The effect of calcium on biofilm formation in dairy wastewater

Global dairy production has increased from 500 million tonnes in 1983 to 769 million tonnes in 2013 (F.A.O. United Nations 2016), resulting in a corresponding increase in wastewater that requires treatment. Generation of wastewater occurs throughout all areas of a processing facility but mainly from cleaning manufacturing equipment. One sustainable treatment for dairy wastewater is irrigation onto pasture where it provides nutrients to support grass growth for cattle feed. However, the nutrients in the wastewater can also support the growth of bacteria in the irrigation system in the form of biofilms (Dixon et al. 2017). This may be problematic as biofilms can lead to odour generation and pipe blockages. Within the dairy plant, acid and alkaline washes incorporated into a Clean In Place system are used to control biofilm formation. Unfortunately, these same processes are unsuitable for the wastewater system. This is especially the case in the final stages of wastewater treatment, such as irrigation systems, as undiluted or untreated chemicals could damage the natural environment and cause pasture ‘burn’ or waterway contamination.

Dairy wastewater has a high Biological Oxygen Demand and a high Chemical Oxygen Demand (Kushwaha et al. 2011) and contains a variety of nutrients from milk residues and chemicals used in cleaning. The ion content of the wastewater can vary greatly over time and this may affect the growth of bacteria present in the system. These ions can have varying effects on the growth and development of bacteria. For example, Ca²⁺ and Mg²⁺ are both essential in the regulation of several steps of cell division (Hepler 1994; Webb 1949). Somerton et al. (2015) showed that supplementation of a milk formula with as low as 2 mM CaCl₂ or 2 mM MgCl₂ increased the biofilm formation of three Geobacillus spp while 100 mM NaCl significantly decreased biofilm formation showing that high free Na⁺ ions and low Ca²⁺ and Mg²⁺ were collectively needed to reduce biofilm formation.

Two mechanisms have been proposed to explain the effect of divalent cations on biofilm development (Somerton et al. 2015). Under the Divalent Cation Bridging (DCB) mechanism the divalent cations bind negatively charged sections of the extracellular polymeric substances (EPS) helping to both stabilize and strengthen the biofilm aggregates. Several other studies have shown that adding Na⁺ ions to the media cause a disruption or weakening of the biofilm by displacing the Ca²⁺ and Mg²⁺ ions within the EPS (Sobeck & Higgins 2002; Kara et al. 2008; Somerton et al. 2015). The second mechanism proposed is that ions present in the media could influence the regulatory pathways within the cells. Na⁺ ions have been shown to influence the integrity of biofilms in wastewater sludge by increasing the negative charge proportion of polymers in the bacterial cell wall while Ca²⁺ and Mg²⁺ ions could bind to regulatory proteins or extracellular DNA (Michiels et al. 2002; Song & Leff 2006).

This study was carried out to understand the effect of ions on the biofilm formation and growth characteristics of four know biofilm formers isolated from a dairy wastewater biofilm. These biofilm formation and growth characteristics are important information for developing a predictive model for the growth and development of bacteria in wastewater systems.

Four known biofilm forming isolates, Citrobacter freundii (DN1), Enterobacter spp (DN3), Raoultella spp (DN5) and Citrobacter werkmanii (DN7), were selected from a biofilm in a dairy factory wastewater system and were confirmed using 16srRNA gene sequencing. These bacteria were associated with biofilm development that completely blocked the irrigators of a primary treated dairy wastewater system. All four bacteria are Gram-negative facultative anaerobes from the Enterobacteriaceae family, which was the predominant family present in the wastewater determine by culture and culture independent methods. Isolates were grown overnight (18 h) in 30 g/L Tryptic Soy Broth (TSB, Bacto™, Difco Laboratories) from stock culture that had been stored at −80 °C.

Monitoring of the wastewater irrigation system from which the four bacteria were isolated, was found to remain at 30 ± 1 °C throughout operation over an 8 hour period. Therefore, 30 °C was selected as the growth temperature.

Monod kinetics parameters (Monod 1942) were determined at different concentrations of ions. At high concentrations, the specific growth rate of bacteria will be independent of the concentration of nutrients and as a result, k_s values are often below chemically detectable limits (Pirt 1975). Ca²⁺, Mg²⁺ and Na⁺ ions were used due to preliminary trials which showed that these were the only ions to exhibit an effect on the biofilm formation of bacteria isolated from a dairy wastewater system.

Growth trials were conducted in a modified Tryptic Soy Broth (mTSB) where the phosphate buffer of TSB was replaced with Tris HCl buffer (20 mM, pH 7) to avoid the precipitation of ions when added to the media and the NaCl content was removed. The mTSB was supplemented with CaCl₂ (Merck, Auckland), MgCl₂ (Ajax Finechem, Thermo Fisher Scientific) or NaCl (Labserv, Thermo Fisher Scientific) at added concentrations of 2, 10, 20, 50 or 100 mM. All trials were conducted with an inoculum of approximately 250 CFU/ml in fresh growth medium. Preliminary trials were conducted to determine the optimal concentration of ions for each bacterium. Triplicate measurements were taken at these optimum concentrations to determine the true value of μ_max.

Aerobic growth trials were conducted with 20 ml aliquots for each concentration and grown in a shaker incubator at 30 °C at 130 RPM. Anaerobic growth trials were conducted with five 10 ml aliquot sets. Each set was kept in a separate anaerobic container to prevent disrupting the atmosphere for subsequent hourly tests.

Aerobic biofilm growth was assessed using 9 ml aliquots containing five 1 cm² 304 stainless steel coupons and grown in a shaker incubator at 30 °C at 70 RPM. Anaerobic trials were conducted in 9 ml aliquot sets, each containing one 1 cm² 304 stainless steel coupon with each set of hourly tests kept in a separate anaerobic container to prevent disrupting the atmosphere. Anaerobic atmosphere was generated using BD BBL™ CO₂ generators, which reduces the O₂ level to less than 1% within 30 min, anaerobic test strips were used in all containers to ensure an anaerobic atmosphere was generated.

Growth curves were analysed using an impedance system (SyLab, BacTrac). Samples (1 ml) were taken over a five-hour time span (hours three to seven after inoculation inclusive) and added into the BacTrac measuring vials containing 10 ml TSB which were stored at 4 °C. The BacTrac was set to measure impedance at 30 °C with a 1.5 hour warm up time to stabilize media from 4 °C to 30 °C. Impedance measurements were recorded over a 24-hour period. A threshold value of 3% was used (early-mid exponential growth) to generate a calibration curve for each microorganism to predict the number of cells per millilitre in each sample. Values of K_s were calculated using ion concentrations of 2, 10 and 20 mM. The Langmuir plot is a linear plot of s (concentration) against s/μ_max with intercept -K_s (Owens & Legan 1987).

The total yield of bacteria was calculated by growing cultures overnight to the start of the stationary phase (10 hours) at varying ion (1, 2, 10, 20, 50, 75 and 100 mM) or nutrient levels (1, 2, 10, 20, 50, 75 and 100% TSB). Total CFU/ml measurements were determined using the BacTrac (3% threshold). Total yield was calculated from the gradient of a plot of log CFU/ml vs nutrient or ion concentration and reported as CFU/ml/g mTSB or CFU/ml/mM.

All growth rate studies were performed in at least triplicate, and results were expressed as mean μ_max. Tukey's Honest Significant Differences (HSD) was used to determine differences between mean μ_max (α = 0.05). Values of μ_max exhibiting no common letters were considered to be significantly different according to Tukey's grouping. Graphed results are expressed as mean values with error bars representing the 95% confidence interval. Error bars were calculated from the standard deviation of the sample replicates and the Students t-distribution value. For the growth rates, the combined aerobic and anaerobic environment provided six total replicates while the yields were conducted in triplicate.

TSB was chosen as the nutrient source, rather than milk (representing a dairy system) as milk fouled the plastic test plates giving false results. Lindsay et al. (2002) utilized 0.1% TSB to represent a washed surface in a milk plant. In the present trial, TSB (30 g/l) favoured planktonic growth while TSB (3 g/l) favoured biofilm.

Ca²⁺ showed the biggest effect on the biofilm formation (Figure 1). In most cases the results of the individual isolates showed the same pattern as that of the mixed cultures. Low concentrations of Ca²⁺ (0.02–0.1 M) added as CaCl₂, increased the amount of biofilm of both individual and mixed bacteria. However, when the concentration of the Ca²⁺ ions increased above 0.5 M, biofilm formation was inhibited to the point of no detection above the negative control wells. Ca²⁺ present in the wastewater of the plant can vary greatly over time due to calcium naturally present in the milk, as well as the practice at the plant of addition of calcium hydroxide [Ca(OH)₂] to neutralize the pH of the wastewater.

Growth experiments were conducted over a range of added ion concentrations (2, 10, 20, 50 and 100 mM). These preliminary trials determined the optimum concentration (highest μ_max, shown in Table 1) at which the individual bacteria would grow. Figure 2 shows the maximum specific growth rates (μ_max) for all bacteria in each environmental condition, planktonic (free floating) or biofilm (attached to surface) growth, and ion content.

All bacteria and ion combinations had at least one similar grouping in the Tukey's HSD test for aerobic and anaerobic environments (data not shown), therefore the reported growth rates are combined aerobic and anaerobic results. As the growth rates for the aerobic and anaerobic growth rates were statistically similar an average μ_max is appropriate. The three slowest growth rates, which had significant difference (p-value <0.05) to the three fastest growth rates (Figure 3), consisted of growth only in the presence of Ca²⁺ (DN1 planktonic and DN5 and DN7 biofilm). The three fastest growth rates were all in the presence of Na⁺ (DN1 planktonic and DN5 planktonic and biofilm). The two Citrobacter spp (DN1 and DN7) isolates exhibited no biofilm growth in the presence of Mg²⁺ and Na⁺. Biofilm growth of C.werkmanii (DN7) in the presence of Ca²⁺ exhibited the slowest μ_max (1.10 h⁻¹) out of all bacteria while planktonic growth of C.freundii (DN1) exhibited the greatest μ_max (1.67 h⁻¹) in the presence of Na⁺ ions.

The saturation constant is the concentration (of nutrient/ion) at half the maximum specific growth rate. The bacteria tested showed higher saturation constants than other reports of similar bacteria (Pirt 1975). In most cases, the calculated saturation constant was less than 10% of the ion concentration in the media.

In the planktonic environment, k_s values ranged between bacteria (Table 2), with the largest k_s values recorded in the presence of Ca²⁺ for planktonic cultures. Raoultella spp (DN3) exhibited some of the lowest k_s values (in the presence of Mg²⁺) of the four bacteria tested. This suggests that the Raoultella spp is less sensitive to the presence of Mg²⁺ than the other bacteria. C. freundii (DN1) also exhibited the highest planktonic k_s values of all bacteria (3.23 mM Ca²⁺ and 1.49 mM Mg²⁺). This would indicate that C. freundii has increased sensitivity to ions present in the wastewater compared to the other bacteria. DN1 and DN7 did not grow in the presence of Mg²⁺ and, therefore, the results do not show any k_s values.

Total yield is the maximum number of cells achieved in the growth medium. The overall yield of the bacteria tested varied depending on both ions and nutrients (mTSB) present (Figure 4). The highest yield for DN1 and DN3 was in mTSB with no added ions, while for DN5 and DN7 the highest yield was in the presence of Ca²⁺. However, the yield for DN5 and DN7 in Ca²⁺ was not significantly different to the yields in mTSB. In three out of the four bacteria, Ca²⁺ produced a significant (p-value <0.05) increase in the yield over all other ions tested. C.freundii (DN1) was the only bacteria that showed no significant differences between all ions. Raoultella spp (DN3) showed not only significant differences (p-value <0.05) between all ions tested but also the mTSB concentration. It is of note that no yields were recorded from the triplicate test of Raoultella (DN3) and Enterobacteriaceae (DN5) in the presence of Na⁺ ions. Of the four tested bacteria, Enterobacteriaceae spp (DN5) demonstrated the highest yield with the yield in the presence of Ca²⁺ being significantly different to all but the yield recorded by itself in mTSB.

Ca²⁺ along with other divalent cations, have the potential to bind to proteins and the EPS causing DCB (Donlan 2002; Michiels et al. 2002; Song & Leff 2006; Flemming & Wingender 2010; Somerton et al. 2015). However, the effect of Ca²⁺, Mg²⁺ and Na⁺ ions on the growth rates in both the planktonic and biofilm tests suggests that ions play more complicated roles than just (DCB).

Due to no net biofilm growth of DN1 and DN7 in the presence of Mg²⁺, an environment containing this ion would favour the growth of the planktonic rather than biofilm growth. It is possible that Mg²⁺ is negatively influencing the production and excretion of extracellular proteins, polysaccharides, RNA and extracellular DNA (eDNA). These EPS are likely to play a role in the attachment of bacteria to the surface of the coupons, and therefore Mg²⁺ could potentially control the biofilm formation. Oknin et al. (2015) showed that 50 mM Mg²⁺ ions significantly inhibited the biofilm formation of a Bacillus subtilis and could be affecting single transduction in matrix gene expressions. However, Mg²⁺ limitation has also been shown to increase biofilm formation by repressing biofilm formation repressor proteins (RetS) in a P. aeruginosa isolate (Mulcahy & Lewenza 2011) while Song & Leff (2006) showed Mg²⁺ increased the bacterial adhesion of Pseudomonas fluorescens, surface colonization and depth of biofilm, but did not affect the growth of the planktonic cells. The range of results shows that different bacteria respond differently to the presence of Mg²⁺.

Typical total Ca²⁺ levels in unprocessed milk (sum of bound and free) are 0.026–0.032 M (∼1,130 mg/L) (Gaucheron 2005). In the present trial, the additions of Ca²⁺ ions showed that low concentrations increased the biofilm formation while the high concentrations inhibited biofilm formation. Previous research has shown that Ca²⁺ ions can affect the biofilm formation of bacteria found throughout the dairy industry (Teh et al. 2015). Somerton et al. (2012) showed that Ca²⁺ ions in milk were important for biofilm formation of Geobacillus spp isolates from a milk powder plant. Guvensen et al. (2012) showed that Ca²⁺ at 1.0 × 10⁻⁵ M (log CFU/cm² = 2.3 × 10⁷), 2.5 × 10⁻⁵ M (log CFU/cm² = 8.3 × 10⁷) and 5.0 × 10⁻⁵ M (log CFU/cm² = 5. × 10⁷) increased the biofilm formation of Sphingomonas paucimobilis on stainless steel coupons over media with no Ca²⁺ present (log CFU/cm² = 1.2 × 10⁷). This was explained by an increase in the hydrophobicity of the cell surface increasing bacterial attachment. Analysis of the Ca²⁺ ion content in the dairy wastewater is approximately 1–4 mM (∼100 mg/L).

Ca²⁺ ions not only influence the biofilm formation due to DCB but can also affect the binding of bacteria to surfaces. Geesey et al. (2000) show that Ca²⁺ can be involved in both specific and non-specific interactions with the EPS adhesion molecules at the cell surface as well as affecting the interactions of cells with the substratum. P. aeruginosa showed increased irreversible attachment when concentrations of NaCl or CaCl₂ increased from 0.1 to 10 mM (Stanley 1983). In sterile salt medium it was found that deficiencies in both Ca²⁺ and Mg²⁺ resulted in less EPS being produced and reduced attachment of cells to glass slides. Ca²⁺ only deficient media had little effect on the amount of cells attached while Mg²⁺ only deficient media had an intermediate effect between that of Ca²⁺ deficient media only and media deficiency in both ions (Allison & Sutherland 1987).

Das et al. (2014) showed with naturally occurring eDNA, Gram negative bacteria (Pseudomonas aeruginosa, Aeromonas hydrophilla and Excherichia coli) experienced greater aggregation and settling than that of the Gram positive (Staphylococcus aureus, Staphylococcus epidermidis and Enterococcus faecalis) bacteria tested. Additionally, with the removal of eDNA, Gram negative bacterial strains exhibited decreases in aggregation. When eDNA and Ca²⁺ were added, four out of the six strains showed increased aggregation, which was hypothesised to be due to cationic bridging mediated by the Ca²⁺. The range of results shows different bacteria are affected in different ways by the addition of Ca²⁺. Allison & Sutherland (1987) showed an interaction between Mg²⁺ and Ca²⁺ taking place with both ion deficient media reduced the amount of EPS produced and bacterial attachment. Dairy wastewater contains both ions, allowing for potential increased EPS production and attachment to take place in the system.

Na⁺ ions, due to being a monovalent cation, can affect both the formation of biofilms and the growth of bacteria. In a milk formulation with high concentrations of monovalent to divalent cations, biofilm inhibition was observed, possibly due to the decrease in electrostatic forces within in the biofilm. The high abundance of monovalent ions can displace the divalent cation preventing DCB from taking place (Somerton et al. 2015). Na⁺ addition to an activated sludge system especially when the ratio of monovalent to divalent cations exceeded 2:1, resulted in deterioration of the settling and dewatering properties. However, this could be restored when the divalent cations (Ca²⁺ and Mg²⁺) were increased reducing the ratios back below 2:1 (Higgins & Novak 1997). Xu et al. (2010) investigated the effect of varying concentrations of NaCl addition (0–10%) of the growth and biofilm formation of four foodborne pathogens. There were two distinct patterns of the growth and development of biofilms. At low added amounts of NaCl (0%, 2% and 4%) the populations of attached cells increased until day 4, after which they decreased until day 10. Higher added concentrations (6%, 8% and 10%) exhibited no decrease in the attached cells. The maximum growth rate of attached L. monocytogenes decreased with increasing NaCl concentrations. The fastest growth was recorded at 0% added NaCl (1.25 d⁻¹) while the slowest (0.16 d⁻¹) was found at 10% NaCl addition (Webb 1949).

The k_s values of the bacteria taken from the dairy wastewater system are higher than reported elsewhere. Dean & Rogers (1967) determined the k_s value of Klebsiella in the presence of Mg²⁺ to be 2.3 μM. In this study, the k_s values for the wastewater isolates in the presence of Mg²⁺ ranged from 0.01 mM to 3.72 mM. The wastewater system has a high abundance of all ions tested throughout normal operation and these higher k_s values could indicate that the wastewater system is selective for bacteria that are dependent on the ions tested for optimal growth.

The growth rates of four known biofilm formers associated with the blockage of a primary treated dairy wastewater irrigation system were assessed. Biofilm growth of the four harvested bacteria, C.freundii, Enterobacteriaceae spp, Raoultella spp and C.werkmanii, was on average slower than growth in the planktonic phase. Ca²⁺ was of interest due to the increased biofilm formation with the addition of calcium at 0.1 M while the addition of 0.5 M or greater completely inhibited biofilm formation. The three slowest growth rates were determined to be in the presence of Ca²⁺. However, Ca²⁺ significantly increased the overall yield with three out of the four isolates. The different effects of ions on the growth rates, yields, saturation constants and biofilm formation suggest that more than one mechanism is involved in the utilization of these ions. These ions could have an effect on the excretion and production of EPS, metabolic pathways or DCB. However, current studies show that the effect of these ions on biofilm formation varies greatly. Saturation constants, while higher than found in the literature, were less than 10% of the added ion concentration. The high k_s values and low growth rates but high yields could imply that the dairy wastewater system is selective for bacteria that are dependent on the ions, especially Ca²⁺ for growth and biofilm formation. These results would allow the development of a mathematical model to predict the initial biofilm growth and development in the system. This would allow operators or technical personal to enact cleaning or reduction techniques, such as ozone treatment, before biofilm growth becomes a problem.

This work has no funding to declare

None declared.