Renewable and sustainable energy resources have received much renewed interest as a solution to the heavy reliance on fossil fuels since global petroleum supplies have diminished and serious environmental problems have arisen from greenhouse gas emissions. Recently, photosynthetic microorganisms have gained particular interest as a new source for biofuel feedstock because they are able to convert carbon dioxide (CO2) to a variety of potent biofuels [3, 7, 11]. In particular, microalgae have become an attractive candidate for liquid transport fuel production due to their higher photosynthetic efficiency and oil yield compared to terrestrial energy crops [12, 16].
However, there are still a number of technical barriers that must be overcome and one of the major challenges in algae-based biodiesel production is to reduce the cost of lipid extraction that accounts for up to 90% of the total energy consumption . Algal biodiesel is mainly obtained by transesterifying algal oil triglycerides with an alcohol in the presence of catalysts [13, 17]. Methanol is the most commonly used alcohol in this process due to its low cost , yet it is also responsible for 26% of the total energy consumption in biodiesel production . It has been reported that some freshwater and marine cyanobacteria could naturally produce alkanes such as pentadecane (C15H32) and heptadecane (C17H36) [25, 26]. The biological function of alkane in cyanobacteria is still unclear, but it was hypothesized that the hydrocarbons blended into the lipid bilayers may enhance permeability, flexibility, and fluidity against curvature and oxidative stresses . As petroleum-based diesel is made up of a mix of alkanes with 8-20 carbon atoms, these cyanobacteria-derived alkanes can be directly used as biodiesel without converting triglycerides to liquid hydrocarbons and thus may serve as a possible candidate to replace fossil fuels.
In this study, a filamentous cyanobacterium, Limnothrix sp. KNUA012 was axenically isolated from a summer bloom sample in Lake Hapcheon, Korea and its potential as biofuel feedstock was investigated.
Materials and Methods
Sample collection and isolation
Bloom samples were collected in September 2010 from Lake Hapcheon, Bongsan-ri, Bongsan-myeon, Hapcheon-gun, Gyeongsangnam-do (35° 37’N, 128° 02’E). Samples were then taken to the laboratory and 1 ml aliquots of these samples were inoculated into 100 ml BG-11 medium  with cycloheximide (Sigma, St. Louis, MO, USA) at a concentration of 250 μg ml−1. The flasks were incubated at 25℃ with shaking at 160 rpm on an orbital shaker (VS-202D, Vision Scientific, Bucheon, Korea) until cyanobacterial growth was apparent. Well-grown cyanobacterial cultures (1.5 ml) were centrifuged at 3,000 g for 15 min (Centrifuge 5424, Eppendorf, Hamburg, Germany). Resulting pellets were streaked onto BG-11 agar and filamentous growth was monitored daily. When emerging cyanobacterial filaments were macroscopically visible, they were aseptically transferred to fresh BG-11 plates to separate cyanobacteria from contaminating bacteria. Cyanobacterial filaments were then streaked onto R2A and LB agar plates (Becton, Dickinson and Company, Sparks, MD, USA) and incubated in the dark to check the axenic status of the culture for 14 days. The stain obtained in this study was deposited in the Korean Collection for Type Cultures (KCTC) under the accession number KCTC 12064BP.
Morphological and molecular identification
The isolate was grown in BG-11 medium for 20 days. Live cells were harvested by centrifugation at 3,000× g for 5 min, washed twice with sterile distilled water, and examined at 1,000× magnification under a Nikon Eclipse E100 Biological Microscope (Tokyo, Japan). For molecular analysis, genomic DNA was extracted using a DNeasy Plant Mini kit (Qiagen, Hilden, Germany). For the amplification of 16S rRNA gene fragments, the primer set, CYA106F and CYA781R(a) and CYA781R(b) described by Nübel et al.  was used. Due to the highly conserved nature of the 16S rRNA gene, three other genetic markers, the phycocyanin encoding operon intergenic spacer (PC-IGS), ribosomal 16S-23S intergenic spacer (ITS), and RuBisCO (ribulose-1,5-bisphosphate carboxylase/oxygenase) rbcLX, were employed. The PC-IGS was amplified using the primer pair, PCβF and PCαR specific for cyanobacteria . The ribosomal 16S-23S ITS region was evaluated using primers, 16S1407F and 23S30R  and region RuBisCO rbcLX was also amplified with primers, CW and CX, described by Rudi et al. , respectively. DNA sequences obtained in this study were deposited in the database of the National Center for Biotechnology Information (NCBI) under accession numbers JQ653272, KU297785, KU662322, and KU662323 (Table 1).
Table 1.aClosest cultured match: Geitlerinema unigranulatum BCCUSP94 (FJ545642)
Phylogenetic analysis using 16S rRNA gene
The sequence obtained from this study was compared to 16S rRNA gene sequences in the NCBI GenBank database using the BLASTN algorithm . Its closely related Limnothrix sequences were downloaded and aligned in the Molecular Evolutionary Genetics Analysis (MEGA) software, ver. 6.0 , with the ClustalW tool. The best-fit nucleotide substitution model (K2) was selected by means of MEGA ver. 6.0 on the basis of the Bayesian information criterion. This model was used to build a maximum likelihood (ML) phylogenetic tree with 1,000 bootstrap replicates . Synechococcus elongatus PCC 6301 (NR_074309) was used as an outgroup.
Late-exponential phase cultures of Limnothrix sp. KNUA 012 (1 ml each) were inoculated into BG-11 medium in triplicate and incubated for 20 days. Survival and growth of KNUA012 cells maintained at temperatures ranging from 5℃ to 25℃ (at intervals of 5℃) were examined to determine the optimum culture temperature. The cell biomass was separated from the bottom of the flask and thoroughly mixed by pipetting and samples taken from each time point were homogenized by sonification for 15 sec on an ultrasonic cell disruptor (Model 550, Fisher Scientific, Pittsburgh, PA, USA). Cyanobacterial cell density was determined by measuring the optical density (OD) of the cultures at 750 nm with an Optimizer 2120UV spectrophotometer (Mecasys, Daejeon, Korea).
Gas chromatography/mass spectrometry (GC/MS) analysis
The isolate was autotrophically grown in BG-11 medium for 20 days under its optimal conditions and cells were harvested for lipid analysis. The samples were freeze-dried to enhance the extraction efficiency. Then, 100 mg of each dried sample was blended with a mixture of chloroform:50% methanol (1:1) and extracted for 16 hr at 25℃. After extraction, the chloroform extract was isolated and dried in a rotary evaporator. The dried extract was treated with a pre-made solution of methanol and potassium hydroxide to facilitate transesterification. Next, 2 ml of hexane was added to the mixture to isolate the fatty acid methyl esters (FAMEs). The whole mixture was heated to 30℃ and stirred for 10 hrs. The mixture was then cooled and the methanol and hexane layers were separated. The yellow hexane layer was isolated for further analysis. GLC-90 (Supelco, Bellefonte, PA, USA) was used as external standard to calculate the FAME concentrations. The FAME composition was analyzed using a 6890N gas chromatograph (Agilent, Santa Clara, CA, USA) equipped with a 5973N mass selective detector and a HP-5MS capillary column (30 m ×0.25 mm ID ×0.25 μm film thickness). The initial oven temperature of the gas chromatograph was 120℃. It was maintained for 2 min, increased to 300℃ at a rate of 5℃ min−1, and held for 22 min. The injection volume was 1 μl with a split ratio of 20:1. Helium was used as carrier gas at a constant flow rate of 1 ml min−1. The mass spectrometer parameters were as follows: injector and source temperatures were 250℃ and 230℃, respectively, and the electron impact mode at an acceleration voltage of 70 eV was used for sample ionization, with an acquisition range from 50-550 m z−1. Wiley/NBS libraries were used as reference databases.
The remaining freeze-dried biomass samples were pulverized with a mortar and pestle and sieved through ASTM No. 230 mesh (opening = 63 μm). Ultimate analysis was conducted in order to determine the carbon (C), hydrogen (H), nitrogen (N), and sulfur (S) contents using a Flash 2000 elemental analyzer (Thermo Fisher Scientific, Milan, Italy). Higher heating value (HHV) was estimated by the following equation developed by Friedl et al. : [HHV = 3.55C2 − 232C−2,230H + 51.2C × H + 131N + 20,600 (MJ kg−1)].
Proximate analysis was carried out on a DTG-60A thermal analyzer (Shimadzu, Kyoto, Japan). Platinum pans were used to contain 30 mg of α-alumina (α-Al2O3) powder (Shimadzu, Kyoto, Japan) as a reference material and approximately 10 mg of each sample, respectively. Nitrogen (>99.999%, N2) was supplied as the carrier gas at a rate of 25 ml min−1 to protect the microalgae powder from oxidation. Samples were heated from 50℃ to 900℃ at a rate of 10℃ min−1. Thermogravimetric analysis (TGA) data were analyzed by ta60 Ver. 2.21 software (Shimadzu, Kyoto, Japan).
Phycocyanin (PC) concentration in the cyanobacterium was quantified by a slightly modified method of Silveira et al. . Briefly, 0.16 g of dried biomass in 2 ml of distilled water (at a biomass-solvent ratio of 0.08 g ml−1) was mixed with shaking at 100 rpm on an orbital shaker at 25℃ for 4 hr. Then, the tubes were centrifuged at 10,000 g for 15 min and the OD of the supernatant was measured at 615 and 652 nm. The PC concentration was calculated by the following equations: [PC (mg ml−1) = (OD615−0.474 × (OD652)) / 5.34]. The extraction yield was estimated by the following formula: [Yield (mg g−1) = PCC × V / DB], where V is the volume of solvent (ml) and DB is dried biomass (g).
Detection and identification of genes involved in alkane biosynthesis
Two primer sets were designed according to the fully sequenced Nostoc sp. PCC 7120 genome to amplify an acyl-acyl carrier protein (ACP) reductase and an aldehyde decarbonylase in Limnothrix sp. KNUA012. Partial fragments of both genes in Limnothrix sp. KNUA012 were successfully amplified by the primer sets, alr5284F and Lim_red_4R, and ADDP1F and ADDP6R, respectively. The downstream and upstream regions of the known sequences were further obtained using genome walking-PCR based on the manufacturer′s protocol (Clontech, Mountain View, CA, USA). Finally, the full length acyl-ACP reductase and aldehyde decarbonylase including open reading frames (ORFs) in Limnothrix sp. KNUA012 were amplified by the primer sets, KNUA012_RED_F and KNUA012_RED_R, and KNUA012_aldehyde_decarbonylase_F and KNUA012_aldehyde_decarbonylase_R, respectively. All the primers used in this study were designed by using Primer3Plus software  and listed in Table 2.
Table 2.List of the primers designed and used in this study
Results and Discussion
Identification of the strain KNUA012
As shown in Fig. 1, the cyanobacterium was filamentous with straight trichomes composed of cylindrical cells. Cells were 1-2 μm wide and 20-50 μm long and not attenuated towards ends. End cells were slightly rounded and gas vacuoles were observed near the cross walls and sometimes distributed within the cells. Overall, strain KNUA012′s morphological characteristics suggested that this isolate belonged to members of the genus Limnothrix.
Fig. 1.Light microscopy of Limnothrix sp. KNUA012 (scale bar represents 20 μm).
Molecular characterization inferred from sequence analysis of the genes for 16S rRNA also indicated that the isolate belonged to the Limnothrix group (Fig. 2). Therefore, the isolate was tentatively identified as Limnothrix sp. KNUA 012. However, the PC-IGS region sequence comparison revealed that uncultured cyanobacterium clone DC07PC-38 (HM020449) was the closest match for strain KNUA012 (Table 1). The 16S-23S ITS sequence analysis showed that the closest sequence match was Geitlerinema unigranulatum BCCUSP 94 (KJ735459) with a query coverage of only 82%. The rbcLX region sequence comparison also showed that uncultured Limnothrix planktonica KLL-C001 clone (KP 698043) was the closest match, but again the coverage was only 23%. This may be due to the lack of sequence data in GenBank for Limnothrix PC-IGS, 16S-23 ITS, and rbcLX genes, so no identification could be made with these data.
Fig. 2.The phylogenetic relationship of strain KNUA012 and its closely related species inferred from the 16S rRNA sequence data. The tree was generated by the maximum likelihood method with 1,000 bootstrap replicates. The scale bar represents a 1% difference in nucleotides sequences.
Optimal growth temperature
As shown in Table 3, strain KNUA012′s maximal growth was obtained at ambient temperatures (20 and 25℃). However, no growth was observed at lower temperatures (5 and 10℃).
Table 3.Growth of strain KNUA012 at various temperatures
GC/MS analysis of strain KNUA012
It was found that strain KNUA012 was able to autotrophically synthesize alkanes (CnH2n+2) such as pentadecane and heptadecane. Since C15H32 and C17H36 are 15- and 17-carbon alkane hydrocarbons known as the major components of petrodiesel , this cyanobacterium-derived alkanes can be directly used as a biodiesel component without having to convert triglycerides into liquid hydrocarbons. In addition, other algal biodiesel constituents, C14:1, C16:0, C16:1 were also produced by strain KNUA012. The GC/MS results are summarized in Table 4.
Table 4.Lipid profile of strain KNUA012
In proximate analysis by TGA, the moisture content (MC) is determined by the mass loss before 110℃ under N2 atmosphere, the volatile matter (VM) refers to the mass loss between 110 – 900℃ under N2 as a result of thermal decomposition, and the remaining mass represents fixed carbon (FC) and ash . The moisture, VM, and FC and ash contents of strain KNUA012 were 6.6%, 86.0%, and 7.4%, respectively. The VM is defined as the part of solid fuel that is driven-off as a gas by heating and typical biomass generally has a VM content of up to 80%(crop residue: 63-80%; wood: 72-78%). The VM content of the cyanobacteria used in this study was higher than those of wood-based biomass feedstocks.
Fig. 3.TGA profiles of Limnothrix sp. KNUA012. The mass change in percentage is on the y-axis and temperature (℃) is on the x-axis.
The HHV was also calculated to understand the potential of cyanobacterial biomass as a biofuel feedstock (Table 5). The result showed that the HHV was within the range of the terrestrial energy crops (17.0-20.0 MJ kg−1) . Given the higher photosynthetic efficiency and biomass productivity of microalgae , strain KNUA012 holds promise as a potential source for biomass feedstocks over crop plants. As high carbon content is a desirable property for fuel, if the higher concentration of CO2 in the medium is available, the higher HHV are possible.
Table 5.Proximate and ultimate analysis results of Limnothrix sp. KNUA012
Strain KNUA012 was also able to autotrophically produce 21.4 mg g−1 PC. Because PC occurs as the major phycobiliprotein in cyanobacteria, it has been extracted from cyanobacteria such as Spirulina platensis and widely used in cosmetic and food industries. Recent studies have demonstrated the hepatoprotective, anti-inflammatory, and antioxidant properties of PC [18, 27]. This makes the isolate also desirable for commercial applications in the pharmaceutical industry.
Alkane biosynthesis genes
Partial sequences of an acyl-ACP reductase and an aldehyde decarbonylase in Limnothrix sp. KNUA012 were successfully obtained by the primer sets, alr5284F and Lim_red_4R, and ADDP1F and ADDP6R, respectively (Table 2). The downstream and upstream regions of the known sequences were further obtained using genome walking-PCR. The translated amino acid sequence of the hypothetical acyl-ACP reductase in Limnothrix sp. KNUA012 was most closely related to the long-chain fatty acyl-ACP reductase (WP_035997292) [Leptolyngbya sp. JSC-1] with 73% identity. In case of the hypothetical aldehyde decarbonylase, the closest match turned out to be the long-chain fatty aldehyde decarbonylase (WP_017659189) [Geitlerinema sp. PCC 7105] with an identity of 73%. The results are presented in Table 6.
Table 6.Acyl-ACP reductase and aldehyde decarbonylase genes in Limnothrix sp. KNUA012
The alkane biosynthesis pathway in cyanobacteria was first described by Schirmer et al. . They demonstrated that fatty acyl-ACP was reduced to fatty aldehyde by a fatty acyl-ACP reductase and then converted into alkane by a fatty aldehyde decarbonylase, and it was found that both genes were required for alkane biosynthesis in cyanobacteria. Therefore, it seemed that these two hypothetical proteins in Limnothrix sp. KNUA012 may have resulted in the production of odd-chain alkanes (C15H32 and C17H36). However, it should be stated that the alkane yields from wild type cyanobacteria are still insufficient to compete with petroleum derived fuels [4, 10] and the over expression of alkane synthesis pathways is necessary [33, 34]. Future work could be required to see whether the gene set from strain KNUA012 would confer alkane production to heterologous hosts such as E. coli or yeast.
In conclusion, this Korean indigenous cyanobacterium could serve as potential biological resource to produce various compounds of biochemical interest. The real potential of the isolate described in this paper should be evaluated through further cultivation studies at molecular, laboratory, and field scales.