Introduction

Polyhydroxyalkanoates (PHAs) is a biopolyester that can be produced naturally by various bacteria and is biodegradable in soil and marine environment (Lopez-Llorca et al. 1993; Doi et al. 1995; Sudesh et al. 2000). Its naturally occurring, biodegradable, and environment-friendly properties have made PHA a good alternative as a bioplastic compared to petroleum-based plastics, which take years to degrade, in addition to the petroleum depletion concern.

PHAs are accumulated in the cytoplasm as their energy reserve and carbon source in response to stress, for example, under nutrient limiting and excess carbon conditions (Anderson and Dawes 1990). Many different carbon sources, such as palm oil, sludge palm oil, crude palm kernel oil, sugar, and many more, were reported to be fed to bacteria for PHA production (Riedel et al. 2012; Hassan et al. 2013; Thinagaran & Sudesh 2019). Among the different carbon sources, fatty acids are preferable as they contain more carbon content compared to sugar. Besides that, there are also reports on waste like frying oil used as feedstock for PHA production using Pseudomonas fluorescens S48 (Gamal et al. 2013). The utilisation of waste as feedstock for PHA production not only reduces the production cost of PHA, but also converts these wastes into useful biodegradable plastics (Lee & Choi 1999; Chee et al. 2010). Depending on the carbon sources, different PHA will be accumulated in the bacteria. For instance, poly(3-hydroxybutyrate-co-3-hydroxyhexanoate), P(3HB-co-3HHx) will be produced when the bacteria are fed with oil, while only poly(3-hydroxybutyrate), P(3HB) will be accumulated when fed with sugar such as fructose and sucrose (Lee et al. 2008; Ng and Sudesh 2016).

Among the different types of PHAs, polyhydroxybutyrate (PHB) is the most common type of polymer produced by bacteria. However, its stiffness, brittle, low thermal stability and high crystallinity properties of PHB have limited its potential application. Therefore, the incorporation of other monomers such as 3-hydroxyhexanoate (3HHx) and 4-hydroxybutyrate (4HB) to produce copolymers such as P(3HB-co-3HHx) and poly(3-hydroxybutyrate-co-4-hydroxybutyrate), P(3HB-co-4HB) is crucial to improve its thermostability, flexibility and biocompatibility to replace the conventional single-used plastics and widen its potential application, especially in the medical field (Jendrossek and Handrick 2002; Chee et al. 2010).

Each bacterium has its unique PHA biosynthetic genes, including phaA, phaB, phaC, phaJ, and phaP. These PHA biosynthetic genes are closely associated with synthesising PHA and significantly affect the amount and types of PHA synthesised in bacteria. For example, PHA synthase (PhaC) is responsible for polymerising monomeric hydroxyalkanoate (HA) substrates into PHA, and its substrate specificity determines the types of PHAs produced by the bacteria (Steinbüchel & Valentin 1995; Rehm 2003). In brief, the discovery of each new biosynthetic gene provides us with a better insight and understanding of the PHA biosynthesis mechanism in bacteria, which may aid us in improving and producing a better and wider application of bio-based plastics.

The genus Aquitalea is a member of Betaproteobacterium in the family Chromobacteriaceae. To date, Aquitalea magnusonii, Aquitalea denitrificans, Aquitalea pelogenes, and Aquitalea aquatilis are the four members of this genus that have been reported (Lau et al. 2006; Lee et al. 2009; Sedláček et al. 2016; Ngo et al. 2020). Compared to the genus Chromobacterium, Aquitalea can be considered understudied as there are very few studies on its production of bio-products such as PHA (Steinbüchel et al. 1993; Brito et al. 2004; Bhubalan et al. 2011; Ling et al. 2011), since the first Aquitalea species was only reported in 2006 (Lau et al. 2006). According to GOLD database, a total of 22 Aquitalea genome sequences are publicly available, for example A. magnusonii H3, Aquitalea sp. strain MWU14-2217, A. pelogenes CCM 7557 and A. aquatilis THG-DN7.12, but the ability to produce PHA has not been reported (Sedláček et al. 2016; Ishizawa et al. 2017; Ebadzadsahrai & Soby 2018; Ngo et al. 2020) (the whole-genome sequences or whole-genome shotgun sequences of Aquitalea strains are available in NCBI and BV-BRC and their accession numbers are listed in Table S2).

To further analyse the potential of the Aquitalea species in PHA production, we found that some of the Aquitalea strains, for example, A. denitrificans strain 5YN1-3, A. aquatilis strain THG-DN7.12 and A. magnusonii strain H3, do have genes that are necessary for PHA synthesis such as phaA, phaB, phaP, phaJ, and phaC, but neither has been reported to produce PHA. Aquitalea sp. USM4 was first isolated by Ng and Sudesh (2016) from freshwater in Perak, Malaysia, and is the first Aquitalea strain reported to produce PHA. Its PhaC (PhaCAs) exhibits high activity (863 U/g protein) (Ng & Sudesh 2016). The ability to produce PHA by Aquitalea sp. USM4 has triggered our interest in elucidating its whole genome, especially its PHA biosynthetic genes, as previously isolated Aquitalea strains are non-PHA producers. Hence, this study provides insight into the first whole genome of Aquitalea sp. USM4 and its PHA biosynthetic genes for future research on Aquitalea species.

Materials and methods

Whole-genome sequencing

Paired-end and mate-pair libraries were prepared from the bacterial genomic DNA using Illumina TruSeq DNA PCR-free and Nextera mate-pair sample prep kit, respectively. One paired-end library (insert size, 600 bp; read length, 250 bp; total, 1.3 Gb) and three mate-pair libraries (insert size, 3 kb, 6 kb, 6.5 kb; read length, 250 bp; total, 912 Mb) were sequenced by Illumina MiSeq sequencer. Adaptor sequences and low-quality regions in raw reads were trimmed using Platanus_trim (version 1.0.7; http://platanus.bio.titech.ac.jp/pltanus_trim), and de novo assembly was done with Platanus (version 1.2.1) (Kajitani et al. 2014), resulting in three scaffolds (size > 1 kb). One scaffold (size, 5,518 bp) was excluded as the PhiX genome was used as a control for Illumina sequencing. The others were considered as a chromosome and a plasmid. Finally, the gap-close module of MetaPlatanus (pre-release version) (Kajitani et al. 2021) was additionally applied, and all gaps were filled. The bacterial genome and the protein-coding sequences (CDSs) were predicted and annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP) (Tatusova et al. 2016; Haft et al. 2018).

Genome analysis

The whole-genome-based pairwise average nucleotide identity (ANI) was done between Aquitalea sp. USM4 and other Aquitalea genomes. Different ANI calculating tools were used independently to calculate the ANI, including the ANI calculator (http://enve-omics.ce.gatech.edu/ani/index), EzGenome (http://www.ezbiocloud.net/ezgenome/ani) and Jspecies (http://www.imedea.uib.es/jspecies) where the genome sequences were retrieved from NCBI (https://www.ncbi.nlm.nih.gov/) and BV-BRC (https://www.bv-brc.org/) (Richter et al. 2015; Yoon et al. 2017). ANI values obtained were based on either BLASTn (ANIb) or MUMMER software (ANIm) (Richter & Rosselló-Móra 2009). The EzGenome was used to calculate OrthoANIb, which is ANI by orthology and calculated using the BLASTn program (Lee et al. 2016). The ANI between Aquitalea sp. USM4 and two other strains from the same clique cluster, A. pelogenes CCM7557 and A. magnusonii SM6, were also analysed using Genomes OnLine Database (GOLD) v.8 (Chen et al. 2020; Mukherjee et al. 2020). Digital DNA–DNA hybridisation (dDDH) values and confidence intervals were calculated using the recommended settings of the GGDC3.0 (https://www.dsmz.de/services/online-tools/genome-to-genome-distance-calculator-ggdc) paired with the BLAST + alignment tool and formula 2 (Meier-Kolthoff et al. 2013).

Bacterial strain and plasmids

Bacterial strains and plasmids utilised in this study are listed in Table 1. Strains of Aquitalea sp. USM4 and Cupriavidus necator were routinely cultured on nutrient-rich agar (NR) containing 10 g/L peptone (HiMedia, India), 10 g/L meat extract (HiMedia, India), and 2 g/L yeast extract (HiMedia, India) at pH 7.0 and 30 ℃. Escherichia coli strains were grown in Luria–Bertani (LB) (HiMedia, India) medium at 37℃. For the cultivation of recombinant strains with the plasmid, the medium was supplemented with 50 µg/mL of kanamycin (Sigma Aldrich).

Table 1 List of plasmids and bacterial strains used in this study
Full size table

Cloning of PHA synthase genes from Aquitalea sp. USM4

In this study, phaC1 of Aquitalea sp. USM4 (phaC1As) was amplified using polymerase chain reaction (PCR). The sequences of two of the oligonucleotide primers used in this study are ATCAAGCTTAGGAGGAGGCATGTTAAGGAGATGTACCTTGA (PhaCAs_HindIII_RBS_Forw) and ATCGGGCCCATTTAAATTTATTGCAGGCTGGCTACCGTGCT (PhaCAs_ApaI_SwaI_Rev).

The PCR products were digested using FastDigest restriction enzymes (Thermo Scientific, USA) and ligated with the digested plasmid using DNA Ligation Kit Ver.2.1 (Takara Bio Inc. Japan) in accordance with the manufacturer’s protocol. The broad-host-range pBBR1MCS-2 plasmid (Kovach et al. 1995) was used and transformed into C. necator PHB4 through transconjugation with E. coli S17-1 (Friedrich et al. 1981). Colony PCR with plasmid-specific primers and DNA sequencing were used to verify the successful transformants.

PHA biosynthesis

PHA accumulation in C. necator PHB−4 harbouring phaC1As was carried out through one-stage shake flask cultivation as described by Budde et al. (2011). The ability of the Aquitalea sp. USM4 to incorporate different PHA monomers was studied by adding structurally related carbon sources or precursors into the PHA biosynthesis medium. In this work, 10 g/L of crude palm kernel oil (CPKO) or fructose was used as the carbon source. To examine the substrate specificity of the PhaC, a total 10 g/L of fructose was added as the carbon source with 2 g/L of sodium 3-hydroxyvalerate (Na3HV), sodium 4-hydroxybutyrate (Na4HB), sodium 5-hydroxyvalerate (Na5HV), sodium 3-hydroxyhexanoate (Na3HHx) and 1 g/L of isocaproic acid (3H4MV) as structurally related precursors.

After 48 h of incubation, the culture was harvested by centrifugation at 8000 rpm for 10 min at 4℃ and then lyophilised for 2 days. Then the cells were further analysed using gas chromatography (GC) (Braunegg et al. 1978). The P(3HB-co-3HHx) copolymer was extracted and further verified using proton nuclear magnetic resonance (1H-NMR).

Nucleotide sequence accession number

The genomic sequence of chromosome and plasmid of Aquitalea sp. USM4 has been submitted in GenBank with accession numbers NZ_CP029539 and NZ_CP029540, respectively. The strain is accessible through the Japan Collection of Microorganisms (JCM) with accession number JCM 19919.

Results and discussion

The genome attributes of Aquitalea sp. USM4

The Aquitalea sp. USM4 genome was sequenced from 600 bp paired-end library and 3 kb, 6 kb, and 6.5 kb mate-pair libraries using Illumina MiSeq sequencer, yielding 1.3 Gb and 912 Mb of sequence data, respectively. A complete genome of Aquitalea sp. USM4 was obtained after the sequenced data were trimmed and assembled as contigs by performing de novo assembly. The Aquitalea sp. USM4 genome consisted of one circular chromosome and one plasmid. The size of the circular chromosome is 4,291,790 bp, with a G + C content of 59.4% (Table 2), whereas the plasmid size is 54,370 bp, and its G + C content is 56.3%.

Table 2 Genome features of Aquitalea sp. USM4
Full size table

The bacterial whole-genome assemblies were annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP). There is a total of 87 tRNAs, 25 rRNA operons, and 3952 protein-coding sequences (CDSs) predicted in Aquitalea sp. USM4. The distribution proportion of Clusters of Orthologous Groups (COGs) functional categories of Aquitalea sp. USM4 were analysed using eggNOG (Table S1). Aquitalea sp. USM4 possesses an exceptionally high number of genes classified by COG as “poorly characterised”, which includes “function unknown” (19.21%) and “general function prediction only” (4.17%), indicating that there are many genes and mechanisms in Aquitalea remain largely unknown. The higher proportion of known protein-coding genes are categorised under the “metabolism” group, especially “amino acid transport and metabolism” (9.14%) and “energy production and conservation” (6.62%). It is followed by “transcription” (7.13%) and “signal transduction mechanisms” (6.54%) under the group “Information storage and processing” and “cellular processes and signalling”.

Previously, the 16S rRNA phylogenetic analysis showed that Aquitalea sp. USM4 was closely related to the A. magnusonii TRO-001DR8(T) and A. denitrificans 5YN1 (Ng and Sudesh 2016). To further investigate the phylogeny relationship of Aquitalea sp. USM4 with other Aquitalea sp., we performed ANI and dDDH analysis using the whole genome of Aquitalea sp. USM4. The ANI values were calculated using different tools, including the ANI calculator, EzGenome, Jspecies and GOLD database. The ANI result showed Aquitalea sp. USM4 has higher similarity (97.64–98.01%) against A. pelogenes CCM7557 (supplementary data, Table S2). The recommended threshold ANI value for the demarcation of species is 96.5% when using complete or nearly complete genomes (Varghese et al. 2015). In addition to that, the classical standard species delineation threshold dDDH value is 70%, from the dDDH analysis result showed 81.6% between Aquitalea sp. USM4 and A. pelogenes CCM7557 which is above the threshold value (supplementary data, Table S2) (Goris et al. 2007; Richter & Rosselló-Móra 2009). Hence, the ANI and dDDH values showed that the Aquitalea strain isolated indicated that it belongs to the A. pelogenes species and can be referred to as Aquitalea pelogenes USM4.

PHA biosynthesis-related genes in Aquitalea sp. USM4

Genes involved in PHA biosynthesis were identified from the whole genome of Aquitalea sp. USM4. There are a total of 19 genes encoded for proteins have been identified (Table 3): three acetyl-CoA C-acyltransferase (phaA), acetoacetyl-CoA reductase (phaB), three phasin family proteins (phaP), seven enoyl-CoA hydratase (phaJ), two MaoC family dehydratase, repressor protein (phaR), and two PHA synthase (phaC).

Table 3 Putative PHA-associated genes in the genome of Aquitalea sp. USM4 with its function and closest organism match based on BLASTx
Full size table

The biosynthetic pathway of P(3HB) consists of three main enzymatic reactions that are catalysed by phaA, phaB, and phaC. These three genes are usually clustered together in the bacterial genome and form the phaCAB operon generally found in the PHA-accumulating bacteria, as it promotes the synthesis of SCL-PHA (Rehm and Steinbüchel 1999; Reddy et al. 2003). In the Aquitalea genome, we found the phaCA operon, while the phaB is located elsewhere in the chromosome. However, this does not affect its ability to produce PHA, as similar gene arrangements can also be found in Chromobacterium violaceum and Jeongeupia sp. USM3 (Kolibachuk et al. 1999; Zain et al. 2020). This is contrary to the C. necator, which is the model organism in PHA studies, where the phaC, phaA, and phaB are clustered together (Peoples and Sinskey 1989). Interestingly, two PhaCs, PhaC1As and PhaC2As were identified in Aquitalea sp. USM4. Class I PhaC is known to incorporate SCL-PHA monomer only; however, our PhaC1As is classified into a special class of class I PhaC where it is capable of incorporating both SCL-PHA and MCL-PHA (Rehm and Steinbüchel 1999; Neoh et al. 2022). This characteristic of PhaC is similar to previously isolated PhaCs, such as PhaCCs from Chromobacterium sp. USM2, PhaCAc from Aeromonas caviae and PhaC isolated from mangrove soil metagenome (PhaCBP-M-CPF4) which are also class I PhaC that are able to incorporate the MCL-PHA monomer (Doi et al. 1995; Bhubalan et al. 2011; Foong et al. 2017).

PhaJ and MaoC family dehydratase were the enzymes involved in supplying monomers for PHA biosynthesis through the β-oxidation pathway (Tsuge et al. 2000; Wang et al. 2013). PhaJ plays a vital role in 3HHx accumulation, as it creates a pathway for supplying (R)-3-hydroxyhexanoate-CoA, [(R)-3HHx-CoA] monomer units from fatty acid β-oxidation (Fukui & Doi 1997). The co-expression of phaJ and phaC can enhance the accumulation and incorporation of 3HHx into the P(3HB-co-3HHx) (Budde et al. 2011; Wang et al. 2013; Tan et al. 2020). The PhaP is a group of amphiphilic proteins that consists of both hydrophobic and hydrophilic surface that binds to the surfaces of PHA granules accumulated in the form of inclusion body in the bacterial cells (Fukui et al. 2001; Zhao et al. 2016).

Production of PHA

Our previous research has shown that Aquitalea sp. USM4 can accumulate up to 1.5 g/L of PHA (Ng and Sudesh 2016) when cultivated in MM (Doi et al. 1995). Depending on the type of carbon sources and precursors added into the culture, the PHA copolymers accumulated were composed of 3HB, 3HV, 4HB, and 3H4MV.

In this study, PhaC1As was cloned into plasmid pBBR1MCS-2 with the phaC1 promoter from C. necator and transconjugated into C. necator PHB−4 (Foong et al. 2017). The transformant was cultured in MM to induce PHA production (Budde et al. 2010). Compared to the previous result from Ng and Sudesh (2016), there is an improvement in the dry cell weight of the culture. The transformant increased in dry cell weight from 2.7 to 3.8 g/L when cultivated using fructose co-fed with sodium 3-hydroxyvalerate. The transformant also showed increments of 4.4–4.6 g/L in dry cell weight when CPKO was used as the carbon source.

The marked improvement is probably due to the effect of the promoter derived from phaC1 of C. necator (PphaC1) exhibiting better expression of phaCAs for PHA biosynthesis as compared to the lacZ promoter (PlacZ). Fukui et al. (2011) reported that different promoters could affect the expression of phaC gene, as they reported that the strain with PphaC1 had better P(3HB) accumulation when cultured using fructose and soybean oil compared to the strain with PlacZ. A better expression of PhaCAs using PphaC1 probably allows the C. necator PHB−4 to accumulate more PHA and better dry cell weight. Furthermore, the addition of ribosome-binding site (RBS) [AGGAGG] is also suspected to increase the expression of PhaCAs and, hence, increase the concentration of PhaCAs in bacterial cells for better PHA accumulation. A conceptually similar work was also carried out by Arikawa and Matsumoto (2016) using the RBS from C. necator [AGAGAGA] with trc promoter (Ptrc), lacUV5 promoter (PlacUV5), and trp promoter (Ptrp), which showed remarkably improved synthase activity due to better gene expression, hinting at the importance of RBS in the expression of PhaC.

Compared to the previous result by Ng and Sudesh (2016), there is a stark improvement in terms of substrate specificity of the PhaCAs, which is the ability to incorporate 3HHx and 5HV in the copolymer produced when supplemented with CPKO, fructose with Na3HHx, and fructose with Na5HV, respectively. A total of 2.2 mol% and 3.3 mol% of 3HHx monomer were successfully incorporated when the transformant was fed with fructose with Na3HHx and CPKO, respectively, whereas 26 mol% of 5HV monomer was successfully incorporated into P(3HB-co-5HV) copolymer (Table 4). The P(3HB-co-3HHx) copolymer was then subjected to additional verification by 1H NMR (Figs. 1 and 2) to confirm the GC result for 3HHx in the P(3HB-co-3HHx) copolymer. The results were consistent as compared to the GC result, so it proved that PhaC1As could incorporate 3HHx as well.

Table 4 PHA biosynthesis C. necator transformant harbouring pBBR1MCS2-CAs using different carbon sources and precursors in shake flask experiments
Full size table
Fig. 1
figure 1

Proton nuclear magnetic resonance spectroscopy (1H NMR) spectrum of P(3HB-co-3HHx) copolymer consisting of 98 mol% 3HB and 2.2 mol% 3HHx synthesised by C. necator PHB−4 transformant harbouring pBBR1MCS2-CAs from 10 g/L fructose and 2 g/L sodium 3-hydroxyhexanoate

Full size image
Fig. 2
figure 2

Proton nuclear magnetic resonance spectroscopy (.1H NMR) spectrum of P(3HB-co-3HHx) copolymer consisting of 97 mol% 3HB and 3.3 mol% 3HHx synthesised by C. necator PHB−4 transformant harbouring pBBR1MCS2-CAs from 10 g/L crude plum kernel oil (CPKO)

Full size image

P(3HB-co-3HHx) is a PHA copolymer reported to have a close resemblance to commercial polypropylene (PP) and low-density polyethylene (LDPE) (Doi 1990). This suggests that P(3HB-co-3HHx) is capable of replacing single-use petrochemical plastic. Besides that, P(3HB-co-3HHx) is also reported to be applied in the field of tissue engineering, where it can be moulded into a scaffold for bone tissue engineering (Ang et al. 2020). As for P(3HB-co-5HV), the copolymers or terpolymers with 5HV monomers were reported to have the potential as biomaterial (Chuah et al. 2013; Lakshmanan et al. 2019). Lakshmanan et al. (2019) reported that lipases could degrade co- and terpolymer of 5HV. Chuah et al. (2013) also reported that terpolymer of 5HV was less cytotoxic and good for cell proliferation.