Effect of Limonene on the Heterotrophic Growth and Polyhydroxybutyrate Production By

Effect of limonene on the heterotrophic growth and polyhydroxybutyrate production by Cupriavidus necator H16

Authors
Guzman Lagunesa, F., Winterburna*, J.B.

a School of Chemical Engineering and Analytical Science, The Mill, The University of Manchester, Manchester, M13 9PL, UK

*Corresponding author:

Tel: +44(0)161 306 4891

Abstract

The inhibitory effect of limonene on polyhydroxybutyrate (PHB) production in Cupriavidus necatorH16 was studied. Firstly, results demonstrate the feasibility of using orange juicing waste (OJW) as a substrate for PHB production. An intracellular PHB content of 81.4 % (w/w) was attained for a total dry matter concentration of 9.58 g L−1, when the OJW medium was used. Later, a mineral medium designed to mimic the nutrient levels found in the complex medium derived from OJW was used to study the effect of limonene on the production of PHB. Results showed a drop in specific growth rate (μ) of more than 50% when the initial limonene concentration was 2% (v/v) compared to the limonene free medium. This work highlights the importance of a limonene recovery stage prior to fermentation, to maintain levels below 1 % (v/v) in the medium, adding value to the OJW and enhancing the fermentation process productivity.

Keywords:

Limonene; inhibition; Cupriavidus necator; PHB; orange peel.

1.Introduction

The wider uptake and utilisation of microbially produced biopolymers is dependent on our ability to efficiently and economically produce these polymers, ultimately to either give significant benefits in applications, as biomaterials, or to be available at a price comparable to oil derived polymers. In order to achieve either of these aims a low cost, a widely available source of carbon substrate is required, along with sufficient biorefining and bioprocessing techniques. Approximately 1.5 million tonnes of orange peel waste from the juicing industry are available every year as a source of fermentable fructose for PHB production (USDA, 2016).

The production of polyhydroxyalkanoates (PHAs) has been studied during the last few decades, as an alternative to petrochemical polymers. PHAs are a group of polyesters that can be synthesised by a range of different microbial strains as carbon and energy reservoir under stress conditions. Cupriavidus necatoris recognised as the model microorganism for PHA production due to its capacity of accumulate up to 80% of total dry weight in biopolymer and the simplicity of the metabolic pathway that involves three enzymatic reactions (Choi and Lee, 1999). PHAs not only have similar mechanical properties to polyethylene terephthalate (PET) or polypropylene (PP) (Koller et al., 2010; Lee, 1996), but they are also susceptible to biodegradation and can be produced from renewable raw materials thus reducing environmental impact and lowering petroleum dependency. However, production costs are still a major disadvantage to a wider use of PHAs and implementation of production strategies on a large scale (Braunegg et al., 2004; Lee and Choi, 1998; Sudesh et al., 2000).

Many studies have been performed in order to improve productivities and reducing production costs involving different fermentation strategies, microbial strains, upstream and downstream processing (Chanprateep, 2010; Koller and Braunegg, 2015; Wang et al., 2014). Moreover, the use of purified carbon sources accounts around 40% of the total cost of production leading to several investigation groups to find alternative culture media that can provide the conditions for the production of PHAs (Lee, 1996; Urtuvia et al., 2014). Numerous lignocellulosic materials have been studied as potential raw materials for the production of a medium rich in the nutrients necessary for fermentation processes (Jain and Tiwari, 2015). These materials are often obtained as by-products from other processes and are generally used as fuel for heat generation (Castilho et al., 2009). The approach of such studies is to recover the fermentable sugars, to be used for biopolymer production, through a pre-treatment step and then the solid residue can still be used as a heat source (Kawaguchi et al., 2016; Oh et al., 2015; Tripathi et al., 2011).

One of the main lignocellulosic materials worldwide is produced by the citrus processing industry, with orange juice being the main product (Angel Siles López et al., 2010; Boukroufa et al., 2014). Approximately 50 million tonnes of orange fruit are produced annually, with 3 million tonnes being used to produce fruit juice of which about 1.5 million tonnes of unwanted waste material (OJW) are produced every year.This waste includes the skin, pulp and seeds and accounts for about 50% (w/w) of the total amount of oranges processed for juice, making it a very wasteful process and representing a disposal challenge for the industries involved (USDA, 2016; Boukroufa et al., 2014; Choi et al., 2013). Different strategies for the valorisation this waste as a whole have been tested, from burning it as heat and power source to the use as hard metals sieve in water treatment (Balu et al., 2012; Bampidis and Robinson, 2006; Santos et al., 2015). The use of OJW as starting material for different biotechnological processes is currently being assessed; however, the complexity of the composition, and the presence of toxic substances, the process has not been successfully established (Pourbafrani et al., 2007). The use of an autohydrolysate of orange peel (Rivas et al., 2008) as complex medium rich in fructose for the production of PHAs by C. necator H16 represents a new approach for the use of this carbon-rich material that has obtained promising results. However, our preliminary studies revealed that even when fructose concentration was close to the optimum found for a mineral medium (Aramvash et al., 2015), a significant drop in the specific growth rate is triggered by increasing the concentration of OJW solids at the beginning of the media preparation process. This suggested that some inhibitory substance was accumulating in the medium.

OJW can contain up to 1.6 % (w/w) of orange essential oil (OEO), with important applications in several industries, including food, cosmetics and pharmaceutical. This essential oil accumulates in small oil sacs of 0.4 to 0.6 mm in diameter and is located at irregular depths in the flavedo at the outer peel of the fruit (Angel Siles López et al., 2010) and in addition to its characteristic smell it also has shown inhibitory effects on the growth of several pathogenic strains (Zahi et al., 2015) (Muthaiyan et al., 2012; Subramenium et al., 2015). Approximately 90% of the OEO consists of limonene, a naturally occurring monoterpene; consequently, studies on the antimicrobial effect of orange essential oil have focused on the limonene titration. According to literature, concentrations as low as 0.05% can inhibit cell growth for bioethanol production (Choi et al., 2013; Joshi et al., 2015). Furthermore, different approaches focused on the holistic implementation of citrus wastes have highlighted the importance of recovery of the OEO prior its biotechnological processing, enhancing the productivities of the microbiological stage, and adding value to the starting material(Lohrasbi et al., 2010; Ruiz and Flotats, 2014).

In this work, the biomass and PHB accumulation by C. necator H16 using an OJW autohydrolysed medium were evaluated.The effect of limonene on the strain’s growthkineticswas as well assessed with the objective todetermine its tolerance to the main component of OEO.

2.Materials and Methods

2.1.Microbial strain

Freeze dried C. necator H16, from the DSMZ-German Collection of Microorganisms and Cell Cultures, DSM No. 428, was purchased and activated according to supplier instructions. Master and working stock were created using MicroBankTMcryovial system (Pro-Lab Diagnostics, UK) and kept at the −80 °C. Short term storage plates consisting of nutrient agar (Sigma-Aldrich, UK) were prepared every time a batch of experiments was started.

2.2.Media preparation

2.2.1.Orange juicing waste medium

The feasibility of using OJW as starting material for the production of PHAs was assessed. OJW was obtained from a local juicing bar. Materialwas storedas received at −20° C until used. The process proposed by Rivas et al. (2008) (Rivas et al., 2008) to produce sugar rich medium from OJW was followed. After defrosting, the OJW was submitted to a drying stage at 60° C for 48h and then ground using a standard food processor.An extra run was performed using fresh material, whole and ground, to measure the effect of the drying stage over the carbohydrates extraction process. Two initial ratios of OJW solids to distilled water were used, 1:8 and 1:12(w:w). A hydrolysis stage was then performed using an autoclave where thetemperaturewas maintained at 121° C for 20 minutes. In order to determine the effect of the grinding stage on the sugar concentration in the medium, both options, ground OJW and whole OJW were tested during the extraction step. Solids were spun down using a centrifuge Sigma 6-16S (Sigma, Germany) at 7000 rpm and supernatant was separated by decantation. A 10M NaOH solution was used to adjust the initial pH of the media to a value of 7.0 ± 0.2. Finally, solutions were sterilised by filtration using 0.2 μmpolyethersulfone (PES) membrane filtration units (Thermo Fisher Scientific Inc., UK) and transferred to shake flasks for the fermentation experiments.

2.2.2.Mineral medium

The effect of limonene over the cell growth of C.necator H16 was studied adding different concentrations of the terpene to the mineral media developed by Aramvash et al. (2015) for the production of PHB. A basal mineral salt medium was prepared with the following composition: KH2PO4 1.75 g L−1; MgSO4.7H2O 1.2 g L−1; NH4CL 2 g L−1; citric acid 1.7 g L−1; trace elements solution10 ml L−1. The trace elements solution was composed of ZnSO4.7H2O 2.25 mg L−1; FeSO4.7H2O 10 mg L−1; CaCl.2H2O 2 mg L−1; Na2B4O7.7H2O 0.23 mg L−1; (NH4)6Mo7O24 0.1mgL−1; CuSO4.5H2O 1 mg L−1; MnSO4.5H2O 0.6 mg L−1; HCl (35%) 10 mL L−1. Fructose was used as the carbon source at a concentration of 25 g L−1. Salts, traceelements and fructose solutions were prepared separately. All solutions were autoclaved at 121 °C during 20 mins; once they reached room temperature, the tree solutions were mixed. The initial pH of the medium was adjusted to 6.8. Limonene (Thermo Fisher Scientific, UK) was filtered using a 0.2 μm PET membrane filter to assure sterility; the corresponding quantity was then added to the mineral media to reach the concentration required. Concentrations of 0, 0.5, 1, 1.5 and 2 % (v/v) of limonene were tested for this work.

2.3.Cultivation conditions and inoculum preparation

For every experiment, a single colony from the short term storage stock was taken, keeping aseptic conditions, and inoculated into 10 mL of nutrient broth No. 2 (Sigma-Aldrich, UK) contained in 50mL falcon tubes. Tubeswere placed on an orbital shaker where conditions were maintained at 30° C and 200 rpm. After 24h of cultivation, an adaptation stage was performed taking 2 mL of the broth to inoculate 20 ml of, either, OJW-based medium or limonene-free mineral media contained in 50 ml falcon tubes and cultivated for 48 h under the same conditions. Finally, for the limonene effect experiments, 10 ml of the limonene-free medium were used to inoculate 100 ml of the mineral media with limonene added, using 500 ml Erlenmeyer flasks. When working with the OJW-based medium 10 ml of the adaptation stage broth were used to inoculate 100 ml of identical medium in 500 ml shake flasks. All experiments were run in triplicate; results are presented as the mean, with error bars showing ± 1 standard deviation.

2.4.Analytical methods

2.4.1.Partial OJW characterization

Total carbohydrates, crude protein, crude fibre and water content measurements were carried out to the OJW in order to characterise the material. The phenol-sulphuric acid method described by Nielsen was used for total carbohydrate determination (Nielsen, 2010). Standard procedures 954.01, 962.09 described in the Official Methods of Analysis for the AOAC (AOAC, 1990) were followed for the determination of crude protein and fibre. A protein factor of 6.25 was used to calculate the protein content. Water content was determined by measuring the weight difference between fresh material and the material after dried. Samples of fresh OJW were located into a drying oven at 60°C during a period of 48 h.

2.4.2.Biomass measurements

Samples were taken periodically throughout the experiments and the cell density was evaluated by optical density measurements at a wavelength (λ) of 600 nm (OD600), using a spectrophotometer UVmini-1240 (Shimadzu,USA). The dry matter content of fermentation media was measured by transferring approximately 2 mL of cell containing broth into a pre-weighed 2 mL micro test tube (Eppendorf, DE), cells were then spun down at 13,000 rpm for 10 minutes using an Eppendorf MiniSpin centrifuge (Fisher Scientific, UK). The resulting supernatant was decanted and frozen to be used in residual nutrient determinations. The remaining cell pellet was washed twice using distilled water and then dried at 60C until constant weight was reached, 48 hours after. Residual biomass concentration was calculated by the subtraction of the PHB concentration from the total dry matter.

2.4.3.PHB determination

Gas chromatography (GC) was employed for PHB quantification according to the method developed by Riis and Mai (1988) (Riis and Mai, 1988). A gas chromatography system model 7820A (Agilent Technologies,USA) coupled with an autosampler Combi/Pal from Varian was used for this study. A Poraplot Q-HT 10×32 mm column was used and the detection system selected was a flame ionization detector (FID) set at 200° C. The injection volume and temperature were 1μL and 230° C respectively. Temperature program started at 120°C to be gradually increased during 3 minutes until 230° C, temperature was then held until finish the analysis. Helium was used as the carrier gas. A calibration curve was prepared using purified PHB as a standard (Sigma-Aldrich, UK) at different known concentrations. Peak areas of the samples were then correlated to concentration using the calibration curve obtained.

2.4.4.Carbohydrate measurement

The concentration of fructose, glucose and sucrose, in the supernatant collected from TDM samples, was determined using a Dionex Ultimate 3000 HPLC equipment. The refractive index intensity of the samples was measured using a RefractoMax 521 (ThermoFisher Scientific, UK) detector, set at 50 C, peak area and concentration were correlated using a calibration curve constructed by running standards of known concentration. An Aminex HPX-87C Column was used to achieve the separation at a temperature of 50 C. The mobile phase used was 5mM sulphuric acid at a flow rate of 0.6 mL min−1. Samples were diluted 10 times to assure a good column performance using HPLC grade water and filtered using nylon syringe filters 0.45 μm pore size prior analysis.

2.4.5.Total nitrogen measurement

Total nitrogen quantification was performed using a Shimadzu TOC-VC equipment coupled with both an ASI-Vautosampler unit and a TNM-1 total nitrogen detector. A calibration curve was created by the equipment, from a master solution of NH4Cl at a concentration of 50 mg L−1. An aliquot of 750 μL of free solids supernatant was taken to a final volume of 15 mL, required for the machine, and filtered through nylon syringe filters 0.45 μm pore before injection.

3.Results and discussion

3.1.OJW as starting material for PHA production

Results for the water content of the material show a solids content around 20 ± 0.6 % (w/w) for the OJW tested, this is similar to that observed by Pourbafrani et al.(2010), 20± 0.8 % of solids content, when working with citrus waste coming from a juice factory (Pourbafrani et al., 2010). The difference in the water content removed from the ground OJW and the whole “as juiced” material was around only 6.5 % (w/w) after 48 h of drying, leading to the decision of grinding the material after the drying stage. The composition of the OJW material was found to be similar to those reported in other studies on the valorisation of this by-product. Results for the total carbohydrates and crude fibre assays were 18.4 and 66. 31 %, respectively.This corresponds to those reported by Rivas et al. [29] a total soluble sugars content of 16.9 % and a crude protein of 63.05 %. The protein determination by Kjendhal digestion yielded a content of 7.22 %, slightly above the 6.50 % reported by Rivas et al.; this variation can be expected when analysing natural materials of different origin.

The concentrations of carbohydratesmeasured in the supernatants obtained for the different conditions tested are showed in table 1.The fructose concentration in the aqueous extracts was improved by almost 15 % comparing the whole material to the ground OJW, after the drying step. The maximum concentration of fructose obtained was 24.74 g L−1 when ground OJW was used. The drying strategy simplifies the handling of OJW, reducing the risk of microbial growth and, as the results show, concentrating the target compounds in the solid fraction. The results also confirmed the observations made in previous reports that have studied the effect of the particle size on hydrolysis and extraction processes for citrus by-products, namely that the grinding stage enhances carbohydrate recovery by increasing the surface area in contact with the aqueous fraction (Agbor et al., 2011; Choi et al., 2013; Lopresto et al., 2014). The process proposed in this contribution only focused on the effect of the grinding stage, not taking into account the resulting particle size. Nevertheless, previous studies focused on lignocellulosic materials show that reduction of particle size below 0.400 mm has little impact on the rates and yields of hydrolysis process (Agbor et al., 2011).

Three carbohydrate peaks were identified by HPLC analysis, glucose, fructose and sucrose; consumption over time was determined from the difference in the peak areas. Initial solids loading during media preparation led to a corresponding difference in the fructose extracted from the peels. The autohydrolysis treatment proved effective for fructose recovery, where Rivas et al. (2008) (Rivas et al., 2008) reported maximum concentrations of 16 g L−1 of fructose, this study obtained 23gL−1. Treatment with an initial ratio of orange peel of 1:12 (w) lead to an initial fructose concentration of 14 g L−1 and complete consumption was achieved after 72 h of fermentation. The depletion of fructose for treatments with ratio 1:8, initial fructose concentration of 23 g L−1, was not achieved for the frame time of the study, indicating that other nutrients were limiting.

Figure 1shows the cell growth curves as well as the PHB concentration time course for the extraction treatments studied. The specific growth rate value for the media with aninitial solids load of 1:12(w:v), (figure 1.b) reached the highest value for the different treatments studied, 0.18 h−1, with an intracellular PHB percentage above 80%. These values are similar to those obtained for C. necatorH16 when grown in a phosphate buffered medium, using organic acids as carbon source reaching a maximum PHA content of 83.7 % (Yang et al., 2010). Other efforts have focused on the implementation of glycerol as carbon source for PHA production, as this by-product of the biodiesel process is available in great quantities. In 2012, Tanadchangsaeng and Yu growing C. necator H16 in a mineral media added with 20 g L−1 of glycerol achieving a μmax of 0.11h−1 and 70% of PHB accumulation (Tanadchangsaeng and Yu, 2012). The treatment with an initial ratio of solids of 1:8 (figure 1.a) exhibited slower growth for a higher concentration of fructose, this can be related to some inhibitors present in the broth as result of the extraction process conditions (Mohan et al., 2015; Talebnia et al., 2007). A recent study showed that methanol contained in the crude glycerol can inhibit the cell growth of C.necator DSM4058, while a μmax of 0.47 h−1was obtained when 50 g L−1 of glycerol were used as carbon source in inhibition free conditions (Salakkam and Webb, 2015).