Inhibitory Effects Of Organic And Inorganic Salts Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Twenty-one organic and inorganic salts were tested in vitro for their inhibitory effect on the growth of the bacteria Erwinia carotovora subsp. carotovora and Erwinia carotovora subsp. atroseptica, the causal agents of soft rot in stored potato tubers. In nutrient broth medium, at 0.2 M concentration, eleven of the salts tested exhibited strong inhibition of the growth of both bacteria. Among them, sodium carbonate, sodium metabisulfite, trisodium phosphate, aluminum lactate and aluminum chloride were bactericidal after one hour of exposure and exhibited low minimal inhibitory concentration (MIC) values (≤10 mM). Sodium bicarbonate, sodium propionate and ammonium acetate exhibited intermediary MIC values (20 mM) and a slower bactericidal effect except sodium propionate, which was bacteriostatic. Aluminum dihydroxy acetate, potassium sorbate and sodium benzoate were bactericidal but exhibited MIC of 100 mM. It is shown that the inhibitory action of salts relates to the water-ionizing capacity of the constituent ions of the salts (low apparent pKa or pKb) and to their lipophilicity (partition coefficient, Po/w), the latter mainly contributing to the action of preservative salts, potassium sorbate and sodium benzoate. The relationships between the inhibition of bacterial growth and apparent pK, and between the inhibition of bacterial growth and the addition parameter (pK' + pPo/w) were sigmoidal. Results of this study suggest a possibility for exploiting some of the effective salts as antimicrobial agents to control the development of E. carotovora in stored potatoes.

2.2 Introduction

Soft rot is a major bacterial disease affecting potato (Solanum tuberosum L.) tubers. The disease causes significant economic losses worldwide especially during storage (Sharga and Lyon, 1998; Stevenson et al., 2001) where it spreads from infected tubers to healthy tubers (Howard et al., 1994). Soft rot infected tissues appear wet, cream to tan in colour with a soft and slightly granular consistency (Howard et al., 1994; Stevenson et al., 2001). In North America, potato soft rot is mainly caused by the Gram-negative bacteria Erwinia carotovora subsp. carotovora (Jones) Bergey et al. and Erwinia carotovora subsp. atroseptica (van Hall) Dye (Stevenson et al., 2001).

Currently, no chemicals are available for an effective control of potato soft rot (Rousselle-Bourgeois and Priou, 1995; Stevenson et al., 2001), and hence, control and management of this disease rely mainly on cultural practices such as planting disease-free seed tubers and maintaining proper harvesting, handling, and storage conditions (Howard et al., 1994; Stevenson et al., 2001). However, in most of the cases, these practices are not sufficient to control the disease effectively. Given the economic importance of losses due to soft rot, and the growing concerns of hazardous effects of the synthetic chemicals used for fresh fruits and vegetable preservation, there is an urgent need to develop new strategies for an effective, safer and reliable control of soft rot during storage.

One approach consists in the exploitation of the antimicrobial properties of different organic and inorganic salts used in food preservation. The inhibitory effect at relatively low concentrations (0.05-0.2 M) of many salts including sorbate, propionate, metabisulfite, aluminum, bicarbonate and carbonate salts, on bacterial or fungal growth has been demonstrated either in vitro or on various vegetable and other food commodities (Smilanick et al., 1999; Hervieux et al., 2002; Mecteau et al., 2002; Yaganza et al., 2004). Different biochemical mechanisms have been put forth to explain the antimicrobial activity of salts including inhibition of several steps of the energetic metabolism (benzoate, bicarbonate, propionate, sorbate and sulfite salts) (Bosund, 1960; Bosund, 1962; Freese et al., 1973; Kritzman et al., 1977; Wedzicha, 1984; Kabara and Eklund, 1991), and complexion to DNA and RNA (aluminum and sulfites) (Johnson and Wood, 1990; Gould and Russell, 1991; Wood, 1995). Even though there is information available on the biochemical mechanisms of antimicrobial action of certain salts, little is known on the physico-chemical basis of the general antimicrobial action of salts.

Salts are attractive as postharvest disease control agents because of their broad-spectrum antimicrobial activity (Olivier et al., 1998), biocompatibility (Horst et al., 1992), low cost (Mecteau et al., 2002) and ease of application. However, their use for the control of potato soft rot has been little explored. Wyatt and Lund (1981) and Bartz and Kelman (1986) observed some control of soft rot following the application of sodium benzoate and citric acid, respectively.

The objectives of this work were (1) to evaluate the inhibitory effect of different organic and inorganic salts on E. carotovora subsp. carotovora and E. carotovora subsp. atroseptica, and (2) to gain an understanding on the physico-chemical basis of their action in the inhibition of bacterial growth.

2.3 Materials and methods

2.3.1 Bacteria

Strains of Erwinia carotovora subsp. carotovora (Ecc 1367) and Erwinia carotovora subsp. atroseptica (Eca 709), were provided by the Laboratoire de diagnostic en phytoprotection (MAPAQ, Québec, Canada). The bacteria were maintained on Nutrient Agar (NA; Difco Laboratories, Becton Dickinson, Sparks, MD) slants at 4 °C, and served as stock cultures.

2.3.2 Chemicals

All salts were purchased from Sigma Chemical Co. (St. Louis, MO), except for ammonium acetate, sodium chloride and sodium bicarbonate (BDH Inc., Toronto, Canada), and aluminum lactate (Aldrich Chemical, Milwaukee, WI).

2.3.3 Effect of salts on bacterial growth

Bacteria were grown in 250 mL flasks containing 50 mL of 20% tryptic soy broth (TSB; Difco) amended with salts (200 mM) or unamended (control), by incubating at 24 °C with agitation (150 rpm; Lab-Line Instruments Inc., Melrose Park, IL) for 24 h. The pH of the media were not adjusted unless stated otherwise, and varied with the type of salts. Flasks were inoculated with 100 µL of each bacterial suspension (1x107 CFU/mL). Bacterial growth was determined by turbidimetry at 600 nm with a UV/Visible spectrophotometer (Ultrospec 2000, Pharmacia Biotech Ltd, Cambridge, UK), using appropriate blanks. Results were expressed as the percentage of growth inhibition compared with the control. A completely randomized experimental design with three replicates was used, the experimental unit being a flask.

2.3.4 Lethal effect of salts

The lethal effect of salts was determined according to the method of Lin et al. (2000) with slight modification. Bacteria (1x106 CFU) were suspended in 250 mL flasks containing 50 mL of the different salt solutions (200 mM) or isotonic NaCl solution (control). Triplicate flasks were incubated at 24 °C with agitation (150 rpm). After different periods of incubation (1, 2, 4, 8, 12 or 24 h), samples (1 mL) were withdrawn. The bacterial cells were recovered by centrifugation (2360 x g for 10 min; Kendro Laboratory Products, Hanau, Germany), washed with isotonic NaCl solution, concentrated by centrifugation once again, and resuspended in 1 mL of isotonic NaCl solution. The viability of bacteria was then determined by plate count on NA after 36-48 h of incubation at 24 °C. Lethality was expressed as percentage: {[viable cells (control) - viable cells (salt solution)] / viable cells (control)} x 100. A salt was considered bactericidal when it caused 100% lethality (irreversible killing of bacteria).

2.3.5 Effect of salt concentration on growth inhibition

For the salts exhibiting a strong inhibitory activity at 200 mM, bacterial growth was evaluated in triplicates as previously described in 20% TSB containing different concentrations (0-200 mM) of each salt. Growth inhibition (expressed as the percentage of growth inhibition compared with the control) was then determined and plotted against concentration.

2.3.6 Effect of medium initial pH on bacterial growth

Bacterial growth was carried out in 250 mL flasks containing 50 mL of 20% TSB adjusted to pH in the range from 3 to 11 at 0.5 unit intervals, using HCl or NaOH. Triplicate flasks were inoculated, and incubated at 24 °C with agitation (150 rpm) as previously described. After 24 h, the growth was evaluated as previously described.

2.3.7 Determination of the octanol/water partition coefficients (Po/w) of the salts

The octanol/water partition coefficient is an indicator of the lipophilic character of a compound. Its measurement is based on the principle that, in a mixture of two immiscible solvents (like the non-polar solvent 1-octanol, and polar solvent water), at equilibrium a hydrophobic compound will migrate more in the non-polar solvent, whereas hydrophilic one will rather move into the aqueous phase. Octanol/water partition coefficients (Po/w) of the salts were determined using the general solvent-solvent separation procedure (Fessenden and Fessenden, 1993). Equal volumes (50 mL) of 1-octanol (Sigma Chemical Co.) and bi-distilled water were poured into a separating flask and thoroughly shaken for 5 min, so that both immiscible solvents are saturated with the other. Four grams of each salt were then added, and the flask thoroughly mixed for 5 min, three times with a rest period of 5 min after each agitation. After complete separation (20-24 h, room temperature), the two phases were recovered separately in different flasks, and the accompanying ion of the salt was directly measured from the aqueous phase by atomic absorption (Perkin-Elmer, Model 3300, Ueberlinger, Germany). The same ion was measured in the organic phase as follows: octanol was distilled-off at 150 °C (48 h), the dry residual crystalline material ashed at 500 °C overnight, and the ash was recuperated with 10% HCl. Partition coefficient (Po/w) was calculated as the ratio of the concentration of ion in 1-octanol to the concentration of ion in aqueous phase. The measurements were performed in duplicate.

2.3.8 Osmotic pressure of salt solutions, and acidity (pKa) and basicity (pKb) constants of ions

The osmotic pressure (Π) of salt solutions was estimated using van't Hoff equation (White, 2000):

Π = iRTc

where R is the gas constant, T is the absolute temperature (°K), c the concentration of the salt (mol/L), and i is the number of ions into which the salt dissociates in solution. The values of the acidity constants (pKa) and the basicity constants (pKb) of the acidic and basic ions respectively, were obtained from literature (Snoeyink and Jenkins, 1980; Weast, 1984; Dean, 1985). For ionic strength > 0.1 M, pKa and pKb values of the ions are more accurate when defined as apparent constants in terms of the activities of hydronium and hydroxyl ions, and ionic species concentrations and activity coefficients (Edsall and Wyman, 1958).

For the acidic ions,

pK'a = pKa + log(γB- / γHB),

and for the basic anions,

pK'b = pKb + log(γHB / γB-),

where pK'a and pK'b are the apparent acidity constant and basicity constant respectively, γB- is the activity coefficient of the conjugate base (B-), and γHB is that of the acidic (HB) ion. The activity coefficient (γ) of the species i can be expressed as a function of ionic strength (μ), using the Güntelberg approximation of the Debye-Hückel equation (Snoeyink and Jenkins, 1980):

-log γ i = [(0.51Z i 2 μ1/2)/(1 + μ1/2)]

where Z i is the charge on the species i, and µ is the ionic strength. Thus,

log(γB-/ γHB) = [(0.51 μ 1/2)/(1 + μ 1/2)] (2ZHB-1),

and log(γHB / γB-) = -[(0.51 μ 1/2)/(1 + μ 1/2)] (2ZB-+1)

Polytropic acid potentiating ions (bicarbonate, carbonate, monohydrogen phosphate, phosphate, sulfite and tartrate) in aqueous solution can exist as (n+1) possible species for which the parent acid is HnA. These species may co-exist in equilibrium under certain pH conditions. For these ions, the apparent acidity or basicity constants were expressed as the mean of the co-existing species at a specified pH.

2.3.9 Statistical analysis

Analysis of variance was carried out with the GLM procedure of SAS (SAS Institute, Cary, NC), according to a randomly experimental design. When significant (P<0.05), treatment means were compared using Fisher's protected LSD.

2.4 Results

Several of the 21 salts tested inhibited the growth of the bacteria E.c. atroseptica and E.c. carotovora significantly at 200 mM concentration (Table 2.1). Aluminum salts (dihydroxy acetate, chloride and lactate), ammonium acetate, sodium benzoate, sodium bicarbonate, sodium carbonate, sodium metabisulfite, sodium propionate, trisodium phosphate, and potassium sorbate completely inhibited the growth of the bacteria. Calcium chloride, sodium formate, sodium acetate, ammonium hydrogen phosphate and sodium hydrogen phosphate exhibited a moderate inhibitory effect; and sodium lactate and tartrate had no effect. On the other hand, ammonium chloride, potassium chloride and sodium chloride stimulated the growth of E.c. atroseptica. The inhibitory effect of the anions with common sodium or potassium cation followed the order: PO4 3- = CO3 2- = HSO3 - = benzoate- = propionate- = sorbate- = HCO3 - > HPO4 2- ≈ acetate- > formate- > lactate- = tartrate- > Cl-. Inorganic anions were found generally stronger inhibitors than the organic ones, except the preservative organic anions (benzoate-, propionate- and sorbate-). With common NH4 + cation, the inhibitory effect was in the order: acetate- > HPO4 2- > Cl-. For the cations with common chloride anion, the inhibitory effect was in the order: Al3+ > Ca2+ > NH4 + > K+ = Na+, whereas with common acetate anion, the order was Al(OH)2 + = NH4 + > Na+; and with common HPO4 2- anion, it was Na+ > NH4 +.

The effect of the medium initial pH on bacterial growth was also evaluated. The pH range for optimal growth was between 7.0 and 8.0, and the growth was maximal at pH 7.5. No bacterial growth occurred either below pH 4.0 or above pH 10.5 (Figure 2.1). Based on this result, the effect of the highly acidic or alkaline salts (which strongly affected the pH of the medium) on the growth of E.c. atroseptica was evaluated at pH 7.5. Sodium carbonate (initial pH of 10.6) and sodium metabisulfite (initial pH of 4.5) completely inhibited E.c. atroseptica growth at pH 7.5, as they did at their initial pHs (Figure 2.2). But, trisodium phosphate (initial pH 11.9) exhibited a lower inhibitory effect (83.2%) at pH 7.5 (Figure 2.2). Since aluminum salts precipitate at pH 7.5 (due to formation of hydrated aluminum hydroxide), it was not possible to test their inhibitory effect at pH 7.5.

For the salts showing strong inhibitory effect at 200 mM concentration, the time course of the lethal effect on E.c. atroseptica is presented in Figure 2.3. Aluminum salts, sodium carbonate, sodium metabisulfite and trisodium phosphate exhibited 100% lethality after one hour (generation time for the growth of E. c. atroseptica under the growth conditions) of exposure. Potassium sorbate, sodium benzoate, sodium bicarbonate and to a lesser extent ammonium acetate, were moderately bactericidal, exhibiting 100% lethality after 4, 8, 12 and 24 h of exposure, respectively. Sodium propionate exhibited only a partial killing effect, with a lethality of 41.3% after 24 h of exposure. The lethality of these salts on E.c. carotovora was similar to E.c. atroseptica (data not shown).

The bacterial growth inhibition as a function of salt concentration is presented in Figure 2.4. It shows that the minimal growth inhibitory concentration (MIC) of the salts were: 5 mM for aluminum lactate and sodium metabisulfite (Figure 2.4 A); 10 mM for aluminum chloride, sodium carbonate and trisodium phosphate (Figure 2.4 A); 20 mM for ammonium acetate, sodium bicarbonate and sodium propionate (Figure 2.4 B); and 100 mM for aluminum dihydroxy acetate, sodium benzoate and potassium sorbate (Figure 2.4 C). With calcium chloride and sodium hydrogen phosphate (Figure 2.4 D), growth inhibition remained under 90% even at the highest concentration (200 mM). Inhibition by these last two salts became appreciable only at concentration higher than 100 mM. Except calcium chloride and sodium hydrogen phosphate, which displayed positive curvilinear inhibition pattern, almost all the other salts showed a rather negative curvilinear pattern, some with an initial lag phase.

To determine whether lipophilicity of salts contribute to their inhibitory activity, the octanol/water partition coefficients (Po/w) of salts with common Na+ or K+, and those with common Cl- were determined. Sodium benzoate was found the most lipophilic (Po/w = 1.41 x 10-2), followed by potassium sorbate (Po/w = 7.6 x 10-3) and sodium metabisulfite (2.0 x 10-4). Most other salts, sodium chloride (reference salt), sodium bicarbonate and carbonate, sodium propionate, sodium acetate, calcium chloride and aluminum chloride mainly remained in the aqueous phase (Po/w = 2.0-5.0 x 10-5).

2.5 Discussion

Eleven of the 21 salts tested exhibited significant inhibition of the growth of both E.c. atroseptica and E.c. carotovora. Among the inhibitory salts, sodium carbonate, sodium metabisulfite, trisodium phosphate, and aluminum chloride, lactate and acetate, were highly bactericidal. They also exhibited low MICs values of up to 10 mM, except aluminum dihydroxy acetate (MIC = 100 mM). Rapid killing effect of these salts, as well as significant inhibition at low concentrations, suggest that their cellular target(s) are readily accessible by salts components. Potassium sorbate (MIC = 100 mM), sodium benzoate (MIC = 100 mM) and sodium bicarbonate (MIC = 20 mM), were moderately bactericidal, causing 100% mortality after 4, 8 and 12 h, respectively. Although ammonium acetate exhibited a MIC of 20 mM, its bactericidal effect was slower leading to complete mortality after 24 h; whereas sodium propionate (MIC = 20 mM) was not bactericidal.

Difference in the inhibitory patterns of salts as a function of concentration indicates their different potencies and different modes of action. Except sodium metabisulfite for which no lag phase was perceivable, the other salts showed lag phases of different extents in the inhibition-concentration relationship (Figure 2.4). The strong and rapid action of sodium metabisulfite may reflect the rapid diffusion of this salt species (SO2·H2O) through the cell envelope (Yaganza et al., 2004; Chapiter 4). On the other hand, the presence of lag in the concentration effect of the other salts is indicative of a cooperative action on the inhibition of bacterial growth. These salts need to accumulate sufficiently in the cell before causing adverse effect.

Although there is information available on the biochemical mechanisms of antimicrobial action of certain salts (Bosund, 1962; Freese et al., 1973; Kritzman et al., 1977; Freese, 1978; Freese and Chernov-Levin, 1978; Salmond et al., 1984; Wedzicha, 1984; Eklund, 1985a and b; Johnson and Wood, 1990; Doores, 1993; Yaganza et al., 2004-Chapter 4), little is known on the physico-chemical basis of the general antimicrobial action of salts. Elevated osmolarity due to salt addition, may trigger the osmoregulatory process, causing an increased maintenance metabolism, and leading thus to reduction in bacterial growth. However, examination of table 2.1 (salts with comparable osmolarities displayed complete or no bacterial growth inhibition) does not indicate that osmotic stress or reduction in Aw alone may have brought about the inhibition of the bacterial growth, even though E. carotovora is a Gram-negative bacterium and the turgor pressure of the cytoplasm of Gram-negative species (0.8-5.0 atm) is generally lower than that of Gram-positive species (15-20 atm) (38). Therefore, other factors may play a role.

The acidity or alkalinity of the medium as a result of the addition of some salts can have profound adverse effects on bacterial growth. Extreme pH conditions can lead to denaturation of proteins like enzymes present on the cell surface, depolarization of transport for essential ions and nutrients, modification of cytoplasmic pH and DNA damage (Langworthy, 1978). Examination of table 2.1 shows that the addition of aluminum lactate, aluminum chloride and sodium metabisulfite strongly acidified the medium, whereas the addition of sodium carbonate and trisodium phosphate strongly increased the pH. Except for ammonium acetate and the common food preservative salts (potassium sorbate, sodium benzoate and sodium propionate) whose ΔpH (ΔpH = |7.5 - pH of salt-amended medium|) is < 1, all the other inhibitory salts generally display ΔpH values ≥ 1 (Figure 2.5). The dissociation of salts in aqueous medium generates ionic species which can participate in proton exchange reactions with water molecules. While an acidic ion donates a proton to water, a basic ion abstracts proton from water molecules, thereby contributing to changes in the pH of the aqueous medium.

Results in Figure 2.2 suggest that growth inhibition by sodium carbonate and sodium metabisulfite can not be attributed solely to extreme pH and passive proton transfer (extreme pH) across the bacterial membrane. At pH 7.5, HSO3 -, SO3 2- and to a lesser extent SO2·H2O species are present in the medium amended with sodium metabisulfite, whereas HCO3 - species predominates in that amended with sodium carbonate (Snoeyink and Jenkins, 1980); all these species appear inhibitory to the bacteria. In contrast, the inhibition of bacterial growth by trisodium phosphate (pH 11.9), which decreased to 83 % when pH was adjusted to 7.5, indicates that PO4 3- is the highly inhibitory species, since its concentration is decreased by lowering the pH to 7.5, where HPO4 2- predominates. This latter species was found less inhibitory to the bacteria (Table 2.1). Evidently, the inhibitory activity of the salts depends on the nature of their constituent ions.

The capacity of an ion to dissociate water is an intrinsic characteristic, determined by its pK value (pKa for acidic species or pKb for basic ones), which needs to be corrected (apparent values pK'a or pK'b; Table 2.2) for the ionic strength of the solution when the salt concentration is higher than 0.1 M (Huss and Eckert, 1977; Snoeyink and Jenkins, 1980). Figure 2.6 shows a sigmoidal relationship between the inhibitory effect of salts on the bacterial growth and the pK'b of the basic ions (with common sodium or potassium cation in the salt) and the pK'a of the acidic ions (with common chloride anion in the salt). The plot exhibits a sharp linear relationship in the pK' range of 8.0 to 12.0, displaying a half-maximal inhibition (I pK'50) at pK' value of 9.8. Below the pK' value of 8.0, inhibition is maximal whereas above the pK' value of 12.0, ions appear to stimulate growth. Exceptions to the sigmoidal relation were the food preservative anions (benzoate, propionate and sorbate) and Ca2+ which display higher inhibition than expected from their pK' value, but NH4 + was less effective than one would expect from this relation. The sigmoidal pattern of the curve with a sharp slope is a reflection of a positive cooperativity of ionization of water molecules that precedes the inhibition of bacterial growth.

The capacity of the constituent ions of the salts to either donate or abstract protons to water molecules either in the growth environment (reflected in pH modification of the medium) or in the developing cells, generally plays a role in their inhibitory action. The consequent transmembrane pH gradient generated leads to a passive H+ transport across the membrane, and the acidification (in the case of ions with low pK'a) or alkalinization (ions with low pK'b) of the cytoplasm, once the capacity for proton-coupled active transport is outstripped. In both cases, proton exchange with outer membrane proteins will destabilize these proteins, their interaction with membrane lipids and ultimately their function of solute transport, leading to growth inhibition. The modification of cytoplasmic pH can also alter nucleic acids structure and function, and contribute to growth inhibition (Langworthy, 1978).

However, the water-ionizing capacity of salt constituent ions and the consequent modification of the pH, is not the sole factor accounting for growth inhibition, as suggested by the exceptional inhibitory action of benzoate, propionate and sorbate (Figures 2.5 and 2.6). These ions provide greater inhibition than expected from their pK' (pK' = 10.0, 9.3, and 9.4, respectively), while the pH of their solution is optimal for bacterial growth (pH 7.4, 7.4, and 7.7, respectively). This suggests that they possess additional characteristics mediating their action, in addition to their water-ionization property. In fact, these preservative agents have been shown to be active either as undissociated acids (like other weak acid) or as anions (Eklund, 1983; Eklund, 1985a and b; Warth, 1991 a and b), due to their hydrophobic nature, which would allow them to interact with lipidic constituents of the cell envelope of Gram-negative bacteria such as Erwinia, and to modify their functionality (Doores, 1993), resulting in growth inhibition. They can also cross the cell envelope due to their hydrophobicity, and their acidification inside the cell can cause additional adverse effects.

Thus, we determined the octanol/water partition coefficients (Po/w) of those effective salts with common sodium or chloride ions. Among the salts, sodium benzoate and to a lesser extent potassium sorbate were found to be lipophilic, which would allow the negatively charged benzoate and sorbate ions to effectively penetrate through cell membrane (Doores, 1993). The lipophilic character of these ions is attributable to their low charge density, since the negatively charged carboxyl group is directly attached to conjugated double bonds, as in aromatic ring (benzoate), and trans, trans 2, 4 hexadiene (sorbate), where the charge can be delocalised effectively and distributed over a large volume of the molecule (Freese and Chernov-Levin, 1978). Propionate, the other preservative salt, was found to be less lipophilic, presumably because its charged group is attached to an aliphatic chain. An addition parameter, pK' + pPo/w, which combines the two properties of constituent ions, i.e., water ionizing capacity (pK') and lipophilicity (pPo/w = -log Po/w), appears to provide a more general basis for the inhibitory effect of salts (Figure 2.7). This suggests that while the dissociation constant of ions plays a major role in growth inhibition as portrayed in Figure 2.6, the lipophilic character of the preservative-salt ions confers to them an added ability to interact with cell membranes. The addition parameter (pK'a or pK'b + pPo/w) value for half-maximal inhibition [I(pK' + pPo/w)50] was 14.2.

Yet three ions (ammonium, calcium and propionate) do not fall within the sigmoidal pattern portrayed in Figure 2.7. The effectiveness of ammonium ion is lower than expected from its pK'a value when the accompanying anion is chloride, presumably because NH4 + is known to form hydrogen bonds with water molecules (Bonner, 1977), and hence its ability to ionize water is probably reduced. However, the H-bonding ability of NH4 + depends on the accompanying conjugate base, and H-bonding is more significant in the presence of Cl- than acetate, which can also form H-bonds, thereby increasing the availability of free NH4 + and its ability to exert its action by ionizing water (Bonner, 1977). The contribution of NH4 + ions in the inhibitory activity becomes apparent when the bacterial growth inhibition of NH4Cl is compared with NaCl (Table 2.1). Similarly, the contribution of NH4 + ions to the inhibitory activity is reduced when the accompanying conjugate base is HPO4 2-, an amphoteric ion, which is less competitive than acetate ion for H-bonding. The effectiveness of Ca2+ is higher than expected from its pK'a value, which may be explained by its ability to destabilize membrane proteins and enzymes (Vaara, 1992). Such a destabilizing effect on the membrane proteins should affect solute transport and other membrane functions. Propionate exhibits a low lethality (Figure 2.3), yet it exhibits a high bacteriostatic effect in spite of its high pK'b and low lipophilicity values. It suggests that propionate per se may not be the inhibitory agent, but a more effective metabolic intermediate of propionate may be involved. Maruyama and Kitamura (1985) have shown that propionyl-CoA, the intermediate of propionate inhibits pyruvate dehydrogenase in Rhodopseudomonas sphaeroides. Propionate is also known to inhibit the synthesis of β-alanine and pantothenic acid (Maruyama and Kitamura, 1985).

The most effective inorganic ions (Al3+, HSO3 -, HCO3 -, CO3 2-, HPO4 2-, PO4 3-) are also known to display specific effects. Aluminum ions possess a high capacity to polarize and dissociate water, producing protons and acidifying the medium. Their ability to ionize water molecules, cationic charge and ability to form aquo-hydroxo complex with lower charge density confers on aluminum ions the ability to destabilize membrane proteins and other membrane molecules. Aluminum ions can also affect many cellular targets, including DNA, RNA and ATP by complexion (Haug, 1984; Johnson and Wood, 1990; Martin, 1992; Wood, 1995; Yaganza et al., 2004-Chapitre 4).

The additional effects of metabisulfite salt rely on the multifunctionality of its generated species (SO2·H2O, HSO3 - and SO3 2-), which are reactive with a vast range of potential targets both outside and inside the microorganisms. These targets include biologically important molecules such as proteins (including enzymes and membrane transporters) containing disulfide bonds (Anacleto and van Uden, 1982; Wedzicha, 1984), coenzymes and cofactors (e.g. NAD+ and NADP+, folic acid, pyridoxal and thiamine) (Williams et al., 1935), as well as nucleic acids, quinones and sugars (Shapiro and Weisgras, 1970; Stratford and Rose, 1985; Molander et al., 1993). Passage of sulfite across the microbial membrane can take place by either free diffusion of SO2 which predominates at lower pH as demonstrated for Saccharomyces cerevisiae (Stratford and Rose, 1986), or by active transport of sulfite and sulfate (SO3 2-, SO4 2- and S2O3 2-) near neutral pH. The transport of bisulfite ion is biphasic depending on the pH; at lower pH, it operates by passive transport, while the active transport predominates at neutral pH (Gould and Russell, 1991). Sulfite can either cause dissipation of the proton motive force across the plasma membrane, thereby inhibiting active transport of solutes, or denaturation of transport proteins located in the outer membrane of Gram-negative bacteria. Once in the cell, the imported species can modify the pH, and also cause cell disfunctioning.

Some evidence shows that HCO3 - can modify membrane structure and function. Indeed, this species is found to decrease the interfacial tension between lipid and aqueous phases, and to increase ion permeability (Sears and Eisenberg, 1961). Thus, lipid-protein interactions could be modified, resulting in decreased activity of membrane-bound proteins. In addition, HCO3 - is known to inhibit TCA cycle enzymes (Kritzman et al., 1977). PO4 3- and HPO4 2- are chelating agents capable of sequestrating Ca2+ or Mg2+. Sequestration of Mg2+ and Ca2+ can lead to structural changes in the outer membrane and expose the cell to the medium (Vaara, 1992).

Rapid mortality by strong inorganic salts indicates that membrane destabilization, accumulation of the ions in the cell, consequent ionization of water and dissipation of membrane potential as well as additional inhibition of intracellular metabolic processes are rapid. Preservative acid anions being amphiphiles and somewhat lipid permeable (Bosund, 1960), they have to bind to outer membrane proteins before transmembrane transport can occur. Since their action by direct proton transfer is low (low capacity of water ionization), their accumulation in the cell will increase the membrane potential. In addition, the preservative salts may inhibit the proton motive force across the cell wall (Salmond et al., 1984; Eklund, 1985 a and b), the transport of nutrient (Brown and Booth, 1991), as well as the energy production pathway (decrease in ATP production) (Bosund, 1962; Krebs et al., 1983; François et al., 1986; Kabara and Eklund, 1991; Warth, 1991 a and b).

NH4Cl, NaCl and KCl at a 200 mM concentration were shown to stimulate bacterial growth. Growth promoting effect of these salts can be achieved either by promoting the assimilation of nutrients by the bacteria (NH4Cl and KCl), or by contributing to the regulation of pH homeostasis (NaCl) (Booth, 1985).

In conclusion, the in vitro tests carried out in this study have shown that several salts inhibit completely the growth of the bacteria E.c. carotovora and E.c. atroseptica, the causal agents of potato soft rot. All the aluminum salts, sodium metabisulfite, sodium carbonate as well as trisodium phosphate were highly bactericidal. The inhibitory action of salts relates to the water-ionizing capacity of the constituent ions of the salts and to their lipophilicity. The constituent ions of the effective salts (Al3+, CO3 2-, PO4 3- and HSO3 -) display a high capacity to ionize water molecules (low pKa or pKb), which appears to be the the primary contributor to the inhibitory action of salts. These salts also display different modes of action, as evidenced from the patterns of their inhibition of bacterial growth as a function of concentration. The preservative salts potassium sorbate and sodium benzoate exhibit a slower bactericidal effect, and their antibacterial mode of action appears to be associated with their hydrophobicity, which allows them to perturb the cell envelope structure and function. Results from in vitro studies suggest a possibility for exploiting some of the effective salts as antimicrobial agents to control the development of E. carotovora in potato tubers. Further work is needed to determine whether the inhibitory action of salts will persists in a more complex system such as potato tuber and contribute to the control of the bacteria.

2.6 Acknowledgements

This study was supported by Conseil des Recherches en Pêche et en agroalimentaire du Québec (CORPAQ), Cultures H. Dolbec Inc. and Propur Inc. We wish to thank Dr. K. Belkacemi for his assistance with data analysis.