Listeria Monocytogenes is the causative agent of human listeriosis, an important foodborne disease with a high fatality rate particularly in new-born infants, pregnant woman, elderly and immunocompromised patients [1-3]. Listeriosis is almost entirely transmitted through the ingestion of contaminated foods [4].

      The ability to colonize food environments, enduring an extensive variety of physicochemical conditions and different processing hurdles, is due to L. monocytogenes physiological and ecological traits [5,6].

      After gaining access to a food facility, through incoming raw materials and ingredients, packaging materials, or even food handlers, L. monocytogenes is able to persist for months or years within the food premises, especially in food contact surfaces [7,8].

     L. monocytogenes can adhere to different surfaces within the food industry, such as plastic, polypropylene, rubber, stainless steel, glass and produce biofilms [7,9]. In the biofilm, bacteria are embedded by an extracellular matrix able to function as a structural scaffold and defense barrier [8]. Once established, biofilms confer protection against harsh environmental conditions, enabling to sustain the survival of bacteria and tolerance to food environment related hurdles [10]. These persistent strains have been linked to recurring contamination of finished products [10-12]. Although this cross-contamination transfers low levels of L. monocytogenes onto food, its psychrotrophic nature enables growth during refrigerated storage, reaching levels that might represent an increased risk to the consumer [13-15]. This is even more concerning if temperature fluctuation occurs in any of the production and distribution stages, or even at the household level, and if the food is a ready-to-eat product, not requiring a listericidal treatment before consumption [16-18].

      This work aims to assess L. monocytogenes growth in planktonic suspension and in mono-species biofilms, depicting food and food-producing environment conditions.

      The resulting data in culture media and different temperatures will provide important initial information to further assist predictive growth modeling in food matrices and environments, also enabling subsequent quantitative microbial risk assessments.

Selection and revival of L. monocytogenes strains

      To account for variation in growth and survival among Listeria Monocytogenes strains and to have representatives of the three serogroups more frequently related to human disease, three reference strains were assessed: L. monocytogenes CECT 4031 (serogroup IIa), L. monocytogenes CECT 935 (serogroup IVb) and L. monocytogenes CECT 937 (serogroup IIb).

      For strains’ revival, stock cultures stored at -80°C in preservation cryotubes containing Brain Heart Infusion (BHI) broth (Scharlab, S.L, Barcelona, Spain) supplemented with 15% glycerol (Merck KGaA, Darmstadt, Germany) were thawed and 100 µl of inoculum was transferred into 5 ml of BHI broth. After 24 hours (h) of incubation at 37°C, a loop (10 µl) of inoculum was streaked onto BHI agar (Scharlab, S.L.) and incubated at 37°C for 24 h.

L. monocytogenes isothermal growth in BHI broth

      An isolated colony of L. monocytogenes was suspended in 5 ml of BHI broth (Scharlab, S.B.). The suspension was incubated at constant temperature of 12° and 37 °C. Each sample was periodically examined to assess the growth of L. monocytogenes. For that, bacterial suspensions’ optical density at 600 nm (OD_600nm) was measured on a spectrophotometer Ultrospec 2000 (Pharmacia Biotech, Cambridge, England) at regular time intervals. Together with OD_600nm, enumeration of viable bacterial cells (VCC) at regular time intervals was also performed. Three independent growth experiments were performed for each temperature condition (37°C and 12°C), in each of the considered sampling time points (Table 1).

Curve fitting and growth parameters estimation

     Growth curves were fitted to Baranyi and Roberts primary predictive model [19] (Equations 1-3), using DMFit online (Quadram Institute, Norwich, United Kingdom), to estimate maximum specific growth rate (μ_max), lag time (λ), initial and final concentration (C₀ and C_f , respectively); R-square (R² ) and standard error of fit (SE) were used to evaluate the performance of the models built in this study. A fitting method for repeated measures was applied considering the different replicates analyzed in each time point.

      Where: N(t)= log of cell concentration (C_fu/ml(g)) at time t (h); N0= log of initial cell concentration (C_fu/ml(g)); μ_max= maximum specific growth rate (log C_fu/ml(g)/h); Nmax= log of maximum cell concentration; q0= parameter expressing the physiological state of cells when t= t0; λ= lag time (h). In this work, μ_max was based on the inflection of the growth curve slope in the exponential phase [20].

      The resulting growth curves and parameters were compared to predicted values generated by ComBase Predictor Growth Model (ComBase, Hobart, Australia). ComBase model was run with the following selected parameters: Initial level= 4.5 log C_fu/ml; pH= 7.4 and aw= 0.997 were BHI broth data.

Calibration curves

      To study OD_600nm and cell count (C_fu/ml) relation, calibration curves were prepared. For that, L. monocytogenes CECT 4031 were cultured on BHI agar (Scharlab, S.B.) for 18 h at 37^o C. Afterwards, cultures were transferred to 10 ml of tryptone salt broth (Scharlab, S.B.), resulting in a suspension with an OD_600nm of 0.4 – 0.5 and serial dilutions were prepared. Serial dilutions OD_600nm were measured and correlated with C_fu/ml obtained in plate counts on BHI agar (Scharlab, S.B.). This assay was performed in triplicate.

Biofilm formation assay

      The protocol proposed by Romanova, Gawande, Brovko and Griffiths [21] was used with some modifications to obtain a 5-day L. monocytogenes mono-cultural biofilm. A single colony of each selected strain was inoculated in Buffered Peptone Water (BPW) (Scharlab, S.B), incubating overnight at 30°C. Bacterial suspension OD_600nm was adjusted to 0.1 approximately, in Ultrospec 2000 (Pharmacia Biotech), to obtain a concentration of 8 log C_fu/ml, according to the above mentioned calibration curve. For each strain, 4 μl were transferred into three separate wells of polystyrene flat-bottomed microtiter plates (Normax, Marinha Grande, Portugal) filled with 200 μl of BPW. Three wells were used as negative controls, containing only BPW. The plates were statically incubated at 30°C for 5 days. The solution was then removed from the wells that were rinsed with sterile distilled water to remove loosely associated bacteria and the attached biofilms were evaluated by Viable Cells Counts (VCC) and Crystal violet staining (cvOD). This assay was performed in triplicate, with three replicates for each strain. L. monocytogenes CECT 4031 and CECT 935 were further assess for biofilm formation at 12^o C and 20^o C. For that, the abovementioned procedure was followed and the plates were statically incubated at 12ºC and 20ºC for 5 days.

Biofilm assessment by viable cells counts

      The biofilm was detached from the well surface with 100 μl of BPW using a mini cell scraper and sonicated (Ultrasonic bath MXB14, Grant Instruments, England) for 5 min to detach and collect sessile cells. Another 100 μl of BPW were pipetted into each well, serial 10-fold dilutions were prepared and 10 μl samples were dropped onto the surface of a Tryptone Soy Agar (TSA) (Scharlab, S.B) plate. Colonies were enumerated after overnight incubation at 30°C in a stereoscopic magnifier (Nikon SMZ645, Tokyo, Japan).

Biofilm assessment by crystal violet staining

      The microtiter plate was left air drying for 45 min in the laminar flow hood. Biofilm was stained using 220 μl of 0.1% crystal violet (bioMérieux, France) solution for 15 min at room temperature. After stain removal, the wells were washed three times with sterile distilled water and left air drying for 30 min in the laminar flow hood. To quantify adhered cells, 220 µl of detaining solution (ethanol: Acetone 80:20 v/v) were added to each well for 15 min at room temperature. The microtiter plate was then shaken (Ultrasonic bath MXB14, Grant) for 5 min and the Crystal Violet OD (cvOD) was measured in SpectraMax 340PC (Molecular Devices, California, USA). Each absorbance value was corrected by subtracting the average absorbance readings of the blank control wells.

      Adherence capability was based on the cvOD exhibited by bacterial biofilms, according to Stepanović, Cirković, Ranin, and Svabić-Vlahović [22]. The cut-off cvOD (cvODc) was defined as three standard deviations above the negative control mean cvOD. The strains were classified as no biofilm producers (cvOD ≤ ODc), weak biofilm producers (cvODc ≤ OD ≤ 2 x ODc), moderate biofilm producers (2 x cvODc ≤ cvOD ≤ 4 x cvODc) and strong biofilm producers (4 x cvODc < cvOD).

Data analyses

      All quantitative data are presented as mean values with Standard Deviation (SD) from three independent experiments.

      OD_600nm and VCC data were used to fit a linear regression in Microsoft Excel 2016 software (Microsoft Corporation, Redmond, USA).

      Pearson’s correlation analysis was performed in GraphPad software Prism 5 (GraphPad Software, La Jolla, USA) to relate OD_600nm and experimental VCC values.

      The calibration curve was obtained from the following equation:

      (4) C_fu/ml = slope × OD_600nm + interception.

      For L. monocytogenes growth characterization, VCC experimental results were adjusted using DMFit Online (Quadram Institute, Norwich, United Kingdom), which was also used to estimate growth related parameters.

      To assess L. monocytogenes biofilm formation parameters at different temperatures (12^o C, 20ºC and 30^o C), Pearson’s correlation analyses were used to evaluate the interdependency of cvOD and VCC. Two-way ANOVA was used to investigate the temperature effect on biofilm formation.

Listeria Monocytogenes isothermal growth in BHI broth

      Growth rates of L. monocytogenes were studied at different temperatures in a defined medium (BHI). L. monocytogenes CECT 4031 was chosen since it is the type strain for this specie [23].

      The selected temperatures were 37^o C, corresponding to L. monocytogenes’ optimal growth temperature [24,25], and 12^o C, which is used in food producing rooms at industrial facilities [26].

      Figure 1 shows the resulting growth curves based on L. monocytogenes CECT 4031 OD_600nm values in BHI at each sampling time point, incubated for 26 hours at 37°C (A) and for 12 days at 12°C (B).

      L. monocytogenes CECT 4031 growth occurred at both temperatures, although differences were observed (Figure 1). At 37°C a lag phase of approximately 10 hours was observed, followed by an exponential growth phase from 10 h to 18 h. From then on until the end of the incubation time (26 h), stationary phase was observed. The maximum OD_600nm value was 0.999 ± 0.280 at 18 h. Mytilinaios et al. (2012), when studying growth rate of L. monocytogenes in tryptone soya broth at 37ºC, obtained an average maximum optical density of 0.99.

     At 12°C the lag phase lasted approximately 48 hours, and from 48 h to 120 h the exponential growth phase was observed. Stationary phase seems to have been reached at 120 hours. The maximum OD_600nm value was 0.902 ± 0.017 at 168 h.

      Figure 2 presents the obtained growth curves for L. monocytogenes CECT 4031 in BHI considering total Viable Cell Counts (VCC) for the assessed sampling time points, at 37°C (A) and at 12°C (B).

      At 37°C, a stationary phase can be observed in the first 18 h of incubation, however the maximum value of VCC 9.484 ± 0.678 log C_fu/ml was obtained at 26h. At 12°C, there seems to be a potential lag phase of approximately 48 hours, and from 48 h to 120 h the exponential growth phase can be observed. Stationary phase seems to have been reached at 120 hours after inoculation. The maximum value of VCC reached was 9,277 ± 0,210 log C_fu/ml at 168 h. Castro [26] obtained similar results when studying the growth of L. monocytogenes in packaged raw milk, in which, from initial low counts, L. monocytogenes was able to develop to 4.3 ± 0.4 log C_fu/ml at refrigerated temperatures (10o C).

      Temperature had a considerable influence on L. monocytogenes growth, because although initial and final concentrations are similar for both temperatures, the time needed to reach final concentration was higher for the lower temperature (12ºC). In order to reach approximate maximum concentrations, around 18/20 hours were needed at 37^o C and 5/6 days at 12^o C. This was observed when applying both OD_600nm and VCC measurements.

      The correlation curves obtained for L. monocytogenes 4031 using OD_600nm and VCC are shown in Figure 3A (37^o C) and Figure 3B (12^o C).

      When comparing experimental OD_600nm and VCC results (Table 2), although a good correlation was observed between both methods, Pearson correlation and R² were higher at 120C

      At 37^o C, the R² value (0.6929) revealed a low adjustment/ fit when using OD_600nm to estimate VCC. This difference may be related to the fact that optical density measures the turbidity of a suspension, and because of that its relationship with cell concentration may not be linear (deposits of non-viable cells in suspension are also measured as total number of cells). Some authors defend that the difference between both methods is especially evident when assessing growth parameters of isolates in stressful conditions, as morphological changes in the cell may result in optical density values that do not reflect the actual cell numbers [27-29]. Jones, Gill, and McMullen [30] showed that cold adaptation can sometimes cause cell elongation, as cells further increase in cell length before dividing to normal cell length, strongly affecting the relationship between the optical density levels reached and the estimated log C_fu/ml.

      Nonetheless, the correlation of both methods was high, indicating that VCC values can be reliably inferred through optical density measurements with the use of calibration equations, delivering fast and effective results.

Curve fitting and growth parameters estimation

      For each temperature and using VCC values, growth curves were built by fitting experimental data to the Baranyi’s DMFit online version.

      At 37°C, L. monocytogenes CECT 4031 concentration peaked at 21h reaching 9.184 ± 0.204 log C_fu/ml (Figure 4A) and remaining stable until the end of incubation time (stationary phase). At 12°C, L. monocytogenes CECT 4031 reached a maximum final concentration of 9.117 ± 0.133 log C_fu/ml after 144 h (Figure 4B) and remained stable until the end of incubation time (stationary phase).

      The obtained R² results for both temperatures revealed a good fit of the model to experimental data.

      Considering the obtained estimated growth parameters (Table 3), μ_max at 37^o C was higher than at 12^o C, and a longer lag phase was observed at this temperature (12^o C). The longer lag phase at 12°C could be due to an adaptation period to lower temperatures. Similar growth parameters were obtained by Pla [31], when assessing L. monocytogenes CECT 4031 growth in Tryptic soy broth supplemented with 0.6% yeast extract at 37ºC (μ_max = 0.447 and λ = 1.86), and by Wang [32] when studying the growth of L. monocytogenes in BHI at 10o C (μ_max = 0.066 and λ = 17 h).

      More time was needed for L. monocytogenes to grow at 12^o C and reach the same concentrations as those obtained at 37^o C. However, final concentrations of the pathogen were very similar, emphasizing the ability of L. monocytogenes to grow at refrigerated temperatures, as the ones used in food producing rooms at industrial facilities. In fact, after 5/6 days of incubation at 12^o C, the levels of L. monocytogenes were similar to the ones reached at 37^o C.

      Predictions obtained from Baranyi’s model using L. monocytogenes VCC at both temperatures were compared to the estimated growth using ComBase Predictor Growth model (Figure 5).

      In general, predictions from ComBase Predictor growth model and this study’s results were quite similar. However, when considering the temperature of 37°C, the growth curve based on VCC presented lower values until 14 h of incubation, but higher maximum values when compared with ComBase estimated growth curve. The stationary phase was reached sooner in ComBase estimated growth curve. Also, μ_max obtained with ComBase predictor was of 0.480 log C_fu/ml/h, higher when compared to the experimental data μ_max (0.375 ± 0.072 log C_fu/ ml/h, Table 3).

      At 12^o C, the stationary phase was reached later in the estimated growth curve based on VCC. ComBase growth curve presents lower maximum log C_fu/ml values. Considering the maximum growth rate (μ_max) obtained with ComBase predictor, at 12°C a μ_max of 0.068 log C_fu/ml/h was estimated, which was higher than the experimental data μ_max (0.054 ± 0.001 log C_fu/ ml/h, Table 3).

\

Calibration curves

     A calibration equation for each strain in the study was obtained by performing three independent calibration curves, in which viable cell counts were plotted against OD_600nm data (Figure 6), allowing a given concentration (C_fu/ml) to be determined from the OD_600nm value assessed using a spectrophotometer.

      Regression parameters and goodness of fit (R² ) of the calibration curve are shown in Table 4.

     Results indicate a high correlation between OD_600nm values and C_fu/ml (coefficient of determination, R² = 0.9987), which indicates that the method is reliable for quantifying L. monocytogenes strains. Pearson’s correlation coefficient indicates a strong positive correlation between C_fu/ml and OD_600nm (ρ = 0.9994, p < 0.0001). Similar results were obtained by RipollesAvila [33] for L. monocytogenes CECT 935.

Biofilm formation assay

      The assessed strains in biofilms revealed cvOD values ranging from 0.068 ± 0.001 to 0.1078 ± 0.005 and viable cell counts of 6.013 ± 0.346 log C_fu/ml to 7.391 ± 0.227 log C_fu/ml after 5 days of growth in polystyrene microtiter wells (Figure 7).

      According to Stepanović, Cirković, Ranin, and Svabić-Vlahović [22] classification, all the strains revealed a weak biofilm- forming ability. Meloni [25] obtained similar results when studying L. monocytogenes isolates from fermented sausage processing plants: 65% of all isolates were weak biofilm producers.

      While L. monocytogenes CECT 4031 revealed the lowest values for both VCC and cvOD at 30^o C, L. monocytogenes CECT 935 exhibited the highest biofilm-forming ability, based on both biofilm formation parameters. Similar results were obtained by Ripolles-Avila [33] when studying the quantification of cell density within L. monocytogenes biofilms based on cvOD.

      Considering the selected methods to analyse biofilm formation - VCC (log C_fu/ml) and cvOD, a positive and strong correlation (ρ= 0.7749, p= 0.009) was obtained. This coefficient value indicates that both methods presented a good relation, being reliable to quantifying L. monocytogenes biofilm formation and complementing each other.

      Differences between cvOD and VCC results are due to the nature of each method determination: While cvOD measures the turbidity of a suspension and quantifies total biomass (viable and non-viable cells, and extracellular matrix components), VCC only considers live cells [34]. Although monitoring biofilm formation with VCC is time-consuming, laborious and expensive because the technique is based on serial dilutions and plating methods, it remains the method of reference for monitoring bacterial growth [27,29,35]. On the other hand, cvOD method may overestimate the number of viable, attached cells [34,36]. Additionally, stressful conditions may induce morphological changes in cells, such as cell elongation, strongly affecting the relationship between cvOD and VCC [29].

      For further testing, L. monocytogenes CECT 935 and CECT 4031 were selected based on Tukey’s test results for VCC and cvOD, that presented statistically significant differences (p < 0.05) in both biofilm formation parameters.

      All of the tested L. monocytogenes isolates were able to form biofilm at 30^o C, 20^o C and 12^o C (Figure 8). Although there were no significant differences (p= 0.958) in biofilm formation at different temperatures considering VCC values (Figure 8A), using cvOD (Figure 8B) significant differences (p= 0.0002) were observed in biofilm formation for the considered temperatures.

      It is important to underline that the studied L. monocytogenes strains revealed biofilm-forming ability at refrigerated processing environment temperatures (12^o C and 20^o C), as the ones used in refrigerated producing rooms in the food industry. Still, the highest biofilm formation occurred at 30^o C for all L. monocytogenes tested strains, confirming that temperature influences L. monocytogenes biofilm formation, as has been previously reported. Tomiĉić [37] reported that growth conditions affected biofilm formation, revealing the lowest biofilm formation for the lowest tested temperature. Similar results were obtained by Russo [2] when testing L. monocytogenes biofilmforming ability on polystyrene under different temperatures. In fact, Abeysundara [38] observed that biofilm formation was influenced by temperature, resulting in decreased biofilm formation with decreasing temperature. Di Bonaventura [39] demonstrated that biofilm production on polystyrene surfaces at 37ºC was significantly higher than at 4o C. However, according to Puga [40], some evidence suggests that persistence of certain strains may be enhanced by low temperatures adaptation mechanisms. Temperature may influence flagella formation actively involved in the adhesion to different surfaces [2].

      In this study, when fitting planktonic L. monocytogenes isothermal growth experimental data to Baranyi’s model, a good fit was obtained for both temperatures. The estimated growth parameters confirmed that μ_max at 37^o C was higher than at 12^o C, and a longer lag phase was observed at this temperature, indicating an adaptation period to lower temperatures. Still, L. monocytogenes final concentrations were identical, emphasizing its ability to grow at refrigerated temperatures. Additionally, experimental results from the isothermal growth assay and ComBase Predictor growth model were rather similar, but higher μ_max were estimated for both temperatures by the predictor model.

      The studied L. monocytogenes strains demonstrated biofilmforming ability at 12^o C, 20^o C and 30^o C after 5 days of growth and all revealed to be weak biofilm producers. Although there were no significant differences in biofilm formation at different temperatures considering VCC values, using cvOD significant differences were found and the highest biofilm formation occurred at 30^o C.

      Still, a positive and strong correlation was found between VCC and cvOD results, which complement each other in biofilm formation assessment.

     Overall, this study’s outcomes contribute with important preliminary data on L. monocytogenes growth at different temperatures, whether in the planktonic form or in biofilms. The gathered data will further assist predictive modeling and risk assessment studies, improving possible interventions and mitigation strategies to control this important foodborne pathogen.

Funding

      This work was financed by national funds through FCT - Foundation for Science and Technology, I.P., within the scope of project UIDB/00276/2020.

      The authors gratefully acknowledge Maria Helena Fernandes, Maria José Fernandes and Maria Paula Silva for the technical support and Carla Carneiro for the discussion time. We acknowledge the logistic support of CIISA - Centro de Investigação Interdisciplinar em Sanidade Animal, Faculdade de Medicina Veterinária, Universidade de Lisboa, Avenida da Universidade Técnica, 1300-477 Lisboa, Portugal.