Science of the Total Environment 650 (2019) 2931–2938

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Adaptive evolution and carbon dioxide fixation of Chlorella sp. in

simulated flue gas
Dujia Cheng a,b,c, Xuyang Li a,d, Yizhong Yuan a,b,c, Chengyu Yang a,b, Tao Tang a, Quanyu Zhao a,c,e,⁎, Yuhan Sun a,c
a
Shanghai Advanced Research Institute, Chinese Academy of Sciences, 99 Haike Road, Shanghai 201210, China
b
University of Chinese Academy of Sciences, 19 Yuquan Road, Beijing 100049, China
c
ShanghaiTech University, 100 Haike Road, Shanghai 201210, China
d
School of Life Science, Shanghai University, 99 Shangda Road, Shanghai 200444, China
e
School of Pharmaceutical Science, Nanjing Tech University, 30 Puzhu South Road, Nanjing, China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Chlorella sp. Cv, was obtained by adap-

tive evolution in 138 days.
• Chlorella sp. Cv tolerates 10% CO2,
200 ppm NOx and 100 ppm SOx.
• The maximum CO2 fixation rate was
1.2 g L−1 d−1 from simulated flue gas.
• Upregulations of photosynthesis and
oxidative phosphorylation were identi-
fied.
• Upregulation of extracellular sulfur
transport is helpful to tolerate SOx.

a r t i c l e i n f o a b s t r a c t

Article history: Carbon dioxide and other greenhouse gas emissions leads to global warming. Biological capture through
Received 21 August 2018 microalgae is a potential approach for solving this environmental problem. It is still a technical challenge to en-
Received in revised form 5 October 2018 hance the tolerance of microalgae to flue gas if CO2 is fixed from flue gas directly. A new strain, Chlorella sp. Cv was
Accepted 5 October 2018
obtained through adaptive evolution (46 cycles) against simulated flue gas (10% CO2, 200 ppm NOx and 100 ppm
Available online 6 October 2018
SOx). It was confirmed that Chlorella sp. Cv could tolerate simulated flue gas conditions and the maximum CO2
Editor: Jay Gan fixation rate was 1.2 g L−1 d−1. In a two-stage process, the biomass concentration was 2.7 g L−1 and the carbo-
hydrate content was 68.4%. Comparative transcriptomic analysis was performed for Chlorella sp. Cv under simu-
Keywords: lated flue gas and control conditions (10% CO2). These responses against simulated flue gas uncovered the
Chlorella significant difference between the evolved strain and the original strain. The metabolic responses to flue gas
Adaptive evolution were explored with focus on various specific genes. Upregulation of several genes related to photosynthesis, ox-
Simulated flue gas idative phosphorylation, CO2 fixation, sulfur metabolism and nitrogen metabolism was beneficial for the evolved
CO2 capture strain to tolerate the simulated flue gas. The upregulation of genes related to extracellular sulfur transport and
Comparative transcriptomic analysis
nitrate reductase was essential to utilize the sulfate and nitrate from dissolved SOx and NOx. The results in this
study are helpful to establish a new process for CO2 capture directly from industrial flue gas.
© 2018 Elsevier B.V. All rights reserved.

1. Introduction
⁎ Corresponding author at: School of Pharmaceutical Science, Nanjing Tech University,
30 Puzhu South Road, Nanjing, China. CO2 and other greenhouse gases emissions lead to global warming
E-mail address: zhaoqy@njtech.edu.cn (Q. Zhao). and other environmental issues (Cabello et al., 2017; Kassim and

https://doi.org/10.1016/j.scitotenv.2018.10.070
0048-9697/© 2018 Elsevier B.V. All rights reserved.
2932 D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938
Meng, 2017; Toledo-Cervantes et al., 2018; Wang et al., 2018). Many
physical, chemical and biological methods have been proposed to
solve this problem. Negative CO2 emission pathways could be recon-
structed using microalgae (Moreira and Pires, 2016). NO2 and SO2 can
be dissolved in algal medium that is then available for the growth of
microalgae (Van den Hende et al., 2012). High concentrations of the
trace components in flue gas have lethal effects for most microalgae
with the exception of high concentration of CO2 (Yen et al., 2015).
NOx and SOx are typical components in flue gas (Chen et al., 2015).
Pretreatments of flue gas could be ignored if the flue gas could be
captured by microalgae without pretreatment. In general, it is challeng-
ing to capture CO2 from industrial flue gas directly by microalgae
(Vuppaladadiyam et al., 2018). Above all, the isolation or selection of a
tolerant microalgae to flue gas is the primary requirement. High initial
cell density (Chiu et al., 2011; Dineshbabu et al., 2017), low concentra-
tions of NOx and SOx (Radmann et al., 2011), short exposure time to the
flue gas (Chiu et al., 2011; Duarte et al., 2016; Tastan et al., 2012), low
gas flow rate (Chiu et al., 2011), additions of growth stimulators
(Tastan et al., 2012) and pH control using buffer (Choi et al., 2017)
have been identified as methods to improve the tolerance and growth
of microalgae in flue gas. Chlorella sp. (Chiu et al., 2011; Duarte et al.,
2017; Duarte et al., 2016), Phormidium valderianum BDU 20041
(Dineshbabu et al., 2017), Haematococcus pluvialis (Choi et al., 2017),
Spirulina sp. (Duarte et al., 2017; Radmann et al., 2011), Scenedesmus
obliquus (Radmann et al., 2011) and Synechococcus sp. (Tastan et al.,
2012) are the extensively studied algal species. The largest technical
challenges for carbon capture by microalgae include both low tolerance
to flue gas and low CO2 fixation rate of microalgae. In addition, the re-
lated tolerance mechanism is not fully understood. Comparative
transcriptomic analysis is effective approach to uncover the cellular re-
sponse and metabolic regulation in microalgae under stress conditions
(Li et al., 2017b; Liu et al., 2017; Qian et al., 2018; Qiao et al., 2018;
Zhou et al., 2017). It is helpful to unravel the possible tolerance and deg-
radation mechanisms.
Adaptive evolution was performed to improve the tolerant capabili-
ties against high concentrations of CO2 (Li et al., 2015), phenol (Wang
et al., 2016), salt (Li et al., 2018), metals (Hochmuth et al., 2015) and
glucose (Li et al., 2017a). Selection and adaptation of microalgal consor-
tia for several months in 100% flue gas has also been reported (Aslam
et al., 2017). The final biomass concentration was 0.3 g L−1 under
100% flue gas condition with a phosphate buffer. If the phosphate buffer
was absent, the microalgae did not survive. This study showed that a
long term adaption may be effective in enhancing the tolerance of
microalgae. Adaptive evolution is widely used for phenotype improve-
ment of Escherichia coli, Saccharomyces cerevisiae and Yarrowia lipolytica
(Dai et al., 2018; Shaw et al., 2016; Yu et al., 2018).Compared to model
microorganisms in fermentation industry, only a few studies have been
performed using microalgae. The growth rate of E. coli is much higher
than that of Chlorella sp. and other microalgae so that the generations
of evolution in the former is greater than that of latter over the same
time period. This indicates that the evolution efficiency of E. coli is
higher than Chlorella sp. The typical difference between E. coli and Chlo-
rella sp. is that microalgae have photosynthetic systems. This means
that microalgae are much more complicated than bacteria. It is critical
to construct an effective approach of adaptive evolution for microalgae.
CO2 is a weak environmental stress for microalgae so 10% or 20% CO2
was adopted as an environmental stress to enhance the CO2 tolerance
of Chlorella sp. (Li et al., 2015). Adaptive evolution was confirmed to
be effective. Flue gas is a much heavier environmental stress than 10%
CO2. Microalgae would be dead if exposed to flue gas in the early
stage of the adaptive evolution process. A new methodology must be
proposed if adaptive evolution is performed for microalgae in flue gas
without a buffer (Aslam et al., 2017).
In the current study, adaptive evolution was performed for a green
microalgae, Chlorella sp. AE10 in simulated flue gas conditions. Compar-
ative transcriptomic analysis was carried out for resulting strain under
simulated flue gas and control conditions. The global response and pos-
sible tolerance mechanism of the resulting strain against simulated flue
gas was discussed.
2. Methods
2.1. Experimental organism and cultivation conditions
Chlorella sp. AE10 is a high CO2 concentration tolerant strain that can
grow in 10–30% CO2 (Li et al., 2015). In this work, Chlorella sp. AE10 was
grown in a 330 ml tubular photobioreactor using BG11 medium. The
light intensity was controlled to be 100 μmol photons m−2 s−1 (LL) or
1000 μmol photons m−2 s−1 (HL). The culture temperature was care-
fully maintained at 28 ± 0.05 °C. The input gas was 10% CO2 (v v−1)
or simulated flue gas (10% CO2, 200 ppm NOx and 100 ppm SOx). The
10% CO2 was mixed with pure CO2 and air. The simulated flue gas was
prepared using two gas resources. One was 10% CO2 and another was
1% NOx, 0.5% SOx and 98.5% N2 (v v−1). These gases were made by
Shangha Pujiang Special Gas Co. The gas flow rate was 0.1 L min−1.
The contents of NOx and SOx were determined by flue gas component
analyzer (Lancom 4, USA).
2.2. Adaptive laboratory evolution
Flue gas is a complex gas mixture. The typical data from coal-fired
power plant are 10–12% CO2, 200 ppm NOx, 10% SOx and other trace
components. In this study, simulated flue gas containing 10% CO2,
200 ppm NOx and 10% SOx was used. Adaptive evolution using a strat-
egy of an increasing simulated flue gas concentration was performed
for the original strain, Chlorella sp. AE10. There were five stages during
the whole adaptive evolution process. The input gas was 10% CO2,
50 ppm NOx and 25 ppm SOx from cycle 1 to cycle 4. In the second
stage (from cycle 5 to cycle 7), the gas was increased to 10% CO2,
75 ppm NOx and 38 ppm SOx. Then, the gas was set to 10% CO2,
100 ppm NOx and 50 ppm SOx from cycle 8 to cycle 11 as the third
stage. In the fourth stage (cycle 12 to cycle 21), the concentrations of
NOx and SOx were changed to 150 ppm and 75 ppm, respectively. The
strain was cultivated under 10% CO2, 200 ppm NOx and 100 ppm SOx
from cycle 22 to cycle 46. The inoculum cell density was approximately
0.3–0.4 g L−1. The biomass concentration at the end of each cycle was
measured using dry weight method (Li et al., 2015). Each cycle of adap-
tive evolution has the same time interval, 3 days. All experiments were
carried out in triplicate. Finally, a resulting strain was obtained after
138 days and was referred as Chlorella sp. Cv.
2.3. Characterization of the resulting strain
Both Chlorella sp. AE10 and Chlorella sp. Cv were cultivated under
10% CO2, 0.5 FG (10% CO2, 100 ppm NOx and 50 ppm SOx) and simulated
flue gas (FG) for 5 days. The optical density at 750 nm (HACH DR2800,
USA) (Jebali et al., 2015), the pH in medium and the maximum quantum
yield of photosystem II of the microalgae were determined each day.
The Fv/Fm was measured using the fluorescence monitoring system
(FMS2, Lufthansa Scientific Instruments Co., Ltd., UK) (Cheng et al.,
2017). In another experiment, the growth profile of Chlorella sp. Cv
was determined using the dry weight method for one week (Cheng
et al., 2017). The growth rates of the resulting strain were confirmed
using different methods with three biological replicates. The maximum
biomass productivity was estimated using the equations that were de-
scribed previously (Li et al., 2018). The maximum CO2 fixation rate
was calculated using 1.88 × biomass productivity (g L−1 d−1) (Bagchi
and Mallick, 2016; Dineshbabu et al., 2017; Li et al., 2018; Yen et al.,
2015).
 D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938 2933

2.4. Accumulation of carbohydrate in the two-stage process (A)

250
Chlorella sp. AE10 trends to accumulate carbohydrate not lipid NOx SO2

Gas concentration˄ppmˈv v-1˅

under stress conditions. A two-stage process with cell dilution was
developed (Cheng et al., 2017). In the first stage (3 days), Chlorella 200
sp. AE10 and Chlorella sp. Cv were cultivated both in 10% CO2, me-
dium with full nitrogen concentration (1.5 g L−1 NaNO3) and under
150
low light intensity (100 μmol photons m−2 s−1) conditions. Then,
Chlorella sp. AE10 was cultivated in 10% CO2, medium with 25% nitro-
gen concentration (0.375 g L−1 NaNO3) and under high light inten- 100
sity conditions for the next 5 days. The culture conditions for
Chlorella sp. Cv in the second stage were the simulated flue gas, the
culture medium with 25% nitrogen concentration of normal condi- 50
tion and the light intensity was 1000 μmol photons m−2 s −1 . The
methods used to determine the biomass concentration and total car-
bohydrate content were described in a previous study (Cheng et al., 0
2017). 1 4 7 11 14 18 21 25 29 33 37 41 45
Cycle
3
2.5. Transcriptomic analysis (B) Inial Increased

Biomass concentration˄g L-1˅

The algal cells were collected at day 1 for Chlorella sp. Cv cultivated 2.5
under simulated flue gas and 10% CO2 (control) conditions with three
replicates. The sample treatment, transcriptomic determination, gene 2
annotation and the related bioinformatics analysis were described in
previous studies (Li et al., 2018; Zhou et al., 2017). Two parameters
were selected to investigate the target genes. If false discovery rate 1.5
(FDR) was less than 0.05 and the absolute value of Log2 fold change
(LogFC) was not less than 1, the gene was regarded as significantly
1
expressed one. The global distributions of significantly expressed
genes were made on the whole genome scale using iPath 2.0 (Yamada
et al., 2011). 0.5

2.6. Statistical analysis

0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46
The results are shown as means ± standard deviation of three biolog-
ical replicates. The statistical significance of the biomass concentration,
Cycles
photosynthetic activity, pH and carbohydrate content was evaluated
Fig. 1. NOx and SOx concentrations in the adaptive evolution process (A) and biomass
using Student's t-Test. In all cases, P value less than 0.05 was considered
concentration in each cycle of the adaptive evolution process (B).
significant.

3. Results and discussion
3.1. Adaptive evolution
Inoculum cell density, NOx and SOx concentrations, and the gas flow
rate are very critical for the flue gas capture by microalgae. Chlorella sp.
AE10 It was confirmed to tolerate 10% CO2, 50 ppm NOx and 25 ppm SOx
for 3 days in the preliminary experiments. The initial NOx and SOx con-
centrations were selected to be 50 ppm and 25 ppm, respectively
(Fig. 1A). After 4 cycles, the strain can tolerate 10% CO2, 75 ppm NOx
and 38 ppm SOx. Finally, the resulting strain, Chlorella sp. Cv, was ob-
tained in 138 days. The biomass concentration at the end of each cycle
was 1.0–2.6 g L−1 (Fig. 1B). In the last ten cycles, the final cell density
in each cycle was greater than 1.7 g L−1. These results demonstrate
that Chlorella sp. Cv is tolerating to the simulate flue gas compared
with short-term adaptation under the simulated flue gas conditions.
From the experimental data, there was approximately 110 generations
from the original strain to resulting strain.
The selection of environmental stress in adaptive evolution is closely
connected to the efficiency of the adaptive evolution. If the environmen-
tal stress exceeds the tolerance of microalgae, the adaptive evolution
will failed. If the environmental stress is very low, the efficiency of adap-
tive evolution will be too low and the whole process will take too much
time. In previous studies, one (Li et al., 2015; Wang et al., 2016) or two
(Li et al., 2018) stress conditions were adopted for the whole process of
adaptive evolution although the growth rates were very low under 10%
CO2 or 500 mg L−1 phenol conditions. A high salinity of 30 g L−1 is
harmful to microalgae so that the first stress condition was 26 g L−1
salt in the adaptive evolution of Chlorella sp. AE10 17. After 11 cycles,
Chlorella sp. AE10 was cultivated in 26 g L−1 salt. In the current study,
more than 50 ppm NOx and 25 ppm SOx (CO2 was 10%) led to a negative
outcome in the original strain (data not shown). A strategy of using a
stepwise increase with five stages was implemented. When the NOx
and SOx concentrations were increased, the adaptive evolution took
many more cycles.
3.2. Characterization of resulting strain, Chlorella sp. Cv
The resulting strain, Chlorella sp. Cv, when subject to three gaseous
mixtures conditions (control, 0.5 FG and FG) had similar growth pat-
terns to the growth patterns of the original strain, Chlorella sp. AE10
under 10% CO2 conditions (Fig. 2A). In contrast, Chlorella sp. AE10
could not grow in the simulated flue gas. Changes were also identified
in both the photosynthetic activity and the pH of the medium. Simu-
lated flue gas, especially SOx and NOx, had negative effects on PS I and
PS II of microalgae. As shown in Fig. 2B, the photosynthetic activity of
Chlorella sp. AE10 decreased from 0.7 to approximately 0.2 at day 1 in
FG. Even in 0.5 FG, the photosynthetic activity decreased significantly
from day 2. It was obvious that Chlorella sp. AE could not tolerate simu-
lated flue gas. If the simulated flue gas can be utilized by microalgae, the
pH in the medium will remain in a normal range or it will reach a very
low value. On day 1, the pH was approximately 3 for Chlorella sp.
2934 D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938

biomass concentration of Chlorella sp. Cv at day 7 was approximately

3.4 g L−1 under simulated flue gas conditions (Fig. 3). The maximum
biomass productivity of the resulting strain was approximately
0.64 g L−1 d−1. The culture conditions and maximum CO2 fixation
rates of microalgae in the flue gas are shown in Table 1. The maxi-
mum CO2 fixation rate of Chlorella sp. Cv was 1.20 g L−1 d−1 which
was greater than that of other microalgae.

3.3. Accumulation of carbohydrate in two-stage process

Lipid (Chen et al., 2018; Chen et al., 2016) or carbohydrate (Cheng

et al., 2017; Yuan et al., 2018) can accumulate in microalgae. The two-
stage process was performed for carbon capture and biomass produc-
tion using the resulting strain, Chlorella sp. CV. Both the original strain
and the resulting strain had similar growth rates in the first stage. The
final cell density in terms of the dry weight of Chlorella sp. AE10 was
3.15 g L−1 in the second stage while that of Chlorella sp. Cv was approx-
imately 2.73 g L−1 (Fig. 4A). Although the carbon fixation of Chlorella sp.
Cv was inhibited partially by SOx and NOx, CO2 capture from flue gas
was also possible. The carbohydrate content of Chlorella sp. Cv in stage
1 was greater than that of Chlorella sp. AE10. Carbohydrate accumula-
tion under 10% CO2 conditions by Chlorella sp. Cv was greater as com-
pared with Chlorella sp. AE10. The greatest carbohydrate content, 75%,
was achieved at day 6 for Chlorella sp. AE and was 68.4% on day 5 for
Chlorella sp. Cv (Fig. 4B). Carbohydrate was effectively accumulated by
Chlorella sp. Cv with the two-stage process from flue gas. NOx and SOx
inhibited the growth of and carbohydrate accumulation in Chlorella sp.
Cv.

3.4. Comparative transcriptomic analysis in simulated flue gas and 10% CO2

A total of 59,958 unigenes were identified in the current study. The

significantly expressed genes are shown in Fig. 5. It should be men-
tioned that there were only 495 genes that were significantly expressed.
The number of significantly expressed genes was distinctively less than
the number reported in previous studies (Li et al., 2018; Zhou et al.,
2017). The number of the significantly expressed genes were 5756 in
a comparative transcriptomic analysis of the salt tolerant strain, Chlo-
rella sp. S30, under 30 g L−1 salt and control conditions (Li et al.,
2018). There were 3318 significantly expressed genes in comparative
transcriptomic analysis of the phenol tolerant strain, Chlorella sp. L5,
under 500 mg L−1 phenol and control conditions (Zhou et al., 2017).
The rapid response of Chlorella sp. Cv to flue gas was limited to the
genes in pathways of photosynthesis, oxidative phosphorylation, CO2
fixation, sulfur metabolism, nitrogen metabolism and amino acid me-
tabolism. The resulting strain has high tolerance to NOx and SOx. It is
Biomass concentration˄g DW L-1˅

3.5
3
Fig. 2. Characterization of the growth profile (A) and the maximum quantum efficiency of 2.5
photosystem II (Fv/Fm) (B) and the pH of the medium (C) of Chlorella sp. AE10 and
Chlorella sp. Cv under 10% CO2, 0.5 FG (10% CO2, 100 ppm NOx and 50 ppm SOx) and FG 2
(10% CO2, 200 ppm NOx and 100 ppm SOx). The data are shown as mean ± SD (n = 3).
1.5
AE10 in FG (Fig. 2C). The pH changed from 7 to 3.3 for Chlorella sp. AE10 1
from day 0 to day 5 in 0.5 FG which inhibits the growth of microalgae.
Chlorella sp. Cv maintained high Fv/Fm value, 0.7, for the whole culture 0.5
process. The pH in the medium of Chlorella sp. Cv exceeded more than 6 0
in FG or 0.5 FG. 0 1 2 3 4 5 6 7
The capability of the evolved strain, Chlorella sp. Cv, to tolerate
flue gas was demonstrated to be significantly enhanced by the adap-
tive evolution. The constructed stepwise increasing strategy was ef- Fig. 3. Growth profile of Chlorella sp. Cv under simulated flue gas (10% CO2, 200 ppm NOx
fective in improving the microalgae s tolerance to flue gas. The and 100 ppm SOx) conditions. The data are shown as mean ± SD (n = 3).
 D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938 2935

Table 1
Comparison of CO2 fixation rate from flue gas by microalgae.

Microalgae species Flue gas flow Flue gas conditions PBR type and Light intensity Maximum biomass MaximumCO2 Ref.
rate (vvm) working volume (μmol m−2 s−1) productivity fixation rate
(g L−1 d−1) (g L−1 d−1)

Chlorella sp. 0.05 25% CO2, 4% O2, 80 ppm NO, 90 Bubble column, 0.8 L 300 0.52 0.98 Chiu et al., 2011
ppm SO2
Haematococcus pluvialis 0.1 3–6% of CO2, 11.99 ± 0.73% O2, Bubble column, 5 L 30–50 0.07 0.125 Choi et al., 2017
21.72 ± 3.72 ppm
NOx, 1.43 ± 4.03 ppm CO,
Chlorella fusca 0.05 10–12% CO2, 400 ppm NOa Flask-type, 1.8 L – 0.36 0.68 Duarte et al., 2017
Spirulina sp. 0.05 10–12% CO2, 400 ppm NOa Flask-type, 1.8 L – 0.135 0.25 Duarte et al., 2017
Chlorella fusca 0.05 10% CO2, 200 ppm NO, 200 ppm Tubular 41.6 0.11 0.21 Duarte et al., 2016
SO2b photobioreactor, 1.8 L
Phormidium valderianum 0.35 15% CO2, +NOx + SOx Open tank, 500 L – 0.03 0.0564 Dineshbabu et al., 2017
BDU 20041
Synechococcus nidulans 0.3 12% CO2, 60 ppm SO2,100 ppm Tubular-type 3200 Lux 0.07 0.13 Radmann et al., 2011
NO photobioreactor, 1.8 L
Chlorella vulgaris 0.3 12% CO2, 60 ppm SO2,100 ppm Tubular-type 3200 Lux 0.09 0.17 Radmann et al., 2011
NO photobioreactor, 1.8 L
Spirulina sp. 0.3 12% CO2, 60 ppm SO2,100 ppm Tubular-type 3200 Lux 0.08 0.15 Radmann et al., 2011
NO photobioreactor, 1.8 L
Scenedesmus obliquus 0.3 12% CO2, 60 ppm SO2,100 ppm Tubular-type 3200 Lux 0.06 0.11 Radmann et al., 2011
NO photobioreactor, 1.8 L
Chlorella sp. Cv 0.3 10% CO2, 200 ppm NOx, 100 ppm Bubble tube, 0.33 L 100 0.64 1.20 This study
SOx

–, not provided.
a
Flue gas was inputted every 2 h of the light period (set between 7 h and 19 h) for 10 min.
b
Flue gas was inputted every 40 min of the light period for 1 min.

4 possible that the gene expressions related to antioxidant enzymes are

(A) high under control condition such that they were not significantly
3.5 changed in the comparative transcriptomic analysis. Additional assays
g DW L-1

LL-AE-CO2
3 HL-AE-CO2 of antioxidant enzymes should be performed in the future.
Flue gas was thought to inhibit photosynthesis and energy metabo-
LL-Cv-CO2
2.5 lism of microalgae and decrease the CO2 fixation dramatically
HL-Cv-FG
Biomass concentration

(Vuppaladadiyam et al., 2018). Just as shown in Table 2, petE in the pho-

2 tosynthetic electron transport chain, the gene responsible for phospho-
enolpyruvate carboxykinase (4.1.1.49) in Calvin cycle and the gene
1.5
coding carbonic anhydrase were downregulated while other genes in
1 oxidative phosphorylation and photosynthesis were upregulated. This
is the reason why the CO2 fixation rate of Chlorella sp. Cv in the flue
0.5 gas was not significantly reduced. The upregulation of gene of nitrate re-
ductase (NR) helped decrease the harmful effect of high concentration
0
of NOx. The upregulations of CysA and CysU CysPUWA increased the ex-
0 2 4 6 8
Culture time (d) tracellular sulfur transport into the algal cells. This upregulation is also
helpful for the evolved strain to tolerate the toxicity of high concentra-
90 tion of SOx.
(B) NOx and SOx are toxic compounds in flue gas that lead to acidic en-
80
vironments or a low pH for microalgae. One choice is to improve the tol-
70 erance of microalgae to NOx and SOx. A tolerant strain, Chlorella sp. Cv,
was obtained in this study. Carbon capture is the first step in the devel-
%

60 opment of negative emission technologies. Determining how to achieve

Carbohydrate content

a large-scale cultivation of microalgae and how to develop algal prod-

50
ucts are essential problems (Chen et al., 2015). The major target of
40 this study is the carbon capture. SO2 is easily dissolved in water while
LL-AE-CO2 the solubility of the major component of NOx, NO, is very low. The re-
30 moval efficiencies of SOx, NO and NO2 of Chlorella sp. CV on day 5
HL-AE-CO2
20 were approximately 98%, 37% and 81%, respectively, calculated from
LL-Cv-CO2
their outlet concentrations and input concentrations. Although NO
10 HL-Cv-FG was inputted into the photobioreactor, the utilization efficiency was
very limited. The NO could be converted into NO2, and the latter could
0
be used as substrate for microalgae cultivation (Chen et al., 2015;
0 2 4 6 8
Chen et al., 2018; Chen et al., 2016). The removal capabilities of the
Culture time (d)
resulting strain for NO, NO2 and SOx in addition to CO2 will be optimized
in the future.
Fig. 4. Characterization of the growth profile (A) and the carbohydrate content (B) of
Chlorella sp. AE10 and Chlorella sp. Cv during the two-stage process. Low light intensity
NOx and SOx in flue gas are absorbed in the culture medium. The
(LL) was 100 μmol photons m−2 s−1 and high light intensity (HL) was 1000 ions, NO32−, NO2−, SO42− and SO32−, lead to increase in acidity and de-
μmol photons m−2 s−1. The data are shown as mean ± SD (n = 3). crease pH. Based on the comparative transcriptomic analysis above,
 2936
D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938
Fig. 5. Global distribution of the significantly upregulated (red) and downregulated (blue) genes in the whole metabolic network of Chlorella sp. Cv under the simulated flue gas (10% CO2, 200 ppm NOx and 100 ppm SOx) and 10% CO2 (control)
conditions. The red line indicates significantly upregulated genes and the blue line means significantly downregulated genes. The black lines indicate there were no significant changes.
 D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938 2937

Table 2 NOx Nitrogen oxides

Comparative transcriptomic analysis for the genes of Chlorella sp. Cv under simulated flue NR Nitrate reductase
gas (10% CO2, 200 ppm NOx and 100 ppm SOx) and control conditions.
PS I Photosystem I
Pathways Gene or enzyme LogFC FDC PS II Photosystem II
Oxidative phosphorylation 1.6.5.3 2.37 3.21e-13 SOx Sulfur oxides
1.6.99.3 3.42 4.99e-12
3.6.3.14 3.44 2.19e-32
3.18 7.03e-44 Conflicts of interests
3.75 6.52e-54
3.01 3.13e-13 The authors declare no conflict of interests. The funding sponsors
1.56 2.25e-2
have no role in the design of the study.
3.87 8.57e-72
3.9 9.83e-65
3.23 1.46e-30
ND2 2.90 3.78e-12
Acknowledgements
ND4 2.35 1.15e-6
Ndufs2 3.42 4.99e-12 This study is supported by National Natural Science Foundation of
Cytb 2.72 2.07e-9 China (21576278).
COX3 3.2 2.05e-14
COX1 3.50 6.12e-21
COX2 2.77 8.92e-8 References
Photosynthesis PsaA 3.80 1.21e-51
Aslam, A., Thomas-Hall, S.R., Mughal, T.A., Schenk, P.M., 2017. Selection and adaptation of
PsaB 4.04 6.16e-56
microalgae to growth in 100% unfiltered coal-fired flue gas. Bioresour. Technol. 233,
PsbA 4.26 1.59e-79 271–283.
PsbB 2.95 2.46e-60 Bagchi, S.K., Mallick, N., 2016. Carbon dioxide biofixation and lipid accumulation potential
PsbC 3.92 5.56e-87 of an indigenous microalga Scenedesmus obliquus (Turpin) Kützing GA 45 for biodie-
PsbE 2.89 6.91e-42 sel production. RSC Adv. 6, 29889–29898.
PsbO 4.14 4.40e-11 Cabello, J., Morales, M., Revah, S., 2017. Carbon dioxide consumption of the microalga
PsbZ 1.30 3.32e-2 Scenedesmus obtusiusculus under transient inlet CO2 concentration variations. Sci.
PetB 3.24 1.83e-37 Total Environ. 584-585, 1310–1316.
PetA 1.97 3.62e-15 Chen, H., Qiu, T., Rong, J., He, C., Wang, Q., 2015. Microalgal biofuel revisited: an
PetC 4.98 1.06e-18 informatics-based analysis of developments to date and future prospects. Appl. En-
petE −1.28 2.53e-10 ergy 155, 585–598.
PetJ 7.75 0 Chen, W., Zhang, S., Rong, J., Li, X., Chen, H., He, C., Wang, Q., 2016. Effective biological
deNOx of industrial flue gas by the mixotrophic cultivation of an oil-producing
4.2.1.1 (carbonic anhydrase) −2.59 6.74e-4
green alga Chlorella sp. C2. Environ. Sci. Technol. 50, 1620–1627.
−1.23 4.32e-3
Chen, H., Wang, J., Zheng, Y., Zhan, J., He, C., Wang, Q., 2018. Algal biofuel production
Nitrogen metabolism NR 2.06 6.03e-17
coupled bioremediation of biomass power plant wastes based on Chlorella sp. C2 cul-
ABC transporters CysU 1.35 2.39e-4 tivation. Appl. Energy 211, 296–305.
Sulfur metabolism CysA 1.27 2.02e-2 Cheng, D., Li, D., Yuan, Y., Zhou, L., Li, X., Wu, T., Wang, L., Zhao, Q., Wei, W., Sun, Y., 2017.
CO2 fixation 4.1.1.49 −1.48 3.35e-5 Improving carbohydrate and starch accumulation in Chlorella sp AE10 by a novel
two-stage process with cell dilution. Biotechnol. Biofuels 10, 75.
Positive value of LogFC(Cv-FG/Cv-control) means upregulated and negative value means
Chiu, S.Y., Kao, C.Y., Huang, T.T., Lin, C.J., Ong, S.C., Chen, C.D., Chang, J.S., Lin, C.S., 2011.
downregulated. Control condition is 10% CO2.
Microalgal biomass production and on-site bioremediation of carbon dioxide, nitro-
gen oxide and sulfur dioxide from flue gas using Chlorella sp. cultures. Bioresour.
the possible tolerance mechanism of Chlorella sp. Cv involves the con- Technol. 102, 9135–9142.
Choi, Y.Y., Joun, J.M., Lee, J., Hong, M.E., Pham, H.M., Chang, W.S., Sim, S.J., 2017. Develop-
version of these ions to ammonia or the improvement of the utilization ment of large-scale and economic pH control system for outdoor cultivation of
of sulfur through upregulations of NR and cysA. The microalgae could microalgae Haematococcus pluvialis using industrial flue gas. Bioresour. Technol.
eliminate these negative effects by themselves. 244, 1235–1244.
Dai, Z., Huang, M., Chen, Y., Siewers, V., Nielsen, J., 2018. Global rewiring of cellular metab-
olism renders Saccharomyces cerevisiae Crabtree negative. Nat. Commun. 9, 3059.
4. Conclusions Dineshbabu, G., Uma, V.S., Mathimani, T., Deviram, G., Ananth, D.A., Prabaharan, D., Uma,
L., 2017. On-site concurrent carbon dioxide sequestration from flue gas and calcite
formation in ossein effluent by a marine cyanobacterium Phormidium valderianum
Chlorella sp. Cv was obtained using adaptive evolution (46 cycles)
BDU 20041. Energy Convers. Manag. 141, 315–324.
against simulated flue gas (10% CO2, 200 ppm NOx and 100 ppm SOx). Duarte, J.H., Fanka, L.S., Costa, J.A., 2016. Utilization of simulated flue gas containing CO,
Its maximum CO2 fixation rate was 1.2 g L−1 d−1. At the end of the SO, NO and ash for Chlorella fusca cultivation. Bioresour. Technol. 214, 159–165.
two-stage process, the biomass concentration was 2.7 g L−1, and the Duarte, J.H., de Morais, E.G., Radmann, E.M., Costa, J.A.V., 2017. Biological CO2 mitigation
from coal power plant by Chlorella fusca and Spirulina sp. Bioresour. Technol. 234,
carbohydrate content was 68.4%. A comparative transcriptomic analysis 472–475.
was performed for Chlorella sp. Cv under the simulated flue gas and con- Hochmuth, J.D., De Meester, L., Pereira, C.M., Janssen, C.R., De Schamphelaere, K.A., 2015.
trol conditions. Several genes involved in photosynthesis, oxidative Rapid adaptation of a Daphnia magna population to metal stress is associated with
heterozygote excess. Environ. Sci. Technol. 49, 9298–9307.
phosphorylation, CO2 fixation, sulfur metabolism and nitrogen metabo- Jebali, A., Acien, F.G., Gomez, C., Fernandez-Sevilla, J.M., Mhiri, N., Karray, F., Dhouib, A.,
lism were upregulated. These results supported the hypothesis that the Molina-Grima, E., Sayadi, S., 2015. Selection of native Tunisian microalgae for simul-
evolved strain can tolerate the simulated flue gas besides the rapid taneous wastewater treatment and biofuel production. Bioresour. Technol. 198,
424–430.
growth rate. The upregulations of genes responsible for extracellular Kassim, M.A., Meng, T.K., 2017. Carbon dioxide (CO2) biofixation by microalgae and its po-
sulfur transport was essential to tolerate high concentration of SO2. tential for biorefinery and biofuel production. Sci. Total Environ. 584-585,
There is a potential for the further development of negative carbon 1121–1129.
Li, D.J., Wang, L., Zhao, Q.Y., Wei, W., Sun, Y.H., 2015. Improving high carbon dioxide tol-
emission technology using microalgae.
erance and carbon dioxide fixation capability of Chlorella sp by adaptive laboratory
evolution. Bioresour. Technol. 185, 269–275.
Abbreviations Li, X., Pei, G., Liu, L., Chen, L., Zhang, W., 2017a. Metabolomic analysis and lipid accumula-
tion in a glucose tolerant Crypthecodinium cohnii strain obtained by adaptive labora-
tory evolution. Bioresour. Technol. 235, 87–95.
FDR False Discovery Rate Li, Y., Wang, H.P., Liao, C.-Y., Zhao, X., Hsiung, T.-L., Liu, S.-H., Chang, S.-G., 2017b. Dual al-
FS Flue gas kali solvent system for CO2 capture from flue gas. Environ. Sci. Technol. 51,
HL High Light intensity 8824–8831.
Li, X., Yuan, Y., Cheng, D., Gao, J., Kong, L., Zhao, Q., Wei, W., Sun, Y., 2018. Exploring stress
LL Low light intensity tolerance mechanism of evolved freshwater strain Chlorella sp. S30 under 30 g/L salt.
LogFC Log2 Fold Change Bioresour. Technol. 250, 495–504.
2938 D. Cheng et al. / Science of the Total Environment 650 (2019) 2931–2938
Liu, S., Guo, C., Lin, W., Wu, F., Lu, G., Lu, J., Dang, Z., 2017. Comparative transcriptomic ev-
idence for Tween80-enhanced biodegradation of phenanthrene by Sphingomonas sp.
GY2B. Sci. Total Environ. 609, 1161–1171.
Moreira, D., Pires, J.C.M., 2016. Atmospheric CO2 capture by algae: negative carbon diox-
ide emission path. Bioresour. Technol. 215, 371–379.
Qian, H., Xu, J., Lu, T., Zhang, Q., Qu, Q., Yang, Z., Pan, X., 2018. Responses of unicellular alga
Chlorella pyrenoidosa to allelochemical linoleic acid. Sci. Total Environ. 625,
1415–1422.
Qiao, F., Lei, K., Li, Z., Wei, Z., Liu, Q., Yang, L., He, J., An, L., Qi, H., Cui, S., 2018.
Transcriptomic responses of the freshwater snail (Parafossarulus striatulus) following
dietary exposure to cyanobacteria. Sci. Total Environ. 624, 153–161.
Radmann, E.M., Camerini, F.V., Santos, T.D., Costa, J.A.V., 2011. Isolation and application of
SOX and NOX resistant microalgae in biofixation of CO2 from thermoelectricity plants.
Energy Convers. Manag. 52, 3132–3136.
Shaw, A.J., Lam, F.H., Hamilton, M., Consiglio, A., MacEwen, K., Brevnova, E.E., Greenhagen,
E., Latouf, W.G., South, C.R., van Dijken, H., Stephanopoulos, G., 2016. Metabolic engi-
neering of microbial competitive advantage for industrial fermentation processes.
Science 353, 583–586.
Tastan, B.E., Duygu, E., Atakol, O., Donmez, G., 2012. SO2 and NO2 tolerance of microalgae
with the help of some growth stimulators. Energy Convers. Manag. 64, 28–34.
Toledo-Cervantes, A., Morales, T., Gonzalez, A., Munoz, R., Lebrero, R., 2018. Long-term
photosynthetic CO2 removal from biogas and flue-gas: exploring the potential of
closed photobioreactors for high-value biomass production. Sci. Total Environ. 640-
641, 1272–1278.
Van den Hende, S., Vervaeren, H., Boon, N., 2012. Flue gas compounds and microalgae:
(bio-)chemical interactions leading to biotechnological opportunities. Biotechnol.
Adv. 30, 1405–1424.
Vuppaladadiyam, A.K., Yao, J.G., Florin, N., George, A., Wang, X., Labeeuw, L., Jiang, Y.,
Davis, R.W., Abbas, A., Ralph, P., Fennell, P.S., Zhao, M., 2018. Impact of flue gas com-
pounds on microalgae and mechanisms for carbon assimilation and utilization.
ChemSusChem 11, 334–355.
Wang, L.B., Xue, C.Z., Wang, L., Zhao, Q.Y., Wei, W., Sun, Y.H., 2016. Strain improvement of
Chlorella sp for phenol biodegradation by adaptive laboratory evolution. Bioresour.
Technol. 205, 264–268.
Wang, Z., Wen, X., Xu, Y., Ding, Y., Geng, Y., Li, Y., 2018. Maximizing CO2 biofixation and
lipid productivity of oleaginous microalga Graesiella sp. WBG-1 via CO2-regulated
pH in indoor and outdoor open reactors. Sci. Total Environ. 619-620, 827–833.
Yamada, T., Letunic, I., Okuda, S., Kanehisa, M., Bork, P., 2011. iPath2.0: interactive path-
way explorer. Nucleic Acids Res. 39, W412–W415.
Yen, H.W., Ho, S.H., Chen, C.Y., Chang, J.S., 2015. CO2 , NOx and SOx removal from flue gas
via microalgae cultivation: a critical review. Biotechnol. J. 10, 829–839.
Yu, T., Zhou, Y.J., Huang, M., Liu, Q., Pereira, R., David, F., Nielsen, J., 2018. Reprogramming
yeast metabolism from alcoholic fermentation to lipogenesis. Cell 174, 1549–1558.
Yuan, Y., Liu, H., Li, X., Qi, W., Cheng, D., Tang, T., Zhao, Q., Wei, W., Sun, Y., 2018. Enhanc-
ing carbohydrate productivity of Chlorella sp. AE10 in semi-continuous cultivation
and unraveling the mechanism by flow cytometry. Appl. Biochem. Biotechnol. 185,
419–433.
Zhou, L., Cheng, D., Wang, L., Gao, J., Zhao, Q., Wei, W., Sun, Y., 2017. Comparative
transcriptomic analysis reveals phenol tolerance mechanism of evolved Chlorella
strain. Bioresour. Technol. 227, 266–272.