2004;64:125e31.

b i om a s s a n d b i o e n e r g y 6 7 ( 2 0 1 4 ) 7 9e8 888

[17] van Maris AJA, Abbott DA, Bellissimi E, van den Brink J,Kuyper M, Luttik MAH, et al. Alcoholic fermentation ofcarbon sources in biomass hydrolysates by Saccharomycescerevisiae: current status. Antonie Leeuwenhoek2006;90:391e418.

[18] Dien BS, Nichols NN, O’Bryan PJ, Bothast RJ. Development ofnew ethanologenic Escherichia coli strains for fermentation oflignocellulosic biomass. Appl Biochem Biotechnol2000;84:181e96.

[19] Turner PC, Yomano LP, Jarboe LR, York SW, Baggett CL,Moritz BE, et al. Optical mapping and sequencing of theEscherichia coli KO11 genome reveal extensive chromosomalrearrangements, and multiple tandem copies of theZymomonas mobilis pdc and adhB genes. J Ind MicrobiolBiotechnol 2012;39:629e39.

[20] Hahn-Hagerdal B, Karhumaa K, Fonseca C, Spencer-Martins I, Gorwa-Grauslund MF. Towards industrial pentose-fermenting yeast strains. Appl Microbiol Biotechnol2007;74:937e53.

[21] Hector RE, Dien BS, Cotta MA, Qureshi N. Engineeringindustrial Saccharomyces cerevisiae strains for xylosefermentation and comparison for switchgrass conversion. JInd Microbiol Biotechnol 2011;39:1597e604.

[22] Van Vleet JH, Jeffries TW. Yeast metabolic engineering forhemicellulosic ethanol production. Curr Opin Biotechnol2009;20:300e6.

[23] Sanchez G, Karhumaa K, Fonseca C, Sanchez Nogue V,Almeida JRM, Larsson CU, et al. Improved xylose andarabinose utilization by an industrial recombinantSaccharomyces cerevisiae strain using evolutionaryengineering. Biotechnol Biofuels 2010;3:13.

[24] Bera AK, Ho NWY, Ihan A, Sedlak M. A genetic overhaul ofSaccharomyces cerevisiae 424A (LNH-ST) to improve xylosefermentation. J Ind Microbiol Biotechnol 2011;38:617e26.

[25] Gerhardt P, Murray RGE, Costilow RN, Nester EW, Wood WA,Krieg NR, et al. Manual of methods for general bacteriology.Washington, DC: American Society for Microbiology; 1981.

[26] Dien BS, Hespell RB, Ingram LO, Bothast RJ. Conversion ofcorn milling fibrous co-products into ethanol by recombinantEscherichia coli strains KO11 and SL40. World J MicrobiolBiotechnol 1997;13:619e25.

[27] Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D,et al. Determination of structural carbohydrates and lignin inbiomass. Technical Report NREL/TO-510e42618, http://www.nrel.gov/biomass/analytical_procedures.html; 2012.

[28] Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D,et al. Determination of ash in biomass. Technical Report

NREL/TO-510e42622, http://www.nrel.gov/biomass/analytical_procedures.html; 2008.

[29] Kuhn A, van Zyl C, van Tonder A, Prior BA. Purification andpartial characterization of an aldo-keto reductase fromSaccharomyces cerevisiae. Appl Environ Microbiol1995;61:1580e5.

[30] Bellissimi E, van Dijken JP, Pronk JT, van Maris AJA. Effects ofacetic acid on the kinetics of xylose fermentation by anengineered, xylose-isomerase-based Saccharomyces cerevisiaestrain. FEMS Yeast Res 2009;9:358e64.

[31] Casey E, Sedlak M, Ho NWY, Mosier NS. Effect of acetic acidand pH on the cofermentation of glucose and xylose toethanol by a genetically engineered strain of Saccharomycescerevisiae. FEMS Yeast Res 2010;10:385e93.

[32] Helle S, Cameron D, Lam J, White B, Duff S. Effect ofinhibitory compounds found in biomass hydrolysates ongrowth and xylose fermentation by a genetically engineeredstrain of S. cerevisiae. Enzyme Microb Technol2003;33(7):786e92.

[33] Wright J, Bellissimi E, de Hulster E, Wagner A, Pronk JT, vanMaris AJA. Batch and continuous culture-based selectionstrategies for acetic acid tolerance in xylose-fermentingSaccharomyces cerevisiae. FEMS Yeast Res 2011;11:299e306.

[34] Parawira W, Tekere M. Biotechnological strategies toovercome inhibitors in lignocellulose hydrolysates forethanol production: review. Crit Rev Biotechnol2011;31:20e31.

[35] Zhang J, Zhu Z, Wang X, Wang N, Wang W, Bao J.Biodetoxification of toxins generated from lignocellulosepretreatment using a newly isolated fungus, Amorphothecaresinae ZN1, and the consequent ethanol fermentation.Biotechnol Biofuels 2010;3:26.

[36] Nichols NN, Szynkarek MP, Skory CD, Gorsich SW, Lopez MJ,Guisado GM, et al. Transformation and electrophoretickaryotyping of Coniochaeta ligniaria NRRL30616. Curr Genet2011;57:169e75.

[37] Nichols NN, Sharma LN, Mowery RA, Chambliss CK, vanWalsum GP, Dien BS, et al. Fungal metabolism offermentation inhibitors present in corn stover dilute acidhydrolysate. Enzyme Microb Technol 2008;42:624e30.

[38] Almeida JR, Modig T, Roder A, Liden G, Gorwa-Grauslund MF.Pichia stipitis xylose reductase helps detoxifyinglignocellulosic hydrolysate by reducing 5-hydroxymethyl-furfural (HMF). Biotechnol Biofuels 2008;1:12.

[39] Agrawal M, Chen RR. Discovery and characterization of axylose reductase from Zymomonas mobilis ZM4. BiotechnolLett 2011;33:2127e33.