Effect of protective agents, rehydration media and initial cell concentration on viability of Pantoea agglomerans strain CPA-2 subjected to freeze-drying


Dr Elena Costa, Postharvest Unit, CeRTA, Centre UdL-IRTA, 177 Rovira Roure Ave., 25198 Lleida, Spain (e-mail: elena.costa@irta.es).


The effect of initial cell density, protective agents and rehydration media on the viability of biocontrol agent Pantoea agglomerans CPA-2 when subjected to freeze-drying was studied. Several additives were tested as protective agents against freeze-drying injury. Maximum viability of the bacterial cells was obtained with disaccharides (survival levels >60%). Freeze-dried samples were rehydrated with several media; the highest percentage viability was obtained with 10% non-fat skim milk (100%+). The effect of initial bacterial load on the final recovery was dependent on protectant but not on rehydration media. Sucrose was an effective protectant when a high initial concentration (1010 cfu ml−1) was used; the opposite occurred with non-fat skim milk. The use of 1010 cfu ml−1 as an initial concentration, sucrose as a protectant and non-fat skim milk as a rehydration medium enabled 100% of P. agglomerans viability to be conserved after freeze-drying. Results suggest the possibility of achieving a good formulation system for the studied biocontrol agent with a high number of viable cells to be used toward pathogens, which is desirable for the industrial development of the product.


Fruit and vegetables suffer significant losses after harvest due to fungal diseases (Eckert 1975; Dennis 1983; Snowdon 1990, 1991). Post-harvest application of synthetic fungicides has been the main method of combating these diseases (Eckert and Ogawa 1988). However, the development of resistance to fungicides (Spotts and Cervantes 1986; Viñas et al. 1991, 1993) and public demand for a reduction in pesticide use have provided the impetus to develop alternative and effective natural methods of controlling post-harvest diseases.

Biological control using microbial antagonists has been considered as a desirable alternative to the use of chemicals. In recent years, success has been achieved through the development of microbial antagonists effective against fungal pathogens on pome (Janisiewicz and Bors 1995; Viñas et al. 1998), stone (Pusey and Wilson 1984) and citrus fruits (Chalutz and Wilson 1990; Smilanick and Denis-Arrue 1992; Arras and D'hallewin 1994; Huang et al. 1995). Some of these are being developed commercially.

Several studies carried out in this laboratory have demonstrated that the strain CPA-2 of Pantoea agglomerans (Gavini et al. 1989), previously classified as Erwinia herbicola, is an effective antagonist to the main fungal pathogens of citrus and pome fruits. Immersion of citrus fruit in solutions of P. agglomerans at 2 × 108 cfu ml−1 reduces the subsequent incidence of post-harvest mould caused by Penicillium digitatum and P. italicum. In the case of pome fruit, apples and pears are drenched with 8 × 107 cfu ml−1 to reduce incidence of Penicillium expansum, Botrytis cinerea and Rhizopus nigricans during storage in packing-houses (Viñas et al. 1999).

In order to use a biological agent as a commercial product it is essential to optimize its formulation. This is necessary to provide the product in a suitable form, and to optimize the efficacy, stability, safety and ease of application of the product (Rhodes 1993). Dehydrated cells have the advantage of not requiring cool temperatures during storage and distribution and thus make the product more economic.

Drying can be accomplished by a number of means including freeze-drying, drying on silica gel and spray drying (Rhodes 1993). Freeze-drying causes little shrinkage and results in a completely soluble product that is easily rehydrated (Powell 1992). Moreover, lyophilization is frequently used to preserve lactic acid bacterial starter cultures involved in dairy and food fermentations (Kearney et al. 1990).

Microbial cell survival during the freeze-drying process is dependent on many factors, including the initial micro-organism concentration (Bozoglu et al. 1987), the protective medium (Font de Valdez et al. 1983) and the rehydration conditions (Sinha et al. 1982; Font de Valdez et al. 1985b). Protective additives have an important role in the conservation of viability. A good protectant should provide cryoprotection to the cells during the freezing process, be easily dried, and provide a good matrix to allow stability and ease of rehydration. Various groups of substances have been tested for their protective action, including polyols, polysaccharides, dissacharides, amino acids and protein hydrolysates, proteins, minerals, salts of organic acids and vitamins-complex media (Berny and Hennebert 1991; Champagne et al. 1991). However, protection afforded by a given additive during these processes will vary with the species of micro-organism (Font de Valdez et al. 1983).

Rehydration is a critical step in the recovery of freeze-dried micro-organisms. Cells that are subjected to sub-lethal injury may not be able to repair the damage that has occurred if they are rehydrated under inappropriate conditions (Champagne et al. 1991). The medium itself, its molarity and the rehydration conditions can significantly affect the rate of recovery (Ray et al. 1971; Font de Valdez et al. 1983).

In order to obtain a suitable commercial product it is necessary to achieve a high density of viable dried cells. Some studies have shown that the initial bacterial load affects the survival rate during treatment. Bozoglu et al. (1987) demonstrated that there were higher survival levels when the highest initial cell densities of lactic acid bacteria were freeze-dried.

The aim of the present study was to evaluate the effect of protective agents, rehydration media and initial cell concentration on viability of the biocontrol agent Pantoea agglomerans CPA-2 when subjected to a freeze-drying process.

Materials and methods


Pantoea agglomerans (strain CPA-2) was obtained from the Postharvest Unit of Centre UdL-IRTA, Lleida, Spain. This strain was originally isolated from an apple surface (Viñas et al. 1999).

Cell preparation

Pantoea agglomerans was sub-cultured weekly on Starch Agar Medium plates (SAM) which contained (g l−1 distilled water): 5, peptone; 5, yeast extract; 3, soluble starch and 15, agar (Atlas 1995). After 24 h at 25 ± 1 °C, plates were stored at 4 °C.

The biocontrol agent was grown at 25 ± 1 °C in a bench top fermentor (Modular Fermenter, Gallenkamp, Loughborough, UK), containing 3 l of Nutrient Yeast Dextrose Broth medium (NYDB) which consisted of (g l−1 distilled water): 8, nutrient broth; 5, yeast extract; 10, anhydride glucose and 15, agar. Cells were harvested at the beginning of the stationary phase (24 h) by centrifugation (6981 g for 10 min at 15 °C). Cell paste was resuspended in 0·05 mol l−1 phosphate buffer (pH 6·5) and a volume of resuspended cells was dispersed into the protective medium in order to obtain the initial concentration. This suspension was incubated for 20 min at room temperature and constantly shaken to allow cell adaptation.


Three vials were filled with 5 ml of bacterial suspension produced as described above and placed at – 20 °C for 24 h. After overnight storage in the freezer, samples were connected to a Cryodos model freeze-drier (Telstar S.A., Terrassa, Spain) operating at 1 Pa pressure and – 45 °C for 24 h.


After freeze-drying, samples were immediately brought to their original volume (5 ml) with each rehydration medium at 25 °C. Then, samples were homogenized for 1 min with a Vortex mixer (SA-5, Stuart Scientific, Redhill, UK) and incubated at room temperature for 9 min. Serial dilutions were spread-plated onto the surface of 9 cm Petri plates containing SAM medium. These plates were incubated for 24 h at 25 ± 1 °C and the viability was then determined.

Survival levels were expressed as the quotient of colony-forming units per millilitre (cfu ml−1) on SAM medium before (N0) and after (Nf) freeze-drying. Viability = (Nf/N0) × 100.

Protectants used in assays

Suspensions of protectants were prepared in water. The additives tested as protective agents against freeze-drying injury were divided into five groups: (i) sugars: trehalose, glucose and fructose (5%), sucrose (10%) and lactose (7%); (ii) amino acids: sodium glutamate (1 mol l−1) and cystine (0·04 mol l−1); (iii) polymers: dextran, MW = 17500 g mol−1 (0·33 mol l−1) and Polyethylene glycol (PEG) MW = 200 g mol−1 (0·05 mol l−1); (iv) polyols: glycerol (5 mol l−1); (v) others: reconstituted non-fat skim milk; NFSM (Sveltesse, Nestle, Vevey, Switzerland) (10%), phosphate buffer (pH 6·5) and water, as controls. These protectants were selected on the basis of previous studies on other micro-organisms (Font de Valdez et al. 1983; Champagne et al. 1991). Protectant solutions were sterilized at 121 °C for 15 min before mixing with a volume of washed cells of antagonist to obtain an initial concentration of 5–8 × 109 cfu ml−1. The general procedure for cell preparation, freeze-drying and rehydration was described above. After freeze-drying, all samples were rehydrated with phosphate buffer to the original volume and the level of survival evaluated. The experiment was repeated twice.

Rehydration assay media

In this assay, the initial concentration used was 1 × 1010 cfu ml−1 and sucrose was used as the protectant. Freeze-dried samples were rehydrated with the following rehydration media: 10% non-fat skim milk, 10% sucrose, 5% sodium glutamate, 10% peptone, water, phosphate buffer, or a combination of 1·5% Peptone, 1% Tryptone, and 0·5% Meat extract (PTM medium, Font de Valdez et al. 1985b). Each rehydration medium was also used for serial dilution plating to assess viability of cells. The general procedure of cell preparation, freeze-drying and rehydration was as described above. The experiment was repeated twice.

Effect of initial concentration of cells on survival of freeze-drying

Washed cells of the tested micro-organism were resuspended in two suspension media used as protectants, 10% sucrose and 10% NFSM, at three concentrations: 108, 109 and 1010 cfu ml−1. Three different rehydration media, 10% sucrose, 10% NFSM and phosphate buffer, were used. Each rehydration medium was also used for serial dilution. The effect of initial bacterial concentration was evaluated for each protective agent in each rehydration medium. The general methodology of cell preparation, freeze-drying and rehydration was as previously described. The experiment was repeated three times.

Statistical treatment of the results

Percentage viability of P. agglomerans cells was estimated in all trials and analysed by a general linear model (GLM) procedure of the Statistical Analysis System (SAS Institute, version 6·03, Cary, NC, USA). Statistical significance was judged at the level P < 0·05. When the analysis was statistically significant, Duncan's Multiple Range Test was used for separation of means. Variables were analysed as fix variables, except for the initial concentration that was considered as an aleatory variable and a random test was performed.


Assay of protectant agents

Using a range of protectants, significant differences in the viability of cells of P. agglomerans after freeze-drying were observed, depending on the protectant used (Fig. 1). The best protection was given by sugars. Sugars could be divided into two groups: on the one hand, disaccharides which gave viabilities >60% and on the other hand, monosaccharides which gave viabilities of between 30 and 50%. The most effective protectant was trehalose at 5% (83% viability), followed by sucrose at 10% concentration (75% viability). The viability of cells of P. agglomerans was less than that obtained with the sugars, with all the other protectants examined.

Figure 1.

Comparative effect of different protective agents on Pantoea agglomerans CPA-2 viability after freeze-drying and rehydrating with phosphate buffer. The separation of means was conducted according to Duncan's Multiple Range Test. Columns with different letters indicate significant differences (P < 0·05).*NFSM: Non-fat skim milk; †M: mol l−1; §PEG: Polyethylene glycol

NFSM or dextran provided a freeze-dried material with a light and porous structure that made rehydration easy, but only around 15% of the cells remained viable. The rest of the assayed substances (amino acids and polyols) showed viabilities <10%, and there was no significant difference from the controls (distilled water and phosphate buffer).

Assay for a rehydration medium

Percentage viability of the P. agglomerans strain with different rehydration media, and 10% sucrose as a protectant, are shown in Fig. 2. All samples were rehydrated to the initial volume, because in previous studies it had been confirmed that an increase in the rehydration medium volume resulted in a decrease in the antagonist viability (data not shown).

Figure 2.

Effect of rehydration media on the viability of Pantoea agglomerans CPA-2 cells freeze-dried in 10% sucrose as protective agent. The separation of means was conducted according Duncan's Multiple Range Test. Columns with different letters indicate significant differences (P < 0·05).*NFSM: Non-fat skim milk; †PTM: Peptone-Tryptone-meat extract medium

The results obtained indicated that there were differences in viability of the bacteria depending on the rehydration media used. Recovery of cells was greater in complex media, such as NFSM or PMT, than in the other media used. The most promising results were obtained where freeze-dried cells were recovered in NFSM or PMT (100% viability). In contrast, <60% of the cells rehydrated with water were able to form colonies. Sucrose also showed a high capacity for injury repair, while sodium glutamate and phosphate buffer gave an intermediate recovery of cells.

Effect of initial concentration of cells on survival of freeze-drying

Statistical analysis of viability results was performed and four factors considered: experimental replication, initial concentration, protectant type and rehydration medium. Experimental replication and its two-way interaction were not significant; therefore, the results for the three different experimental replications were pooled.

Rehydration medium and two-way interaction of rehydration medium × protectant, and protectant × initial concentration, were statistically significant. In order to study the effect of initial concentration and rehydration media on viability, a statistical analysis was consequently performed for each protectant used.

In general, viability obtained when sucrose was used as a protective agent was higher than that obtained with NFSM as protectant (Fig. 3). Moreover, results showed that the effect of initial concentration had a marked dependence on the type of protectant used. There were no significant differences between rehydration media used when NFSM was used as protectant (Table 1). However, initial concentration, and two-way interaction between rehydration medium and initial concentration, were significant. This indicated that depending on the rehydration media used, the effect of initial concentration was different. The highest viabilities in NFSM–protectant medium were obtained at low concentration (108 cfu ml−1) in all rehydration media used. Recovery decreased when the concentration was increased to 1010 cfu ml−1 (Fig. 3a).

Figure 3.

Percentage of viability of Pantoea agglomerans freeze-dried in (a) non-fat skim milk and (b) sucrose as protector and rehydrated in non-fat skim milk (NFSM), sucrose and phosphate buffer at three initial concentrations 108 (▪), 109 (▓) and 1010 (□) cfu ml−1. The separation of means for each protector was conducted according to Duncan's Multiple Range Test (P < 0·05). Columns with different letters indicate significant differences between rehydration medium

Table 1.  Analysis of variance of effect of rehydration media (reh) and initial concentration (con) two-way interaction on viability of Pantoea agglomerans after freeze-drying when non-fat skim milk was used as protectant
SourceDFMSFP > F
  • *

    Significant P < 0·05; NS Not significant.

  • MS = Mean square.

Reh2312·2511·8820·2651 NS
Reh × Con490·6613·4560·0124*

When sucrose was used as a protectant, rehydration medium and concentration were significant (Table 2). With this protectant, final viability was higher after rehydration with NFSM, followed by sucrose and phosphate buffer in the three initial concentrations tested (Fig. 3b). The effect of initial concentration was also noticeable but in this case, the highest recovery was obtained at higher concentrations of 109 and 1010 cfu ml−1.

Table 2.  Analysis of variance of effect of rehydration media (reh) and initial concentration (con) two-way interaction on viability of Pantoea agglomerans after freeze-drying when sucrose was used as protectant
SourceDFMSFP > F
  • *

    Significant P < 0·05;

  • **

    Significant P < 0·005; NS Not significant.

  • MS = Mean square.

Reh × Con4518·5511·0910·3677 NS

For the freeze-dried biocontrol agent, 100% viability was obtained at 1010 cfu ml−1 of initial concentration, using sucrose (10%) as a protectant and NFSM (10%) as the rehydration medium.


Freeze-drying has been studied as a dehydration process for bacteria in order to achieve a solid formulation. This study showed the impact of protective additives, rehydration media and initial concentration of micro-organism on viability after freeze-drying. It was demonstrated that P. agglomerans is highly resistant to freezing, thawing and dehydration during the processes of freezing and freeze-drying, which is useful from a commercial viewpoint. This result was expected because many bacteria are known to survive freeze-drying well, including strains of Erwinia which have been stored for up to 10 years after freeze-drying without loss of viability (Rudge 1991).

The differences exhibited in cell survival in this study indicate that certain additives are more effective than others in protecting P. agglomerans. Maximum protection of cells of P. agglomerans during freeze-drying was achieved with sugars. The use of disaccharides resulted in viabilities >60% while monosaccharides resulted in cell viability of 30–60%. Sugars replace structural water in membranes after dehydration (Clegg 1986; Crowe and Crowe 1986) and prevent unfolding and aggregation of proteins by hydrogen bonding with polar groups of proteins (Hanafusa 1985; Carpenter et al. 1990). Trehalose was the best protective agent for cells of P. agglomerans when subjected to freeze-drying (>80% of viability), followed by sucrose and lactose. Differences exhibited by sugars are connected with their water-binding capacity and prevention of intracellular and extracellular ice crystal formation (Baumann and Reinbold 1964; Burke 1986). Trehalose has less of a tendency to crystallize than sucrose and lactose (Aguilera and Karel 1997). Unfortunately, the cost of trehalose limits its industrial use. However, sucrose, which has elicited viabilities >75%, is a cheap sugar which could be used economically.

The final cell viability obtained using NFSM and dextran as protectants was only 14 and 12%, respectively. However, these compounds provide the freeze-dried material with a light and porous structure which makes rehydration easy. Berny and Hennebert (1991) obtained similar results with Trichoderma viride, Brettanomyces bruxellensis, Saccharomyces cerevisae and Arthrobotrys arthrobotryoides. The amino acids sodium glutamate and cystine tested in this study were not effective in protecting cells of P. agglomerans. In contrast, some authors found sodium glutamate to be effective in conserving viability of streptococci and mesophilic lactobacilli (Font de Valdez et al. 1983) in Gram-negative bacteria (Ashwood-Smith and Warby 1972), and also in protecting yeast cells subjected to a freezing process (Takagi et al. 1997). PEG (polyethylene glycol) and glycerol were found to be completely ineffective in the protection of the studied biocontrol agent. Polymers such as PEG might accelerate drying, and over-drying is detrimental to survival during freeze-drying (Clementi and Rossi 1984; Font de Valdez et al. 1985a). Glycerol is an effective cryoprotectant and is widely used in frozen concentrates. However, it provided no protection for cells of P. agglomerans subjected to freeze-drying.

When using freeze-dried micro-organisms, rehydration has been considered as an important step. High recoveries obtained by complex rehydration media, such as NFSM and PTM, might be related to the rate of hydration of the samples. An environment of high osmotic pressure may control the rate of hydration and avoid osmotic shock. Moreover, these complex media may have a greater ability to repair damaged cells and to improve the final recovery obtained (Ray et al. 1971). In fact, the viability afforded by complex media was close to 100% and in some samples, the calculated values exceeded 100%. Viabilities greater than 100% after freeze-drying were also obtained by To and Etzel (1997) with Streptococcus thermophilus. Sucrose is also a good rehydration medium for cells of P. agglomerans, with similar results reported by Record et al. (1962) for Escherichia coli and other Gram-negative organisms.

As high cell concentrations are necessary for the preparation of commercial biocontrol products, it is useful to increase the initial concentration as much as possible to optimize the industrial process. In the case of P. agglomerans, the effect of initial cell concentration is related to the protective medium used. When sucrose was used as a protective agent, the highest recovery was obtained at a high concentration (1010 cfu ml−1). Bozoglu et al. (1987) suggested that the death of micro-organisms is proportional to their area of contact with the external medium. The shielding effect of the micro-organisms with each other could increase the role of sucrose as a protectant. However, when NFSM was used as a protectant, the lowest viabilities were obtained at high concentration. It could be said that proteins contained in milk could not provide a protective coat for all cells, or maybe 1010 cfu ml−1 would be too concentrated and harmful for the cells because of an unbalanced osmotic pressure. Bozoglu et al. (1987) suggested that 1012 cfu ml−1 was too concentrated for freeze-drying of Lactobacillus bulgaricus and Streptococcus thermophilus. The rehydration medium did not appear to interfere with the initial concentration and the viability was higher in NFSM, followed by sucrose and phosphate buffer in the two protective media tested.

In conclusion, this study has shown that the recovery of cells of P. agglomerans when subjected to freeze-drying is dependent on the protective medium used, the rehydration medium and the initial bacterial cell load. An appropriate selection of these factors seems to be essential for obtaining maximum viability of cells for use as a biocontrol agent. Future research must focus on demonstrating whether such optimized systems of freeze-drying and rehydration media conserve biocontrol efficacy against key post-harvest fungal pathogens of citrus and pome fruits, when compared with freshly-produced cells. It would also be important to study the shelf-life of freeze-dried cells. These findings are necessary for the industrial development of the biocontrol agent formulation.


The authors are grateful to Dr N. Magan from Cranfield University, UK, for his valuable discussion and advice, and to the Spanish Government for its financial support (INIA, Instituto Nacional para la Investigacion Agraria, Madrid, Spain).