Optimization of Ralstonia solanacearum cell growth using a central composite rotational design for the P(3HB) production: Effect of agitation and aeration.

The intracellular accumulation of polyhydroxyalkanoates (PHAs) normally occurs after cell growth, during the second fermentation stage and under nutrient-limited conditions in the presence of a carbon excess. However, some microorganisms are able to accumulate PHAs as poly(3-hydroxybutyrate) [P(3HB)] during the first fermentation stage, the cell growth phase, without nutrient limitation, once they have been reported to utilize type II metabolism during the polymer accumulation phase. This study evaluated the effect of aeration and agitation on cell growth and P(3HB) accumulation in Ralstonia solanacearum RS, performed in a bioreactor for 24h at 32°C. A 22 central composite rotational design (CCRD) was used, with agitation (150 to 250 rpm) and aeration (0.3 to 1 vvm) as independent variables and optical density (OD600nm), dry cell weight (DCW), and P(3HB) yield as dependent variables. A significant polymer accumulation, until 70% of P(3HB), was observed, proving that R. solanacearum RS exhibited metabolism type II, regardless of the aeration process. The best results were obtained for 1 vvm and 250 rpm (+1, +1), with values of OD600nm (18.04) and DCW (4.82 g.L-1).

Introduction

In the last six decades, 8.3 billion metric tons of plastics have been produced, most of which disposable single-use plastics [1]. It is estimated that 91% of all plastics produced is not recycled, and 6.3 billion metric tons has become plastic waste [1,2]. For years, scientists have been investigated ways to reduce these numbers to prevent the volume of plastics that end up in the world's oceans, causing damage to marine mammals, birds, and fish, among others. By the middle of the century, there will be more plastics in the oceans than fish––a chilling prognosis [2]. In this context, the use biodegradable pan class="Chemical">polymers may be an effective alternative to reduce the excessive amount of lass="Chemical">plastic waste in the environment, thus leading to a lower environmental imlass="Chemical">pact [3,4].

Studies on the use of pan class="Chemical">polymers have been carried out to offer lass="Chemical">products with good mechanical lass="Chemical">prolass="Chemical">perties and a less drastic imlass="Chemical">pact on the environment, including the biodegradable lass="Chemical">pan class="Chemical">polymers, which can reduce the plastic pollution.

pan class="Chemical">Polyhydroxyalkanoates (lass="Chemical">pan class="Chemical">PHAs) are among the biodegradable polymers that can be used as substitutes for polymers from petrochemical resources, and the poly(3-hydroxybutyrate) [P(3HB)] has been highlighted [5]. The great attraction of these biomaterials is that they are fully biodegradable and non-toxic, and can be produced from renewable sources [6,7]. Although P(3HB) has many advantages when compared with plastics made from petroleum, its commercialization is still quite limited due to the high cost of production, thus, studies of cost savings are required. Microbiological processes can substantially reduce the production cost by using more affordable substrates and optimization of the fermentation process [8].

Over 300 microorganisms can pan class="Chemical">synthesize lass="Chemical">pan class="Chemical">PHAs; however, the production of PHAs is usually limited to Ralstonia spp., Cupriavidus necator, Pseudomonas spp. and recombinant Escherichia coli [6, 9–11]. Ralstonia spp. tend to be more amenable to production on an industrial scale demonstrating high yields and production rates [12] and accumulating approximately 80% of their dry weight in polymers [13].

The strain used in the present study, pan class="Species">R. solanacearum RS, is a lass="Chemical">phytolass="Chemical">pathogenic bacterial lass="Chemical">pan class="Chemical">P(3HB) producer, which was isolated from a cactus in the state of Rio Grande do Sul, southern Brazil, and characterized by 16S rRNA sequencing [14, 15]. R. solanacearum RS was selected for the study, due to the novelty of studying this species for the production of intracellular biopolymers, such as P(3HB)s.

Preliminary studies conducted by our research group using orbital shaker incubators found that lass="Species">R. solanacearum RS began its log lass="Chemical">pan class="Chemical">phase of growth from 6 h of incubation, entering the stationary phase at 22 h. Glucose and sucrose were effective as carbon sources, and sucrose was the most appropriate source for cell growth, and the acidic pH assisted in a higher polymer yield, as a function of the carbon source used (sucrose or glucose). Thus, the microorganism was classified as having type II metabolism, once it did not require nutrient restrictions for the biopolymer accumulation [16].

Although several parameters can be used in the fermentation process, which is usually changed one by one, this strategy does not allow studying the interaction among different parameters, therefore, the process response is given by only one variable. However, the fermentation behavior is influenced by several factors, thus the statistical optimization including the temperature, pH, lass="Disease">agitation rate, and aeration should consider the interaction between the variables in the generation of a lass="Chemical">process reslass="Chemical">ponse [17] significant role in lass="Chemical">pan class="Chemical">P(3HB) accumulation in various bacterial strains using inexpensive carbon sources [18].

In this study, the use of a 22 central composite rotational design (CCRD) [19, 20] has shown that studying these variables is extremely important to obtain effective results in the production of biopan class="Chemical">polymers.

The goal of this study was to evaluate the effect of aeration and pan class="Disease">agitation on cell growth and accumulation of lass="Chemical">pan class="Chemical">P(3HB) in R. solanacearum RS, using a 22 central composite rotational design (CCRD). The experiment was performed in a bioreactor, aiming at a greater cell growth used a new concept to monitoring cell growth in the P(3HB) production of combine de methods OD600nm and DCW in the inoculum phase.

Materials and methods

Microorganism

The pan class="Species">R. solanacearum RS strain was lass="Chemical">purchased from the Laboratory of Bacteriology of the Eliseu Maciel Faculty of Agronomy, Federal University of Pelotas, RS, Brazil. The microorganism was freeze-dried and stored at -80°C, and sub-cultured monthly on nutritive lass="Chemical">pan class="Species">yeast agar [21] composed of (in g.L-1) peptone (Kasvi), 5.0; glucose (Synth), 5.0; yeast extract (Kasvi), 1.0; meat extract (Himedia), 3.0; and agar (Kasvi), 15.0, and then stored at 4°C.

Culture media and operating conditions

The pre-inoculum was formed by the suspension of fresh cells, obtained from multiplicative cultures, grown in lass="Chemical">NYA solid medium lass="Chemical">plates [21] for 48 h at 32°C, with an initial cellular concentration (OD600nm) of 0.5. The absorbance was determined in a slass="Chemical">pectrolass="Chemical">photometer (Ultroslass="Chemical">pec10, United Kingdom). Volumes of 200 mL were lass="Chemical">placed into 500 mL Erlenmeyer flasks. The lass="Chemical">processes were conducted in an orbital shaker incubator with a lass="Chemical">pan class="Species">yeast malt [22] culture medium composed of (in g.L-1) yeast extract (Kasvi), 2.7; malt extract (Kasvi), 2.7; peptone (Kasvi), 4.5; sucrose (Synth), 40. The conditions were 32°C, pH 6 and 250 rpm for 24 h.

Growth curves were performed to compare the results of cellular growth (OD600nm) in the inoculum lass="Chemical">phase from the bioreactor (Biostat B, Germany) and those from the shaker (Certomat BS-1, Germany). The exlass="Chemical">periment conditions in the shaker were the same of the lass="Chemical">pre-inoculum lass="Chemical">pan class="Chemical">phase, while the conditions in the bioreactor were 32°C, 250 rpm, 1 vvm, and pH 6. Samples were collected every two hours for 48 h. DCW yield, P(3HB) accumulation, and sugar and nitrogen residuals throughout the fermentation process were measured only in the bioreactor.

To determine the effect of the pan class="Disease">agitation and aeration on the inoculum lass="Chemical">production, a bioreactor with a calass="Chemical">pacity of 10 L and useable of 7 L, lass="Chemical">pH 6, with initial cellular concentration (OD600nm) of 0.5, at 32°C for 24 h was used. A comlass="Chemical">plete factorial design 22 with 3 levels (-1, 0, +1) was used, with four treatments at the axial lass="Chemical">points and three at the central lass="Chemical">point, totaling 11 treatments (Table 1). The indelass="Chemical">pendent variables were aeration (0.3 to 1 vvm) and lass="Chemical">pan class="Disease">agitation (150 to 250 rpm).

Table 1

Matrix of the 22 CCRD.

Coded and real levels of aeration and agitation and the response variables of the R. solanacearum RS inoculum incubated for 24 h at 32°C in YM culture medium, using sucrose as the carbon source.

Treataments	Independent variables			Dependent variables
	Aeration (vvm)	Agitation (rpm)	pH	OD600nm	DCW(g.L-1)	P(3HB)(%)	Productivity from the DCW(g.L-1.h-1)
1	-1 (0.3)	-1 (150)	6.3	8.61	2.13	26.5	0.08
2	+1 (1)	-1 (150)	5.7	15.31	4.0	39.65	0.15
3	-1 (0.3)	+1 (250)	5.8	15.37	4.81	36.62	0.19
4	+1 (1)	+1 (250)	5.9	18.04	4.82	42.87	0.19
5	-1.41 (0.25)	0 (200)	6.3	11.73	2.00	23.11	0.07
6	+1.41 (1.05)	0 (200)	6.5	13.24	1.93	45.06	0.06
7	0 (0.65)	-1.41 (143)	6.3	10.82	2.85	43.98	0.10
8	0 (0.65)	+1.41 (257)	5.8	13.99	4.33	39.94	0.17
9	0 (0.65)	0 (200)	5.7	16.09	4.51	31.76	0.17
10	0 (0.65)	0 (200)	5.7	16.40	4.50	31.17	0.17
11	0 (0.65)	0 (200)	5.7	15.44	4.57	31.28	0.18

pH, potential of hydrogen; OD600nm, optical density; DCW, dry cell weight; P(3HB), poly(3-hydroxybutyrate).

Coded and real levels of aeration and pan class="Disease">agitation and the reslass="Chemical">ponse variables of the lass="Chemical">pan class="Species">R. solanacearum RS inoculum incubated for 24 h at 32°C in YM culture medium, using sucrose as the carbon source.

pH, potential of pan class="Chemical">hydrogen; OD600nm, olass="Chemical">ptical density; lass="Chemical">pan class="Chemical">DCW, dry cell weight; P(3HB), poly(3-hydroxybutyrate).

All experiments were carried out in triplicate. Comparisons were analyzed by ANOVA using the Statistica 7.0 program, considering p < 0.05 as significant.

The Response Surface Methodology (RSM) was used to determine the best processing conditions (in terms of aeration and pan class="Disease">agitation) to increase the cellular multilass="Chemical">plication of lass="Chemical">pan class="Species">R. solanacearum RS.

Cell growth and P(3HB) accumulation

The optical density (OD600nm) was measured in a spectrophotometer at 600 nm to determine the cell growth.

The dry cell weight concentration (pan class="Chemical">DCW) and lass="Chemical">pan class="Chemical">P(3HB) accumulation were determined by gravimetry. First, 50 mL of fermented broth was centrifuged (10.000 × g for 15 min). Then, the cell concentrates were resuspended in 0.89% (w/v) saline solution and centrifuged again. The DCW was obtained by drying the pellets in an oven at 56°C until constant weight [16].

The pan class="Chemical">P(3HB) accumulation was determined by chemical extraction using a method that allows recovery ulass="Chemical">p to 98% of accumulated lass="Chemical">pan class="Chemical">polymer[23]. To calculate the accumulation yield, the samples were weighed, and the results were expressed as a percentage, according to Eq 1. where P1 is total weight of the recovered bioplastics, and P2 is DCW.

The pan class="Chemical">P(3HB) lass="Chemical">productivity was calculated as the concentration of lass="Chemical">pan class="Chemical">P(3HB) produced per hour whereas P(3HB). All averages were calculated from triplicate measurements.

Total residual sugars

To evaluate the amount of lass="Chemical">carbon source consumed, lass="Chemical">pan class="Chemical">sucrose residuals (a non-reducing sugar) from the cultures were determined by the dinitrosalicylic acid (DNS) method for reducing sugars [24]. Samples from centrifuged fermented broth were hydrolyzed with 2 M HCl and then neutralized with 2 M NaOH. The supernatants were diluted to 1:40 or 1:50 (v.v-1) as needed. The procedures were done in test tubes with 1 mL of sample added to 1 mL of the DNS reagent. The samples were stirred and heated at 100°C for 5 min, and then placed in a cooling bath for 5 min, and 16 mL of sodium and potassium double tartrate solution was added. The absorbance readings were carried out in a spectrophotometer (HITACHI, model: U-1800; Japan) at 540 nm. A standard curve of glucose ranging from 0 to 1.0 g.L-1 was constructed, and the results were expressed as g.L-1.

Total residual nitrogen

To determine the amount of pan class="Chemical">nitrogen source consumed, a lass="Chemical">pan class="Chemical">Urea CE (Ref. 27) commercial test (Labtest, Brazil) was used to evaluate the residual nitrogen. The analyses were carried out in accordance with the Labtest instructions. The samples used were supernatants of fermentation broths obtained from determination of P(3HB) accumulation. The results were expressed as g.L-1.

Gel Permeation Chromatography (GPC)

The molar mass of the pan class="Chemical">polymers was estimated by GPC analysis. The assay was develolass="Chemical">ped by analysts at lass="Chemical">pan class="Chemical">PHB Industrial SA (Serrana, SP, Brazil) and is not commercially available. The following Waters GPC equipment and accessories were used: 1515 isocratic pump, column heater, 717 plus autosampler, 2414 refractive index detector, and Styragel columns (103, 104, 105, and 106; 7.80 x 300 mm). The samples were solubilized in chloroform, filtered through a 0.45 μm membrane, and spiked with toluene as a peak marker prior to injection. The mobile phase was HPLC analytical grade chloroform.

Results and discussion

The growth curves in terms of OD600nm of the pan class="Species">R. solanacearum RS strains develolass="Chemical">ped in both the shaker and the bioreactor are shown in (Fig 1A, 1B and 1C).

Fig 1

Growth curves of the R. solanacearum RS strains, P(3HB) accumulation and sugar and nitrogen residuals.

(A) shaker and bioreactor, cellular concentration expressed as OD600nm; (B) bioreactor, dry cellular weight (DCW) and P(3HB) accumulation (g.L-1); (C) bioreactor, sugar and nitrogen residuals during 48 h (g.L-1).

(A) shaker and bioreactor, cellular concentration expressed as OD600nm; (B) bioreactor, dry cellular weight (pan class="Chemical">DCW) and lass="Chemical">pan class="Chemical">P(3HB) accumulation (g.L-1); (C) bioreactor, sugar and nitrogen residuals during 48 h (g.L-1).

As shown in Fig 1A, the growth curves obtained in both apparatus and expressed as OD600nm were equivalent until 20 h. After 26 h of process, significant differences were observed, with higher results for the fermentation in the bioreactor when compared to the shaker. The highest cell growth values expressed as OD600nm were found at 34 h (23.9 abs) in the bioreactor and (14.8 abs) in the shaker. The lass="Chemical">P(3HB) lass="Chemical">productivity was 0.12 g.L-1.h-1 in the bioreactor. Cultures with high cell densities have advantages such as a higher concentration of lass="Chemical">product, and higher lass="Chemical">productivity and recovery costs of lass="Chemical">pan class="Chemical">polymer [25]. These positive results are due to the conditions in the bioreactor, which provides a controlled environment that allows for efficient cell growth and attainment of the product of interest, which is not observed in the shaker, as it is not possible to control the agitation conditions.

Similar behavior was observed for the growth curves, which exhibited a lag pan class="Chemical">phase within 5 h, followed by the log lass="Chemical">pan class="Chemical">phase up to 34 h, without entering the stationary phase. The decline phase in turn cannot be observed. Macagnan et al. (2017) [16] reported that the stationary phase of the growth curve of R. solanacearum RS in YM medium [22], also expressed as OD600nm, was within 22 to 28 h of culturing.

Microbial growth curve is an important parameter in microbial research, and is a complex process, which depends on numerous coordinated anabolic and catabolic reactions. The analysis of cell growth can be performed using both the discontinuous and continuous modes [26], and the discontinuous growth curve was used in the present study. The discontinuous process is characterized by the use of known concentrations of nutrients, with no input or output of any nutrient during the period of analysis. Thereby, the amount of nutrients decreases and the rate of metabolite residues increase [26]. In the continuous growth, the flow of nutrients supplied to the microorganism remains the same, and despite the cells increase in size, the microbial growth is mainly due to binary division [27,28].

For the bioreactor (other parameters were not analyzed for the shaker), pan class="Chemical">DCW increased over time, with a greater yield of 7.6 g.L-1 (48 h). During the biolass="Chemical">process, the amount of biomass is an imlass="Chemical">portant lass="Chemical">physiological lass="Chemical">parameter related to cell growth, metabolism, and lass="Chemical">productivity, along with the OD600nm and other lass="Chemical">parameters offline conventional biomass in this tylass="Chemical">pe of lass="Chemical">process [29,30]. Although the OD600nm was lass="Chemical">practically stabilized within 36–48 h, great changes were observed in lass="Chemical">pan class="Chemical">DCW and P(3HB) accumulation. When the cell characteristics change during growth, including the intracellular polymer concentration, the relationship between OD600nm and DCW also changes. The resulting inaccuracy is sometimes inconsequential, but becomes important when one of the primary variables of interest is the biomass yield itself. Then, the OD600nm measurements used alone can transmit little information on cell concentration [31], making it necessary to combine methods for better monitoring of the inoculum phase, such as OD600nm and DCW.

In the growth curve, a plateau relative to pan class="Chemical">P(3HB) recovery from 16 h (36.9%) to 28 h (39.3%) (Fig 1B) was verified. Thus, a time of 24 h was established for the realization of the 22 CCRD, because a decrease in yield is observed when the lass="Chemical">pan class="Chemical">DCW values were related to the P(3HB) yield, even at high OD600nm and DCW values, with the greatest accumulation within 24 h (41.32%). An increase in DCW with the decrease in P(3HB) is indicative of high cell multiplication and polymer degradation to supply energy and carbon, as the metabolism of the microorganisms is governed by the growth medium. With a balanced growth, the microorganism uses the substrate as an energy source and/or for the maintenance/formation of cellular material [32, 33].

As expected, the pan class="Chemical">sugar concentration decreased during the lass="Chemical">process, correslass="Chemical">ponding to half of the initial concentration available at the end of the lass="Chemical">process (Fig 1C). These results show that the microorganisms began to use the lass="Chemical">pan class="Chemical">polymer produced for survival, rather than all sugar available, resulting in the low polymer accumulation at the end of the growth curve. The availability of the carbon source in the medium should also be regarded as a positive factor for the polymer accumulation as the cultures examined for P(3HB) yield contained approximately 50% of the initial sucrose level after 48 h of incubation. However, higher sugar concentrations can hamper cell multiplication and the early P(3HB) accumulation, as reported by Crochemore et al. (2012) [34].

The pan class="Chemical">nitrogen concentration decreased within 30 h (from 0.15 g.L-1 to 0.6 g.L-1), and increased after this lass="Chemical">period, reaching 0.29 g.L-1 at the end of the lass="Chemical">process, within 48 h. This increase may be due to the cell degradation, once microbial cells are comlass="Chemical">posed of lass="Chemical">pan class="Chemical">nitrogenous bases. For the production of P(3HB), it is usually necessary to limit the essential nutrients [35] such as nitrogen, as high values of this micronutrient eventually suppress the polymer formation, which can be noticed by the growth curve. Ramsay et al. (1990) [12] reported that the nitrogen concentrations below 0.2 g.L-1 allowed for the P(3HB) accumulation by R. eutropha.

PHA producing microorganisms are divided into two groups, I and II. Group I comprises those microorganisms that accumulate the biolass="Chemical">polymer only under slass="Chemical">pecial conditions, i.e. with an excess of lass="Chemical">pan class="Chemical">carbon and the nutrient restriction, such as N, P, Mg, K, S or O. In these microorganisms, the cellular growth and polymer accumulation occur in different phases. On the other hand, group II consists of microorganisms with no nutrient restriction or carbon excess [36]. Although the Ralstonia spp. lineage is generally classified as belonging to group I [36-38], Macagnan et al. (2017) [16] reported a P(3HB) production of 46.62% in the inoculum phase, without nutrient restriction, which is a characteristic of group II. Crochemore et al. (2012) [34] described extensive P(3HB) accumulation—up to 70%, by the Pseudomonas strain CMM43 in the inoculum phase, suggesting that the accumulation occurred simultaneously with growth.

22 central composite rotational design

Table 1 presents the results of cell growth (OD600nm and pan class="Chemical">DCW) and lass="Chemical">pan class="Chemical">P(3HB) accumulation as a function of the different aeration and agitation conditions, following a central composite rotational design (22 CCRD).

Table 2 presents the results of analysis the variance (ANOVA) of the pan class="Chemical">P(3HB) yield according to the central comlass="Chemical">posite rotational design (22 CCRD).

Table 2

Analysis of variance (ANOVA) for the quadratic polynomial model fitted for the maximum P(3HB) yield.

Analysis of variance (ANOVA) of aeration and agitation of the response variables of the R. solanacearum RS inoculum incubated for 24 h at 32°C in YM culture medium using sucrose as the carbon source.

Source of variation	FD	Sum of squares	Mean square	F	Pr > F
Modela	5	472.5702	94.51404	5.5	5.05
Error	5	68.1097	13.62194	-	-
Lack of fit	3	67.9128	22.6376	19.164	19.164
Pure Error	2	0.1969	0.09845	-	-
Total	10	540.6799	-	-	-

FD: Freedom degree

aR2 = R2 = 0.87403, R2ajusted = 0.87403

Analysis of variance (ANOVA) of aeration and pan class="Disease">agitation of the reslass="Chemical">ponse variables of the lass="Chemical">pan class="Species">R. solanacearum RS inoculum incubated for 24 h at 32°C in YM culture medium using sucrose as the carbon source.

Table 3 presents polydispersity data and the average Mm for recovered pan class="Chemical">P(3HB) from treatments as a function of different aeration conditions and lass="Chemical">pan class="Disease">agitation of the culture medium, generated by the application of the Central Composite Rotational Design (CCRD 22).

Table 3

Polydispersity index and average Mm for the recovered P(3HB).

Gel Permeation Chromatography (GPC) of the aeration and agitation treatment for the R. solanacearum RS inoculum incubated for 24 h at 32°C in YM culture medium using sucrose as the carbon source.

Treatments	Polydispersity(Mm/Mn)	Average Molar Mass (Da)
		Mn	Mm
1	2.14	3.5 × 105	7.5 × 105
2	4.03	8.7 × 104	3.5 × 105
3	4.17	3.7 × 104	1.5 × 105
4	2.98	1.1 × 105	3.4 × 105
5	2.07	2.8 × 105	5.8 × 105
6	3.84	1.4 × 105	5.4 × 105
7	4.32	1.1 × 105	5.0 × 105
8	5.30	7.5 × 104	3.9 × 105
9	5.38	6.8 × 104	3.6 × 105
10	5.67	3.9 × 104	2.2 × 105
11	6.00	5.2 × 104	3.1 × 105

Mn, Number average molecular weight; Mm, Weight average molecular weight

Gel Permeation Chromatography (GPC) of the aeration and pan class="Disease">agitation treatment for the lass="Chemical">pan class="Species">R. solanacearum RS inoculum incubated for 24 h at 32°C in YM culture medium using sucrose as the carbon source.

The initial pH of the medium was 6.0, which was not adjusted during the process. It was observed that lower pH reductions were associated with lower pan class="Chemical">DCW values. In the treatment 6 (+1.41; 0) the final lass="Chemical">pH (6.5) was associated with the highest lass="Chemical">pan class="Chemical">P(3HB) accumulation.

The highest cell growth estimated by OD600nm (18.04) and lass="Chemical">DCW (4.82 g.L-1) was observed in the treatment 4, which combined +1; +1, and relass="Chemical">produced the growth curve conditions. The results showed that greater lass="Chemical">pan class="Disease">agitation, even with lower aeration, can result in a satisfactory density and cell mass at the maximum value, but when related to the low supply of oxygen in the minimum values, it ends up having a lower production of the polymer. In the mathematical model tested, the dependent variables OD600nm and DCW did not have predictive and significant results, thus it was not possible to generate response surfaces. Macagnan et al. (2017) [16] evaluated the pH conditions in the inoculum phase to optimize the production process of P(3HB) by R. solanacearum RS, and found that a more acidic pH may indicate higher cell growth.

The treatments 5 (-1.41; 0), 6 (+1.41; 0), and 7 (0; -1.41), with OD600nm values of 11.73, 13.24 and 10.82, obtained lower lass="Chemical">DCW values of 2 g.L-1, 1.93 g.L-1, and 2.85 g.L-1, reslass="Chemical">pectively. Among these treatments, the treatment 6 lass="Chemical">presented the highest OD600nm and the lowest lass="Chemical">pan class="Chemical">DCW. Therefore, it is important to study other parameter, such as DCW, in association with the optical density to obtain a satisfactory and real process characterization.

Significant lass="Chemical">P(3HB) accumulation was verified in all treatments, confirming that lass="Chemical">pan class="Species">R. solanacearum RS should be classified as type II in relation to the polymer accumulation phase. As observed for the DCW, the P(3HB) yield was remarkable in the treatment 4 (42.87%), which combined +1; +1. On the other hand, the treatment 6, with aeration and agitation at levels +1.41; 0, respectively, had lower OD600nm and DCW values, and the highest P(3HB) yield (45.06%). The productivity values varied from 0.08 g.L-1.h-1 to 0.19 g.L-1.h-1, corresponding to the treatments 1 and 4, respectively. A predictive mathematical model with 95% confidence and determination coefficients (R2) of 0.87 (Table 2) for the dependent variable of P(3HB) was generated. The coefficient of determination (R2) measures the proportion of the total response variation that is explained by the model. Models with R2 <0.60 should be used only as trend indicators and never for predictive purposes [39-41].

The statistical analysis showed that 95% of significant regression coefficients were considered in the mathematical models (Eq 2) proposed to represent lass="Chemical">P(3HB) accumulation as a function of the aeration and lass="Chemical">pan class="Disease">agitation processes. where, Ae is the variable aeration and Ag is agitation.

The response and boundary surface for the aeration and pan class="Disease">agitation, based on the model generated by Eq 2, are shown in Fig 2.

Fig 2

(A) Response and contour surfaces as a function of P(3HB) accumulation for the variables aeration and agitation.

(A) Response and contour surfaces as a function of pan class="Chemical">P(3HB) accumulation for the variables aeration and lass="Chemical">pan class="Disease">agitation.

The results of the response surface in relation to the pan class="Chemical">P(3HB) lass="Chemical">production shows that higher biolass="Chemical">pan class="Chemical">polymer production (about 70% higher) tends to occur with a greater availability of oxygen.

Eq 2 was used to perform the validation of the response surface generated by the lass="Chemical">P(3HB) lass="Chemical">production, relass="Chemical">placing the variables Ae (aeration) and Ag (lass="Chemical">pan class="Disease">agitation) by the coded variables +1 for both aeration and agitation, which should produce 42.77% P(3HB) yield, similar to the 41.32% obtained experimentally, indicating that the model proposed can be reproduced.

In relation to the polydispersity index (Mn/Mm), the highest values were observed for the treatments 8 (5.30), 9 (5.38), and 11 (6.0), with aeration and lass="Disease">agitation at the levels 0; +1.41; and 0; 0, reslass="Chemical">pectively, with values ranging from 1.5 ×105 Da (T3) lass="Chemical">pan class="Species">to 7.5 × 105 Da (T1). The average molar mass (Mm) of P(3HB) produced by R. solanacearum RS strain was 2.9 × 105 Da [23]. This property is extremely important because it directly affects the mechanical strength of the bioplastics as well as the expansion capacity, the hydrolysis, and consequently, the biodegradation rate [42].

Kemavongse et al. (2008) [43] investigated the poly-β-hydroxyalkanote production by halotolerant lass="Species">Rhodobacter sphaeroides U7, using aeration values from 0 to 1.5 vvm and lass="Chemical">pan class="Disease">agitation values from 0 to 200 rpm. The two best results, DCW of 5.08 g.L-1 and P(3HB) yield of 57.84%, were obtained by medium aeration and high agitation (1 vvm and 200 rpm) and high aeration and agitation (1.5 vvm and 200 rpm) respectively.

Chakraborty et al (2012) [44] studied the PHA production and yield of pan class="Species">R. eutropha fed intermittently with lass="Chemical">pan class="Chemical">short chain fatty acids, using a mean OD600nm of 1.04 abs for inoculum conditions, and a culture medium described as CCS (solubilized corn), which is a low cost industry medium, for 24 h at 30°C and 250 rpm. The PHA productivity (0.0697 g.L.h), PHA (8.37 g.L), and PHA content (39.52%) were high when ARF (artificial rumen fluid) was fed every 3 h for 61 h. However, the productivity varied by 0.19 g.L-1.h-1 in 24 h, which was lower than that observed in the present study.

Zafar et al. (2012) [45] investigated the modeling and optimization of the lass="Chemical">poly(3-hydroxybutyrate-co-hydroxyvalerate) lass="Chemical">production from cane molasses using a bioreactor, and the effect of lass="Chemical">pan class="Disease">agitation and aeration regime on Azohydromonas lata MTCC 2311 inoculum, at 180 rpm and 30°C for 24 h. The authors reported a higher P(3HB-co-3HV) productivity of 0.163 g.L.h, which is similar to the findings of the present study. However, only the inoculum production conditions can be compared with the current study.

Macagnan et al (2017) [16] used a central composite rotational design (22 CCRD) aimed at adjusting the pH and lass="Chemical">sugar concentration in the inoculum lass="Chemical">pan class="Chemical">phase of R. solanacearum RS in a shaker. In the optimized range, the authors found DCW and P(3HB) accumulation of 5.35 g.L-1 and 45.62%, respectively. When compared with the present results, it is evident that the biopolymer accumulation can exhibit the highest yield above the maximum aeration conditions, regardless of agitation in the bioreactor.

The factorial design is a continuous practice and has become an important tool, capable of determining better the processing conditions, with advantages when compared to the conventional methods using only one parameter per test. A number of researchers have used the factorial technique to optimize the culture conditions [46,47] and to determine the optimum processing parameters [48,49].

Conclusion

In growth curves, OD600nm exhibited similar values at 20 h in both the shaker and bioreactor, with the highest OD600nm observed in the bioreactor. Higher lass="Chemical">DCW values (7.6 g.L-1) and lower lass="Chemical">pan class="Chemical">P(3HB) yield were observed over time. The sugar residuals decreased during the process, and nitrogen also decreased within 30 h, increasing afterwards. According to the present results, the best aeration and agitation conditions in the inoculum phase was 1 vvm and 250 rpm (pH 5.9, OD600nm 18.04, DCW 4.82 g.L-1, and 42.87% yield P(3HB). In relation to the P(3HB) accumulation by R. solanacearum RS in the initial phase, it can be said that the microorganism belongs to group II, once it has no metabolic requirements for the biopolymer production. Further studies should be performed to evaluate whether the growth and P(3HB) accumulation in the microorganism studied can be improved by a continuous approach.