Two-stage fermentation process for alginate production by Azotobacter vinelandii mutant altered in poly-β-hydroxybutyrate (PHB) synthesis


  • M.Á. Mejía,

    1.  Departamentos de Ingeniería Celular y Biocatálisis y de 2 Microbiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Morelos, México
    Search for more papers by this author
  • D. Segura,

    1.  Microbiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Morelos, México
    Search for more papers by this author
  • G. Espín,

    1.  Microbiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Morelos, México
    Search for more papers by this author
  • E. Galindo,

    1.  Departamentos de Ingeniería Celular y Biocatálisis y de 2 Microbiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Morelos, México
    Search for more papers by this author
  • C. Peña

    1.  Departamentos de Ingeniería Celular y Biocatálisis y de 2 Microbiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Morelos, México
    Search for more papers by this author

Carlos Peña, Departamento de Ingeniería Celular y Biocatálisis, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Apdo. Post. 510-3, Cuernavaca, 62250, Morelos, México. E-mail:


Aims:  A two-stage fermentation strategy, based on batch cultures conducted first under non-oxygen-limited conditions, and later under oxygen-limited conditions, was used to improve alginate production by Azotobacter vinelandii (AT6), a strain impaired in poly-β-hydroxybutyrate (PHB) production.

Methods and Results:  The use of sucrose as carbon source, as well as a high oxygen concentration (10%), allowed to obtain a maximum biomass concentration of 7·5 g l−1 in the first stage of cultivation. In the second stage, the cultures were limited by oxygen (oxygen close to 0%) and fed with a sucrose solution at high concentration. Under those conditions, the growth rate decreased considerably and the cells used the carbon source mainly for alginate biosynthesis, obtaining a maximum concentration of 9·5 g l−1, after 50 h of cultivation.

Conclusion:  Alginate concentration obtained from the AT6 strain was two times higher than that obtained using the wild-type strain (ATCC 9046) and was the highest reported in the literature. However, the mean molecular mass of the alginate produced in the second stage of the process by the mutant AT6 was lower (400 kDa) than the polymer molecular mass obtained from the cultures developed with the parental strain (950 kDa).

Significance and Impact of the Study:  The use of a mutant of A. vinelandii impaired in the PHB production in combination with a two-stage fermentation process could be a feasible strategy for the production of alginate at industrial level.


Alginates are linear polysaccharides composed of (1-4)-β-d-mannuronic acid and its epimer, α-l-guluronic acid, of both technological and scientific interest (Gacesa 1998). Currently, commercial alginates are extracted from marine brown algae and are used for a variety of food and pharmaceutical applications, and recently, new potential applications in the medical field has been reported (Brownlee et al. 2005; Remminghorst and Rehm 2006; Galindo et al. 2007).

The alginate production by fermentation using bacteria as Azotobacter vinelandii could be a feasible strategy for the synthesis of this polymer; however, the alginate concentrations in batch cultures reported so far in the literature are very low, finding values of maximal alginate concentration in the range of 3–5 g l−1 (Peña et al. 1997, 2000; Parente et al. 1998; Sabra et al. 2000). It is important to point out that the final alginate concentration is the most important parameter in determining the economics of the process, as it is related to the recovery costs, especially during the initial isolation of the polymer by precipitation (Sutherland 1990).

The low volumetric yield of alginate is partly because of the simultaneous synthesis of poly-β-hydroxybutyrate (PHB) in A. vinelandii, which is produced under oxygen-limited conditions at the same that the alginate (Peña et al. 1997; Sabra et al. 2000). PHB content may comprise up to 70% of cellular dry weight in certain strains of A. vinelandii (Brivonese and Sutherland 1989; Peña et al. 1997). PHB synthesis thus constitutes a waste of substrate when seeking to optimize alginate production. An interesting strategy is to block the synthesis of PHB (by mutation), and therefore the alginate production could be improved, overcoming one impediment for the production of alginate by A. vinelandii. Three enzymes encoded by the phbBAC operon are responsible for the synthesis of PHB in A. vinelandii (Segura et al. 2003b).

The mutant strain AT6 is an ATCC 9046 derivative carrying a phbB::Tn5-lacZ mutation, polar on phbA and phbC, that totally impairs PHB synthesis (Segura et al. 2003a). Although a fourfold increase in the alginate yield (mg alginate/mg of protein) was obtained from shake flasks with respect to the parental strain (ATCC 9046), the volumetric production of alginate using the mutant strain was very low (2·0 mg ml−1), because of its poor growth. In order to overcome this fact, it is necessary to implement a different approach. One of these strategies might consist in the use of multistage fermentation processes or fed-batch cultures that are able to achieve high biomass concentration, in order to take advantage of the higher specific-alginate production capacities of strains such as the AT6.

Fermentation processes involving two or three stages have been widely used for the production of biodegradable polymers, such as the polyhydroxyalkanoates (PHAs) (Song et al. 1993; Chen and Page 1997; Ruan et al. 2003; Rocha et al. 2008). In all those reports, the PHAs production has been improved, reaching up to 36 g l−1 of polymer in a 2·5-l fermentor (Chen and Page 1997). In the case of the alginate, there is no information about the application of a multistage fermentation system. Recently, Priego-Jiménez et al. (2005) reported that by using an exponentially fed-batch culture of an A. vinelandii mutant unable to synthesize alginase (enzyme that degrade the alginate), it was possible to obtain alginates with a mean molecular mass (MMM) 15 times higher (1300 kDa) than that obtained from batch culture. In addition, the alginate yield on biomass was about twice that observed in the batch culture.

In the present work, a two-stage fermentation process was developed to improve the alginate concentration using a mutant strain of A. vinelandii, carrying a mutation that impairs its PHB biosynthetic capacity and which have a stimulatory effect on the production of alginate. The behaviour of the mutant was contrasted with that observed with the parental strain ATCC 9046.

Materials and methods


The A. vinelandii strains used in this study were: ATCC 9046 and its derivate, AT6, which contains a phbB::mini-Tn5-lacZ (Segura et al. 2003b) and therefore does not produce polyhydroxybutyrate (PHB). The cells were maintained by monthly subculture on Burk’s agar slopes and stored at 4°C.

Culture medium and growth conditions

A. vinelandii was grown on a modified Burk’s medium (Peña et al. 1997), containing (in g l−1) sucrose (20), yeast extract (3), K2HPO4 (0·66), KH2PO4 (0·16), CaSO4 (0·05), NaCl (0·2), MgSO4·7H2O (0·2), Na2MoO4·2H2O (0·0029), FeSO4 (0·027). The initial pH level was adjusted to 7·2, using NaOH (2 N).

Preparation of inoculum

In order to start with a high concentration of cells in the batch cultures, the inoculum was prepared as follows. A. vinelandii cells were grown at 29°C in a New Brunswick G-25 shaker (NJ, USA) in four 500-ml Erlenmeyer flasks, containing 100 ml of modified Burk’s medium for 36 h at 200 rev min−1. Flasks were incubated until they reached a biomass concentration of 1·8 g l−1 (measured by dry weight). The cells obtained (0·72 g) were transferred to a bioreactor containing 2 l of Burk’s medium and incubated for 24 h at 5% of dissolved oxygen tension (DOT), 700 rev min−1 and pH was kept constant at 7·2. After that time, a biomass concentration of 4 g l−1 (measured by dry weight) was obtained. The biomass obtained was aseptically collected by two centrifugation steps at 15 500 g, for 10 min and resuspended in 200 ml of fresh Burk’s liquid medium. This suspension was inoculated into the fermenter containing 1800 ml of fresh culture medium.

Batch cultures

Batch cultures were carried out in a 3.0-L Applikon bioreactor (Applikon, Netherlands) using 2·0 l of working volume. This was operated at 700 rev min−1 and 29°C at an airflow of 1 l min−1. The pH was measured with an Ingold (ADI 1010; Applikon) probe and controlled by an on/off system using a peristaltic pump and a 4 N NaOH solution. DOT was measured with an Ingold polarographic probe and controlled at 10% of the saturation value (by gas blending) using a system based on a PID control that has been previously described (Trujillo-Roldán et al. 2001).

Two-stage fermentations

Two stage fermentations were conducted in the same bioreactor described earlier. A vessel in which sucrose (400 g l−1) and nutrients of Burk’s medium were stored was incorporated into the fermentation system. Two-stage fermentation was started in batch mode, and sucrose feed began in the prestationary phase of growth. The sucrose solution was fed to an average flow rate of 36 ml min−1. At the same time, the DOT was changed from 10% to 1% by means of the manipulation of the partial pressure of nitrogen and oxygen inflowing to the fermentor. In order to avoid the expression of alginases, which degrade the alginate, the agitation rate was decreased from 700 to 300 rpm (Peña et al. 2000).

Analytical determinations

Determination of biomass, alginate and sucrose

Biomass and alginate concentration was determined gravimetrically as described previously by Peña et al. (1997). Sucrose was assayed for reducing power with DNS reagent (Miller 1959). The specific growth rate was calculated by the logistic equation, as described by Klimek and Ollis (1980), and the specific production rate was calculated by linear regression using the available data points between the beginning of the culture and the time of maximal production of alginate.

Poly-β-hydroxybutyrate (PHB) determination

For analysis, PHB was first extracted from cells debris as described by Peña et al. (1997). The PHB was hydrolysed using concentrated H2SO4 in a boiling water bath for 10 min. The quantification of PHB (as crotonic acid, a product of PHB hydrolysis) was assayed using a high performance liquid chromatography (HPLC) system with a UV detector and an Aminex HPX-87H ion-exclusion organic acid column (Karr et al. 1983). Elution was performed with 0·014 N H2SO4 at a flow rate of 0·7 ml min−1 at 50°C.

Characterization of the alginate

Molecular mass distribution

The molecular mass of alginate was estimated by gel permeation chromatography with a serial set of Ultrahydrogel columns (UG 500 and Linear Waters), using a HPLC system with a differential refractometer detector (Waters, 2414). Elution was performed with 0·1 mol l−1 NaNO3 at 35°C at a flow rate of 0·9 ml min−1. Calibration of the columns was performed via a standard calibration method using pullulans of Aureobasidium pullulans as molecular mass standards with a range from 360 to 780 000 Da.


First stage, cultures under non-oxygen-limited conditions

Azotobacter vinelandii AT6, a strain impaired in PHB synthesis, exhibits a high production of alginate but a poor growth when cultivated in shake flasks (Segura et al. 2003a). To improve alginate production of this strain, a two-stage fermentation strategy was implemented. In the first stage, the objective was to reach a high concentration of biomass to achieve a higher concentration of the polysaccharide during the second stage.

To establish the conditions for the first stage, A. vinelandii was cultivated in batch mode using sucrose as carbon source and supplementing with yeast extract as nitrogen source. Cultures were conducted at DOT of 10%, as under that condition A. vinelandii grows efficiently (Peña et al. 2000; Sabra et al. 2000). The DOT was controlled satisfactorily by gas blending at 10% of DOT. As shown in Fig. 1a, the high DOT in the culture promoted the cellular growth of AT6, reaching a maximum production of biomass of 7·5 g l−1 at 24 h of cultivation with a specific growth rate (μ) of 0·13 h−1 (doubling time = 5·3 h).

Figure 1.

 Biomass growth kinetics (a), alginate production (b), and sucrose uptake (c), of cultures of Azotobacter vinelandii AT6 under nonnitrogen-fixing conditions in batch cultures.

Figure 1b shows the production of alginate during the course of the culture. It is interesting to point out that alginate production was also promoted under the conditions tested, obtaining a maximum concentration of 2·2 g l−1. Under this condition, the alginate production rate was 0·11 g l−1 h, and the alginate production was totally growth associated (Fig. 1b).

As shown in Fig. 1c the sucrose was not exhausted at the end of the cultivation (24 h); this indicates that sucrose was not the limiting substrate, thus, indicating that other nutrient can be the factor responsible for the limitation of biomass growth. The biomass yield (Yx/s) was 0·5 gbiom/gsuc.

The two-stage fermentation (TSF)

For the TSF, the former conditions were used in the first stage, and the second stage was started by sucrose feeding in the prestationary phase of growth (18 h) with the AT6 mutant. At the same time, the DOT was changed from 10% to 1%. Figure 2 shows the results obtained during the cultivation of A. vinelandii in a two-stage fermentation process. As it is shown in Fig. 2a, at the end of the first stage, about 7·5 g l−1 of biomass and 2·2 g l−1 of alginate were produced. In the second stage (after 18 h of cultivation) the μ decreased significantly up to 0·07 h−1. Figure 2b shows that, in the second stage of the fermentation, the sucrose uptake rate decreased (from 0·83 g l−1 h to 0·35 g l−1 h), as a result of the shift in the DOT. With respect to the alginate concentration, the results are shown in Fig. 2c. During the second stage, A. vinelandii AT6 produced 7·3 g l−1 of alginate and therefore, at the end of the cultivation, 9·5 g l−1 were reached. In this stage, the alginate production rate was of 0·18 g l−1 h and therefore 1·65-fold times higher than the calculated rate from the data of the first stage of fermentation.

Figure 2.

 Two-stage process for alginate production using strains Azotobacter vinelandii AT6 (inline image) and ATCC 9046 ( inline image ). Biomass growth (a), sucrose uptake (b), alginate production (c), PHB production (d), mean molecular mass (e).

In order to compare the behaviour of strain AT6 with that of the parental strain (ATCC 9046), this latter strain was cultured under the same conditions used for the mutant. Figure 2a shows that the wild-type strain (ATCC 9046) reached a maximal biomass concentration of 8·2 g l−1 after 12 h during the first stage of the process, with a specific growth rate (μ) of 0·2 h−1. A difference with the results observed with the AT6 mutant was that the biomass concentration remained constant throughout the second stage of the culture.

During the second stage, the maximal alginate concentration obtained from the wild-type strain was 4·8 g l−1 after 56 h of cultivation (Fig. 2c). This value is considerably lower than that obtained from the cultures conducted using the AT6 strain (9·5 g l−1). The low alginate production by the ATCC 9046 strain is associated with the high production of PHB, as it is shown in Fig. 2d, confirming that PHB and alginate synthesis compete for the available carbon source.

In order to establish a comparative analysis between the distribution of carbon along the culture, the sucrose yield on biomass (Yx/s), alginate (Yalg/s) and PHB (YPHB/s) and the sucrose global yield (equivalent to the sum of the previous yields) were calculated for each stage of the culture (Table 1). In the first stage, under high DOT (10%), both strains exhibited a higher yield of biomass on sucrose (Yx/s) with respect to that observed in the second stage. However, during this stage, the decrease in the DOT allowed a higher alginate yield on sucrose (Yalg/s) for the mutant AT6 (0·74 g alginate/g sucrose) with a consequent decrease in the biomass yield. On the contrary, under the same condition (low DOT), the strain ATCC 9046 increased the PHB yield from 0·03 to 0·44 g PHB/g sucrose.

Table 1.   Biomass yields on sucrose (Yx/s), alginate yield on sucrose (Yalg/s), PHB yield on sucrose (YPHB/s) and global yield (Yglobal) calculated for the cultures of AT6 mutant and ATCC 9046
ATCC 9046First stage0·35 ± 0·040·04 ± 0·0060·03 ± 0·0010·42
Second stage0·11 ± 0·020·34 ± 0·040·44 ± 0·020·89
AT6First stage 0·5 ± 0·040·15 ± 0·030·65
Second stage 0·2 ± 0·040·74 ± 0·20·94

Mean molecular mass of the alginate

Figure 2e shows the evolution of the MMM of the alginate obtained from both A. vinelandii strains. Although a high alginate volumetric production (9·5 g l−1) was obtained from the mutant AT6, the MMM of the polymer was lower than that determined in the alginate isolated from the wild-type cultures. Analysing the polymer isolated after 55 h of culture using the wild-type strain, a maximal molecular mass of the polymer of 950 kDa was determined. In contrast, the maximum molecular mass of alginate isolated from cultures carried out with mutant AT6 was considerably lower (400 kDa) after 70 h of cultivation.

During the first stage of the culture (when the DOT was 10%), the MMM of the alginate obtained from the cultures of both strains was practically constant, having values around 100 kDa. During the second stage (DOT close to cero), the molecular mass increased along the culture as μ decreased, reaching values of 400 and 950 kDa for the AT6 and parental strain respectively.


The blockage in the PHB biosynthetic pathway in strain AT6 together with a two-stage fermentation strategy allowed improving the synthesis of alginate by A. vinelandii. The maximum biomass concentration obtained under nitrogen nonfixing conditions in the bioreactor (7·5 g l−1; Fig. 1a) was higher than that previously reported by Segura et al. (2003a) in shake flasks (0·7 g l−1), and even to that reported for the cultivation of the wild-type strain, ATCC 9046 (5·0 g l−1), in batch culture at high DOT (5%) (Peña et al. 2000). During the second stage of the culture, the specific growth rate decreased dramatically, because of the oxygen limitation imposed in the culture (oxygen close to 0%); a linear-type growth, characteristic of cultures limited by oxygen (Díaz-Barrera et al. 2007), was observed. The cells grew to reach a biomass concentration of 9 g l−1 after 40 h of cultivation, and after 70 h a slight decrease was detected, probably because of cellular lysis.

The alginate concentration reached with the AT6 mutant (9·5 g l−1) in the second stage is the highest reported so far in the literature using wild-type or other mutant strains of A. vinelandii (Parente et al. 1998; Sabra et al. 1999; Peña et al. 2000, 2002; Galindo et al. 2007). For example, this value was nearly three times higher than that previously reported by Peña et al. (2002), using strain DM, a mutant altered in its PHB biosynthetic capacity and in the regulation of the algD gene, which codes for the GDP-mannose dehydrogenase, the key enzyme of the alginate biosynthetic pathway (Campos et al. 1996).

From a practical point of view, the data show that the cultures of the AT6 strain in a two-stage fermentation system would allow to obtain a high alginate concentration and therefore will make the process more competitive. As the sucrose is not limiting for the alginate production (Fig. 2b), it is possible that other nutrient(s), unknown at this time, could be responsible for the end of the alginate synthesis. Therefore, an approach to further improve the process would be the new supply (before 60 h of cultivation) of salts and trace elements of the Burk’s medium to the fermentor.

Cultures using the wild-type strain showed in the second stage that the PHB yield increased from 0·03 to 0·44 g PHB/g sucrose (Table 1). These results confirm that PHB synthesis in A. vinelandii ATCC 9046 is enhanced under oxygen limitation, a situation that clearly does not occur in the mutant AT6 under the same conditions. Our results are in agreement with previous studies (Peña et al. 1997; Page et al. 2001), which indicated that, under low oxygen concentration, the bacterium uses the carbon source mainly for PHB production. On the other hand, the sucrose global yield is very similar in both strains (Table 1), indicating that the amount of PHB accumulated by the wild type is equivalent to the additional concentration of alginate produced by the mutant strain AT6.

The data shown in Table 1 reveal a better distribution of carbon towards the synthesis of alginate in the AT6 mutant (as a result of the impaired PHB synthesis) in combination with a low DOT lead to the A. vinelandii to produce a high alginate concentration. The high efficiency of conversion of sucrose to alginate, 0·74 g of alginate per g sucrose consumed, could lead to a significant decrease in the costs of the substrate, which in the case of production of biopolymers such as alginate can account for about 30% of the final product cost (Sutherland 1990). Thus, the production of bacterial alginate by AT6 strain of A. vinelandii in a two-stage fermentation process could be technically feasible for the production of this polymer at industrial level.

The results shown in Fig. 2e reveal important differences in the molecular mass of the alginate depending upon the strain employed. An alginate having a low molecular mass (400 kDa) was obtained from the cultures with the AT6 mutant; whereas the alginate isolated from cultures with the parental strain exhibited a molecular mass of 950 kDa. Although no conclusive explanation can be given, it is possible that the transcription of alg8, alg44, algX or algK (as parts of polymerase complex) or the activity of the enzymes encoded by these genes had been affected during the construction of AT6 mutant. In previous studies, it has been reported that the Alg L protein (alginate-lyase activity) is not necessary to modify the final molecular mass of the alginate produced. This task should be restricted to the polymerase itself (Trujillo-Roldán et al. 2004). However, further studies will be necessary to elucidate in detail the mechanism.

It is important to point out that the alginates produced in the second stage include fractions with a low molecular mass (produced under non-oxygen-limited conditions), as well as polymers having a high MMM (synthesized under oxygen-limited conditions). Therefore, the MMM of the alginate determined during the second stage is likely to be underestimated because of a ‘dilution effect’. For both strains, a drop in the MMM of the polymer was observed at the end of the second stage of the culture, possibly as a consequence of de-polymerization by the activity of alginate lyases (Peña et al. 2000; Trujillo-Roldán et al. 2004).

Overall, employing the mutant AT6, which has a blockage in the synthesis of PHB, together with the use of a two-stage process, it was possible to improve the alginate production. The volumetric production of alginate using A. vinelandii AT6 is the highest reported in the literature; however, the molecular mass of the alginate was lower than that produced by the parental strain.


Financial support of DGAPA-UNAM (grant IN230407) is gratefully acknowledged. The authors thank Arturo Ocádiz for computer support. Miguel A. Mejía thanks CONACyT for an M.Sc. scholarship.