Bioflocculant production from untreated corn stover using Cellulosimicrobium cellulans L804 isolate and its application to harvesting microalgae

Isolation and identification of bioflocculant-producing strains

Twelve cellulase-producing and xylanase-producing strains were isolated from corn
farmland soil samples. Among them, one strain, named L804, was identified as a bioflocculant-producing
strain with high flocculating activity. Strain L804 was Gram positive, rod shaped,
and aerobic strain. The colony of L804 was circular and moist. The 16S rRNA of strain
L804 was sequenced after PCR amplification and deposited into GenBank database (accession
number: KT280277). A total of 1400 bp of 16S rRNA was determined and compared with
the sequences of the GenBank database. The highest level of 16S rRNA sequence similarity
to Cellulosimicrobium cellulans was 99 %. Therefore, strain L804 and its bioflocculant product were named C. cellulans L804 and MBF-L804, respectively. In previous studies, C. cellulans has been reported to produce an array of plant cell-wall-degrading enzymes, such
as endo-?-1,3-glucanases, proteases, and mannanases 49]. C. cellulans has also been reported to secrete the lignocellulolytic enzymes, such as cellulase
and xylanase, and thus has the capacity for degrading lignocellulosic biomass 44], 50], 51]. In this study, C. cellulans L804 was also found to be able to secrete cellulase and xylanase (Fig. 1). More importantly, it was observed for the first time that C. cellulans produces bioflocculant. Therefore, strain L804 has potential to produce bioflocculant
by degrading the lignocellulosic biomass directly.

Fig. 1. Cellulase (a) and xylanase (b) of C. cellulans L804 analyzed using the medium containing CMC or Xylan. Images were taken at the
culture time of 72 h. The bar indicates 1 cm

In previous studies, the hydrolyzates of some lignocellulosic biomasses, such as rice
stover 44] and corn stover 29], were used as carbon source to produce the bioflocculants. However, these biomasses
were hydrolyzed under high-temperature and strong acidic conditions (121 °C and 1.7 %
sulfuric acid) 29], 44]. After acidic hydrolysis, a neutralization of pH is
necessary for the downstream fermentation processes 29], 44], 45], and the acidic hydrolysis process always produces toxic by-products, such as phenolic
compounds, furan derivatives, and carboxylic acids, which are major fermentation inhibitors
for the activities of bioflocculant-producing strain 52], and are difficult to remove in the extraction process, and thus contaminate the
bioflocculant product. In this study, C. cellulans L804 was found to be able to utilize untreated lignocellulosic biomasses as carbon
sources to produce bioflocculant, which can achieve the resourceful utilization of
lignocellulosic biomass.

MBF-L804 production in the medium with different initial pH

The initial pH of the medium was adjusted using Na
₂
CO
₃
and HCl solution. The effects of pH variation in the range of 6.0-10.5 on the cell
growth and MBF-L804 production were analyzed. As shown in Table 1, C. cellulans L804 produced bioflocculant MBF-L804 in the initial pH range of 7.4–9.8, and the
final pH range was 8.5–9.2. Therefore, C. cellulans L804 is an alkali-resistant strain. Previous studies reported that weak alkaline
treatment opens up the cell wall of lignocellulosic materials, leading to an increase
in internal surface area, a decrease in the degree of polymerization, and a decrease
in crystallinity; and the separation of structural linkages between lignin and carbohydrates
45]. Therefore, weak alkaline culture condition can improve the degradation efficiency
of lignocellulosic biomass, and the highest flocculating activity of 90.88 % was achieved
at initial pH 8.2 (Na
₂
CO
₃
concentration was 0.4 g/L). Therefore, 0.4 g/L Na
₂
CO
₃
was selected to adjust initial pH of the fermentation medium in the following experiments.

Table 1. Effects of initial pH on cell growth and flocculating activity

Effects of nitrogen sources on MBF-L804 production

The effects of nitrogen sources on MBF-L804 production were investigated. Figure 2a shows the effects of different nitrogen sources on MBF-L804 production. Among the
nitrogen sources investigated, ammonium sulfate, sodium nitrate, and urea resulted
in poor cell growth and flocculating activity. Comparatively, yeast extract, casein,
trypepton, beef extract, and peptone were significantly better sources for MBF-L804
production. Yeast extract was selected as the optimal nitrogen source in the following
experiments because it was favorable for the MBF-L804 production.

Fig. 2. Effects of nitrogen sources (a) and carbon sources (b) on flocculating activity and cell growth. The control sample indicated that strain
L804 cultured in FSS medium. 100 ?L of fermentation broth of 48 h was used for flocculating
activity assay

Effects of carbon sources on MBF-L804 production

The effects of various carbon sources on MBF-L804 production were then studied when
yeast extract was used as nitrogen source. As shown in Fig. 2b, flocculating activity over 80 % was achieved when maltose, galactose, sodium acetate,
carboxymethyl cellulose (CMC), and microcrystalline cellulose (CM) were used as carbon
sources. More exciting it was to find that strain L804 is able to produce bioflocculant
using CMC and CM as carbon sources, confirming that strain L804 has the potential
in conversion of lignocellulosic materials into bioflocculants.

Selection of lignocellulosic biomass to produce bioflocculant

Seven kinds of lignocellulosic biomasses, including corn stover, corn cob, rice hull,
potato residues, wheat bran, wheat straw, and peanut shell, were used as carbon sources
of the fermentation medium. The results showed that over 80 % flocculating activities
were achieved when the corn stover, corn cob, potato residues, and peanut shell were
used as carbon sources, which were much better than for rice hull, wheat bran, and
wheat straw (data not shown). Then, the flocculating efficiencies of the broth-containing
corn stover, corn cob, potato residues, and peanut shell were compared at different
time intervals. As shown in Fig. 3, the flocculating activities of broth with four kinds of lignocellulosic biomasses
were much higher than those of the control without added biomasses, and the highest
flocculating activity was achieved when corn stover was used as carbon source. Flocculating
efficiency over 90 % was observed after the culture had been grown for 48 h. Therefore,
the corn stover was selected as optimal carbon source, and the highest titer of MBF-L804
achieved at the culture time of 48 h was 4.75 g/L, which is much higher than 0.13 g/L
of the control broth without added biomasses. After 48 h of culture, the titer decreased
gradually with the increase of culture time. Therefore, the culture time of 48 h was
selected to extract MBF-L804 in the following experiments. In addition, 1.2 g/L soluble
sugar was detected in the medium containing 20 g/L corn stover before inoculating
with strain L804. These soluble sugars can promote the initial growth of L804 cells
and the secretion of lignocellulolytic enzymes, which further degrade the lignocellulosic
biomasses. Soluble sugar of 1.2 g/L was much lower than the titers of bioflocculant
MBF-L804 4.75 g/L, suggested that bioflocculant MBF-L804 was mainly produced from
the degradation of lignocellulosic biomasses. The bioflocculant titer of 4.75 g/L
obtained in this study is much higher than the reported 2.4 g/L bioflocculant secreted
by Rhodococcus erythropolis in the medium using the hydrolyzates of rice stover as carbon source 44], but is lower than 6 g/L bioflocculant produced by Ochrobactrum cicero W2 using the hydrolyzates of corn stover 53]. Although O. cicero W2 obtained a higher titer, the acidic hydrolysis of corn stover was under high-temperature
and strong acidic conditions (121 °C and 1.7 % sulfuric acid), and the acidic hydrolyzates
required the neutralization of pH before the downstream fermentation processes, and
the toxic by-products produced during the acidic hydrolysis process contaminated the
bioflocculant product.

Fig. 3. Production of MBF-L804 using different lignocellulosic biomasses as carbon source.
Strain L804 was cultured in the mediums (added with 3 g/L yeast extract as nitrogen
source and with different biomasses as carbon sources) and the control medium (added
with 3 g/L yeast extract, but without added biomasses). 100 ?L of fermentation broth
of 48 h was taken for flocculating activity assay

Time curves of pH, flocculating activity, cellulase, and xylanase

The effects of pH on the activities of xylanase and cellulase present in the supernatant
were determined. Figure 4a showed that the optimal pH for these two hydrolytic enzymes was around 5.2–6.0,
which was similar to a previous report 50]. Higher pH inhibited their activity. As mentioned above, the initial pH 8.2 of fermentation
medium was selected because the bioflocculant was not produced under acidic conditions.
Although weak alkaline condition was not the optimal pH for xylanase and cellulase,
more than 70 % xylanase activity and 50 % cellulase activity relative to their highest
activities remained at the alkaline pH.

Fig. 4. Effects of pH on the activities of cellulase and xylanase produced by C. cellulans L804 (a) and variation curves of pH, flocculating activity, cellulase, and xylanase during
cell growth in fermentation medium with corn stover as carbon source (b). Error bars indicate standard deviation of at least three replicates

The time profiles for pH, flocculating activity of fermentation broth, cellulase,
and xylanase were analyzed (Fig. 4b). The pH value decreased sharply from 8.2 to 7.4 in the first 12 h. This decrease
may be caused by the release of organic acids during the cell growth 50]. After 12-h culture, the pH increases may be due to the utilization of organic acids
as carbon source. A similar change trend of pH was observed in a previous study 50]. Time profiles of cellulase and xylanase activities showed that enzymatic activities
were not detectable at the beginning of the fermentation, suggesting that the enzyme
content in the lignocellulosic biomass was negligible and that the enzymes were produced
only by the microorganism. Figure 4b shows sharp increases in the activities of two enzymes in the early stage of incubation:
Xylanase was higher than cellulase. Over 0.6 IU/mL of xylanase activity was achieved
after 24-h culture. The highest cellulase activity of 0.046 IU/mL was achieved at
36 h, which then decreased with time. No flocculating activity in fermentation broth
was observed at the beginning, and the trends observed for enzyme activities were
similar to that of the flocculating activity, suggesting that the conversion from
corn stover into bioflocculant was dependent on the activity of hydrolytic enzymes
produced by C. cellulans L804.

Characterization of the bioflocculant MBF-L804

The components of MBF-L804 were determined. The results showed that the MBF-L804 contained
68.6 % polysaccharides and 28.0 % proteins. Polysaccharides were the major components
of MBF-L804. Gel permeation chromatography analysis indicated that the approximate
molecular weight (MW) of the purified bioflocculant L804 was 229 kDa. To reveal the
functional groups involved in the flocculating activity of MBF-L804, the FTIR spectrum
of the MBF-L804 was analyzed (Additional file 1; Figure S1). The results showed that the MBF-L804 displayed a broad peak at around
3300 cm
^?1
, indicating the presence of hydroxyl groups, and the spectrum also displayed a stretching
band at 1680 cm
^?1
and a weak symmetric stretching band near 1420 cm
^?1
, which are indicative of carboxyl groups. The absorption around 1080 cm
^?1
is known to be a characteristic for all sugar derivatives. The FTIR spectrum was consistent
with the results of most bioflocculants produced by other organisms 28], 32], 54], 55].

Flocculating properties of the bioflocculant MBF-L804

The effects of temperature, metal ion, bioflocculant dosage, and pH on the flocculating
activity were evaluated when the Kaolin clay was used as solid phase. The temperature
is an important factor influencing the flocculating activity 23]. As shown in Fig. 5a, MBF-L804 showed good heat stability. Over 85 % flocculating activity was achieved
at all the tested temperatures, and the highest flocculating activity of 92.68 % was
achieved at 35 °C. This could be due to the main components of MBF-L804 being polysaccharides
which are more heat stable compared with proteins or nucleic acids 22].

Fig. 5. Effects of temperature, metal ion, dosage, and pH on the flocculating activity of
MBF-L804

The effects of various metal ions on the flocculating acitivity of MBF-L804 were also
investigated (Fig. 5b). It was found that the additions of Ca
²⁺
and Mg
²⁺
evidently enhanced the flocculating efficiency of MBF-L804. Metal ions are important
in the process of flocculation 23]. The bioflocculant generated by Enterobacter aerogenes required the presence of Zn
²⁺56]. The flocculating activity of bioflocculant secreted by Nannocystis sp. Nu-2 depended strongly on cations 57]. Ca
²⁺
ion can enhance the flocculating efficiency of the bioflocculant produced by Bacillus agaradhaerens C9 28]. Cations stimulate the flocculating activity by neutralizing and stabilizing the
residual negative charge of functional groups and by forming the bridges between particles
54]. In this study, it was also found that MBF-L804 shows a good flocculating efficiency
of 91.67 % without adding any ion (Blank sample), and Ca
²⁺
and Mg
²⁺
can further improve the flocculating activity of MBF-L804.

As shown in Fig. 5c, the flocculating activity over 90 % was achieved in the concentration range of
4.0–14.0 mg/L. The solution with lower or higher MBF-L804 concentration showed poor
flocculating activity. The bridging phenomenon between particles formed insufficiently
when the MBF-L804 dosage was lower than 4 mg/L. When the concentration was higher
than 14.0 mg/L, the decrease of flocculating activity could be explained by the repulsion
between particles with the same negative charge due to the excessive introduction
of charged polysaccharides. The similar relationship between bioflocculant concentration
and flocculating activity was observed in other reported extracellular bioflocculants
28], 58]. As shown in Fig. 5d, the flocculating activity of Kaolin suspension was over 90 % in a wide pH range
from 3 to 12. The very wide pH and temperature ranges adaptable by MBF-L804 indicated
the potential of using it in field applications.

Application of MBF-L804 in harvesting two microalgae

Chlamydomonas reinhardtii and Chlorella minutissima are well known microalgae used for biodiesel production research 59], 60]. Compared with traditional methods, such as centrifugation, filtration, and gravity
sedimentation, flocculant is a low-cost option to harvest microalgae 9], 18], 61]. Therefore, the feasibility of flocculating C. reinhardtii and C. minutissima cells using MBF-L804 was investigated in this study. As shown in Fig. 6, the flocculating efficiencies of two microalgae enhanced with the increasing MBF-L804
concentration, and the flocculating efficiency 99.04 % of C. reinhardtii was achieved when the culture supernatant of L804 was mixed with C. reinhardtii at a ratio of 1/3. The flocculant of C. minutissima required more L804 culture supernatant, and flocculating efficiency of 93.83 % was
observed when the L804 supernatant was mixed with C. minutissima culture in a ratio 1/2, which is significantly lower than the dosage used in previous
studies. Wan et al. 3] reported that the bioflocculant produced by Solibacillus silvestris W01 can harvest 90 % marine microalga Nannochloropsis oceanica DUT01 when the culture supernatant of W01 was mixed with the microalgal culture at
a ratio of 3:1. Manheim and Nelson 27] reported that the bioflocculant secreted by Burkholderia cepacia could settle microalgae Scenedesmus sp. and Chlorella vulgaris when algae and bacteria suspensions were mixed in a 2:1 ratio (v/v). In this study,
strain L804 can produce bioflocculant by utilizing untreated lignocellulosic biomass
as carbon source, which can avoid the capital expenditure intensive pretreatment step,
and promote its application in harvesting of microalgae.

Fig. 6. Flocculating efficiencies of C. reinhardtii and C. minutissima in different volume ratios of L804 fermentation broth/microalgae culture