Appl Microbiol Biotechnol DOI 10.1007/s00253-014-6288-0

BIOENERGY AND BIOFUELS

Granulation, control of bacterial contamination, and enhanced lipid accumulation by driving nutrient starvation in coupled wastewater treatment and Chlorella regularis cultivation Dandan Zhou & Yunbao Li & Yang Yang & Yao Wang & Chaofan Zhang & Di Wang

Received: 25 October 2014 / Revised: 27 November 2014 / Accepted: 29 November 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract Bacterial contamination and biomass harvesting are still challenges associated with coupling of microalgae and wastewater treatment technology. This study investigated aggregation, bacterial growth, lipid production, and pollutant removal during bacteria contaminated Chlorella regularis cultivation under nutrient starvation stress, by supposing the C/N/P ratios of the medium to 14/1.4/1 (MB2.5) and 44/1.4/1 (MB4.0), respectively. Granules of 500–650 μm were formed in the bacteria contaminated inoculum; however, purified C. regularis were generally suspended freely in the medium, indicating that bacterial presence was a prerequisite for granulation. Extracellular polymeric substance (EPS) analysis showed that polysaccharides were dominant in granules, while protein mainly distributed in the outer layer. Denaturing gradient gel electrophoresis (DGGE) results revealed Sphingobacteriales bacterium and Sphingobacterium sp. are vital organisms involved in the flocculation of microalgae, and nitrifiers (Stenotrophomonas maltophilia) could co-exist in the granular. Both EPS and DGGE results further supported that bacteria played key roles in granulation. C. regularis was always dominant and determined the total biomass concentration during co-cultivation, but bacterial growth was limited owing to nutrient deficiency. Starvation strategy also contributed to enhancement of lipid Yang Yang contributed equally to this work with Yunbao Li Electronic supplementary material The online version of this article (doi:10.1007/s00253-014-6288-0) contains supplementary material, which is available to authorized users. D. Zhou (*) : Y. Li : Y. Yang : Y. Wang : C. Zhang : D. Wang Key Lab of Groundwater Resources and Environment, Ministry of Education, Jilin University, Changchun 130021, China e-mail: [email protected] D. Zhou School of Environment, Northeast Normal University, Changchun 130024, China

accumulation, as lipid content in MB4.0 with a greater C/N/P led to the greatest increase in the starvation period, and the maximum lipid productivity reached 0.057 g/(L·day). Chemical oxygen demand and nitrogen removal in MB4.0 reached 92 and 96 %, respectively, after 3 days of cultivation. Thus, cultivation of microalgae in high C/N/P wastewater enabled simultaneous realization of biomass granulation, bacterial overgrowth limitation, enhanced lipid accumulation, and wastewater purification. Keywords Granulation . Bacterial contamination . Lipid accumulation . Nutrient starvation

Introduction Microalgae are promising biodiesel feedstocks with great potential to displace fossil fuels and meet world’s energy demands (Stephens et al. 2010). In recent years, there has been renewed interest in the cultivation of microalgae for biofuel production owing to their ability to synthesize and accumulate high amounts of lipids using C and N sources. However, the cost of employing microalgae as an alternative fuel source has remained high, especially with respect to their cultivation and harvest. A number of studies have been conducted to overcome the aforementioned bottlenecks, and the results have identified coupled microalgae cultivation and sewage treatment as a promising process with the potential to dramatically reduce the cost of microalgae cultivation (Clarens et al. 2010; Zhou et al. 2012). Microalgae can rapidly take up pollutants in wastewater, so the coupled processes are expected to achieve simultaneous wastewater purification and microalgal biomass production. Unfortunately, bacterial contamination always occurs during this process. This contamination is generally considered to have a negative effect on the microalgae growth,

Appl Microbiol Biotechnol

as the competition for substrates represses the growth of microalgae. This competition in turn decreases biodiesel production (Zhang et al. 2012). Moreover, several studies have reported that some bacteria damage the algae by releasing soluble enzymes (Fergola et al. 2007) or releasing chemicals that cause lysis of microalgal cells (Chen et al. 2014). However, avoiding bacterial contamination by wastewater sterilization is not practical during industrial-scale microalgal production (Cho et al. 2011; Gentili 2014; Zhou et al. 2012). Bacteria and microalgae might also be mutually beneficial to each other. Bacteria degrade intractable compounds to ammonium, nitrogen, phosphate, and carbon dioxide, which can then be used by algae. Microalgae also supply nutrients to bacteria for synthesis (Croft et al. 2005). Moreover, the lipid content in Chlorella vulgaris increased approximately fourfold when co-immobilized in alginate beads with bacteria Azospirillum brasilense (de Bashan et al. 2002). Another positive effect of bacteria on microalgae is enhanced harvesting efficiency (de Bashan et al. 2002). Bacteria are the most efficient self-aggregation microorganisms and can form bioaggregates with other microorganisms, resulting in the production of activated sludge, biofilms, aerobic granules, and anaerobic granules (Lee et al. 2010; More et al. 2014). Bioaggregates usually have good settling capability and are easily separated from treated wastewater. Indeed, microalgae harvesting following co-cultivation of bacteria and microalgae is more efficient than chemical flocculation from the aspects of avoiding chemical contamination and water reuse (Castrillo et al. 2013). As mentioned above, the dose of bacteria could improve algal harvesting during co-cultivation; however, as there are competition between bacteria and microalgae, how to avoid the overgrowth of bacteria is the most critical issue in coupled wastewater treatment and microalgae cultivation. According to the universal structural formula of bacteria, C5H7O2NP1/2, a C/N/P of 10/2/1 is required for bacterial cell synthesis, and the C/N/P ratio, which could not stratify the demand of bacterial growth is defined as nutrient starvation. However, the C/N/P ratio required for heterotrophic cultivation of microalgae has never been identified, and the ratios in the cultivation mediums were reported in a wide range from 27/10/1 to 177/4/1 (Espinosa Gonzalez et al. 2014; Su et al. 2011; Zhang et al. 2012). This may indicate that the N and P contents in microalgae are self-adjusted via synthesis of either lipids or proteins according to the nutrients present in the medium (Espinosa Gonzalez et al. 2014). Nutrient starvation generally stimulates lipid accumulation in microalgae (Prathima Devi et al. 2012). Moreover, starvation plays a crucial role in the aggregation process via induction of changes in the characteristics of the cell surface and morphology that favor the formation of microbial aggregates (McSwain et al. 2004; Tay et al. 2001). Therefore, starvation is a potential strategy to enhance lipid accumulation and microalgae aggregation and to control

bacterial contamination during coupled microalgae cultivation and wastewater treatment. However, little work has been conducted to investigate the role of nutrient starvation in microalgae aggregation and bacterial growth control. In this study, heterotrophic co-cultivation was conducted under nutrient-limiting conditions, which resulted in formation of microalgae–bacteria granules. To the best of our knowledge, this is the first report of the formation of such granules. In addition, the effects of starvation on microalgae–bacteria granulation, lipid accumulation, and bacterial growth were identified. Overall, the results of this study can be used to drive research into coupled microalgae cultivation and wastewater treatment.

Materials and methods Microorganisms, medium, and cultivation Chlorella regularis var. minima (C. regularis, FACHB-729) was obtained from the Freshwater Algae Culture Collection at the Institute of Hydrobiology in Wuhan, China. Bacterial contamination occurred during subculture under heterotrophic cultivation, after which the contaminated C. regularis was inoculated further for a long period in modified BG11 medium, and then contaminated microalgae were inoculated in cocultivation. Heterotrophic cultivation was conducted using modified BG11 medium, MB2.5 and MB4.0, which was prepared by adding 2.5 and 4.0 g/L glucose into the BG11 medium (Zhou et al. 2012), respectively. The nitrogen and phosphorus concentrations were assumed to differ owing to varying degrees of starvation. The MB2.5/MB4.0 media, which were based on BG11 medium, consisted of the following (g/L): glucose (2.500/4.000), NaNO3 (1.500/0.375), K2HPO4 (0.040/0.010), and a trace metal solution. The characteristics of the modified media are provided in Table 1 (Pancha et al. 2014). C. regularis was rejuvenated in the classic BG11 for 20 days by the method used to maintain the freshwater algae culture collection at the Institute of Hydrobiology in Wuhan, China. The purified C. regularis and the bacteriacontaminated C. regularis were then inoculated (10 %,v/v) into the modified BG11 medium. Both rejuvenation and culture were performed in 250 mL flasks with 150 mL working

Table 1

Characteristics of MB2.5 and MB4.0 used in the study

Chemical oxygen demand (COD g/L) Theoretical C/N/P

MB2.5

MB4.0

2.6 14/1.4/1

3.9 44/1.4/1

Appl Microbiol Biotechnol

volumes. Each operation was conducted at 21±1 °C with agitation (ZWY-240; Zhicheng Shanghai, China) at 120 rpm under 2000 lux light supplied by fluorescent lamps with 12-h/ 12-h light/dark cycles for 6 days.

8000 rpm for 10 min, after which the organic phase (bottom) was collected into a 15-mL tube and dried at 60 °C using a nitrogen evaporator (Allsheng, Voshengkd 2000, China). Crude oil was weighed for lipid content calculation:

Analytical procedures and methods

CL ðg=gÞ ¼ ðm1 −m0 Þ=m2

Analytical methods Chemical oxygen demand (COD) and NH4+-N, NO2−-N, and NO3−-N concentrations were determined according to the APHA Standard Methods (APHA 1998). The consortia morphology was observed microscopically using a microscope (Olympus, Bx40, Japan) equipped with a Pro-MicroScan system. A 5-mL culture solution was sampled and sonicated for 20 min before microalgae and bacteria counting. The algal cell concentration was measured by counting the cell number using a hemocytometer under a fluorescent microscope (AMG, EVOS, USA) at 20× magnification (Guo and Tong 2014). For bacteria counting, the samples after pretreatment were diluted in tenfold serial. Thereafter, six replicates of 10 μL from each of the six selected dilutions were plated on to agar medium for incubation. Afterward, the bacteria cell concentration was measured using modified MPN enumeration (Chen et al. 2003). To determine the biomass concentration (g/l), 100 ml culture solution was harvested by centrifugation at 10,000 rpm for 10 min, then dried at 90 °C for 5 h, and finally weighed after cooling at room temperature (Araujo et al. 2013). All analyses were carried out in triplicate, and the results represent the average of three independent analyses. Extracellular polymeric substance (EPS) extraction and measurement EPS was extracted following the procedure described by Frolund et al. (1996). Quantitative analysis of the polysaccharides in EPS was conducted according to Dubois et al. (1956), using glucose as a standard. The method described by Lowry et al. (1951) was used to determine the protein concentration with bovine serum albumin (MYM Biological Technology Company, China) as a standard.

Where CL is the lipid content. m1 is the weight of 15 mL tube with dried lipids; m0 is the weight of empty 15 mL tube; m2 is the weight of the algae powder. The lipid productivity PL (g/(L·day)) was calculated according to the following equation (Lv et al. 2010): PL ¼ CL ˙

ΔDW t

where ΔDW is the accumulated dry cell weight per liter from inoculation to harvest, and t is the culture time. Scanning electron microscopy (SEM) The interior microstructure of the bacteria–microalgae consortia was observed by SEM (Akashi-SX-40, USA) coupled with a cryosectioning method. Briefly, the consortia were fixed with 2.5 % glutaraldehyde in paraformaldehyde for 3 h to reduce the damage to the granular morphology caused by cryosectioning. The consortia were then frozen at −20 °C and embedded in Tissue-Tek O.C.T. (Sakura Finetek, CA, USA) for cryosectioning using a cryomicrotome (Leica, CM1900, Germany). Relatively well-preserved samples were selected for SEM imaging. Fluorescence staining and confocal laser scanning microscopy (CLSM) imaging Granule samples were collected, pretreated, and stained as described in our previous study (Zhou et al. 2013). Protein and β-polysaccharides in the samples were stained with fluorescein isothiocyanate (Sigma–Aldrich, USA) and calcofluor white (Sigma–Aldrich, USA), respectively. Confocal laser scanning microscopy (Olympus, FV1000, Japan) was employed to probe the internal cells and EPS distributions of the consortia. The consortia were imaged using a ×10 objective and analyzed with the equipped Zen software.

Lipid extraction and measurement Bacteria community analysis Algal total lipid was extracted using a modified version of the method described by Bligh and Dyer (Iverson et al. 2001). Briefly, 40 mg of dried biomass powder was mixed with 3.6 mL 5 % NaCl and then sonicated for 5 min, after which 6 ml of CHCl3: methanol (2:1, v/v) was added and vortexed well using a mix smart vortexer (Allsheng, Mix-100, China). Next, 2 ml of methanol was added and centrifuged at

Biomass samples were collected regularly during the culture periods, washed twice with PBS buffer, and then subjected to DNA extraction, polymerase chain reaction amplification, and denaturing gradient gel electrophoresis (DGGE) analysis. The DGGE profile and UPGMA dendrogram generated from the DGGE profile were subsequently analyzed by Quantity One

Appl Microbiol Biotechnol

4.6.2 (Bio-Rad), and the 16S ribosomal DNA (rDNA) sequences were compared with those available in the National Center for Biotechnology Information database by conducting BLAST searches (Zhou et al. 2014).

Results Bacteria-contaminated C. regularis granulation Microscopic observation A series of microscopic photographs (Fig. 1) shows the evolution of bacteria-contaminated C. regularis granules. These photographs were compared with those showing the natural aggregation of purified C. regularis. The inoculated biomass suspension was small, ranging from 2 to 6 μm (Fig. 1a, day 0). The results indicated that algal–bacterial flocs were formed within 1 day after inoculation (Fig. 1a, day 1), after which compact granules with a size of 500–650 μm and a clear and regular outer shape were common in the culture medium (Fig. 1a, day 4). To the best of our knowledge, this is the first report of such compact and regular algal–bacterial granules forming under heterotrophic cultivation conditions. However, the granules gradually broke following day 4, after which the algal–bacterial flocs with fluffy and irregular morphology were dominant (Fig. 1a, day 6). Conversely, granules in MB4.0 (Fig. 1b, day 6) were still dominant by then, implying that organic carbon abundance rather than sufficient nutrient

levels determined the stability of the granules. It should be noted that considerably large granules never appeared (Fig. 1c, day 6) during the cultivation of purified C. regularis, indicating that bacteria played a key role in granulation. Biomass concentrations increased during granulation, with the maximum concentrations reaching 1.2 g/L during the steady stage in MB 4 . 0 (Fig. 2) for both bacteriacontaminated C. regularis cultivation and purified C. regularis cultivation. This value was much higher than that observed in MB2.5 during the steady stage (0.4 g/L, Fig. 2). These differences were the result of promotion of cell growth by the high concentrations of organic carbon.

Settling ability of bacteria-contaminated C. regularis Consistent with the granule development results, bacteriacontaminated C. regular showed good settling ability in both MB2.5 and MB4.0 (Figure S1). A trend toward settling was apparent in the bacteria-contaminated C. regularis culture during the settling term. Most granules settled to the bottom of the 100-ml cylinder within 5 min, whereas the algal–bacterial flocs, which were less compact and irregularly shaped, settled to the bottom gradually during the following 10 min. It should be noted that aggregates grown in MB2.5 showed better settling ability than those grown in MB4.0. This was likely because the microalgae in MB4.0 had a higher negative surface charge, which impeded spontaneous agglomeration and settling (Valigore et al. 2012). Moreover, the greater biomass

Fig. 1 Microscopic observation of the granule evolution. Bacteria-contaminated C. regularis granulation in MB2.5 (a) and in MB4.0 (b) and purified C. regularis granulation in MB4.0 (c), Bar=200 μm

Appl Microbiol Biotechnol 1.2

transfer and supported microbial metabolism in the cores of the granules.

Biomass concentration (g/L)

1.0

0.8

EPS content and distribution

0.6

The polysaccharide and protein contents in the biomass of bacteria-contaminated C. regularis (in MB2.5 and MB4.0) and the purified C. regularis (in MB4.0) cultures are shown in Fig. 4. Polysaccharides were dominant in all cultures. The concentrations of polysaccharides generally stabilized at 60 and 50 mg/L in MB2.5 and MB4.0, respectively. Additionally, the polysaccharide concentrations of the purified C. regularis culture were basically over than 60 mg/L, but this value underwent a large fluctuation and decreased dramatically to 28 mg/L at day 4. The obvious dominance of polysaccharides over proteins in the EPS was opposite to a previous report of EPS secretion by microalgae (Sawayama et al. 1992; Zeng et al. 2013), mainly due to high C/N ratio in the study (Sheng et al. 2010). According to the evolution of research, concentration fluctuated slightly and granules were observed when contaminated by bacteria, which indicates bacteria are prerequisite for microalgae granulation while EPS were not the determinant factors. The spatial distribution of polysaccharides and protein in the granules was analyzed by a fluorescent staining method (Fig. 5). β-D-glucopyranose polysaccharide was clearly the main component of the EPS, being evenly distributed through the granule (Fig. 5b). However, protein was only distributed at the boundary of the granule. The EPS distribution of the bacteria-contaminated C. regularis granule was consistent with the polysaccharide and protein concentrations given in Fig. 4. EPS with a large proportion of polysaccharides is a bioglue, facilitating cell-to-cell interaction and further strengthening microbial structure in granules (Liu et al. 2004a). Furthermore, surface hydrophobicity is considered a triggering force for granulation (Liu et al. 2004b) and thus proteins surrounding the granule generate cell walls with high surface hydrophobicity. The distribution characteristics of EPS enhanced granulation and promoted the ability of granules to settle.

0.4

bacterial contaminated C. regularis in MB2.5

0.2

bacterial contaminated C. regularis in MB4.0 purified C. regularis in MB4.0

0.0 0

1

2

3

4

5

6

Time (day)

Fig. 2 The variation of biomass concentrations during cultivation

concentrations in the MB4.0 likely resulted in such differences. However, the purified microalgae culture always showed much poorer settling ability, in which few microalgal flocs settled to the bottom, whereas the majority were suspended for at least 15 min. Figure 3 shows the SEM images of the microcharacteristics of the bacteria-contaminated C. regularis granule (sampled from MB4.0 at day 4). The distribution density of microalgae and bacteria was almost uniform in the granule. The formation of a bacteria–microalgae consortium was previously reported to be based on the attachment of bacteria to algal cells, which formed a skeletal structure in the algal– bacterial flocs (Salim et al. 2011; Van Den Hende et al. 2011). However, the granulation in this study was quite different from that observed in previous investigations. Specifically, both microalgae and bacteria grew in clusters (Fig. 3c, d), implying that contaminated bacteria were potentially symbiotic colonies with C. regularis. Moreover, the numerical dominance of C. regularis was maintained, in part owing to limited nutrient because of the rapid bacterial growth. EPS was also found to be uniformly distributed inside of the granules and was considered to be the bridge connecting bacteria and microalgae. Furthermore, broad holes present among the clusters provided channels that promoted mass

Fig. 3 SEM images of the bacteria-contaminated C. regularis granules in MB4.0. Universal view of the granule (a), microstructure of the granule (b), microalgae cluster in the granule (c), bacteria cluster in the granule (d)

Appl Microbiol Biotechnol bacterial contaminated C. regularis in MB2.5 bacterial contaminated C. regularis in MB4.0

bacterial contaminated C. regularis in MB2.5 bacterial contaminated C. regularis in MB4.0

40

purified C. regularis in MB4.0

80

purified C. regularis in MB4.0

Protein (mg/L)

Polysaccharide (mg/L)

100

60

40

30

20

10

20

0 1

4

2

0

6

1

Time (day)

2

4

6

Time (day)

Fig. 4 Polysaccharides and protein contents in the biomass during the cultivations

Bacteria and C. regularis growth profile

COD and nitrogen removal

The competition between microalgae and the bacteria influences lipid production and is a key issue when developing methods of harvesting by granulation. In this study, C. regularis and the bacterial cells were counted during cultivation to evaluate competition (Fig. 6). In general, the growth trends of C. regularis were similar to the biomass concentrations during cultivation of the bacteria-contaminated C. regularis in both MB2.5 and MB4.0, and the bacterial concentrations appeared to be very stable during the 6-day cultivation period. These findings indicate that C. regularis induced the increase in biomass during co-cultivation, which subsequently favored lipid accumulation in the biomass. The concentration of C. regularis was slightly higher under purified cultivation, probably attributed to its greater initial amount than others.

To evaluate purification of synthetic wastewater, variations in COD and nitrogen in MB4.0 were evaluated during cultivation. Changes in phosphorus were not tracked owing to its low initial concentration. The concentration of COD decreased from 3900 to 311 mg/L during bacterial–algal co-cultivation in MB4.0, reaching a removal efficiency of 92 % by the end of the cultivation period (Fig. 7). The COD removal efficiency of the purified C. regularis and bacteria was 88 and 14 %, respectively, indicating that bacteria did not play a significant role when its growth was limited by nutrient starvation. These findings indicate that C. regularis, combined with bacteria, was very efficient at the removal and transformation of organic compounds. The positive effects of bacteria likely occurred through two pathways. Specifically, the bacteria degraded and then used a portion of the organic compounds, increasing the removal efficiency. Additionally, the bacteria decomposed

Fig. 5 CLSM images of bacteria-contaminated C. regularis granule in MB4.0 at day 3 (bar=100 μm). Optical microscopy photograph (a), CLSM image of β-D-glucopyranose polysaccharide (b), (blue, calcofluor white), and CLSM image of protein (c) (green, FITC)

Appl Microbiol Biotechnol

8.5

8.0

7.5

7.0

bacterial contaminated C. regularis in MB2.5 bacterial contaminated C. regularis in MB4.0 purified C. regularis in MB4.0

6.5 1

2

3

4

5

lg(bacterial cell numbers)

lg( Microalgae cell number)

8.5

8.0

7.5

bacterial contaminated C. regularis in MB2.5

7.0

bacterial contaminated C. regularis in MB4.0 6.5

6.0 1

6

2

3

Time (day)

4

5

6

Time (day)

Fig. 6 Count of bacteria and C. regularis during the cultivations

complex organics that were not readily available into smaller molecules. These smaller molecules could be used by algae (Zhang et al. 2012). The concentration of COD was stable from day 3, indicating that the residual organic matter could not be synthesized by cells or further decomposed owing to nutrient starvation, which was consistent with the results of biomass and cell concentrations shown in Fig. 6. However, the organic materials in the biomass continued to be transferred to lipids according to the lipid production results. Nitrogen evolution via bacterial and microalgal bioactivity occurred through a complex pathway. NO3−-N, which is known to be the best N source for microalgal growth, was the major nitrogen source in the BG11 medium (Isleten Hosoglu et al. 2012). Both bacteria and microalgae adapted to the environment gradually at the beginning of cultivation, and nitrification appeared to occur via nitrifying bacteria from day 1. By day 3, the nitrogen was almost completely

exhausted, resulting in starvation conditions. The total inorganic nitrogen (sum of NH4+-N, NO2−-N, and NO3−-N) removal efficiency reached 96 % at the end of the cultivation (MB4.0). Lipid accumulation Nutrient starvation is not only critical to cell growth, but is also directly relevant to lipid accumulation. There were similar trends observed in the growth and harvesting of biomass between bacterial–algal co-cultures grown in MB2.5 and MB4.0; however, higher lipid content (Fig. 8a) was present in MB4.0. This difference was attributed to the high initial concentration of organic carbon and therefore more serious nitrogen starvation. Overall, the lipid content increased to 0.31 and 0.34 (g/g) gradually at day 5 for MB2.5 and MB4.0, respectively, whereas the lipid content in MB4.0 was greater

5000

COD concentration (mg/L)

4000

3000

bacterial contaminated C. regularis purified C. regularis bacteria

2000

1000

0

0

1

2

3

Time (day)

4

5

6

NH+ 4 -N/ NO2-N/ NO3-N (mg/L)

60

a

b

50

+ NH4 -N NO2-N NO3-N

40 30 20 10 0 0

1

2

3

4

5

6

Time (day)

Fig. 7 Variation of COD during the cultivations of bacteria, purified C. regularis, and bacteria-contaminated C. regularis in MB4.0 (a); variation of nitrogen during the cultivations of contaminated C. regularis in MB4.0 (b)

Appl Microbiol Biotechnol

Lipid content (g/g)

0.3

bacterial contaminated C. regularis in MB2.5 bacterial contaminated C. regularis in MB4.0

a

purified C. regularis in MB4.0

0.07

0.2

0.1

bacterial contaminated C. regularis in MB2.5

b

bacterial contaminated C. regularis in MB4.0

Lipid productivity (g/(L d ))

0.4

0.06

purified C. regularis in MB4.0

0.05 0.04 0.03 0.02 0.01

0.0

3

4

5

6

0.00

Time (day)

3

4

Time (day)

5

6

Fig. 8 Lipid accumulation in the biomass during the cultivations lipid content (a); lipid productivity (b)

than that in MB2.5. The lipid accumulation then increased owing to the nutrient-limiting conditions during the stationary phase. The lipid content (0.31 g/g) in the MB2.5 culture was greater than that (0.23 g/g) in the MB4.0 culture after inoculation on day 4 owing to its rapidly undergoing nitrogen starvation relative to MB4.0. According to the results, the higher biomass in MB4.0 led to higher lipid productivity. Thus, the maximum lipid productivity of 0.057 g/(L·day) for the MB4.0 culture at day 5 was much higher than that in the MB2.5 culture (0.015 g/(L·day)). Overall, the productivity obtained in this study was higher than that reported in previous studies, and the bacteria appeared to have little effect on lipid accumulation (Lv et al. 2010). Thus, a sensible balance between biomass concentration and lipid accumulation was necessary to achieve a microalgae–bacteria consortium with high efficiency for wastewater treatment and lipid productivity. Bacterial community in granules The 16S rDNA gene sequence of the DGGE bands and their NCBI nucleic BLAST results are shown in Table 2 (DGGE band sequences are shown in the supplementary file). The majority of bacterial 16S rDNA sequences grouped with species that promote growth, enhance lipid accumulation, and contribute to flocculation. Sphingobacteriales bacterium (band 1) and Sphingobacterium sp. (band 4) are vital organisms involved in the flocculation of microalgae (Lee et al. 2013). The microalgal granulation observed in the present study clearly depended on these adhered bacteria, rather than the EPS secretion. e. Rhizobium sp. (bands 2 and 5) has been attributed to the promotion of microalgal growth (Lee et al. 2013), and its role in C. regularis domination could not be ruled out. It should be noted that Rhizobium sp. is known to enhance lipid productivity at the same time. Furthermore, the appearance of

Stenotrophomonas maltophilia (band 3) confirmed that autotrophic bacteria (nitrifier) could co-exist in the granules (Xu and Han 2013).

Discussion Mechanism of bacteria-contaminated microalgae granulation Harvest of microalgae is currently recognized as a challenge during microbial biofuel production. Co-culture of microalgae with bacteria under heterotrophic conditions has the potential to improve microalgae harvesting. This is because bacteria usually present as bioflocs, providing opportunities for microalgal aggregation. However, microalgae are susceptible to competition from bacteria, and the interaction between bacteria and microalgae is complex (Liu et al. 2012). Nevertheless, the bacterial strains evaluated in the present study formed a symbiotic relationship with the microalgae, playing a key role in bacteria–microalgae granulation. Chlorella are considered to be non-flocculating species owing to their low level of EPS excretion (Vandamme et al. 2012). However, the content of EPS excreted by purified C. regularis was similar to that generated by co-cultures. This was likely related to the limitation of bacterial growth because of nutrient starvation. It was interesting that such EPS levels did not induce the purified C. regularis harvesting, but a large amount of granules could be formed when contaminated with bacteria, and then biomass was efficiently harvested in only 5 min (Figure S1). Taken together, the above results indicate that EPS was not the sole and determinate factor influencing microalgae harvesting and that negative charges on the surface likely played a dominant role in the inhibition of harvesting/flocculation. The strong negative surface charge of C. regularis was weakened when bacteria adhered to the algae

Appl Microbiol Biotechnol Table 2

DGGE profile and bacteria identification of the selected bands

DGGE Bands profile 1 2 3

4

5

%

Closest relatives (accession no.)

Function

no. identity 1

94

Sphingobacteriales bacterium

2

99

Rhizobium sp.

3

99

Stenotrophomonas maltophilia (KJ784477)

Nitrification

4

99

Sphingobacterium sp. (EU580525)

Flocculation

5

100

Rhizobium sp. (KJ720681)

microalgae growth & lipid production

AB851173

microalgae growth & lipid production

KF551166

and granulation is expected to occur through a bridging flocculation mechanism; thus, the participation of bacteria in the microalgae cultivation enhanced the affinity among C. regularis. The distribution of bacteria and C. regularis clusters was basically uniform in the granules (Fig. 3), supporting the finding that algae and bacteria were in a symbiotic relationship. Algal exudates were the main carbon source for bacteria during the stationary phase, whereas algae benefited from the products of bacterial decomposition. However, this balance was subject to the activity of bacteria and microalgae, in which bacteria were more susceptible to nutrient starvation. As a result, bacterial cell numbers remained stable during co-cultivation, whereas the granules degraded at the end of the cultivation period in MB2.5 (Fig. 1). Overall, C. regularis could only be granulized if bacteria participated in the process during cultivation. C. regularis and the bacteria formed a symbiotic relationship under starvation conditions. However, to ensure stability of the formed granules, the cultivation time should be appropriately controlled to prevent over-starvation of the bacteria. Lipid productivity under starvation conditions Lipid productivity is useful in evaluation of biodiesel production capacity, which is determined by both biomass concentration and lipid content. It was previously assumed that bacterial contamination could either inhibit or stimulate algal biomass accumulation (Fergola et al. 2007; Zhang et al. 2012), but that negative effects of bacterial contamination on microalgae reproduction were more common due to the higher growth rate of bacteria. However, the effects of bacteria on algal growth were not significant under nitrogen starvation conditions in the present study (Fig. 6a). The increase of microalgae cell numbers during bacteria–algae co-cultivation indicated that the algal cell concentration was similar to that

Flocculation;

under purified cultivation and that bacterial growth was inhibited by starvation strategy. As shown in Fig. 6, algae were superior competitors to bacteria during nutrient starvation, resulting in promotion of their growth by enhanced organic concentrations. The bacterial concentration tended to be stable in the co-cultivation system, which was related to their nutrient starvation during cultivation. However, the C/N/ P was only 14/1.4/1 and 44/1.4/1 for MB2.5 and MB4.0, respectively. The C/N/P should at least be 10/2/1 for bacterial cell synthesis. For microalgae reproduction, the required N and P contents could vary elastically according to synthesis clues or to lipids and proteins (Espinosa Gonzalez et al. 2014). It is worth noting that lipids and carbohydrates are preferred storage products under various stress conditions (Siaut et al. 2011), including nutrient starvation. As shown in Fig. 8, evaluation of the relationship between nitrogen evolution and changes in COD and lipid accumulation revealed that lipid accumulation was affected by environmental stress and could be promoted by nitrogen deficiency (Gentili 2014; Zhu et al. 2013). The total inorganic nitrogen (sum of NH4+-N, NO2−-N and NO3−-N) was depleted to no more than 10 mg/L as a result of microbial synthesis and metabolism at day 4 for MB4.0 because COD was rapidly reduced during the exponential phase. Moreover, the lipid content increased during starvation in response to all three protocols (Fig. 8a), with the lipid contents of the bacteria–microalgae consortium in MB4.0 showing the greatest increase. These findings clearly demonstrate that nitrogen depletion initiates a shift in metabolic pattern that favors the accumulation of oils in the form of lipids in microalgal cells. Lipid storage also requires carbon skeletons; therefore, a dramatic reduction in COD can also occur after the accumulation and evolution of lipids by microalgae. Because of the enhanced biomass concentration in MB4.0, the consortia in MB4.0 had a high lipid productivity throughout the cultivation period (Fig. 8b).

Appl Microbiol Biotechnol

Nutrient starvation caused by a high C/N/P is a good strategy to weaken competition between microalgae and bacteria. It can also enhance lipid accumulation. In this study, this method led to simultaneous removal of pollutants and lipid accumulation in the coupling of wastewater treatment and microalgae cultivation technology, without sterilization. Insight into coupling of wastewater treatment and bacteria– microalgae co-cultivation In this study, a bacterial–algal consortium was found to have good ability to remove COD and nitrogen from wastewater. The effluent COD of the coupling technology met the requirements for high organic carbon-polluted wastewater treatment. Evolution of nitrogen tends to lead to the production of revenue-generating products such as biofuel instead of traditional nitrification and denitrification during the wastewater treatment process. Bacterial contamination is a primary cause of process failure during cultivation under heterotrophic conditions. To date, sterilization was believed to be the only method preventing process failure. However, the results of the present study indicate that appropriate starvation cultivation is an effective strategy for reducing bacterial growth while promoting the growth of microalgae. Many studies have recently been conducted to investigate bacteria–microalgae consortia, but most have only focused on the removal rate of pollutants and the settleability of microbial biomass while ignoring lipid accumulation and microalgae dominance (Valigore et al. 2012; Wang et al. 2013; Zhang et al. 2012). Moreover, few studies have investigated the various means of ensuring the dominant position of microalgae for lipid storage. The results of the present study indicated that bacterial growth can be controlled, biomass harvesting realized by settling alone, and lipid accumulation promoted when cultivation is driven by nutrient starvation. Overall, these findings indicate that high C/N/P wastewater could be another choice for microalgae cultivation in addition to the effluent of secondary wastewater treatment processes or sterilized industrial wastewater.

Acknowledgments The authors are grateful to the financial support from the Science and Technology Development Program of the Jilin province, China (20140101006JC) and the Graduate Innovation Fund of Jilin University (Project 2014103).

References APHA (1998) Wef. Standard methods for the examination of water and wastewater 20th ed Washington DC: American Public Health Association, American Water Work Association, Water Environment federation 252

Araujo GS, Matos LJ, Fernandes JO, Cartaxo SJ, Goncalves LR, Fernandes FA, Farias WR (2013) Extraction of lipids from microalgae by ultrasound application: prospection of the optimal extraction method. Ultrason Sonochem 20(1):95–98 Castrillo M, Lucas Salas L, Rodríguez Gil C, Martínez D (2013) High pH-induced flocculation–sedimentation and effect of supernatant reuse on growth rate and lipid productivity of Scenedesmus obliquus and Chlorella vulgaris. Bioresour Technol 128:324–329 Chen CY, Nace GW, Irwin PL (2003) A 6×6 drop plate method for simultaneous colony counting and MPN enumeration of Campylobacter jejuni, Listeria monocytogenes, and Escherichia coli. J Microbiol Methods 55(2):475–479 Chen Z, Zhang J, Lei X, Zhang B, Cai G, Zhang H, Li Y, Zheng W, Tian Y, Xu H (2014) Influence of plaque-forming bacterium, Rhodobacteraceae sp. on the growth of Chlorella vulgaris. Bioresour Technol 169:784–788 Cho S, Luong TT, Lee D, Oh YK, Lee T (2011) Reuse of effluent water from a municipal wastewater treatment plant in microalgae cultivation for biofuel production. Bioresour Technol 102(18):8639–8645 Clarens AF, Resurreccion EP, White MA, Colosi LM (2010) Environmental life cycle comparison of algae to other bioenergy feedstocks. Environ Sci Technol 44(5):1813–1819 Croft MT, Lawrence AD, Raux Deery E, Warren MJ, Smith AG (2005) Algae acquire vitamin B12 through a symbiotic relationship with bacteria. Nature 438(7064):90–93 de Bashan LE, Bashan Y, Moreno M, Lebsky VK, Bustillos JJ (2002) Increased pigment and lipid content, lipid variety, and cell and population size of the microalgae Chlorella spp. when coimmobilized in alginate beads with the microalgae-growthpromoting bacterium Azospirillum brasilense. Can J Microbiol 48(6):514–521 Dubois M, Gilles KA, Hamilton JK, Rebers P, Smith F (1956) Colorimetric method for determination of sugars and related substances. Anal Chem 28(3):350–356 Espinosa Gonzalez I, Parashar A, Bressler DC (2014) Heterotrophic growth and lipid accumulation of Chlorella protothecoides in whey permeate, a dairy by-product stream, for biofuel production. Bioresour Technol 155:170–176 Fergola P, Cerasuolo M, Pollio A, Pinto G, Della Greca M (2007) Allelopathy and competition between Chlorella vulgaris and Pseudokirchneriella subcapitata: experiments and mathematical model. Ecol Model 208(2):205–214 Frølund B, Palmgren R, Keiding K, Nielsen PH (1996) Extraction of extracellular polymers from activated sludge using a cation exchange resin. Water Res 30(8):1749–1758 Gentili FG (2014) Microalgal biomass and lipid production in mixed municipal, dairy, pulp and paper wastewater together with added flue gases. Bioresour Technol 169:27–32 Guo Z, Tong YW (2014) The interactions between Chlorella vulgaris and algal symbiotic bacteria under photoautotrophic and photoheterotrophic conditions. J Appl Phycol 26(3): 1483–1492 Isleten Hosoglu M, Gultepe I, Elibol M (2012) Optimization of carbon and nitrogen sources for biomass and lipid production by Chlorella saccharophila under heterotrophic conditions and development of Nile red fluorescence based method for quantification of its neutral lipid content. Biochem Eng J 61:11–19 Iverson SJ, Lang SL, Cooper MH (2001) Comparison of the Bligh and Dyer and Folch methods for total lipid determination in a broad range of marine tissue. Lipids 36(11):1283–1287 Lee DJ, Chen YY, Show KY, Whiteley CG, Tay JH (2010) Advances in aerobic granule formation and granule stability in the course of storage and reactor operation. Biotechnol Adv 28(6):919–934

Appl Microbiol Biotechnol Lee J, Cho DH, Ramanan R, Kim BH, Oh HM, Kim HS (2013) Microalgae-associated bacteria play a key role in the flocculation of Chlorella vulgaris. Bioresour Technol 131:195–201 Liu YQ, Liu Y, Tay JH (2004a) The effects of extracellular polymeric substances on the formation and stability of biogranules. Appl Microbiol Biol 65(2):143–148 Liu Y, Yang SF, Tay JH, Liu QS, Qin L, Li Y (2004b) Cell hydrophobicity is a triggering force of biogranulation. Enzyme Microbiol Tech 34(5):371–379 Liu H, Zhou Y, Xiao W, Ji L, Cao X, Song C (2012) Shifting nutrientmediated interactions between algae and bacteria in a microcosm: evidence from alkaline phosphatase assay. Microbiol Res 167(5): 292–298 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193(1):265–275 Lv JM, Cheng LH, Xu XH, Zhang L, Chen HL (2010) Enhanced lipid production of Chlorella vulgaris by adjustment of cultivation conditions. Bioresour Technol 101(17):6797–6804 McSwain B, Irvine R, Wilderer P (2004) The effect of intermittent feeding on aerobic granule structure. Water Sci Technol 49(11–12):19–25 More T, Yadav J, Yan S, Tyagi R, Surampalli R (2014) Extracellular polymeric substances of bacteria and their potential environmental applications. J Environ Manag 144:1–25 Pancha I, Chokshi K, George B, Ghosh T, Paliwal C, Maurya R, Mishra S (2014) Nitrogen stress triggered biochemical and morphological changes in the microalgae Scenedesmus sp. CCNM 1077. Bioresour Technol 156:146–154 Prathima Devi M, Venkata Subhash G, Venkata Mohan S (2012) Heterotrophic cultivation of mixed microalgae for lipid accumulation and wastewater treatment during sequential growth and starvation phases: effect of nutrient supplementation. Renew Energy 43: 276–283 Salim S, Bosma R, Vermuë MH, Wijffels RH (2011) Harvesting of microalgae by bio-flocculation. J Appl Phycol 23(5):849–855 Sawayama S, Minowa T, Dote Y, Yokoyama S (1992) Growth of the hydrocarbon-rich microalga Botryococcus braunii in secondarily treated sewage. Appl Microbiol Biol 38(1):135–138 Sheng GP, Yu HQ, Li XY (2010) Extracellular polymeric substances (EPS) of microbial aggregates in biological wastewater treatment systems: a review. Biotechnol Adv 28(6):882–894 Siaut M, Cuiné S, Cagnon C, Fessler B, Nguyen M, Carrier P, Beyly A, Beisson F, Triantaphylidès C, Li Beisson Y (2011) Oil accumulation in the model green alga Chlamydomonas reinhardtii: characterization, variability between common laboratory strains and relationship with starch reserves. BMC Biotechnol 11(1):7 Stephens E, Ross IL, King Z, Mussgnug JH, Kruse O, Posten C, Borowitzka MA, Hankamer B (2010) An economic and technical evaluation of microalgal biofuels. Nat Biotechnol 28(2):126–128

Su Y, Mennerich A, Urban B (2011) Municipal wastewater treatment and biomass accumulation with a wastewater-born and settleable algalbacterial culture. Water Res 45(11):3351–3358 Tay JH, Liu QS, Liu Y (2001) Microscopic observation of aerobic granulation in sequential aerobic sludge blanket reactor. J Appl Microbiol 91(1):168–175 Valigore JM, Gostomski PA, Wareham DG, O’Sullivan AD (2012) Effects of hydraulic and solids retention times on productivity and settleability of microbial (microalgal-bacterial) biomass grown on primary treated wastewater as a biofuel feedstock. Water Res 46(9): 2957–2964 van den Hende S, Vervaeren H, Desmet S, Boon N (2011) Bioflocculation of microalgae and bacteria combined with flue gas to improve sewage treatment. New Biotechnol 29(1):23–31 Vandamme D, Foubert I, Fraeye I, Meesschaert B, Muylaert K (2012) Flocculation of Chlorella vulgaris induced by high pH: role of magnesium and calcium and practical implications. Bioresour Technol 105:114–119 Wang H, Zhang W, Chen L, Wang J, Liu T (2013) The contamination and control of biological pollutants in mass cultivation of microalgae. Bioresour Technol 128:745–750 Xu CY, Han HJ (2013) Research on the treatment of coal chemical wastewater by activated sludge process with fixed biological media process and strains. Adv Mater Res 610:2006–2011 Zeng X, Danquah MK, Halim R, Yang S, Chen XD, Lu Y (2013) Comparative physicochemical analysis of suspended and immobilized cultivation of Chlorella sp. J Chem Technol Biol 88(2):247–254 Zhang Y, Su H, Zhong Y, Zhang C, Shen Z, Sang W, Yan G, Zhou X (2012) The effect of bacterial contamination on the heterotrophic cultivation of Chlorella pyrenoidosa in wastewater from the production of soybean products. Water Res 46(17):5509–5516 Zhou W, Min M, Li Y, Hu B, Ma X, Cheng Y, Liu Y, Chen P, Ruan R (2012) A hetero-photoautotrophic two-stage cultivation process to improve wastewater nutrient removal and enhance algal lipid accumulation. Bioresour Technol 110:448–455 Zhou D, Dong S, Gao L, Liu M, Niu S (2013) Distribution characteristics of extracellular polymeric substances and cells of aerobic granules cultivated in a continuous-flow airlift reactor. J Chem Technol Biol 88(5):942–947 Zhou D, Niu S, Xiong Y, Yang Y, Dong S (2014) Microbial selection pressure is not a prerequisite for granulation: dynamic granulation and microbial community study in a complete mixing bioreactor. Bioresour Technol 161:102–108 Zhu L, Wang Z, Shu Q, Takala J, Hiltunen E, Feng P, Yuan Z (2013) Nutrient removal and biodiesel production by integration of freshwater algae cultivation with piggery wastewater treatment. Water Res 47(13):4294–4302