A Forward-Design Approach to Increase the Production of Poly-3-Hydroxybutyrate in Genetically Engineered Escherichia coli

Biopolymers, such as poly-3-hydroxybutyrate (P(3HB)) are produced as a carbon store in an array of organisms and exhibit characteristics which are similar to oil-derived plastics, yet have the added advantages of biodegradability and biocompatibility. Despite these advantages, P(3HB) production is currently more expensive than the production of oil-derived plastics, and therefore, more efficient P(3HB) production processes would be desirable. In this study, we describe the model-guided design and experimental validation of several engineered P(3HB) producing operons. In particular, we describe the characterization of a hybrid phaCAB operon that consists of a dual promoter (native and J23104) and RBS (native and B0034) design. P(3HB) production at 24 h was around six-fold higher in hybrid phaCAB engineered Escherichia coli in comparison to E. coli engineered with the native phaCAB operon from Ralstonia eutropha H16. Additionally, we describe the utilization of non-recyclable waste as a low-cost carbon source for the production of P(3HB).


Introduction
Conventional oil-derived polyolefin plastics exhibit useful characteristics that have extensive commercial applications. However, the accumulation of plastics in the environment and the non-renewable source of polyolefin plastics have stimulated interest in sustainable sources of plastic production. Biopolymers, such as poly-3-hydroxybutyrate (P(3HB)), are produced as a carbon store in an array of organisms and exhibit characteristics which are similar to oilderived plastics [1,2]. Furthermore, P(3HB) has the added advantages of biodegradability, biocompatibility and, based upon several life cycle analyses, P(3HB) production is more environmentally sustainable than polyolefin plastic production [1,3,4]. Despite these advantages, P(3HB) production is currently more expensive than the production of oil-derived plastics, and therefore, more efficient P(3HB) production processes would be desirable [5].
Genetic engineering approaches-in which the P(3HB)-producing operon, phaCAB, is cloned into Escherichia coli-have pioneered the industrial production of P(3HB) [6]. More recently, synthetic biology approaches involving the rational engineering of the phaCAB operon [7], and metabolic engineering strategies [8] have continued to increase the efficiency of P (3HB) production. In this paper we report on the forward-design and experimental validation of a rationally engineered phaCAB operon with a hybrid promoter design. Additionally, we describe the utilization of non-recyclable waste as a low-cost carbon source for the production of P(3HB).
In order to increase P(3HB) production, this study used a forward-design approach that is based upon the engineering principles of synthetic biology [10]. Several phaCAB operons were rationally engineered (Fig. 1B). The constitutive phaCAB operon (BBa_K1149052) was The constructs used in this study. Abbreviations: Pwt (wildtype promoter; green arrow), J23104 (Anderson constitutive promoter, BBa_J23104; red arrow), B0034 (ribosomal binding site, BBa_B0034; red half-circle), phaC (PHA synthase), phaA (3-ketothiolase) and phaB (acetoacetyl-CoA reductase). Green half-circles denote native ribosomal binding sites. Construct symbols are based on the Synthetic Biology Open Language Visual (SBOLv) v1.0.0 guidelines [21]. designed such that the native promoter and RBS were replaced with the constitutive promoter J32104 from the Anderson collection [11] and the RBS B0034. The hybrid phaCAB operon design (BBa_K1149051) was constructed in parallel to the constitutive operon and was noted for its dual promoter and RBS combinations (Fig. 1B). Design considerations were based upon modeling simulations of the PhaCAB pathway in engineered E. coli. The model was comprised of the glycolysis pathway, the tricarboxylic acid (TCA) cycle and the phaCAB synthetic pathway (S1 Supporting Information; S2 Supporting Information). These pathways were coupled in order to reflect the metabolic flux [12] of several metabolites and species between them and to provide a qualitative indication of their influence on P(3HB) production.
A sensitivity analysis of species within the synthetic pathway revealed that increasing expression of phaB would be critical for directing the reaction flow towards the end product-an increase in P(3HB) production (S1 Supporting Information; S2 Supporting Information). In order to increase expression of phaB, further simulations predicted that of the several simulated phaCAB operon designs, an operon that incorporates the constitutive promoter, J23104, would result in the greatest increase in P(3HB) production (Fig. 2).
To experimentally validate our forward-design approach, E. coli MG1655 carrying either the empty vector (BBa_K608002), native phaCAB (BBa_K934001), constitutive phaCAB (BBa_K1149052) or hybrid phaCAB (BBa_K1149051) operons were cultured with Lysogeny broth (LB) media, supplemented with 3% glucose (w/v) for either 24 or 48 hours. P(3HB) was purified from each of the engineered populations using sodium hypochlorite and were weighed for comparison (Fig. 3A). Average P(3HB) production at 24 hours was around two-fold higher in constitutive phaCAB-engineered E. coli (0.49 g/L S.D. ± 0.06) compared to native phaCAB (0.22 g/L S.D. ± 0.18), while average P(3HB) production was six-fold higher in hybrid phaCABengineered E. coli (1.47 g/L S.D. ± 0.48). At 48 hours, P(3HB) production was around three-fold higher in constitutive phaCAB-engineered E. coli (1.05 g/L S.D. ± 0.26) and hybrid phaCABengineered E. coli (0.94 g/L S.D. ± 0.14), compared to native phaCAB (0.30 g/L S.D. ± 0.14) (Fig. 3A). In hybrid phaCAB-engineered E. coli we observed a decrease in P(3HB) content, expressed as a percentage of the cell dry weight (CDW) from 50% at 24 hours to 32% at 48 hours Simulated P(3HB) production in phaCAB-engineered E. coli. In order to simulate P(3HB) production in phaCAB-engineered E. coli, a P(3HB) synthesis model was constructed using the Simbiology toolbox of Matlab. Using this model the flux of several metabolites and species were simulated in order to identify aspects of the system that could be selectively tuned to increase the production of P(3HB). From these analyses, several novel phaCAB operons were designed. These data show the simulated P(3HB) production across several different phaCAB operon designs, where phaCAB expression is under the control of the indicated Anderson constitutive promoters.  Forward-Design of Engineered phaCAB Operons (Fig. 3B). Similar observations have been previously reported, and is understood as a consequence of competition for cellular metabolites between cell growth and P(3HB) production [13][14][15]. The empty vector-engineered E. coli did not produce detectable levels of P(3HB).
Qualitative analysis of Nile Red stained E. coli also confirmed that the constitutive and hybrid phaCAB operon designs increased P(3HB) production (S3 Supporting Information). It is likely that the dual promoter and RBS design (Fig. 1B) of the hybrid system results in a higher level of mRNA transcript production and/or ribosome recruitment and thus an increase in the expression of the PhaCAB enzymes. qRT-PCR analysis, confirmed that both the hybrid and constitutive phaCAB operon designs result in higher levels of phaCAB transcripts than that of the native phaCAB operon (S6 Supporting Information).
Interestingly, as shown in Fig. 3A, P(3HB) production from the hybrid operon at 24 h was greater than the combined P(3HB) production of the native and constitutive operons. Together, these data suggest that the hybrid promoter performs a multiplicative, rather than an additive combination of the native and constitutive promoter designs. Unlocking the design rules of the hybrid promoter may have applicability that extends beyond the phaCAB operon. For instance, Li et al. 2012, suggest that engineered promoter clusters may be a useful metabolic engineering approach for increasing metabolite production. Additionally, model-guided optimisation of the PhaCAB pathway led to the generation of the constitutive operon design, which in combination with the hybrid promoter, represent an emerging family of rationally engineered P(3HB)-producing operons.
The broader project aim was to exploit non-recyclable waste as a low-cost carbon source that could be utilized by engineered E. coli tasked with the de novo synthesis of P(3HB). Industrial processing of non-recyclable mixed waste into solid and liquefied fractions of mixed cellulosic (~80%) and plastic (~20%) streams is currently possible [16]. However, these processes could be made more efficient through the use of natural and genetically engineered organisms that can both degrade and use these waste streams as a low-cost carbon source for the production of bioplastics. Beneficially, this approach may also result in the diversion of non-recyclable waste away from environmentally damaging activities such as landfill or incineration [3]. To this end, phaCAB-engineered E. coli were cultured in waste-media, which is analogous to industrially processed liquefied cellulosic waste; and from these cultures, P-(3HB) was purified according to a modified sodium hypochlorite protocol, where wastemedia cultures were filtered prior to centrifugation to remove residual waste material. Purified P(3HB) from each culture was depolymerized with recombinant phaZ1 into monomeric 3-hydroxybutyrate (3HB), which could then be detected as a yellow colour change when these samples were analysed with the β-Hydroxybutyrate colorimetric assay kit (S4 Supporting Information). PhaZ1-treated P(3HB) from E. coli carrying either the empty vector or native phaCAB operon resulted in a negligible color change, suggesting that little or no P(3HB) was produced in those cultures. Whilst, PhaZ1-treated P(3HB) from E. coli carrying either the constitutive phaCAB or hybrid phaCAB operon, depolymerized into 3HB, as indicated by a detectable, yellow color change. Additionally, flow cytometry analysis of Nile Red stained, [17,18] phaCAB-engineered E. coli from waste-media cultures, confirmed an increase in P-(3HB) content (Fig. 4).
In conclusion, our study used a forward-design approach to rationally engineer novel phaCAB operon designs that were experimentally validated to increase the production of P-(3HB) in genetically engineered E. coli. Additionally, we have described the utilization of non-recyclable waste as a low-cost carbon source for the production of P(3HB). With further development, these strategies, in combination with additional synthetic biology approaches [7,8], will further increase the efficiency and commercial viability of P(3HB) production.

Materials and Methods Modeling
The P(3HB) synthesis model was constructed and simulated using the Simbiology toolbox of Matlab. The model is comprised of the glycolysis pathway, the tricarboxylic acid (TCA) cycle and the phaCAB synthetic pathway. Full details are provided as supplementary information (S1 Supporting Information).

Construct assembly
Empty vector (BBa_K608002), and native phaCAB (BBa_K934001) constructs were sourced from the 2013 distribution of the iGEM Registry of Standard Biological Parts (partsregistry. org). The constitutive phaCAB operon (BBa_K1149052) was generated via PCR, with the native phaCAB operon (BBa_K934001) as the template. Primers Pha_Fw 5 0 -cgcttctagagatggctactgggaaaggagccg-3 0 and BBa_G1005 5 0 -gtttcttcctgcagcggccgctactagta-3 0 , were utilised to generate a PCR product containing the phaCAB operon but excluding the native promoter and RBS. The PCR product was cloned into the destination vector (BBa_K608002) to generate the final constitutive phaCAB operon construct. The hybrid phaCAB (BBa_K1149051) construct was generated in parallel to the constitutive phaCAB operon. The entire native phaCAB operon including the native promoter and RBS was cloned into the destination vector (BBa_K608002) to generate the final hybrid phaCAB operon construct. All constructs were generated using the iGEM submission backbone, pSB1C3 and all restriction digests utilized the standard iGEM prefix and suffix restriction sites. Full sequences are provided as supplementary information (S5 Supporting Information).

P(3HB) purification
This method is scaled down from an existing P(3HB) purification protocol [19]. Briefly, 1-liter production cultures were centrifuged at 4000 rpm (Beckman J2-M1) for 15 minutes. Wastemedia cultures were filtered through Whatman filter paper in order to remove residual waste material before centrifugation. Post-centrifugation, bacterial cell pellets were re-suspended in phosphate-buffered saline (PBS), transferred into 50 ml tubes and centrifuged as described above. Cell pellets were washed with PBS and incubated for 30 minutes at room temperature in 1% (v/v in PBS) Triton-X 100. For the final purification of P(3HB), cells were centrifuged (4000 rpm-Beckman J2-M1 for 10 minutes), washed with PBS and incubated in aqueous sodium hypochlorite for 60 minutes at 30°C. The resultant purified P(3HB) granules were centrifuged (4000 rpm-Beckman J2-M1 for 10 minutes), washed with distilled water and dried overnight at 37°C. To determine the weight of the purified P(3HB), the weight of the 50 ml tube was subtracted from the combined weight of the 50 ml tube and P(3HB).

Gene expression analysis with quantitative real-time PCR (qRT-PCR)
To analyse phaCAB gene expression we adapted a previously described qRT-PCR protocol [7]. Briefly, 1 ml of each phaCAB-engineered E. coli strain was harvested after 24 h of growth in 100 ml LB, supplemented with 3% glucose (w/v) and 34 μg/mL Chloramphenicol. To stabilise RNA, 400 μl of RNA Protect Bacteria reagent (Qiagen, Hilden, Germany) was added to each cell pellet. Cell pellets were stored at −80°C until RNA extraction. ZR Fungal/Bacteria RNA Microprep kits (Zymo Research, CA, USA, # R2010) were used to harvest total RNA. To eliminate any remaining plasmid or genomic DNA, the samples were treated with RNase-free DNase I (NEB, Beverly, MA, USA) for 10 minutes, during the RNA extraction process. RNA purity and concentration was determined using a Nanodrop ND-1000 spectrophotometer (Thermo Scientific). cDNA synthesis from 500ng of total RNA was performed using the iScript Select cDNA synthesis kit (Bio-Rad, Richmond, CA, USA).
cDNA templates were used to determine gene expression levels by quantitative real-time PCR. Primers for phaA, phaB and phaC were designed according to previously published sequences [7]. To perform real-time quantitative PCR, the iQ SYBR green Supermix kit (Bio-Rad) was used and the reactions were run on a Mastercycler ep. realplex (Eppendorf, Hamburg, Germany). The PCR reaction conditions were: 3 minutes at 95°C and then 40 cycles of 15 seconds at 95°C, 1 minute at 60°C. To confirm the specificity of the PCR primers, a melting curve analysis was carried out. Data were analysed via the relative standard curve method and subsequently normalised to native phaCAB samples. Data were analysed from three biological replicates.
Supporting Information S1 Supporting Information. P(3HB) Modeling. This supplementary file includes a 31-page document that contains complete details about the P(3HB) synthesis model we constructed to forward-design the phaCAB operons that increase P(3HB) production. phaCAB constructs were cultured in 5 ml LB media, supplemented with 3% glucose (w/v) for 24 h. Analysis of phaCAB gene expression was carried out using qRT-PCR and analyzed via the relative standard curve method. Experiments were carried out in triplicate and were normalized to native phaCAB-engineered E. coli. Error bars, +/-the standard deviation. Student t-test, ÃP<0.05 and ÃÃP <0.01. (TIFF)