Production of Poly(3-Hydroxybutyrate) by Haloarcula, Halorubrum, and Natrinema Haloarchaeal Genera Using Starch as a Carbon Source

Microbial production of bioplastics, derived from poly(3-hydroxybutyrate) (PHB), have provided a promising alternative towards plastic pollution. Compared to other extremophiles, halophilic archaea are considered as cell factories for PHB production by using renewable, inexpensive carbon sources, thus decreasing the fermentation cost. This study is aimed at screening 33 halophilic archaea isolated from three enrichment cultures from Tunisian hypersaline lake, Chott El Jerid, using starch as the sole carbon source by Nile Red/Sudan Black staining and further confirmed by PCR amplification of phaC and phaE polymerase genes. 14 isolates have been recognized as positive candidates for PHA production and detected during both seasons. The identification of these strains through 16S rRNA gene analyses showed their affiliation to Halorubrum, Natrinema, and Haloarcula genera. Among them, three PHB-producing strains, CEJ34-14, CEJ5-14, and CEJ48-10, related to Halorubrum chaoviator, Natrinema pallidum, and Haloarcula tradensis were found to be the best ones reaching values of 9.25, 7.11, and 1.42% of cell dry weight (CDW), respectively. Our findings highlighted that Halorubrum, Natrinema, and Haloarcula genera were promising candidates for PHB production using soluble starch as a carbon source under high salinity (250 g L−1 NaCl).


Introduction
Plastic is a highly useful material, and its production is growing. Million tons of nondegradable plastics end up in our natural environment every year affecting our health, wildlife, terrestrial, and marine habitats [1]. For this reason, polyhydroxyalkanoates (PHAs) are biodegradable and biocompatible polymers which have been promoted as an alternative to conventional oil-based plastics [2,3]. PHAs are synthesized by a wide variety of bacteria and archaea from various carbon sources and served as intracellular storage compounds to survive under unbalanced conditions [4,5]. Extremely halophilic archaea, inhabiting hypersaline environments containing high salt concentrations, are preferred for various potential applications, including the synthesis of polyhydroxyalkanoates. Poly(3-hydroxybutyrate) (PHB) was the PHB by archaeal strains were obtained using glucose as the sole carbon source [9,12,13] than with other substrates (wastes and pretreated vinasse) [14,15]. However, few studies based on the exploration of PHB production by archaeal species using starch have been reported, despite its wide availability [16,17]. Therefore, it was important to select strains able to use starchy substrates for PHB biosynthesis.
Chott El Jerid, the largest Salt Lake, is located in the south of Tunisia. Furthermore, this lake is the biggest one in the north of Africa (5360 km 2 ) with a salt concentration above 33% NaCl [18]. It has a thalassohaline salt composition, despite its continental origin. It may be flooded in the winter and evaporates to a desert in the dry season. These climatic conditions make the Chott an ephemeral extreme environment. Our previous studies described microbial diversity in Chott El Jerid during wet and dry seasons, using culture and molecular methods. Halophilic anaerobic fermentative bacterial strains were isolated from surface sediments [19,20]. Culture-independent techniques, targeting 16S rRNA and functional markers encoding the dissimilatory sulfite reductase beta-subunit gene (dsrB), methyl coenzyme M reductase (mcrA), showed abundant and diverse prokaryotic communities, sulfate-reducing bacteria (SRB), and methanogens, respectively [20,21]. Additionally, the isolation of halophilic aerobic bacterial and archaeal strains producing extracellular hydrolases have been achieved in Chott El Jerid [22]. Because Archaea outnumbered Bacteria in the studied samples, our purpose, here, is to expand the possibility of PHA synthesis by Archaea from Chott El Jerid. The objectives of this research were (1) enrichment, isolation, and screening PHA-accumulating halophilic archaea in water or water/sediment mix samples collected from Chott El Jerid in both seasons using phenotypic and molecular methods; (2) identification and characterization of PHA-producing isolates, utilizing starch as the carbon source; and (3) identification and quantification of the polymer produced.

Materials and Methods
2.1. Sample Collection. The samples (S1-10) and (S6M-14; S6W-14) were collected in dry (October 2010) and wet seasons (January 2014) from a continental ephemeral lake, Chott El Jerid, respectively. Hypersaline water or water/sediment mix samples were collected at different locations, approximately 0-10 cm from the surface. The samples were collected in sterile bottles, brought to the laboratory within three hours, and kept aseptically at 4°C until analyses. The environmental parameters of sampling sites are listed in Table 1 as previously reported [20]. In order to isolate PHA-producing haloarchaea, samples were serially diluted and 100 μL of each dilution were plated onto agar medium (20 g L -1 ) as described above. Following incubation, 33 colonies from the plates were picked as a result of their pigmentation and/or morphology and were transferred onto fresh plates several times until pure culture was obtained. Colonies were differentiated by colour, shape, and edge appearance. Morphological features of cells were examined using oil immersion at 100x objective (Optika, B-500 ERGO Model, Italy).

Enrichment and
2.3. Extraction of Genomic DNA. When the growth of archaeal cultures reached the exponential phase, the extraction of genomic DNA was performed using the GF-1 Nucleic Acid Extraction Kit (Vivantis Technologies Sdn Bhd, Selangor DE, Malaysia) according to the manufacturer's instructions.
2.4. Identification of Archaeal Strains. PCR amplification was achieved using the primer set 21F (5′-TTCCGGTTGATCC YGCCGGA-3′) [23] and 1492R (5′-GGTTACCTTGTTAC GACTT-3′) [24]. The PCR reaction was realized in a 50 μL mixture containing 1.25 U of Taq polymerase (Fermentas), 1x PCR buffer, 200 μM of dNTP, 0.2 μM of each primer, and 50 ng of genomic DNA. Thirty cycles (1 min 94°C; 1 min 55°C; 2 min 72°C) were carried out using a thermocycler (Applied Biosystems, USA). The amplified PCR products of size 1500 bp were analyzed by electrophoresis in 1% agarose gels and photographed with a Gel Doc XR Imaging system (Bio-Rad). Then, the restriction analysis was performed by digesting 10 μL of PCR products with 10 U of restriction enzymes AluI and MboI while HaeIII (8 U μL -1 ) and the appropriate restriction buffer in a final volume of 20 μL for 3 hours at 37°C. These enzymes (Life Technologies), frequently used in restriction analysis, gave the most significant differences between species. 16S rRNA fragments, acquired after enzymatic digestion, were separated on 3% agarose gels for 4 h at 50 V and photographed with a Gel Doc system. The original PCR products of positive isolates selected by comparing enzymatic restriction patterns and described as PHA-producing isolates in the following section were purified with PureLink® Quick Gel Extraction and PCR Purification Combo Kit (Cat. No. K220001, Invitrogen, Carlsbad, USA) following the manufacturer's instructions prior to cloning. The purified PCR products were ligated into pGEM-T easy (Promega Corporation, Madison, WI) system as recommended by the manufacturer. The ligation mixture was transformed into DH5α competent cells. Recombinant plasmids were verified by EcoRI digestion and chosen for sequencing. Sequencing and phylogenetic analysis were performed as previously reported [25].
2.5. Screening of Potential Halophilic PHA Producers. All isolates were subjected to PHA screening by Sudan Black B (SBB) and confirmed by Nile Red (NR) staining (Sigma), a more specific stain. Staining of colonies was done using 0.3% alcoholic solution of Sudan Black B. PHA-producing colonies appeared black [26]. Nile Red stain (25% (w/v) stock solution in dimethylsulfoxide (DMSO)) was directly inoculated in duplicate (0.5 μg mL -1 (w/v)) in agar medium containing 1% (w/v) starch and growth cells occurred in the presence of the dye. The strain Escherichia coli was used as 2 Archaea a negative control. Natrinema altunense strain CEJGTEA101 [KY129977] was identified as a PHA-producing isolate in our previous work [9] and was used as a positive control in this study. After 15 days of incubation at 37°C, the isolates which revealed orange fluorescence after exposure of plates to UV light were selected as PHA accumulators [27]. Staining of promising cells with Nile Red was done using a fluorescence microscope (Olympus BX51) [28].
2.6. Growth Kinetics of Potential PHA Producers. The strains were cultivated in 50 mL of PHA-producing medium amended with 10 g L -1 starch. The cultures in Erlenmeyer flasks were incubated in duplicate at 37°C, 180 rpm. Absorbance at 600 nm was determined using a UV-visible spectrophotometer (Shimadzu UV-1800, Japan) each 24 h for each PHA-producing strain.

Determination of Cell Dry Weight (CDW).
Cultures grown to late logarithmic phase were centrifuged at 6000 rpm for 30 min; then, cells were washed twice with distilled water. The supernatant was discarded leaving the pellet, which was lyophilized and weighed.

Determination of PHB Content in Potential Halophilic
PHA Producers by Gas Chromatography (GC). After methanolysis, the cellular content of the polymer and its compo-sition were assayed by gas chromatography (GC) using an Agilent Technologies 7890A chromatograph, equipped with a capillary column (30 m × 0:25 mm × 0:25 μm) and a flame ionization detector (FID) as previously reported [31]. Samples were analyzed in duplicate. The standard PHB (Sigma-Aldrich, USA) was used for calibration. The PHB content in the cells was determined as ðmass of PHB/cell dry massÞ × 100%. The peak at 4.4 min represents the 3hydroxybutyrate methylester.

Screening of PHA-Producing Isolates by Staining
Procedures. A large number of orange, red, and pink colonies were picked and purified by repeating subculturing. A total of 33 extremely halophilic strains were isolated from three tested enrichment cultures and were screened for PHA accumulation (Table 2). 11, 20, and 2 isolates were obtained from the water/sediment mix of the sample S1-10, water/sediment mix of the sample S6M-14, and hypersaline water of the sample S6W-14, respectively. 14 positive isolates were selected after staining by Sudan Black B ( Figure S1 (a)). Additionally, they showed high fluorescence intensity with Nile Red when exposed to ultraviolet ( Figure S1 (b)). The positive control strain exhibited orange fluorescence under UV light.

Screening of PHA Synthase Genes by Degenerate
Polymerase Chain Reaction. As a result, the screening of 33 strains showed the detection of 14 strains as PHA producers. The same strains, which showed positive results with phenotypic methods (Sudan Black and Nile Red), gave bands of approximately 230 bp (phaE) and 280 bp (phaC) ( Table 2; Figure S2).

Morphological Characterization of Potential PHA
Producers. The cells of all isolates were rods, cocci, and pleomorph ( Table 2). All selected strains as producers were round. The approximate cell dimensions were 1 to 2 μm ( Figure S3).

Discussion
The development of biodegradable plastics represents an alternative way to respond to problems associated with plastic waste. Polyhydroxyalkanoates are considered to be excellent candidates for biodegradable plastics. Extremely halophilic archaea have the ability to synthesize and accumulate PHA as inclusions in their cells [11,32]   5 Archaea effort has been taken to search the PHB-producing archaea isolated from the hypersaline lake, Chott El Jerid, using a starchy substrate. A total of 33 isolates were obtained from three enrichment cultures. PHB was detected in 14 strains grouped in three haloarchaeal genera Haloarcula, Halorubrum, and Natrinema and clustered within the phylum Euryarchaeota including Haloarculaceae, Halorubraceae, and Natrialbaceae families, respectively. They were screened via staining means (Sudan Black B and Nile Red) which were widely used for halophilic bacteria but also successfully used for Halococci, Haloarcula, Haloferax, Halorubrum, Natronococcus, Halogeometricum, Halobacterium genera, and other haloarchaeal strains [8,13]. In parallel, it was evident that the detection of phaC and phaE genes were shown in the 14 archaeal cells as described above with staining methods, confirming the PHB biosynthesis. To our knowledge, few studies based on molecular characterization of the genes involved in PHB synthesis in the domain of Archaea have been investigated [33]. Importantly, one group of extremely halophilic archaea with great biotechnological importance was Haloar-cula which was genetically well understood. Han et al. [12] identified two adjacent genes phaE Hm and phaC Hm encoding two subunits of PHA synthase (class III) and showed that these genes are required for PHB synthesis in Haloarcula marismortui (cultivated on 2% glucose, production of 21% PHB of CDW) [12]. Later, Han et al. [30] confirmed the detection of phaEC genes in 18 PHB or poly(3-hydroxybutyrate-co-hydroxyvalerate) PHBV producers, including Haloarcula, Halorubrum, Natrinema, and other genera by utilizing carbohydrates either glucose or fructose [30]. More recently, it was reported that these two genes were detected in the genome of Natrinema altunense CEJGTEA101, isolated from Chott El Jerid [9].
The present study is a continuity of our previous work [9], proposing the possibility to enlarge our knowledge about PHB secretion by a large number of haloarchaeal strains from Chott El Jerid due to their several advantages: firstly, their growth at high salinity minimized microbial contamination. Secondly, the high osmotic pressure in their cells facilitated the PHB recovery. Finally, their ability to consume of similarity), and Haloarcula tradensis (97.72% of similarity) have been considered as the best PHB producers 9.25%, 7.11%, and 1.42% of its CDW, respectively. It was important to note that these three genera were found to be PHA accumulators in other hypersaline environments [7], but only a few species were able to utilize starch to secrete large amounts of PHB. With regard to Haloarcula investigations, no PHB was accumulated by Haloarcula species using starch except Haloarcula. sp. IRU1 which could produce 57% PHB/CDW [17]. This species isolated from Urmia lake has been shown to produce important quantities of PHB (63% of CDW) from petrochemical wastewater as a carbon source containing multiple hydrocarbons such as linear alkylbenzenes [34]. Other Haloarcula species such as Haloarcula japonica, Haloarcula amylolytica, and Haloarcula argentinensis can accumulate PHB, and their yields obtained from glucose were 0.5, 4.4, and 6.5% (of CDW), respectively [30,35]. Although production of PHB has been reported from starch, glucose, and waste materials, members of Haloarcula were observed for the first time in southern Tunisian salt lakes as PHB producers [15]. Currently, there is no evidence of PHB accumulation by the Halorubrum species when starch was used as a carbon source [11]. As previously stated, two species affiliated with Halorubrum which produced PHB or PHBV (2.1-12.7% of CDW) were able to use glucose as the sole carbon source [30]. On the other hand, the majority of members of Natrinema were found to be PHBV producers when cultivated on a medium with glucose [30] or starch [36] as the carbon sources.

Archaea
In our previous work [9], 20 extremely halophilic archaea belonging to the genera Halorubrum (17 strains), Natrinema (2 strains), and Haloterrigena (1 strain) and isolated from the sample S1-10 collected from Chott El Jerid in the dry season were screened for PHA production in the same PHAaccumulating medium used in this study. Among them, only two strains belonging to Natrinema and Haloterrigena genera have been shown to accumulate 7% of PHB and 3.6% of poly(3-hydroxyvalerate) (PHV), respectively, in a medium supplemented with 2% glucose. In this current study, enrichment cultures with the same sample (S1-10) and supplemented with starch showed the selection of only one PHB producer related to the genus Haloarcula. However, enrichment cultures with samples collected in the wet season and supplemented with the same carbon source revealed the presence of one species Halorubrum and a high number of Natrinema species (12 strains) with PHB-producing ability from the samples S6W-14 and S6M-14, respectively. These findings displayed that isolation method, carbon source, season, and sampling location could influence the selection of PHB-producing archaeal species.

Conclusions
On the basis of data obtained, different archaeal isolates were obtained in pure cultures from the Tunisian desert and were able to consume starch as the sole carbon source for the PHB biosynthesis. However, the yields of PHB production in flasks batch cultures within the cells were low in comparison with other closest relatives. Therefore, future studies on fermentations optimization in batch, fed-batch, and continuous cultures using starchy and other low-cost feedstock's as well as the metabolic engineering strategies to improve both the quality and PHB productivity will be investigated.

Data Availability
The sequences have been submitted to the GenBank database under accession numbers: MN516817 to MN516830.

Conflicts of Interest
The authors declare no conflict of interest.

Authors' Contributions
Fatma Karray and Manel Ben Abdallah contributed equally to this work.

Acknowledgments
This work was supported by the Tunisian Ministry of Higher Education and Scientific Research. We thank also the reviewers for their helpful comments on this manuscript.