Bioresource Technology 309 (2020) 123323

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Strategy for biological co-production of levulinic acid and T

polyhydroxyalkanoates by using mixed microbial cultures fed with synthetic
hemicellulose hydrolysate
⁎
F. Pinto-Ibietaa,b, M. Ceac,d, F. Cabrerae, M. Abantod, F.E. Felissiaf, M.C. Areaf, G. Ciudadc,g,
a
Doctorate of Engineering Sciences with Specialization in Bioprocess, Universidad de La Frontera, Av. Francisco Salazar #01145, Temuco, Chile
b
Departamento de Procesos Industriales, Facultad de Ingeniería, Universidad Católica de Temuco, Casilla 15-D, Temuco, Chile
c
Departamento de Ingeniería Química, Universidad de La Frontera, Casilla 54-D, Temuco, Chile
d
Scientific and Technological Bioresource Nucleus (BIOREN), Universidad de La Frontera, Temuco, Chile
e
Instituto de Ciencias Químicas Aplicadas, Universidad Autónoma de Chile, Temuco, Chile
f
IMAM, UNaM, CONICET, FCEQYN, Programa de Celulosa y Papel (PROCYP), Misiones, Argentina, Félix de Azara 1552, Posadas, Argentina
g
Instituto del Medio Ambiente (IMA), Universidad de La Frontera, Avenida Francisco Salazar #01145, Temuco, Chile

A R T I C LE I N FO A B S T R A C T

Keywords: Hemicellulose hydrolysates (HH), which could be an interesting carbon source to feed mixed microbial cultures
Levulinic acid biological production (MMC) able to accumulate high value-added compounds. This research focused on the evaluation of a culture
Polyhydroxyalkanoates and levulinic acid co- strategy to achieve the simultaneous biological production of Levulinic Acid (LA) and Polyhydroxyalcanoates
production (PHA) by MMC fed with a synthetic HH (SHH). The culture strategy involves the use of sequential batch reactors
Mixed microbial cultures
(SBR) to select microorganisms capable of producing LA and PHA. This work proved that the cultivation strategy
Hemicellulose hydrolysate
used allowed the biological production of LA, reaching 37%w/w when the SHH was composed of 85% pentoses.
In addition, the simultaneous biological production of LA and PHB was possible when the SHH was enriched
with acetate (45% pentoses − 50% acetate). Finally, this study showed that the composition of the SHH impacts
directly on the selected microorganism genus and the type and quantity of the value-added compounds obtained.

1. Introduction
One of the greatest challenges today is the transition from a fossil
fuel economy to one based on renewable resources (Jeong et al., 2018;
Zhang et al., 2015). In this sense, levulinic acid (LA) and poly-
hydroxyalkanoates (PHAs) are interesting compounds that may poten-
tially be obtained from lignocellulosic biomass through chemical and
biochemical synthesis (Dietrich et al., 2019; Tiong et al., 2019;
Rodriguez-Perez et al., 2018). LA is the building block for at least fif-
teen highly valued chemical compounds. It is considered the eightieth
major value-added compound in the USA due to its broad range of
applications such as fuel additives, flavors, fragrances and polymers
through chemical catalysis (Lappalainen and Dong, 2019). On the other
hand, PHAs are bioplastics that have physicochemical and mechanical
properties similar to petroleum-based plastics and they can also be used
in medical, agroindustrial, pharmaceutical and food applications
(Cassuriaga et al., 2018; Rodrigues et al., 2019). Applications of PHAs
as bioplastics include packaging, textiles and prostheses. However,
PHAs are not used significantly in these products, possibly due to their
high production cost. One way to reduce the cost is to look for alter-
native raw materials, such as hemicellulose extracts (Brodin et al.,
2017).
Wheat straw and sugarcane bagasse, among others, are non-wood
lignocellulosic biomass derived from the agricultural sector, which has
been cited as a source of value-added products such as xylitol, ethanol,
cellulose nanofibers and recently, LA and PHA. In every case, fractio-
nation of the biomass is required to obtain cellulose, hemicelluloses and
lignin (Stoffel et al., 2014). In non-woods, hemicelluloses are rich in
pentoses, and the best way to separate this fraction is through a hy-
drothermal treatment (Vallejos et al., 2017). Hot water HH are then
rich in xylose (Queirós et al., 2014; Vallejos et al., 2016), which can be
used to produce LA and PHA. LA production is commonly based on
thermochemical reactions, which use different acid catalysts to induce
the conversion of carbohydrates to LA (Jeong et al., 2018). Further-
more, the biological production of LA by direct fermentation of pen-
toses using engineered Saccharomyces Cerevisiae and Pichia stipis has
been reported (Zhangelinni, 2017). PHA production from pentoses as a
carbon source has been widely studied using pure cultures from

⁎
Corresponding author at: Departamento de Ingeniería Química, Universidad de La Frontera, Casilla 54-D, Temuco, Chile.
E-mail address: gustavo.ciudad@ufrontera.cl (G. Ciudad).

https://doi.org/10.1016/j.biortech.2020.123323
Received 5 February 2020; Received in revised form 2 April 2020; Accepted 3 April 2020
Available online 06 April 2020
0960-8524/ © 2020 Elsevier Ltd. All rights reserved.
F. Pinto-Ibieta, et al.
different strains such as Ralstonia eutropha and Chlorella fusca,
(Cassuriaga et al., 2018; Yu and Stahl, 2008). However, the main and
common problem of pure or recombinant cultures is the requirement of
sterile conditions and the use of specific carbon sources resulting in
high operating costs (Montiel-Jarillo et al., 2017).
Different alternatives based on the use of mixed microbial cultures
(MMC) have been explored for the replacement of pure cultures. MMC
work in non-axenic environments; therefore, no sterilization process is
needed, allowing the use of wastewaters as feedstock with the resulting
cost reduction (Fra-Vázquez et al., 2018; Queirós et al., 2014). MMC
can perform beneficial cooperation between strains, tending to be more
robust to changes in operating conditions and enhancing the degrad-
ability of complex substrates (Brenner et al., 2008; Rafieenia et al.,
2019).
To the best of our knowledge, LA production using MMC from
pentoses has never been reported. Nevertheless, pentoses could be
transformed into pyruvic acid via the pentose phosphate pathway (Lee
et al., 2019; Miscevic et al., 2019). Subsequently, novel studies have
reported that pyruvic acid could be transformed into LA (Lee et al.,
2019; Zanghellini, 2017). Unlike LA production, the use of MMC to
produce PHA from volatile fatty acid (VFA) has been widely reported
(Wang et al., 2017; Zhao et al., 2019), and there are also some refer-
ences about the direct production of PHA from pentoses using MMC
(Huang et al., 2016; Queirós et al., 2014). The hypothesis here is that
MMC fed with HH (mainly pentoses) could potentially be adapted for
the simultaneous biosynthesis (co-production) of high-added-value
compounds. Therefore, this study aimed to explore the potential use of
MMC for the simultaneous biological production of LA and PHA,
evaluating the impact of the composition of the synthetic HH (SHH) as
a carbon source and its effects on the selection of MMC type during a
SBR operation. The relevance of this research is based on the generation
of novel biological methods to produce LA from HH using MMC; the
possibility of simultaneously biosynthesizing value-added compounds
from renewable resources; and the assessment of different residual
biomasses rich in hemicelluloses.
2. Materials and methods
2.1. Strategy for selection of mixed microbial culture
The selection strategy to generate a mixed microbial culture en-
riched in microorganism producers of LA and PHA was performed in a
sequential batch reactor (SBRs) fed with a synthetic hemicelluloses
hydrolysates (SHH) prepared using commercial reagents. The SBRs
were inoculated with activated sludge obtained from the municipal
activated sludge wastewater treatment plant in Temuco, Chile. To
evaluate the SHH as a carbon source to produce value-added com-
pounds, one reactor was fed with a solution of SHH prepared according
to the composition reported for hydrolyzed hemicellulose from su-
garcane bagasse reported by Vallejos et al. (2017). To study the effect of
the substrate composition on the MMC selection and the production of
compounds, an additional SHH enriched with acetate was also assayed
(Table 1). Both SHH were supplemented with nutrients and minerals to
achieve optimal microorganism growth. The nutrient media consisted
of (mg L−1): peptone 132, beef extract 68, glucose 200, (NH4)2SO4 112,
KH2PO4 49, MgSO4 66, NH4Cl 170, K2HPO4 92, KH2PO4 45, MgSO4
600, EDTA 100, CaCl2 70 and 2 ml of trace element solution (Huang
et al., 2017). Thiourea (10 mg L−1) was also added to prevent ni-
trification. The trace solution was composed of (mg L−1): FeSO4·6H2O
1000, H3BO3 150, CoCl26H2O 150, MnCl24H2O 120, ZnSO47H2O 120,
Na2MoO42H2O 60, KI 30 and CuSO45H2O 30.
2.2. SBR implementation
Two SBR reactors with 2 L of useful working volume were im-
plemented using the following nomenclature and feeding conditions:
2
Bioresource Technology 309 (2020) 123323
Table 1
Synthetic hemicellulose hydrolyzed used as a carbon source.
Composition (%)
Component SHH-1 SHH-2
Celobiose 0.8 0.4
Glucose 2.3 2.3
Xylose 79.7 42
Arabinose 4.8 2.7
FormicAcid 0 0.4
AceticAcid 8.9 50
HMF 0.14 0.09
Furfural 5.6 2.87
SBR1 was fed with the SHH prepared according to the composition
reported by Vallejos et al., (2017) and SBR2 fed with the SHH enriched
with acetate. Both SBRs worked under the same following conditions.
They were aerated constantly with 4L min−1 using an air compressor,
mechanically stirred with a paddle connected to a 12 V motor at 60 rpm
and maintained at 25 °C using a thermo-regulated bath. A C/N/P ratio
of 45/2/1 (in mmol L−1) was used, since it has been reported that the
restriction of nitrogen and phosphorus in the culture could lead to the
accumulation of compounds such as PHA, xanthan gum and tria-
cylglycerols (Chong et al., 2019; Guerfali et al., 2018). Two cycles per
day were carried out in each SBRs until reaching the steady-state. One
cycle consisted of: a feed period (6 min) followed by an aerobic reaction
(704 min) and 10 min of withdrawal (Huang et al., 2017; Wang et al.,
2017). No settling phase was performed, and all excess of biomass was
withdrawn with the mixed liquor (Cabrera et al., 2019; Dionisi et al.,
2005). In this way, the biomass retention time was equal to the hy-
draulic retention time. Each cycle was fed with 0.25 L of SHH, and at
the end of each cycle 0.25 L of the liquid was withdrawn by digital
peristaltic pumps using a Compact DAQ system (cDAQ-9178 chassis,
National Instruments, Austin, TX, USA) and a routine programmed
using the LabView software (National Instruments). To monitor the
evolution of LA, PHA and biomass growth in both SBRs, samples were
taken at the end of the cycle, every 4 cycles. Once the stationary phase
was reached, three cycles in both reactors were analyzed. Samples were
taken every hour per cycle to monitor the evolution of LA, PHA and
substrate uptake (pentoses and acetate) as a function of the time inside
the cycle. Studies about batch production of these compounds are not
included in this manuscript because this paper is mainly focused in
exploring the potential use of hemicellulose hydrolysate as a carbon
source to the recovery of LA and PHA using as culture strategy MMC
selected in an SBR system, not in the optimization process to produce
LA and PHA.
2.3. Analytical methods
LA and PHA quantification and identification were performed ac-
cording to Lappalainen and Dong (2019) and Serafim et al. (2004). A
sample of 2 ml of culture medium was centrifuged and the obtained
pellet was lyophilized. The biomass was resuspended in 1 ml of acid-
ified methanol (20% H2SO4) with 0.65 mg ml−1 of benzoic acid as an
internal standard (Sigma Aldrich). To this mixture, 1 ml of chloroform
was added and the solution was kept in a thermoblock at 100 °C for
3.5 h. After cooling, 0.5 ml of water was used for extraction. The
chloroform phase was collected, and molecular sieves (0.3 nm) were
added for water removal. One ml of the obtained chloroform phase was
injected in-column in a GC/MS column (Clarus 600, PerkinElmer). The
column used to determine PHA concentration was a DB-FFAP
30 m × 0.25 mm × 0.25um (Agilent) while an ELITE 1701
30 m × 0.25 mm × 0.25um was used to determine LA concentration
(PerkinElmer). The calibration curve was obtained by injecting a series
of standards at different concentrations of LA and polyhydroxybutyrate
(PHB) (Sigma Aldrich), previously subjected to the procedure described
F. Pinto-Ibieta, et al.
Fig. 1. LA, PHB, VS and biomass evolution against operating time for SBR1 (A) and SBR2 (B).
for the reactor samples. The substrate consumption was measured by
determining the concentration of reducing sugars and acetate in filtered
samples (0.22 µm pore size PVDF membrane, Merck). To quantify the
total reducing sugars, the dinitrosalicylic acid (DNS) reagent method
was used (Prasertsung et al., 2017; Zhang et al., 2012). The DNS re-
agent was prepared by mixing dinitrosalicylic acid, potassium sodium
tartrate and NaOH in hot deionized water (50 °C). 0.5 ml of the DNS
reagent and 0.5 ml of filtered HH sample were mixed and heated for
5 min, then the mixture was cooled and 5 ml of distilled water was
added. The absorbance intensity of the final solution was measured at
540 nm using a spectrophotometer (Thermo Scientific – GENESYS 10 s).
The calibration curve was performed using glucose (Merck) as a stan-
dard reagent. Acetate was determined by gas-chromatography in a
flame ionization detector (Clarus 400, PerkinElmer), using a Nukol™
capillary column (Sigma-Aldrich, Darmstadt, Germany). Volatile solids
(VS) and soluble chemical oxygen demand (SCOD) were quantified
using a standard technique (AWWA, 2005)
2.4. DNA extraction, amplicon sequencing and reads processing
To quantify the abundance of the different bacteria at different
times, biomass samples were collected in both SBRs at the initial phase
(day 0) and in the stationary state (day 70), respectively, for the culture
selection. The biomass was centrifuged at 12,000g for 10 min. After
removing the supernatant, a 0.25 g pellet was collected for DNA
3
Bioresource Technology 309 (2020) 123323
extraction using the PowerSoil®DNA Isolation Kit (Mo Bio Laboratories)
according to the manufacturer's instructions. After quantification in a
NanoDrop spectrophotometer (Thermo Scientific), DNA samples were
prepared using barcoded library adaptors (400 nM) containing the
bacterial and fungi primers. PCR reactions run in triplicates were
pooled, and the amplicon libraries were purified using Agencourt®
AMpure XP bead protocol (Beckmann Coulter, USA). The concentration
was measured by Quant-iTTM HS DNA assay (Thermo Fisher Scientific,
USA) and quality was evaluated with a Tapestation 2200, using D1K
ScreenTapes (Agilent, USA). Afterward, sequencing libraries were
pooled in equimolar concentrations and sequenced on an Illumina
MiSeq platform (v.3) with 15% PhiX spike-in and at a read-length of
300 bp paired-end by the Nucleus of Scientific and Technological
Development in Bioresources of UFRO (BIOREN-UFRO). Sequenced
sample libraries were processed following the DADA2 bioinformatics
pipeline reported by Callahan et al. (2016) using version 1.3.3 of the
DADA2 R-package. DADA2 allows the inference of exact amplicon se-
quence variants (ASVs), providing several benefits over traditional OTU
clustering methods (Callahan et al., 2017). Illumina sequencing adap-
tors and PCR primers were trimmed before quality filtering. Sequences
were truncated after 280 and 250 nucleotides (16S rDNA) and 190 and
190 (ITS) for forward and reverse reads respectively, the length at
which the average quality dropped below a Phred score of 20. After
truncation, reads with expected error rates higher than 3 and 5 for
forward and reverse reads were removed. After filtering, error rate
F. Pinto-Ibieta, et al.
learning, ASV inference and denoising, reads were merged with a
minimum overlap of 20 bp. Chimeric sequences were identified and
removed. For a total of 12 samples, 77,703 out of 144,741 reads from
16S rDNA were kept on average per sample after processing, re-
presenting 53.7% of the average input reads. Similarly, for ITS analysis,
from the same 12 samples, 154,406 out of 180,384 reads were kept on
average per sample after processing, representing 85.6% of the average
input reads. Taxonomy based on 16S and ITS sequences was assigned
using the SILVA v.132 (Glöckner et al., 2017) and UNITE v.8.0 (Nilsson
et al., 2019) databases, respectively. Adequacy of sequencing depth
after reads processing was corroborated with rarefaction curves at the
ASV level.
3. Results and discussion
3.1. Enrichment reactor evolution over the operation time using formulated
SHH
The SBR1 was operated for 80 days without any interruptions,
finding two value-added intracellular compounds identified as levulinic
acid (LA) and polyhydroxybutyrate (PHB). The evolution of LA, PHB,
VS and biomass during the operation of the SBR1 is shown in Fig. 1.A.
An acclimatization stage during the first 40 days of operation followed
by a stationary stage was observed, reflected by the reach of a constant
concentration of LA and PHB in successive cycle-ends. At the same time,
a cell wash-out was observed since there was a gradual reduction in VS
and biomass content across the acclimatization stage until 2.4 g L−1 of
biomass was reached during the stationary stage. This behavior de-
monstrates that a selection of microorganisms became able to use SHH
as a substrate. This selection of microorganisms has been previously
observed in many studies (Guo et al., 2016; Lee et al., 2015; Oliveira
et al., 2016). For example, Huang et al. (2018) used different VFA to
feed an SBR, provoking a selection of PHB-producing microorganisms,
reaching the stationary stage after 40 days. The required time for
reaching the stationary stage has been reported in a range from 30 to
167 days, depending on the conditions, such as the cycle length, sub-
strate complexity and origin and composition of the microbial biomass,
among others (Albuquerque et al., 2007; Morgan-Sagastume et al.,
2015).
According to the operating conditions in SBR1, it was possible to
select microorganisms able to synthesize intracellularly LA and, to a
lesser extent, PHB. LA increased during the first forty days until
reaching a maximum of 32%w/w at the stationary state, while only a
maximum of 3%w/w of PHB was reached at the same stage. Although
both compounds are synthesized by the microbial biomass, a higher
concentration of LA (30% w/w) than that of PHB (3% w/w) was not
expected. These results are similar to those reported by Zanghellini
(2017), where pentoses were also used as a carbon source for LA
synthesis. On the other hand, it is possible that PHB was synthesized
only from the 9%w/w of the acetate in the SHH since MMC prefer VFA
over pentoses to synthesize PHA (Queirós et al., 2014). Consequently, it
is likely that under the operating conditions used in this study, the
selected microbial biomass preferred to metabolize pentoses to LA,
contrary to what has been reported by other studies, where pentoses are
converted to PHA by MMC (Huang et al., 2016; Queirós et al., 2014).
The evolution of LA, PHB, VS and biomass during the 80 days op-
erated in the SBR2 is shown in Fig. 1.B. An acclimatization stage during
the first 30 days is observed, followed by a stationary stage from day 15
onward. Less time was needed to reach a stationary stage in SBR2
(10 days), probably due to its higher concentration of acetate in the
substrate compared to SBR1, since the metabolic pathway to synthesize
PHB from acetate seems to have fewer steps than the two metabolic
pathways to synthesize LA from pentoses reported in the literature. The
reported possible pathways for intracellular LA production start with
the transformation of the pentoses to pyruvic acid via the pentose
phosphate pathway (Lee et al., 2019; Miscevic et al., 2019)then pyruvic
4
Bioresource Technology 309 (2020) 123323
acid is transformed into LA via two potential pathways: (1) the trans-
formation of pyruvate by redox reactions to 4-oxo-2 hydroxy-pentanoic
acid, which is dehydrated into 4-oxo-2 pentanoic acid that finally is
reduced to LA (Zanghellini, 2017), and (2) the oxidation of pyruvic
acid, producing acetyl-CoA which goes into the Krebs cycle and the
synthesis of LA from succinyl CoA (Lee et al., 2019).
Fig. 1.B. showed that during the first 30 days of operation, the
microorganisms present in SBR2 were able to synthesize both LA and
PHB, reaching a maximum of 14% and 17% w/w respectively, unlike
the SBR1 where the selection of microorganisms was able to synthesize
LA almost exclusively. It has been widely reported that simple carbon
chain molecules like acetic acid can lead to the accumulation of storage
polymers such as PHB, which could explain the greater amount of PHB
synthesized by microbial communities in the SBR2 reactor. The results
obtained showed that LA and PHB can be co-produced when the SHH is
fed with acetate as the co-substrate. The eventual benefits of this co-
production should be evaluated considering the productivity of each
compound, their final properties and the difficulties and costs of their
separation, among others. Some experiences about the recovery of
different value-added compounds coming from co-cultures have been
reported (Rodrigues et al., 2019; Wang et al., 2018).
The initial pentose and acetate concentrations (in the SBR inputs of
the cycle) are positively correlated with the final concentrations of PHB
and LA (at the outputs of the cycle), since LA and PHA production in-
creased proportionally to pentose and acetate concentrations, respec-
tively. SBR1 was fed with twice the amount of pentoses as SBR2 (86%
vs 46%) and produced twice the amount of LA (31% versus 14%).
Moreover, the amount of acetate was 5.5 times lower in SBR1 (9%
versus 50%) and produced 5.7 times less PHB (3% versus 18%).
Pentoses, acetate and soluble COD consumption in both reactors were
monitored at the beginning and the end of some cycles during the
stationary stage (Fig. 2). The pentose and acetate concentration inside
SBR1 at the beginning of each cycle were 41 and 4.3 Cmmol L−−1,
respectively (Fig. 2.A), whereas they were 19 and 25 Cmmol L−1, re-
spectively in SBR2 (Fig. 2.B). At the end of each cycle, neither pentoses
nor acetate were detected in either reactor. Furthermore, the soluble
COD concentrations at the beginning of the cycles remained constant
around 4 g L−1, proving that the added carbon source was not accu-
mulated by degrading all pentoses, VFA and some inhibitors like HMF
and furfural, contained in the feed of the two reactors.
3.2. Evolution of the production of value-added compounds, substrate
consumption and dissolved oxygen in the different SBR operation cycles.
The average value of the studied variables in the three cycles is
shown in Fig. 3. A feast/famine (F/F) dynamics can be identified from
the variations observed in DO in both reactors during one operating
cycle (Dionisi et al., 2005). According to Korkakaki et al. (2017), DO
concentrations may decrease below 2 mg L−1 during the feast phase,
while in the famine phase it increased above 8 mg L−1, indicating no
significant biological activity. It has been strongly demonstrated that F/
F dynamic adapts biomass to stressful conditions and drives the accu-
mulation of valuable compounds (Rodriguez-Perez et al., 2018). The F/
F dynamic was observed in both conditions which, regardless of the
carbon source imposed on the cultures, stimulated the storage of dif-
ferent compounds inside the microorganisms.
The end of the feast phase was reached at 4.8 h of operation in SBR1
(Fig. 3.A), consuming 87% of the SHH. This behavior suggests that the
SHH was used for cell growth and LA production simultaneously. The
intracellular LA increased from 30 to 37% in the feast phase, whereas it
decreased from 37 to 30% in the famine phase. It seems that during the
famine phase the MMC can consume the LA accumulated during the
feast phase, probably due to external carbon source limitation. In the
case of SBR2, 82% pentoses and 96% acetate were consumed during the
feast phase (Fig. 3.B). The simultaneous consumption of both substrates
is indicative of the presence of two groups of microorganisms
F. Pinto-Ibieta, et al.
Fig. 2. Pentose and acetate concentration at the beginning and end of the cycles as a function of the operating time during the stationary stage in SBR1 (A) and SBR2
(B).
synthesizing LA and PHB separately. LA reached 20% w/w at the end of
the feast phase, whereas PHB reached a maximum of 25% w/w at
180 min, prior to the end of the feast phase. The maximum amounts of
PHB and LA were reached when acetate and pentoses were depleted,
respectively. This behavior confirms the results showed in point 3.1,
i.e., the preference of different groups in the selected MMC to meta-
bolize pentoses to LA and acetate to PHB under the operating conditions
used in this study.
3.3. Structure of mixed microbial cultures related to the biological
production of value-added compounds.
The structure of the bacterial community identified for the in-
oculum is presented in Fig. 4. The bacterial microbial community was
conformed mainly by Rhodoferax, Flavobacterium and Arcobacter,
showing a relative abundance of 50%, 25% and 25%, respectively.
Moreover, the genera Paracoccus, Lactococcus and Enterococcus were
found to a lesser extent. As a result of the MMC selection in the sta-
tionary stage, the predominant communities were mainly Paracoccus,
Lactococcus and Enterococcus in the SBR1 (relative abundances of 26%,
28% and 15% respectively). On the other hand, the genus Azospirillum
was the main component in SBR2 with a relative abundance of 90%.
5
Bioresource Technology 309 (2020) 123323
Fungus communities are shown in Fig. 5. The predominant fungi
identified in the inoculum of both reactors were those of the genera
Trichosporon (30%), Cutaneotrichosporon (23%), Apiotrichum (14%) and
Candida (10%). Haglerozyma (87%) predominated in the stationary
stage in SBR1 and Gueomyces (80%) in SBR2. These results suggest that
a highly selective pressure was applied to the microorganisms in gen-
eral that favored the prevalence of the minor fraction of the bacteria
able to resist overstressing conditions.
This result is noteworthy considering that Paracoccus, Enterococuss
and lactococuss of the genus Azospirillum are organisms known to ac-
cumulate carbon as PHA (Huang et al., 2018; Kumar et al., 2018; Mota
et al., 2019; Müller-Santos et al., 2015; Sundaram et al., 2016). Azos-
pirillum is one of the best studied polymer-accumulating rhizospheric
bacteria, which base their survival strategy on PHB generation (Müller-
Santos et al., 2015). By contrast, the fungus haglerozyma belongs to the
family Trichosporoneceae, which has phylogenetic relationships with
apiotrichum and Zooglea, which have been previously associated with
the accumulation of different storage polymers such as PHAs (Queirós
et al., 2017). If the two reactors are compared, the overall micro-
organisms reveal a high variability in their respective community
structures. These changes were principally guided by the feed compo-
sition, indicating a strong feed dependence on the MMC selection
F. Pinto-Ibieta, et al.
Fig. 3. Accumulated compounds (LA and PHB), acetate, reducing sugars and dissolved oxygen as a function of time in one cycle of SBR1 (A) and SBR2 (B).
process. Oliveira et al. (2016) have reported that the selected MMC will
be different as a consequence of changes in the feed strategy. In their
work, when nitrogen and carbon substrates were added simultaneously
at the beginning of the cycle, the communities were dominated by
Corynebacterium. However, when the nitrogen addition was delayed to
the end of the feast phase, Xantomonadaceae were prevailed.
Finally, this study provides a first approach to the biological
synthesis of LA using MMC, which is highly desirable given the need to
generate clean alternatives to increase LA production. Moreover, LA
synthesis cannot be attributed to a specific microorganism, fungus,
bacterium or a synergic relationship between them under the evaluated
systems.
4. Conclusions
MMC in SBR, feed with HH and submitted to feast/famine regimen,
proved to be a good strategy for biological LA and PHB production. The
composition of the hydrolysate impacted directly on the type of mi-
croorganism selected and the type and quantity of the final product.
When the HH was enriched with acetate, LA and PHB were produced
simultaneously. Although the main selected LA-producing micro-
organisms are reported, possible synergies between them are unknown,
so further studies are required to understand the possible synergies,
6
Bioresource Technology 309 (2020) 123323
metabolic pathways and the way in which LA is stored by the cell.
CRediT authorship contribution statement
F. Pinto-Ibieta: Writing - original draft, Conceptualization,
Investigation. M. Cea: Writing - review & editing, Resources,
Supervision, Conceptualization. F. Cabrera: Writing - original draft,
Conceptualization. M. Abanto: Software, Formal analysis. F.E.
Felissia: Writing - review & editing. M.C. Area: Writing - review &
editing. G. Ciudad: Writing - review & editing, Funding acquisition,
Supervision, Conceptualization.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
The authors wish to thank the financial support provided by the
ELAC2015/T03-0715-ValBio-3D. The authors would also like to thank
F. Pinto-Ibieta, et al. Bioresource Technology 309 (2020) 123323

Fig. 4. Profile of bacterial consortium in the original sample (day 0) and in the selected MMC (steady-state) for bacteria.

Fig. 5. Profile of fungus consortium in the original sample (day 0) and in the selected MMC (steady-state) for fungus.

7
F. Pinto-Ibieta, et al. Bioresource Technology 309 (2020) 123323

for providing financial support to the CONICYT PFCHA/DOCTORADO coli.

BECAS CHILE/2018 - 21180926 Montiel-Jarillo, G., Carrera, J., Suárez-Ojeda, M.E., 2017. Enrichment of a mixed mi-
crobial culture for polyhydroxyalkanoates production: Effect of pH and N and P
concentrations. Sci. Total Environ. 583, 300–307. https://doi.org/10.1016/j.
References scitotenv.2017.01.069.
Morgan-Sagastume, F., Hjort, M., Cirne, D., Gérardin, F., Lacroix, S., Gaval, G.,
Karabegovic, L., Alexandersson, T., Johansson, P., Karlsson, A., Bengtsson, S., Arcos-
Albuquerque, M.G.E., Eiroa, M., Torres, C., Nunes, B.R., Reis, M.A.M., 2007. Strategies for
Hernández, M.V., Magnusson, P., Werker, A., 2015. Integrated production of poly-
the development of a side stream process for polyhydroxyalkanoate (PHA) produc-
hydroxyalkanoates (PHAs) with municipal wastewater and sludge treatment at pilot
tion from sugar cane molasses. J. Biotechnol. 130, 411–421. https://doi.org/10.
scale. Bioresour. Technol. 181, 78–89. https://doi.org/10.1016/j.biortech.2015.01.
1016/j.jbiotec.2007.05.011.
046.
AWWA, W.A., 2005. Standard Methods for the Examination of Water and Wastewater.
Mota, M.J., Lopes, R.P., Simões, M.M.Q., Delgadillo, I., Saraiva, J.A., 2019. Effect of high
20th ed., Washington DC, USA.
pressure on paracoccus denitrificans growth and polyhydroxyalkanoates production
Brenner, K., You, L., Arnold, F.H., 2008. Engineering microbial consortia: a new frontier
from glycerol. Appl. Biochem. Biotechnol. 188, 810–823. https://doi.org/10.1007/
in synthetic biology. Trends Biotechnol. 26, 483–489. https://doi.org/10.1016/j.
s12010-018-02949-0.
tibtech.2008.05.004.
Müller-Santos, M., Maltempi de Souza, E., De Oliveira Pedrosa, F., Chubatsu, L.S., 2015.
Brodin, M., Vallejos, M., Opedal, M.T., Area, M.C., Chinga-Carrasco, G., 2017.
Polyhydroxybutyrate in Azospirillum brasilense. In: Handbook for Azospirillum:
Lignocellulosics as sustainable resources for production of bioplastics – A review. J.
Technical Issues and Protocols. Springer International Publishing, pp. 241–250.
Clean. Prod. 162, 646–664. https://doi.org/10.1016/j.jclepro.2017.05.209.
https://doi.org/10.1007/978-3-319-06542-7_13.
Cabrera, F., Torres, Á., Campos, J.L., Jeison, D., 2019. Effect of Operational Conditions on
Nilsson, R.H., Larsson, K.H., Taylor, A.F.S., Bengtsson-Palme, J., Jeppesen, T.S., Schigel,
the Behaviour and Associated Costs of Mixed Microbial Cultures for PHA Production.
D., Kennedy, P., Picard, K., Glöckner, F.O., Tedersoo, L., Saar, I., Kõljalg, U.,
Polymers (Basel). 11, 191. https://doi.org/10.3390/polym11020191.
Abarenkov, K., 2019. The UNITE database for molecular identification of fungi:
Callahan, B.J., McMurdie, P.J., Holmes, S.P., 2017. Exact sequence variants should re-
handling dark taxa and parallel taxonomic classifications. Nucleic Acids Res. 47,
place operational taxonomic units in marker-gene data analysis. ISME J. 11,
D259–D264. https://doi.org/10.1093/nar/gky1022.
2639–2643. https://doi.org/10.1038/ismej.2017.119.
Oliveira, C.S.S., Silva, C.E., Carvalho, G., Reis, M.A., 2016. Strategies for efficiently se-
Cassuriaga, A.P.A., Freitas, B.C.B., Morais, M.G., Costa, J.A.V., 2018. Innovative poly-
lecting PHA producing mixed microbial cultures using complex feedstocks: Feast and
hydroxybutyrate production by Chlorella fusca grown with pentoses. Bioresour.
famine regime and uncoupled carbon and nitrogen availabilities. N. Biotechnol. 37,
Technol. 265, 456–463. https://doi.org/10.1016/j.biortech.2018.06.026.
69–79. https://doi.org/10.1016/j.nbt.2016.10.008.
Chong, J.F., Fadhullah, W., Lim, V., Lee, C.K., 2019. Two-stage cultivation of the marine
Prasertsung, I., Chutinate, P., Watthanaphanit, A., Saito, N., Damrongsakkul, S., 2017.
microalga Chlorella salina for starch and carbohydrate production.
Conversion of cellulose into reducing sugar by solution plasma process (SPP).
Dietrich, K., Dumont, M.J., Del Rio, L.F., Orsat, V., 2019. Sustainable PHA production in
Carbohydr. Polym. 172, 230–236. https://doi.org/10.1016/j.carbpol.2017.05.025.
integrated lignocellulose biorefineries. N. Biotechnol. 49, 161–168. https://doi.org/
Queirós, D., Fonseca, A., Rossetti, S., Serafim, L.S., Lemos, P.C., 2017. Highly complex
10.1016/j.nbt.2018.11.004.
substrates lead to dynamic bacterial community for polyhydroxyalkanoates produc-
Dionisi, D., Majone, M., Vallini, G., Gregorio, S. Di, Beccari, M., 2005. Effect of the ap-
tion. J. Ind. Microbiol. Biotechnol. 44, 1215–1224. https://doi.org/10.1007/s10295-
plied organic load rate on biodegradable polymer production by mixed microbial
017-1951-y.
cultures in a sequencing batch reactor. Biotechnol. Bioeng. https://doi.org/10.1002/
Queirós, D., Rossetti, S., Serafim, L.S., 2014. PHA production by mixed cultures: a way to
bit.20683.
valorize wastes from pulp industry. Bioresour. Technol. 157, 197–205. https://doi.
Fra-Vázquez, A., Pedrouso, A., Palmeiro-Sánchez, T., Moralejo-Gárate, H., Mosquera-
org/10.1016/j.biortech.2014.01.099.
Corral, A., 2018. Feasible microbial accumulation of triacylglycerides from crude
Rafieenia, R., Pivato, A., Lavagnolo, M.C., 2019. Optimization of hydrogen production
glycerol. J. Chem. Technol. Biotechnol. 93, 2644–2651. https://doi.org/10.1002/
from food waste using anaerobic mixed cultures pretreated with waste frying oil.
jctb.5618.
Renew. Energy 1077–1085. https://doi.org/10.1016/j.renene.2019.03.012.
Glöckner, F.O., Yilmaz, P., Quast, C., Gerken, J., Beccati, A., Ciuprina, A., Bruns, G.,
Rodrigues, P.R., Assis, D.J., Druzian, J.I., 2019. Simultaneous production of poly-
Yarza, P., Peplies, J., Westram, R., Ludwig, W., 2017. 25 years of serving the com-
hydroxyalkanoate and xanthan gum: From axenic to mixed cultivation. Bioresour.
munity with ribosomal RNA gene reference databases and tools. J. Biotechnol.
Technol. 283, 332–339. https://doi.org/10.1016/j.biortech.2019.03.095.
https://doi.org/10.1016/j.jbiotec.2017.06.1198.
Rodriguez-Perez, S., Serrano, A., Pantión, A.A., Alonso-Fariñas, B., 2018. Challenges of
Guerfali, M., Ayadi, I., Mohamed, N., Ayadi, W., 2018. Triacylglycerols accumulation and
scaling-up PHA production from waste streams. A review. J. Environ. Manage. 205,
glycolipids secretion by the oleaginous yeast Rhodotorula babjevae Y-SL7: Structural
215–230. https://doi.org/10.1016/j.jenvman.2017.09.083.
identification and biotechnological applications. Bioresour. Technol. Elsevier Ltd.
Serafim, S., Lemos, P.C., Oliveira, R., Reis, M.A.M., 2004. Optimization of poly-
https://doi.org/10.1016/j.biortech.2018.11.036.
hydroxybutyrate production by mixed cultures submitted to aerobic dynamic feeding
Guo, Z., Chen, Z., Wen, Q., Huang, L., Bakke, R., Du, M., 2016. Strategy to reduce the
conditions. Biotechnol. Bioeng. https://doi.org/10.1002/bit.20085.
acclimation period for enrichment of PHA accumulating cultures. Desalin. Water
Stoffel, R.B., Felissia, F.E., Silva Curvelo, A.A., Gassa, L.M., Area, M.C., 2014.
Treat. 57, 29286–29294. https://doi.org/10.1080/19443994.2016.1155177.
Optimization of sequential alkaline-acid fractionation of pine sawdust for a bior-
Huang, L., Chen, Z., Wen, Q., Lee, D., 2017. Enhanced polyhydroxyalkanoate production
efinery. Ind. Crops Prod. 61, 160–168. https://doi.org/10.1016/j.indcrop.2014.06.
by mixed microbial culture with extended cultivation strategy. Bioresour. Technol.
047.
241, 802–811. https://doi.org/10.1016/j.biortech.2017.05.192.
Sundaram, M., Radhakrishnan, L., Chithiraivelu, S., Athiyappa Gounder, P., 2016.
Huang, L., Chen, Z., Wen, Q., Zhao, L., Lee, D.J., Yang, L., Wang, Y., 2018. Insights into
Enhanced production of polyhydroxybutyrate using enterococcus faecium KT722772
Feast-Famine polyhydroxyalkanoate (PHA)-producer selection: microbial community
isolated from cattle rumen fluid by substrates optimization. Polym. - Plast. Technol.
succession, relationships with system function and underlying driving forces. Water
Eng. 55, 1908–1915. https://doi.org/10.1080/03602559.2016.1185616.
Res. 131, 167–176. https://doi.org/10.1016/j.watres.2017.12.033.
Tiong, Y.W., Yap, C.L., Gan, S., Yap, W.S.P., 2019. Optimisation studies on the conversion
Huang, L., Liu, C., Liu, Y., Jia, X., 2016. The composition analysis and preliminary cul-
of oil palm biomass to levulinic acid and ethyl levulinate via indium trichloride-ionic
tivation optimization of a PHA-producing microbial consortium with xylose as a sole
liquids: A response surface methodology approach. Ind. Crops Prod. 128, 221–234.
carbon source. Waste Manag. 52, 77–85. https://doi.org/10.1016/j.wasman.2016.
https://doi.org/10.1016/j.indcrop.2018.10.087.
03.020.
Vallejos, M.E., Chade, M., Mereles, E.B., Bengoechea, D.I., Brizuela, J.G., Felissia, F.E.,
Jeong, H., Park, S.Y., Ryu, G.H., Choi, J.H., Kim, J.H., Choi, W.S., Lee, S.M., Choi, J.W.,
Area, M.C., 2016. Strategies of detoxification and fermentation for biotechnological
Choi, I.G., 2018. Catalytic conversion of hemicellulosic sugars derived from biomass
production of xylitol from sugarcane bagasse. Ind. Crops Prod. 91, 161–169. https://
to levulinic acid. Catal. Commun. 117, 19–25. https://doi.org/10.1016/j.catcom.
doi.org/10.1016/j.indcrop.2016.07.007.
2018.04.016.
Vallejos, M.E., Felissia, F.E., Area, M.C., 2017. Hydrothermal treatments applied to agro-
Korkakaki, E., van Loosdrecht, M.C.M., Kleerebezem, R., 2017. Impact of phosphate
and forest- industrial waste to produce high added-value compounds. BioResources
limitation on PHA production in a feast-famine process. Water Res. 126, 472–480.
12, 2058–2080. https://doi.org/10.15376/biores.12.1.Vallejos.
https://doi.org/10.1016/j.watres.2017.09.031.
Wang, F., Gao, B., Huang, L., Su, M., Zhang, C., et al., 2018. Evaluation of oleaginous
Kumar, P., Jun, H.B., Kim, B.S., 2018. Co-production of polyhydroxyalkanoates and
eustigmatophycean microalgae as potential biorefinery feedstock for the production
carotenoids through bioconversion of glycerol by Paracoccus sp. strain LL1. Int. J.
of palmitoleic acid and biodiesel. Bioresour. Technol. 270, 30–37. https://doi.org/10.
Biol. Macromol. 107, 2552–2558. https://doi.org/10.1016/j.ijbiomac.2017.10.147.
1016/j.biortech.2018.09.016.
Lappalainen, K., Dong, Y., 2019. Simultaneous production of furfural and levulinic acid
Wang, X., Oehmen, A., Freitas, E.B., Carvalho, G., Reis, M.A.M., 2017. The link of feast-
from pine sawdust via acid-catalysed mechanical depolymerization and microwave
phase dissolved oxygen (DO) with substrate competition and microbial selection in
irradiation. Biomass Bioenergy 123, 159–165. https://doi.org/10.1016/j.biombioe.
PHA production. Water Res. 112, 269–278. https://doi.org/10.1016/j.watres.2017.
2019.02.017.
01.064.
Lee, S.Y., Kim, H.U., Chae, T.U., Cho, J.S., Kim, J.W., Shin, J.H., Kim, D.I., Ko, Y., Jang,
Yu, J., Stahl, H., 2008. Microbial utilization and biopolyester synthesis of bagasse hy-
W.D., Jang, Y., 2019. A comprehensive metabolic map for production of bio-based
drolysates. Bioresour. Technol. 99, 8042–8048. https://doi.org/10.1016/j.biortech.
chemicals. Nat. Catal. 2. https://doi.org/10.1038/s41929-018-0212-4.
2008.03.071.
Lee, W.S., Chua, A.S.M., Yeoh, H.K., Nittami, T., Ngoh, G.C., 2015. Strategy for the
Zanghellini, A., 2017. Patente: Fermentation route for the production of levulinic acid,
biotransformation of fermented palm oil mill effluent into biodegradable poly-
levulinate esters and valerolactone and derivates thereof.
hydroxyalkanoates by activated sludge. Chem. Eng. J. 269, 288–297. https://doi.org/
Zhang, G.C., Liu, J.J., Kong, I.I., Kwak, S., Jin, Y.S., 2015. Combining C6 and C5 sugar
10.1016/j.cej.2015.01.103.
metabolism for enhancing microbial bioconversion. Curr. Opin. Chem. Biol. 29,
Miscevic, D., Srirangan, K., Kefale, T., Abedi, D., Moo-young, M., Chou, C.P., 2019.
49–57. https://doi.org/10.1016/j.cbpa.2015.09.008.
Production of cellulosic butyrate and 3-hydroxybutyrate in engineered Escherichia
Zhang, Z., Liu, B., Kent, Z., 2012. Ef fi cient acid-catalyzed hydrolysis of cellulose in

8
F. Pinto-Ibieta, et al.
organic electrolyte solutions. Polym. Degrad. Stab. 97, 573–577. https://doi.org/10.
1016/j.polymdegradstab.2012.01.010.
Zhangelinni, A., 2017. Fermentation route for the production of levulinic acid, levulinate
esters and valerolantone and derivatives thereof. US 2017/0275657 A1.
Bioresource Technology 309 (2020) 123323
Zhao, L., Han, D., Yin, Z., Bao, M., Lu, J., 2019. Biohydrogen and polyhydroxyalkanoate
production from original hydrolyzed polyacrylamide-containing wastewater. Bioresour.
Technol. 287, 121404. https://doi.org/10.1016/j.biortech.2019.121404.