The production of poly(3-hydroxybutyrate) [P(3HB)] by a newly isolated Bacillus sp. ST1C using liquid waste from biodiesel production

Abstract

A newly isolated poly(3-hydroxybutyrate) [P(3HB)] producing strain, ST1C, was identified as Bacillus aryabhattai based on its morphological, biochemical and molecular characteristics. It synthesized and accumulated relatively high amounts of P(3HB). The aim of this work was to establish if it could convert an inexpensive liquid waste product from the production of biodiesel, biodiesel liquid waste (BLW), to P(3HB). Using a mineral salt medium (MSM) containing 2.0 % (v/v) glycerol present in the BLW and both normal batch and a draw and fill culture method, B. aryabhattai ST1C produced a maximum P(3HB) content and biomass concentration of 72.31 % dry cell weight (DCW) and 7.24 g/L, respectively, over a 24 h cultivation period in the draw and fill cultivation method. From 24 h to the end of cultivation at 72 h both the P(3HB) content and the biomass concentrations continuously reduced. Concentrations of glycerol in the BLW in this MSM above 3.0 % (v/v) or from pure glycerol (PG) or with an added NaCl concentration of greater than 3.0 % significantly reduced both the maximum P(3HB) content and the biomass concentrations.

Introduction

Polyhydroxyalkanoates or PHAs are considered to be strong candidates for producing biodegradable polymer materials because they possess many properties similar to various synthetic thermoplastics and elastomers (Din et al. 2008). In addition, upon disposal, they are completely degraded by microorganisms to water and carbon dioxide under aerobic conditions or to methane and water under anaerobic conditions (Volova et al. 2010). Of the large PHA family, a homopolymer of 3-hydroxybutyrate, poly(3-hydroxybutyrate) [P(3HB)], is the most widespread in nature and the best characterized PHA compound (Obruca et al. 2011). P(3HB) is synthesized by numerous bacteria as a reserve intracellular carbon and energy storage material, usually when an essential nutrient such as nitrogen or phosphorus is limiting growth in the presence of an excess source of metabolizable carbon (Sudesh et al. 2000; Merugu et al. 2012). The major problem that has limited the commercialization of P(3HB) as a substitute for conventional petrochemical-based plastics is its high production costs. Hence, much effort has been devoted to lowering the cost of P(3HB) by the isolation of more efficient bacterial strains and/or the development of less costly substrates, production processes and more economical recovery methods (Chee et al. 2010).

One economic evaluation showed that the production expenses for P(3HB) could be reduced by over a half if a renewable waste material and/or a by-product from industry was used as the main carbon substrate for P(3HB) biosynthesis (Mengmeng et al. 2009).

Biodiesel can now replace petroleum diesel as a fuel source as it is produced from renewable sources such as animal fats and vegetable oils, but this process generates about 10.0 % (w/w) glycerol as the main waste byproduct as it is the fatty acids present as glycerol esters that are the required biodiesel products. The excess biodiesel liquid waste (BLW) generated may become an environmental problem, when released as a waste material. One possible way to eliminate this BLW would be to refine the glycerol to a high purity before applying it in the drug, food, beverage, chemical and synthetic material industries (Yang et al. 2012). However, the refining of BLW may be a costly process depending on the economy of production scale and/or the availability of a glycerol purification facility (Thompson and He 2006). One possible application for this BLW would be to use it as a carbon and energy source for supporting microbial growth and their products in industrial microbiology (Silva et al. 2009).

There have been a number of attempts to convert BLW to PHA using: Pseudomonas oleovorans NRRL B-14682 and Pseudomonas corrugata 388 (Ashby et al. 2004), Cupriavidus necator JMP134, Paracoccus denitrificans and a number of different microorganisms (Mothes et al. 2007), Cupriavidus necator DSM 545 (Cavalheiro et al. 2009), Bacillus sonorensis, Halomonas hydrothermalis (Shrivastav et al. 2010), Halomonas sp. KM-1 (Kawata and Aiba 2010), Novosphingobium sp. THA_AIK7 (Teeka et al. 2010), Burkholderia cepacia ATCC 17759 (Zhu et al. 2010), Bacillus aryabhattai S4 (Tanamool and Kaewkannetra 2011). The most successful among these was in a small scale cultivation, in which Burkholderia cepacia ATCC 17759 accumulated a P(3HB) content of up to 81.9 % of its cell dry weight (5.8 g/L) that amounted to a 4.8 g/L P(3HB) concentration using a mineral salt medium (MSM) supplemented with 3.0 % (v/v) of a biodiesel-glycerol as a carbon source (Zhu et al. 2010). The aims of the work reported here were to detect, isolate and characterize bacteria that could convert BLW to relatively high amounts of PHA, and to optimize the culture and operating conditions for the most suitable isolate to produce the maximum PHA yields from the BLW and to determine the composition of the isolated PHA. This is the first report of the isolation of a B. aryabhattai strain able to efficiently utilize all the glycerol present in BLW prepared from waste cooking oil as a feedstock into a relatively high cell mass and P(3HB) production.

Materials and methods

Isolation of the organism

Bacteria were isolated by adding 1 g of soil into 100 mL of mineral salt medium (MSM) consisting in g/L: total glycerol in BLW, 20; (NH₄)₂SO₄, 1.0; KH₂PO_4, 2.0; Na₂HPO₄, 0.6; MgSO₄·7H₂O, 1.0; and trace element solution 1 mL containing in g/L: CaCl₂, 20; ZnSO₄·7H₂O, 1.3; FeSO₄·7H₂O, 0.2; (NH₄)₆Mo₇O₂₄·4H₂O, 0.6; and H₃BO₃, 0.6; (Kulpreecha et al. 2009). The pH in the medium was adjusted to 7.0 before sterilization. After 3 days at 30 °C and shaking at 200 rpm, 0.1 mL aliquots from each enrichment culture were plated onto MSM agar containing 2.0 % (v/v) of glycerol in the BLW. Approximately 0.25 mg of the hydrophobic dye, Nile Blue A stain powder (Sigma-Aldrich) was dissolved in 1 mL of dimethyl sulfoxide (DMSO) and this Nile Blue A solution was added to the MSM agar at a final concentration of 0.5 μg/mL to check for possible PHA deposits (Chaudhry et al. 2011). After 72 h incubation at 30 °C, the agar plates were observed under UV light. The colonies that produced strong fluorescence because of their PHA granules were purified. All selected colonies grew so the accumulation of dye had not interfered with their viability.

The colony that seemed to have the deepest red fluorescent color and, therefore, good PHA production was chosen for further work. Good P(3HB) production was further confirmed by checking for PHA granules using a phase contrast microscope and a fluorescence microscope after staining with Nile Blue A (Online Resource S1).

This new PHA accumulating bacterium was isolated from soil obtained from a dense forest in the Hala-Bala mountain range, Satun province in southern Thailand. The stock culture was kept on a nutrient agar (NA) slant at 4 °C and transferred to other new NA slants every 2 weeks.

Identification of the new isolate using microbiological methods and its 16S rRNA gene

The new PHA accumulating bacterium was first characterized using standard morphological and biochemical tests. From these tests it was identified as a Bacillus species and was named Bacillus sp. ST1C. It was then further identified as Bacillus aryabhattai ST1C using a full-length 16S rDNA sequencing method. The genomic DNA from the Bacillus sp. ST1C strain was prepared following the standard protocol described by Krueger et al. (2012). The 16S rRNA gene amplification was carried out using Taq polymerase with the forward primer: 5’-GAG TTT GAT CCT GGC TCA G-3’ and reverse primer: 5’-GTT ACC TTG TTA CGA CTT-3’. The amplification program employed was the DNA Engine Dyad® Thermal Cycler (Bio-Rad Laboratories) and comprised one cycle at 94 °C for 3 min, 25 cycles of denaturation at 94 °C for 1 min, annealing at 50 °C for 1 min and elongation at 72 °C for 2 min, followed by a final amplification step at 72 °C for 3 min. The amplification products were purified using the Qiagen PCR purification kit and its sequence was determined on an ABI Prism® 3730XL DNA Sequencer (Applied Biosystems, Foster City, California, USA). The 16S rRNA gene sequence analysis was carried out using the NCBI-BLAST (National Centre for Biotechnology Information http://www.ncbi.nml.nih.gov) program. The DNA sequences were determined and aligned for comparison with a program CLUSTAL X (version 1.8) in the BioEdit Program. Alignment gaps and unidentified bases were eliminated. Distance matrices for the aligned sequences were calculated using the Kimura’s two-parameter method (Kimura 1980; Zakaria et al. 2010). A phylogenetic tree of 16S rRNA genes was constructed by the neighbor-joining method of Saitou and Nei (Saitou and Nei 1987; Zakaria et al. 2010). The robustness for individual branches was estimated by 1,000 replication bootstrapping with the program MEGA Version 4.0.

Culture medium and conditions

Preparation of the pre-culture inoculum

Bacillus sp. ST1C was grown on an NA slant for 24 h then suspended in 0.85 % (w/v) sodium chloride to an optical density (660 nm) of 0.5. A 250 mL Erlenmeyer flask containing 50 mL of pre-culture medium was inoculated with 2 mL of cell suspension; Basal Culture Medium (BCM) (Kulpreecha et al. 2009) was adjusted to an initial pH of 7.0 and incubated on a rotary shaker at 200 rpm and 30 °C. This medium, analyzed by a CHNS-O analyzer (CHNS-O analyzer, CE Instruments Flash EA 1112 Series, Thermo Quest, Italy), had a C/N ratio of 4.7. This C/N ratio was probably suitable to promote more cell growth because of its relatively high nitrogen content.

Growth medium for PHA accumulation

A waste product from the manufacturing process for biodiesel production using waste cooking oil as feed stock was named ‘biodiesel liquid waste’ (BLW). This BLW was analyzed by a CHNS-O analyzer as above prior to use as a carbon source in the PHA production medium in order to determine its carbon and nitrogen content. The approximate glycerol content was measured by the free glycerol determination method as described by Bondioli and Della Bella (2005; Teeka et al. 2010).

The PHA production was carried out in shake flasks using the BLW or pure glycerol (PG from Sigma-Aldrich) as a sole carbon source. After the Bacillus sp. ST1C had been grown in BCM pre-culture medium until it reached mid exponential growth, 4.0 % (v/v) of the pre-grown inoculum was inoculated into a 250 mL Erlenmeyer flask containing 50 mL of MSM (Kulpreecha et al. 2009) supplemented with 2.0 % (v/v) of total glycerol in the BLW or 2.0 % (v/v) of PG with an initial pH adjusted to 7.0. Aerobic conditions were maintained by shaking at 30 °C and 200 rpm. Samples were taken every 12 h until 72 h of cultivation and then the biomass concentration, PHA content, and the glycerol utilization in the cell free culture medium were analyzed.

Effect of glycerol concentration in BLW on cell growth and P(3HB) production

The effect of the glycerol concentrations in BLW on the PHA production and growth of Bacillus sp. ST1C was performed in MSM production medium supplemented with various concentrations of the total glycerol in the BLW (0.5, 1.0, 2.0, 3.0, 4.0, and 5.0 % (v/v)). The initial pH of the medium was adjusted to 7.0 and cultivated under the same conditions as described above.

Effects of supplementing the initial growth medium with extra nutrients (draw and fill cultivation method) on cell growth and P(3HB) production

The first draw and fill cultivation was carried out using the same conditions as described above until 12 h of cultivation. The original working volume before feeding was 50 mL. Then 10 mL of MSM containing 2.0 % (v/v) total glycerol in the BLW with 1.0 g/L ammonium sulfate was added to the culture at 12 h and again at every 12 h interval until 60 h. At the same time 10 mL of culture was removed to promote the fresh medium to the cells and wash out some of the older cells from the culture medium. We called this system the draw and fill cultivation method because extra nutrients were supplied and the culture was removed every 12 h. This extra carbon and nitrogen might enhance and prolong the cell growth and P(3HB) accumulation. This did result in a small increase in cell dry weight and P(3HB) but it was not significant.

As there was a possibility that addition of the first extra substrate at 12 h may have exceeded the concentration of glycerol that resulted in some inhibition and perhaps the C/N ratio became less optimal so we changed the strategy. Other workers had also shown that biodiesel-derived crude glycerol at higher concentration of glycerol of 2.0 % (v/v) inhibited cell growth and PHA accumulation (Shrivastav et al. 2010; Wattanaphon and Pisutpaisal 2011; Sindhu et al. 2011). MSM containing 2.0 % (v/v) total glycerol in BLW was then used and the initial nitrogen content was increased to 1.25 g/L to promote a little more cell growth and then the MSM containing 0.5 % (v/v) total glycerol in the BLW and 0.2 g/L ammonium sulfate were added at 12 h and then at 12 h intervals until 60 h of cultivation was reached. The lower amounts of glycerol ensured that its inhibitory concentrations were not reached during the time P(3HB) was accumulating.

Analytical methods

Dry cell weight

The cell pellet obtained after centrifugation (7,155 × g at 4 °C for 20 min) was dried at 80 °C to a constant cell weight.

PHA content

Chloroform and acidified methanol (3.0 % (v/v) H₂SO₄) at 2 mL each were added into 20 mg of dried cells and then heated at 80 °C for 3.5 h. After cooling at room temperature, 2 mL of distilled water was added followed by vigorous shaking and centrifugation (1,006 × g for 10 min), the chloroform portion containing the PHA methyl esters was transferred to a vial for analysis by Gas Chromatography (GC). PHA was determined by the method described by Sun et al. (2009). Benzoic acid was used as the internal standard.

Glycerol utilization

The glycerol concentration was measured by the free glycerol determination method as described by Bondioli and Della Bella (2005; Teeka et al. 2010). The working reagents prepared for the experiment were 1.6 M acetic acid stock solution and 4.0 M ammonium acetate stock solution. The 0.2 M acetylacetone solution and 10 mM sodium periodate solution were prepared in a 1:1 ratio of acetic acid and ammonium acetate stock solution. The extraction solvent was prepared by mixing an equal volume of distilled water with 95 % ethanol. A mixture of hexane and extraction solvent was added to the sample. Upon mixing, two layers were formed and the lower layer was transferred into a new test tube. Then, 1.5 mL of the extraction solvent and 1.2 mL of sodium periodate were added. After that, 1.2 mL of acetylacetone solution was added and the mixture was incubated in a water bath at 70 °C for 1 min. An absorbance measurement was made at 410 nm against a blank sample.

Results and discussion

Identification of the bacterial isolate

Of the 190 isolates obtained, only 11 showed strong fluorescence from accumulated PHA granules. Among them, a Bacillus sp. ST1C isolate, seemed to have the most potential to synthesize and accumulate PHA because it showed the strongest fluorescent color under a fluorescence microscope and the largest PHA granules using phase contrast microscopy (Online Resource S1).

The morphological and physiological characteristics of the isolate were investigated (Online Resource S2). The isolate was grown with various carbohydrates and alcohols as the carbon source (API 50 CHB/20E Medium test kit, bioMerieux, France, and API ZYM test kit, bioMerieux, France). As a result of these observations, the new isolate was determined to be a Bacillus sp. ST1C (with 97.7 % identity to Bacillus megaterium). However, the biochemical tests alone do not provide enough data to make a totally accurate identification.

Further identification was then performed using 16S rDNA analysis. A full length 16S rDNA sequence of 1,482 bps was obtained by PCR. The BLASTX analysis revealed a 99.93 % identity to the sequence of the 16S rRNA gene of Bacillus aryabhattai B8W22^T (accession no. EF114313). Lineage was according to the phylogenetic tree generated (Fig. 1). The isolate was identified as a strain of Bacillus aryabhattai according to all the identification results. The isolate was then deposited in GenBank under the code name Bacillus aryabhattai ST1C (JX 524506).

Fig. 1

Neighbour-joining phylogenetic tree constructed on the basis of 16S rRNA gene sequences showing the phylogenetic relationships between Bacillus sp. ST1C and its close relationship to Bacillus aryabhattai B8W22^T (GenBank accession no. EF114313). The percentages of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) are shown next to the branches

Bacillus aryabhattai sp. nov. was first isolated in 2009 by Shivaji et al. from the upper atmosphere (Shivaji et al. 2009). Tanamool and Kaewkannetra (2011) were the first to report that another isolate of Bacillus aryabhattai S4 from a sugarcane plantation soil produced PHA with a maximum content of 57.62 % of its dry weight and productivity of 0.097 g/L·h when cultivated in 20 g/L of initial total sugar present in a sweet sorghum juice at pH 7.0, and 35 °C.

Comparison of the 16S rDNA of our isolate STIC and the standard strain (Bacillus aryabhattai B8W22^T) showed a 99.93 % similarity which was higher than that of Bacillus aryabhattai S4 that possessed only a 99.7 % similarity (Tanamool and Kaewkannetra 2011).

Analysis of BLW and its effect on cell growth and P(3HB) production

Before using this BLW as a carbon source for P(3HB) production, the BLW was first analyzed by a CHNS-O analyzer in order to determine the carbon and nitrogen contents (Table 1). The carbon and nitrogen element averages were 43.65 % and 0.15 %, respectively and the average glycerol content in the initial BLW sample was 43.16 % (w/v). In general, glycerol was the most single abundant component in the BLW (data from the specialized R & D center for alternative energy from palm oil and oil crops, Faculty of Engineering, Prince of Songkla University, Online Resource S3). From this result it was concluded that this BLW was suitable for use as a carbon source for P(3HB) biosynthesis because of its high carbon and glycerol content whereas it had a very low nitrogen content. The C/N ratio of the MSM containing 2.0 % (v/v) total glycerol in the BLW used in the culture medium was 20. Moreover, the glycerol was likely to be an energetically favorable substrate for the formation of acetyl-CoA (the precursor for P(3HB) synthesis). This can be seen in the stoichiometry of the conversion of glycerol to acetyl CoA (Ashby et al. 2004):

Table 1 The percentages of carbon and nitrogen elements and glycerol content in BLW collected at different times

$$ \mathrm{Glycerol}+3{\mathrm{NAD}}^{+}+\mathrm{ADP}+\mathrm{CoA}\to \mathrm{Acetyl}-\mathrm{CoA}+3\mathrm{NADH}+3{\mathrm{H}}^{+}+\mathrm{ATP}+{\mathrm{CO}}_2+{\mathrm{H}}_2\mathrm{O} $$

The production of PHA by Bacillus aryabhattai ST1C from 2.0 % (v/v) total glycerol in BLW was investigated (Fig. 2).

Fig. 2

Time profiles of cell growth, P(3HB) production and glycerol utilization by Bacillus aryabhattai ST1C grown in MSM containing 2.0 % (v/v) total glycerol in BLW as a sole carbon source. Each data point represents a mean value of three independent experiments and a vertical bar represents the standard deviation

During the period of exponential growth, accumulation of P(3HB) increased gradually and reached a maximum dry cell weight (DCW) and P(3HB) content of 5.68 g/L and 57.76 % DCW, respectively, at 24 h of cultivation simultaneously with the almost complete utilization of glycerol (the most abundant component in BLW) as indicated by the reduction of the glycerol concentration to 0.75 g/L. After 36 h the accumulation of P(3HB) significantly decreased to 40.77 % DCW due presumably to the absence of glycerol and no alternative readily metabolizable carbon compound; therefore, cells started to degrade the intracellular PHA storage for cell maintenance purposes whereas the DCW remained fairly constant to 24 h then decreased to 3.59 g/L at 72 h. In addition, some sporulation was observed but only after 24 h so perhaps some P(3HB) was being used for the production of the spores (data not shown). GC analysis of the polymer produced from this B. aryabhattai ST1C showed that only the 3HB monomer unit was detected (Online Resource S4).

Effect of the glycerol concentration in BLW on cell growth and P(3HB) production using normal batch fermentation

The effect of the glycerol concentration in BLW on cell growth and P(3HB) accumulation were investigated at various concentrations of MSM containing total glycerol in the BLW that ranged from 0.5 % (v/v) to 5.0 % (v/v). Both cell growth and P(3HB) production were expected to increase with an increasing amount of glycerol. However, at concentrations above 3.0 % (v/v) glycerol, the amount of the dry cell weight was reduced with a significant reduction of PHA productivity (Table 2).

Table 2 Maximum cell growth, P(3HB) accumulation and P(3HB) productivity produced from Bacillus aryabhattai ST1C at various concentrations of glycerol in BLW using normal batch fermentation

In comparison to other work the P(3HB) content of B. aryabhattai ST1C (57.76 % DCW) was similar to that of Bacillus aryabhattai S4 (57.62 % DCW). However, the cell mass of 5.68 g/L DCW and productivity of 0.137 g/L·h when using 2.0 % (v/v) of total glycerol in BLW was significantly higher than for B. aryabhattai S4 cultivated with 20 g/L of total sugar in a sweet sorghum juice that produced only 3.02 g/L DCW with a productivity of 0.097 g/L·h (Tanamool and Kaewkannetra 2011). This indicated that the glycerol in BLW was a good substrate for B. aryabhattai ST1C to grow and produce P(3HB) (Ashby et al. 2004). However, a further increase of the total glycerol concentration in BLW to 3.0 % (v/v) produced a reduced cell growth and P(3HB) production (Table 2). The cell biomass and P(3HB) biosynthesis decreased to 2.05 g/L and 46.58 % DCW, respectively, with a PHA productivity of only 0.029 g/L·h that was produced when B. aryabhattai ST1C was cultivated in 5.0 % (v/v) glycerol. A similar observation had been made with Bacillus sonorenis SM-P-1S that grew less on plates with 5.0 % (v/v) of a Jatropha biodiesel byproduct but showed a luxuriant growth on plates containing 1.0 % and 2.0 % (v/v) of the same biodiesel byproduct (Shrivastav et al. 2010) whereas in Bacillus sphaericus NII 0838 the maximum P(3HB) yield was at a 1.0 % (v/v) glycerol concentration (Sindhu et al. 2011). The growth of Pseudomonas oleovorans was not affected by increasing concentrations of a co-product stream from a soy-based biodiesel production (CSBP) but the increased concentration did produce a 100 % increase in the yield of polymer (from 0.2 g/L at 1.0 % (w/v) CSBP to 0.4 g/L at 5.0 % (w/v) CSBP). In contrast, when the CSBP concentration was increased for Pseudomonas corrugata from 1.0 % to 5.0 % (w/v), the cell growth decreased from 2.1 g/L to 1.7 g/L but the polymer yields stabilized at 0.7 g/L with an increase of the initial CSBP media concentration from 2.0 % to 5.0 % (w/v) (Ashby et al. 2004).

One possible reason for the reduction of cell biomass and P(3HB) biosynthesis when using the biodiesel liquid waste as the main carbon source was an increase in sodium ions (a catalyst in the de-esteraification process). From a previous report, crude glycerol from biodiesel production was contaminated with salts, primarily sodium, at approximately 2–3 % (Hansen et al. 2009). Also, sodium was found to have a particularly adverse effect on both the growth rate and polymer yield due to osmoregulation (Cavalheiro et al. 2009). These cell growth and polymer yield results have indicated that the controls on PHA synthesis vary and are species/even strain specific.

To study the effect of NaCl on cell growth and P(3HB) production, 0.5 % and 3.0 % of NaCl were added into MSM containing 2.0 % (v/v) PG. The sodium content in the BLW used in this study was determined to be only 0.5 % by ICP-OES analysis. The growth and P(3HB) content in both the BLW and PG with added 0.5 % NaCl were almost identical with a slightly higher yield with the PG (Fig. 3). From this it was inferred that there was no inhibitory effect of BLW containing 0.5 % sodium ion on cell growth and P(3HB) production. In contrast, MSM with PG and added 3.0 % NaCl had significantly less growth and polymer synthesis than the others and produced only 9.26%DCW of PHA produced and 1.96 g/L of biomass at the end of cultivation (Fig. 3) hence in that case 3.0 % NaCl was inhibitory.

Fig. 3

a Cell growth and b P(3HB) production by Bacillus aryabhattai ST1C when cultivated in MSM containing 2.0 % (v/v) PG, PG adding 0.5 % and 3.0 % of NaCl and 2.0 % (v/v) total glycerol in BLW. Each data point represents a mean value of three independent experiments and a vertical bar represents the standard deviation

Glycerol in some bacteria serves to function as an intracellular osmolyte for balancing external osmotic pressure. It plays important roles in physiological processes such as combating osmotic stress, managing cytosolic phosphate levels, and maintaining the NAD+/NADH redox balance. In spite of this at high concentrations it could suppress cellular metabolism, and decrease the production of PHA. Some mixed bacterial cultures are capable of growth in glycerol concentration up to 50 % and only a further increase above 50 % started to inhibit cell growth with no growth at glycerol concentrations of 60 and 70 % (Wattanaphon and Pisutpaisal 2011). However, this is not the normal response of bacteria.

Pseudomonas corrugate cultures grew to high cell densities in media with glycerol concentration only up to 2.0 % w/v glycerol. This is similar to the results reported here for Bacillus aryabhattai ST1C. It can be assumed that when the concentration of glycerol in the medium was increased above 2.0 % v/v the organisms began to feel the effects of osmotic stress (as evidenced by lower biomass yields) that also decreased polymer production (Ashby et al. 2005). This assumption was further confirmed by determining the reduction of the carbon content in MSM containing 3.0 % (v/v) of total glycerol in BLW during the fermentation process. The percentage of total carbon content in the culture medium with 3.0 % (v/v) glycerol cultures decreased from 1.98 % at the initial time (0 h) to 1.71 % at the end of cultivation (72 h) at the same time the biomass and P(3HB) content continuously decreased from the maximum of 4.35 g/L and 52.40 % DCW at 24 h to 1.41 g/L and 31.38 % DCW at 72 h, respectively. It is interesting that the total carbon content in this culture medium was still in the range of 1.7–1.8 % after 24 h whereas the DCW and P(3HB) production steadily declined. It would seem that this new Bacillus sp. ST1C was not able to readily utilize glycerol at 3.0 % (v/v), hence the accumulated P(3HB) was being slowly utilized. The glycerol content also remained constant in the range of 5–7 g/L after 24 h. In contrast at 2.0 % (v/v) total glycerol in the BLW, the percentage of the total carbon content of the medium decreased from 1.42 % at the initial time to 1.05 % at 24 h and then to 0.25 % at the end of cultivation and the glycerol content decreased from 20 g/L to less than 0.75 g/L at 24 h when the highest DCW and P(3HB) was achieved. The remaining glycerol was then completely consumed and the cells started to degrade P(3HB) for cell maintenance and survival. These results support the above assumption that the new Bacillus sp. ST1C efficiently metabolized the glycerol at 2.0 % (v/v). However, the further increase of glycerol up to 3.0 % (v/v) lowered the biomass and P(3HB) content (Fig. 4). More work would be required to identify the cause of this effect of the higher glycerol content as it inhibits glycerol utilization, cell growth and PHA accumulation.

Fig. 4

Comparison of a cell growth, b P(3HB) production and substrate utilization by Bacillus aryabhattai ST1C when cultivated in MSM containing 2.0 % (v/v) and 3.0 % (v/v) total glycerol in BLW. Each data point represents a mean value of three independent experiments and a vertical bar represents the standard deviation

This, however, did raise the question about what might happen to the growth and P(3HB) synthesis if one started the culture at 2.0 % v/v glycerol concentration then added more glycerol with or without nitrogen when the initial concentration had significantly reduced. Other workers have shown that it was possible to increase the amount of growth and PHA by feeding a fresh culture medium into a typical batch fermentation (Sabra and Abou-zeid 2008; Ibrahim and Steinbüchel 2009; Kulpreecha et al. 2009; Pandian et al. 2010; Obruca et al. 2011). When the normal medium, i.e. 2.0 % (v/v) total glycerol in BLW with 1.0 g/L ammonium sulfate was added to the culture 12 h after normal culture conditions then at 12 h intervals until 60 h, the dry cell weight at 24 h increased from 5.68 to 6.22 g/L and the P(3HB) content from 57.76 to 64.44 % DCW, so more P(3HB) was being produced but it was not a huge gain. However, when MSM containing 0.5 % (v/v) total glycerol in BLW and 0.2 g/L ammonium sulfate was fed 12 h after growth in the normal medium then at 12 h intervals, the cell growth and P(3HB) content at 24 h significantly increased to 7.24 g/L and 72.31 % DCW, respectively, with productivity of 0.216 g/L·h (Fig. 5). From these results it can be concluded that the glycerol in BLW and the nitrogen concentration both affected cell growth and P(3HB) accumulation. Therefore, addition of glycerol concentration at 0.5 % v/v and a little more nitrogen led to an increase in the biomass and P(3HB) content. The maximum cell growth and P(3HB) production achieved from the normal batch fermentation and the draw and fill cultivation method are compared in Table 3. Further work will be required to find the optimum conditions for P(3HB) production in scale up conditions but a yield of P(3HB) from utilizing glycerol of more than 70 % DCW in a 24 h period is probably worth further experimentation. In the normal batch cultivation at 72 h the dry cell weight reduced to 3.85 g/L and the PHA to 25.73 %, whereas in the draw and fill cultivation method with 0.5 % v/v glycerol and 0.2 g/L ammonium sulfate the dry cell weight and PHA content decreased to 5.89 g/L and 53.71 % DCW, respectively. For maximum production of cells and P(3HB) cultivation a time somewhere between 24 and 36 h is probably optimum.

Fig. 5

Comparison of a cell growth and b P(3HB) production by Bacillus aryabhattai ST1C when cultivated in normal batch fermentation: MSM containing 2.0 % (v/v) total glycerol in BLW without extra feeding and a draw and fill fermentation (N-limitation): MSM supplemented with 2.0 % (v/v) total glycerol in BLW containing 1.0 g/L ammonium sulfate or 0.2 g/L ammonium sulfate was fed every 12 h for 72 h. Each data point represents a mean value of three independent experiments and a vertical bar represents the standard deviation

Table 3 Comparison of the maximum biomass and P(3HB) production by different fermentation methods

In conclusion, a newly isolated Bacillus aryabhattai ST1C showed a high efficiency for production of P(3HB) from glycerol present in BLW, a waste product from biodiesel production. However, glycerol accounted for only 43 % of the total carbon in the waste and there was little evidence that the remaining carbon was being utilized (Figs. 4 and 5) for PHA production or growth. So, although the BLW is a suitable cheap and effective source of glycerol for PHA production, there would still be the problem of what to do with the remaining carbon. It would be of interest to determine if the remaining carbon could be utilized by other bacteria even perhaps to supplement the production of PHA.