Enriching Cured Meat Products with Bioactive Compounds Recovered from Rosa damascena and Rosmarinus officinalis L. Distillation By-Products: The Pursuit of Natural Antimicrobials to Reduce the Use of Nitrites

Abstract

Endogenously formed N-nitroso compounds (NOCs) from nitrite use in red meat have been recently linked to the risk of colorectal cancer. Therefore, replacing sodium nitrite (E250) with natural antimicrobials, such as bioactive compounds, is an issue of great industrial and scientific interest. In this research, such compounds were extracted from rose petal (Rosa damascena) and rosemary leaf (Rosmarinus officinalis L.) by-products of the essential oil industry and appropriately introduced in commercial cured meat products, as partial nitrite replacers. Shelf-life tests demonstrated an inhibitory effect of the rose extracts on microbial growth, obtaining similar or reduced growth rates and a prolonged lag phase, when compared to conventionally produced samples (CNT): μ_max,ref = 0.128 vs. 0.166 d⁻¹ and λ_ref = 15.9 vs. 12.4 d at 4 °C, respectively. None of the bacon samples suffered from lipid oxidation in the examined period. The RSE samples (containing the rose extract and half of the nitrites) exhibited a good retention of their red color, receiving an acceptable sensory score throughout storage. Since the proposed partial nitrite replacement leads to an important shelf-life extension, namely 78 d (RSE) vs. 60 d (CNT), our results provide preliminary proof of the potential use and valorization of such side streams as effective natural antimicrobial agents for cured meat products, in order to reduce the use of nitrites.

1. Introduction

Bacon, a popular cured meat, is a perishable product that requires the use of the additive sodium (E 250) or potassium (E 249) salts. These salts act as antioxidants, stabilizing its color and limiting lipid oxidation, and as antimicrobials that prevent the germination of toxicogenic Clostridium botulinum spores and vegetative cell division of spoilage bacteria [1,2]; yet, they also positively contribute to flavor. Briefly, sodium nitrite (NaNO₂) added in meat dissolves easily into nitrogen dioxide (NO₂) and nitrogen monoxide (NO). The latter reacts with the meat protein myoglobin, forming nitroso-myoglobin, a bright-red pigment which is unstable and is denatured during thermal treatment to nitrosyl-hemochrome, an attractive reddish-pink pigment crucial for consumer acceptance. Nitrites also have antimicrobial properties by blocking bacterial metabolic enzymes, like catalase and peroxidase, limiting oxygen absorption, and breaking the gradient of membrane protons [2,3]. As an antioxidant, nitrite similarly affects the color formation, neutralizing free radicals, shutting down oxygen and reactive oxygen species, and blocking lipid and protein oxidation cycles [4].

However, over the years, concerns have been raised about the potential adverse effects of nitrites, specifically N-nitroso compounds (NOCs), on human health. Cured meats are implicated in this issue for their presence of N-nitroso compounds, specifically N-nitrosamines (RR’NNO) and N-nitrosamides [5,6]. Nitroso compounds are derived from the reaction of a nitrosating agent and amines, which are abundant in a meat matrix. Nitrosating agents mainly comprise nitrites added in meat, as well nitrous acid and nitric oxide, which are also produced from nitrites [2]. The compelling data on the involvement of N-nitroso compounds in the development of carcinogenesis in humans have prompted, for several years now, authorities in numerous countries to form regulations [7]. The European Union has lowered the permissible limits of nitrates in meat products due to evidence linking N-nitroso compounds to human carcinogenesis. The EU issued an amendment to Regulations (EC) no. 1333/2008 [8] and 231/2012 [9] concerning nitrites and nitrates in food additives. This amendment gradually reduces the maximum limits for nitrites (E 249–250 [8]) and nitrates (E 251–252 [9]) in various products. A transitional period has been given until 9 October 2025, after which the limits will be significantly reduced. For bacon, the limit for nitrite salts will remain at 150 ppm until 9 October 2025, then it reduces to 80 ppm [10].

This situation has prompted the food industry and scientific community to work together to reduce the use of nitrates and nitrites in cured meats [4] through various strategies, such as the application of natural substances as alternatives to nitrates, bacteria with nitrite-reducing effects, heat treatment, irradiation, and atmospheric plasma treatment [2]. The food industry has shown promising potential in incorporating natural additives derived from plants that can partially or fully substitute nitrites. Two approaches have been adopted for nitrite substitution: using plant-based ingredients that naturally contain nitrites, or using natural extracts of plant origin where the chemical function of nitrites is substituted. Vegetables like celery, spinach, lettuce, parsley, broccoli, and cabbage, which naturally contain significant amounts of NO₃, have been proposed as nitrate-rich sources for meat products. These nitrate-rich sources have been used to create meat products with similar characteristics to conventionally cured products [4]. The second approach involves studying alternative natural antimicrobial/antioxidant preservatives for meat product preservation. Side streams from fruit, vegetable, and aromatic plant processing industries offer a potential reservoir of bioactive components. These by-products have a high concentration of bioactive constituents, which can replace synthetic additives like antioxidants, preservatives, and colorants. Their appeal lies in their status as by-products from industrial processes [11]. Incorporating these by-products back into the food chain would also enhance the endeavor to promote a circular economy.

A variety of plant-derived bioactive compounds have been tested as nitrite substitutes in meat products, and many of these trials are included in the review by Ferysiuk and Wójciak (2020) [12], Karwowska et al. (2022) [13], and Stoica et al. (2022) [4]. Some illustrative cases include grape pomace, a by-product of winemaking, that has been utilized as a colorant and antioxidant in beef sausages at 1% and 2% concentrations in combination with 30 ppm nitrites [14]. Pomegranate peel extracts and green pistachio hulls were also introduced in the formulation of nitrite-free beef sausages compared to sausages that contained 120 ppm nitrite [15]. Extracts from red dragon fruit peel, rich in phytochemicals, were also incorporated at different concentrations in the beef sausages, and the results showed that the peel extract was effective as an antibacterial agent and natural antioxidant [16]. Very recently, Difonzo et al. (2022) [17] used olive leaf extract (OLE) in fermented sausages with a reduced nitrite content. Their results showed that OLE enhanced the antioxidant capacity of sausages, while the antioxidant activity of NO was also enhanced.

Distilled rose petals are a by-product generated during the manufacturing process of distilling rose petal water (Rosa damascena Mill) to produce rose essential oil and rose hydrosol. Rosa damascena Mill (R. damascena), known as the Damask rose, is a rose hybrid derived from Rosa gallica and Rosa moschata. Schiber et al. (2005) [18] analyzed the extract of the petals of Rosa damascena Mill and detected flavanol glycosides, particularly kaempferol and quercetin glycosides, which both exhibit antimicrobial and antioxidant activity [19]. Recently, both of these compounds have been recognized as GRAS by the FDA [20]. In the case of rosemary leaf distillation, the most common technique used is steam distillation. Rosemary steam distillation leads to the production of rosemary essential oil with two side streams, which are rosemary hydrosol and distilled rosemary leaves. The biomass of distilled rosemary leaves is free of volatile compounds while being rich in phenolic compounds [21,22,23], which are kept almost intact after the steam distillation process. Phenolic diterpenes, namely carnosol and carnosic acid, are considered to be the most important phenolic compounds found in distilled rosemary leaves [21] and they possess a high antioxidant activity [24]. The use of rosemary phenolic diterpenes in several food categories (including meat products) and the maximum level of this use is regulated by Regulations (EC) no. 1333/2008 [8] and no. 1129/2011 [25], and the published amendment of 23 December 2020 (02008R133–EN–23.12.2020–046.001 [8]).

The application of plant extracts in meat products is a challenge due to the considerable diversity observed in both their production technologies and their composition. Consequently, it is necessary to investigate the effectiveness of each plant extract in various meat products on a case-by-case basis.

The aim of this study is to investigate the effect of partially replacing nitrites with (i) distilled rose petal extract of the species Rosa damascena, and (ii) a combination of distilled rose petal extract and distilled rosemary leaf extract, on the aerobic microbiota, the population of lactic acid bacteria, and the color and lipid oxidation of bacon slices, made and vacuum-packaged in industrial conditions. This research thus explores strategies to retain the meat product’s quality, focusing not only on reducing the use of synthetic additives, but also on boosting nutritional value through the incorporation of bioactive compounds.

2. Materials and Methods

2.1. Enrichment of Meat Products with Extracts

Two natural extracts, recovered from rosemary leaf and rose petal post-distillation side streams, were used to enrich cured pork meat (bacon). Preparation of the bacon samples was performed in the production facilities of a collaborating meat producer in Greece, following the conventional industrial preparation procedure. Rosemary leaf extract (RSME) was recovered via acetone extraction of post-steam distillation rosemary biomass, in which carnosol and carnosic acid (phenolic diterpenes) were identified through high-performance liquid chromatography analysis with a diode-array detector (HPLC-DAD) as the key bioactive compounds. Rose petal extract (RSE) was recovered via ultrasound assisted extraction of post-water distillation rose biomass, using a 71% v/v ethanol–water solution as an extraction solvent, for 25 min, with 40 mL/g dry sample, and at 53% ultrasound power, as described in detail by Tsiaka et al. (2023) [26]. The total phenolic content in the two extracts was 225 mg/g and 400 mg/g, for RSE and RSME, respectively. In preliminary (unpublished) tests, RSE was found to have antimicrobial activity; whereas, on the other hand, RSME did not show such an activity. Regarding the antioxidant activity, RSME showed higher levels of antioxidant activity against lipid oxidation when compared to RSE. Thus, the groups of bacon samples prepared in this study were selected to study the antimicrobial effect of RSE with or without the use of RSME that could synergistically provide oxidation stability, as well.

Four groups of bacon samples were prepared as described in

Table 1. Description of the bacon group samples.

Sample Code	Nitrites	Rose Phenolic Compounds	Rosemary Phenolic Diterpenes (Carnosol and Carnosic Acid)
CNT: Control samples	150 ppm	-	-
RN: Reduced-nitrite samples	75 ppm	-	-
RSE: Reduced-nitrite samples enriched with rose extract	75 ppm	90 ppm	-
RSE–RSME: Reduced-nitrite samples enriched with rose and rosemary extract	75 ppm	45 ppm	45 ppm

. The extracts were used to enrich the infusion brine (sodium nitrite solution) injected in the raw meat, followed by the rest of the wet curing process steps. The use of nitrites was reduced by 50%, compared to the traditionally produced commercial bacon samples, in the samples enriched either with the RSE or with a mixture of RSE and RSME. In the case of the RSE-enriched bacon samples, the samples contained 90 ppm rose phenolic compounds (90 mg phenolic compounds/kg raw meat), while in the case of the combined RSE and RSME-enriched samples, the samples contained 45 ppm rose phenolic compounds (45 mg phenolic compounds/kg raw meat) and 45 ppm rosemary phenolic diterpenes (45 mg of carnosol and carnosic acid/kg raw meat), corresponding to a 150 mg sum of carnosol and carnosic acid on the basis of the fat content, according to EU regulation (EC Regulations no. 1333/2008 and no. 1129/2011, and the published amendment of 23 December 2020).

2.2. Shelf-Life Study

2.2.1. Experimental Design

Vacuum-packed bacon samples (100–110 g/pack) from all four group categories described in Table 1 were transferred under refrigerated conditions from the production facility to the laboratory. Samples were split and stored at three controlled isothermal conditions (5, 10, and 15 °C) in high-precision low-temperature incubators (POL-EKO SP.K. ST 1B SMART, Warsaw, Poland). At predetermined time intervals, sampling was performed to determine the quality/spoilage indices described in the following sections.

2.2.2. Microbiological Analysis

For microbiological enumeration, a representative sample of 10 g was transferred to a sterile stomacher bag with 90 mL sterilized Ringer solution (Merck, Darmstadt, Germany) and was homogenized for 60 s with a Stomacher (BagMixer ^® Interscience, Paris, France). Samples (0.1 mL) of 10-fold serial dilutions of meat homogenates were spread on the surface of the appropriate media plate count agar (PCA) (Merck, Darmstadt, Germany) in Petri dishes for enumeration of the aerobic plate count (APC). APC was enumerated after incubation at 30 °C for 72 h, whereas for lactic acid bacteria (LAB) enumeration, the pour-plate method was used, and lactic acid bacteria (LAB) were enumerated on de Man–Rogosa–Sharpe agar (MRS) (Merck, Darmstadt, Germany), followed by incubation at 30 °C for 120 h, based on the principles of ISO 6887 (‘Preparation of test samples, initial suspension and decimal dilutions for microbiological examination’) and according to the instructions of the medium manufacturer (MilliporeSigma, Merck Group, Burlington, MA. USA, www.sigmaaldrich.com/deepweb/assets/sigmaaldrich/product/documents/368/486/69964dat.pdf, (accessed on 1 December 2023)). Two replicates of at least three appropriate dilutions were enumerated, while the pH value was measured in the meat homogenates.

2.2.3. Evaluation of Lipid Oxidation

Determination of the thiobarbituric acid reactive substances (TBARs) was performed spectrophotometrically using the method described by Mendes et al. (2009) [27]. A standard curve was prepared by using concentrations that ranged from 2 to 10 μM of 1,1,3,3-tetraethoxypropane (TEP) in a 7.5% trichloroacetic acid (TCA) solution, with TEP serving as the malonaldehyde (MDA) standard. The TBAR values were expressed as mg MDA per kg of bacon. The absorbance was measured at 530 nm using a digital UV-Vis spectrophotometer (HITACHI UV-VIS U-2900, Tokyo, Japan). All chemicals and reagents were purchased from Thermo Fisher (Kandel, Germany).

2.2.4. Evaluation of the Color Change/Sensory Attributes of the Raw Meat Samples

The color of the bacon samples was determined using a 3nh NR110 precision colorimeter (3hn, Zengcheng District, Guangzhou, China). The measurements were expressed in the CIELab* color scale, where L* represents darkness–lightness (on a scale from 0 to 100), a* represents green–red color (negative values indicate a green color and positive values indicate a red color) and b* represents blue–yellow color (negative values indicate a blue color and positive values indicate a yellow color).

Sensory evaluation was conducted by a panel of nine laboratory members trained to assess the appearance and odor of the bacon samples. Appearance was evaluated while the vacuum packaging remained sealed, while odor was assessed as soon as the packaging was opened. The scale for rating this ranged from 1 to 9, with a score of 5 being the minimum threshold for acceptability.

2.3. Microbial Growth Modelling

Mathematical models that describe microbial behavior (growth, inactivation, survival) are generally divided into two types: (i) primary and (ii) secondary [28]. Primary models are developed and used to describe microbial behavior as a function of time and estimate microbial kinetic parameters. Over the years, several primary models (e.g., logistic, Gompertz, Richard equation) have been developed and used to predict and describe sigmoid microbial growth curves [29,30].

The proper fitting of these experimental and mathematical models, used in addition to the estimated kinetic parameters, is critical for selecting the primary model. The model proposed by Baranyi and Roberts (1994) [31], as given by Equations (1) and (2), is one of the most-used models due to its ability for proper kinetic parameter determination even from a limited amount of data; and, as will be described in the following paragraphs, it gives a good fitting of the experimental data of microbial growth for all different bacon categories investigated.

$$y\left(\mathrm{t}\right){=\mathrm{y}}_{\mathrm{o}}{+\mathsf{\mu }}_{\max }·\mathrm{A}\left(\mathrm{t}\right) - \ln \left[1+\frac{\exp \left({\mathsf{\mu }}_{\max }·\mathrm{A}\left(\mathrm{t}\right)\right)-1}{\exp \left({\mathrm{y}}_{\max }{-\mathrm{y}}_{\mathrm{o}}\right)}\right]$$

(1)

and

$$\mathrm{A}\left(\mathrm{t}\right)=\mathrm{t}+\frac{1}{{\mathsf{\mu }}_{\max }}·\ln \left[\exp \left({-\mathsf{\mu }}_{\max }·\mathrm{t}\right)+\exp \left({-\mathsf{\mu }}_{\max }·\mathsf{\lambda }\right)-\exp \left({-\mathsf{\mu }}_{\max }·\left(\mathrm{t}+\mathsf{\lambda }\right)\right)\right]$$

(2)

where y_o (logCFU/g) describes the decimal logarithm of initial microbial population for t = 0; y_max (logCFU/g) is the decimal logarithm of maximum microbial population; μ_max (logCFU/g/((logCFU/g)d⁻¹)) is the maximum specific growth rate; (logCFU/g/((logCFU/g)d⁻¹)), t (d) is time; A(t) is an adjustment function that ranges from 0 to t and is related to the physiological state of microbial density; and λ(d) is the lag phase.

To determine the effects of environmental conditions, such as the temperature, secondary models are used. The most-known and frequently used secondary models are Arrhenius-type models and square root-type models. In this research, to study the effect of temperature on the microbial growth of different bacon samples, both of these secondary models were implemented, comparing preliminary results of the sum of square errors (SSE) between the two secondary models. The Arrhenius model presented the least SSE and was used as the secondary model for this analysis. The temperature dependence of the maximum specific growth rate and lag phase were modelled by the Arrhenius equation, shown in Equations (3) and (4), respectively, and the evaluation of SSE was conducted using Equation (5) [32]. Data analysis and mathematical modeling was performed in Microsoft, Excel 365.

$${\mathsf{\mu }}_{\max }{\left(\mathrm{T}\right)=\mathsf{\mu }}_{\max ,\mathrm{ref}}·\exp \left(\frac{{\mathrm{Ea},}_{\mathsf{\mu }}}{\mathrm{R}}·\left(\frac{1}{\mathrm{T}}-\frac{1}{{\mathsf{{\rm T}}}_{\mathrm{ref},\mathsf{\mu }}}\right)\right)$$

(3)

where μ_max,ref (d⁻¹) describes the pre-exponential factor for maximum specific growth rate; E_a,μ (J/mole) is the activation energy for maximum specific growth rate of the investigated microorganism; (J/mole), R (8.314 J/mole.K) is the universal gas constant; T (K) is the absolute temperature; and T_ref (K) is the reference temperature, here assumed to be equal to 4 °C.

$${\mathsf{\lambda }\left(\mathrm{T}\right)=\mathsf{\lambda }}_{\mathrm{ref}}·\exp \left(-\frac{\mathrm{Ea},\mathsf{\lambda }}{\mathrm{R}}·\left(\frac{1}{\mathrm{T}}-\frac{1}{{\mathrm{T}}_{\mathrm{ref},\mathsf{\lambda }}}\right)\right)$$

(4)

where λ_ref (d) is the pre-exponential factor for the reference temperature, and E_a,λ (J/mole) is the activation energy to express the temperature dependency of the lag phase of the investigated microorganism.

$$\mathrm{SSE}={\sum }_{\mathrm{i}=0}^{\mathrm{n}}{\left({\mathsf{\mu }}_{\mathrm{obs}}{ - \mathsf{\mu }}_{\mathrm{pred}}\right)}^2$$

(5)

where μ_obs (d⁻¹) is the maximum specific growth rate of the experimental measurements; μ_pred (d⁻¹) is the maximum specific growth rate calculated in the investigated model; n (-) is the number of observed data; and i = 0, 1, 2, 3…, n.

3. Results and Discussion

3.1. Shelf-Life Study

3.1.1. Microbial Growth

In this section, the experimental results of the examined bacon samples and mathematical modeling of isothermal microbial growth (lag, exponential, stationary phases) of two different examined microbial groups, (i) aerobic plate count (APC) and (ii) lactic acid bacteria (LAB), are presented.

The number of microorganisms (pathogens and/or spoilage microorganisms) present in foods influences the safety and quality of the food product. Predicting and modeling the number of microorganisms and the rate of microbial growth under specific storage conditions can help to understand and predict food safety risk assessments [33]. Accurately estimating the number of microorganisms present during the “lifetime” of food before consumption requires a quantitative knowledge of the microbial responses to environmental conditions (temperature, nutrients, chemicals, preservatives, other microorganisms), synthesized in a mathematical model, which enables objective evaluation of the effects of processing, distribution, and storage conditions on the safety and quality of food [34]. A mathematical model that quantitatively describes the combined effect of these environmental parameters can be used to predict the growth, survival, or inactivation of a microorganism, thereby providing important information on the product’s safety and shelf-life [29,30].

The growth data for aerobic plate count (APC) and lactic acid bacteria (LAB) in the four different categories of bacon samples examined were fitted using a two-step modeling approach and statistical analysis based on SSE results between the experimental data and the model-predicted data.

For the case of the APC of the bacon samples, the microbial growth data observed from the bacon samples without any added antimicrobial factors (reduced nitrite (RN) samples) presented the highest rates of microbial growth at all three different temperatures tested, with values of μ_max of 0.283 d⁻¹ and 0.666 d⁻¹ at 5 °C and 15 °C, respectively. On the other hand, the antimicrobial effect of the rose extract appeared to be more effective for APC growth, with the RSE samples presenting the lowest APC microbial growth rate, with values of 0.163 d⁻¹ and 0.599 d⁻¹ at 5 °C and 15 °C, respectively (Figure 1). The combined incorporation of rose and rosemary extracts in the RSE–RSME samples showed a better antimicrobial performance when compared to the RN samples, but did not show any further beneficial effect to microbial inhibition when comparing these results to the traditionally produced counterparts. To be more specific, conventionally produced samples (CNT control samples), containing the maximum nitrite content allowed by legislation, appeared to have an effective antimicrobial impact (as anticipated) similar to that of the RSE samples at higher temperatures (10 and 15 °C). Nonetheless, it is worth noting that at the lower temperature of 5 °C (representative of the chilled conditions of the cold storage of this product), the RSE samples, with half the nitrite content, seemed to inhibit microbial growth more effectively when compared to their CNT counterparts.

Figure 1 presents the fitting between the experimental data and the mathematical model curves, showing that Baranyi and Roberts’s 1994 model [31] is a well-fitted primary model for APC microbial growth in bacon samples, something that is confirmed by the estimated SSE values, with the highest SSE value observed at 0.556 for all of the examined APC microbial data. Table 1 summarizes this model’s statistical analysis and provides the parameters of Baranyi and Roberts’s model (1994) [31].

As mentioned in the section above, for the secondary model, examining the effect of temperature on microbial growth, the Arrhenius model was selected based on a comparison of the SSE with other secondary models. As expected, at the lowest temperature of 5 °C (~refrigerated conditions), the lowest APC microbial growth rates and the maximum lag phase (λ) values for all of the examined bacon samples were observed. At the highest temperature tested (15 °C), no significant difference was observed between the samples of different antimicrobial compositions (RN vs. RSE vs. RSE–RSME) in terms of APC growth. The Arrhenius parameters for APC growth were estimated (Figure 2) and summarized in

Table 2. Parameters derived from Arrhenius equations (secondary model) that describe the dependence of the maximum specific growth rate and the lag phase of APC on the temperature in the different bacon samples examined.

Maximum Specific Growth Rate (μmax)
Samples	Ea,μ (J/mole)	μref (d−1), at T = 4 °C	SSE	R2
CNT	86,491.17	0.1426	0.0043	0.9091
RN	70,648.81	0.2024	0.0090	0.8276
RSE	105,976.05	0.1067	0.0033	0.9297
RSE–RSME	76,777.55	0.1673	0.0031	0.9347
Lag phase (d)
Samples	Ea,λ (J/mole)	λref (d), at T = 4 °C	SSE	R2
CNT	−116,783.84	12.7	0.0253	0.9985
RN	−136,132.21	8.3	0.0385	0.9938
RSE	−124,309.74	16.0	0.0076	0.9997
RSE–RSME	−91,577.67	9.4	0.8724	0.9464

. Using these estimated Arrhenius parameters and comparing the maximum APC growth rate calculated from the Arrhenius equation, μ_max.ref at T_ref of 4 °C (an average refrigerator temperature, typical of storage conditions for bacon products), it is confirmed that the antimicrobial effect of replacing half of the nitrite content with rose extract (RSE) appears to be more efficient for APC microbial growth in bacon samples, followed by the traditional recipe containing the full amount of nitrites (CNT) and the rose–rosemary extract bacon samples (RSE–RSME), with the samples containing half of the nitrites (RN) coming in last, with a significant difference.

The results concerning lactic acid bacteria microbial growth data were quite similar to what was observed based on APC data. Moreover, the RN category of samples, having a reduced nitrite content, presented the highest growth rates for all of the temperatures examined, with the value of μmax being equal to 0.213 d⁻¹ and 0.654 d⁻¹ at 5 °C and 15 °C, respectively. On the other hand, the antimicrobial effect of rose extract appeared to be more effective for LAB growth, with the RSE samples presenting the lowest μmax values, equal to 0.163 d⁻¹ and 0.615 d⁻¹ at 5 °C and 15 °C, respectively (Figure 3). For the other two groups tested (CNT/RSE–RSME), no significant differences were observed, with the conventionally produced samples (CNT) being more effective for LAB growth compared to the RSE–RSME samples containing the rose/rosemary extract.

Based on Figure 3, Baranyi and Roberts’s 1994 model [31] is a well-fitted primary model for assessing lactic acid bacteria microbial growth in bacon samples. Regarding the estimated SSE values, the highest SSE value was observed at 0.601 for all of the examined experimental vs. predicted data, which is an acceptable level for model predictions.

For a secondary model, the Arrhenius model was used to quantify the effect of the temperature on LAB growth. As expected, at the lowest temperature, 5 °C, the lowest LAB microbial growth rate and the highest lag phase (λ) was estimated for all of the examined bacon samples. In contrast to the APC data, for the case of LAB at the highest temperatures studied (10 and 15 °C), no significant differences in the growth rate of LAB were observed between the different groups of bacon samples. As shown in Figure 3, at 15 °C, a slight difference was observed between the growth curve and growth rate of the different samples, which is in the range of experimental and statistical error.

Similar to the APC results, the estimated parameters regarding the application of the Arrhenius model to describe the effect of temperature are summarized in

Table 3. Parameters of secondary models derived from Arrhenius equations to describe the temperature dependence of the maximum specific growth rate and the lag phase of lactic acid bacteria in the different bacon samples examined.

Maximum Specific Growth Rate (μmax)
Samples	Ea,μ (J/mole)	μref (d−1), at T = 4 °C	SSE	R2
CNT	103,225.90	0.1213	0.0033	0.9436
RN	86,422.33	0.1578	0.0026	0.9574
RSE	88,792.87	0.1305	0.0004	0.9922
RSE–RSME	91,065.15	0.1366	0.0006	0.9907
Lag phase (λ)
Samples	Ea,λ (J/mole)	λref (d), at T = 4 °C	SSE	R2
CNT	−109,798.65	9.8	0.5782	0.9642
RN	−159,416.30	6.3	0.0394	0.9879
RSE	−117,055.14	10.5	0.7335	0.9589
RSE–RSME	−98,938.67	6.6	0.8305	0.9165

. Once more it is confirmed that in the temperature range of refrigeration (the lowest temperature tested, 5 °C), the antimicrobial effect of rose extract appears to be more efficient on LAB microbial growth inhibition in bacon samples, followed by the traditional recipe containing the full amount of nitrites (CNT) and the rose–rosemary extract bacon samples (RSE–RSME), and last, with a significant difference, are the samples containing half of the nitrites (RN).

Modeling the kinetics of growth of both APC and LAB, it is interesting to mention that the microbial growth rate appears to be very similar for both examined microorganisms, while at the increased (abusive) temperatures tested, the antimicrobials used appear to have no significant impact, an observation that emphasizes the importance of maintaining proper temperature conditions during storage and handling in the chill chain. Furthermore, LAB were identified as the dominant microorganisms of the examined meat products. The presence of the vacuum packaging and the low storage temperature favor the growth of psychrotrophic LAB [35]. LAB naturally constitute the predominant microflora of raw meats and other meat products that are vacuum packed and chill stored, causing their spoilage. Finally, it is essential to mention the potential antimicrobial resistance of rose extract, with the RSE samples presenting the lowest growth rate for all the microorganisms tested under different temperature conditions. The antibacterial properties of rose petal extract can be ascribed to the presence of flavonoids such as naringenin, quercetin, and kaempferol [19,36]. These phenolic compounds were identified as the primary constituents by Tsiaka et al. (2023) [26], who conducted an analysis of the extract utilized in the manufacturing of bacon. Extensive reviews on the antimicrobial mechanism of action of flavonoids is presented by Górniak et al. [37], as well as by Gyawali and Ibrahim, 2014 [38]. Flavonoids operate as antimicrobials through various mechanisms. Quercetin induces cell death through the disruption of microbial cell membranes, which results in increased permeability and the subsequent leakage of vital cellular contents. Additionally, it may hinder the production of proteins and nucleic acids (specifically DNA and RNA). In relation to naringenin, it alters the electrochemical membrane potential by decreasing the fluidity of hydrophilic and hydrophobic regions of the cell membrane. Regarding kaempferol, it exerts antibacterial effects through cell membrane disruption and the inactivation of critical enzymes (gyrase, helicase) for the synthesis of proteins and DNA [39].

3.1.2. Shelf-Life Estimation (t_SL)

The shelf-life estimation was performed using Equation (6) [40], and assuming a limit of 8 logCFU/g for the APC count [35,41]:

$${\mathrm{t}}_{\mathrm{SL}}=\left(\frac{{\mathrm{logN}}_1-{\mathrm{logN}}_0}{{\mathrm{k}}_{\mathsf{\mu }\max ,\mathrm{ref}}\exp \left[\left(-\frac{{\mathrm{E}}_{\mathrm{a},{\mathsf{\mu }}_{\max }}}{\mathrm{R}}\right)·\left[\frac{1}{\mathrm{T}}-\frac{1}{{\mathrm{T}}_{\mathrm{ref}}}\right]\right]}\right)+{\mathsf{\lambda }}_{\mathrm{ref}}\exp \left[\left(-\frac{{\mathrm{E}}_{\mathrm{a},\mathsf{\lambda }}}{\mathrm{R}}\right)·\left[\frac{1}{\mathrm{T}}-\frac{1}{{\mathrm{T}}_{\mathrm{ref}}}\right]\right]$$

(6)

where t_SL is the estimated shelf-life in days; logN₁ is the limit of the APC load (8 log CFU/g); logN₀ is the initial value of APC; µ_max,ref is the maximum growth rate constant of the microorganism at a reference temperature, T_ref = 4 °C; T is the temperature in K; E_a,µ and E_a,λ are the activation energy of the studied action for the maximum growth rate and lag phase, respectively, indicating the temperature-dependence; R is the universal gas constant (R = 8.314 J/(mol·K)); and λ_ref is the lag phase at the reference temperature. Based on Equation (6) and the kinetic parameters in Table 2, the predicted shelf-life of bacon at a representative (4 °C) refrigeration temperature is calculated to be 60 d (CNT), 42 d (RN), 78 d (RSE) and 51 d (RSE–RSME), respectively, demonstrating the beneficial effects of the rose extract.

3.1.3. Chemical Indices—Lipid Oxidation

Thiobarbituric acid reactive substance (TBAR) values remained stable during the 49 days of storage time at refrigeration (5 °C), and all measurements showed a level below 1.0 mg of malonaldehyde (MDA)/kg of the sample, a behavior attributed to the anaerobic conditions in the vacuum-sealed packaging of the bacon slices. Moreover, no significant differences were observed among the samples, whether enriched or not with the extracts (RSE and RSME samples). It has been demonstrated that pro-oxidants have little effect on oxidation during storage when oxygen is absent [42]. The average initial TBAR values on day 0 of storage were 0.49 ± 0.02, 0.50 ± 0.02, 0.47 ± 0.01 and 0.51 ± 0.02 mg MDA/kg of bacon for the CNT, RN, RSE and RSE–RSME samples, respectively, and after 49 d of storage, the average values were 0.51 ± 0.02, 0.53 ± 0.03, 0.51 ± 0.02 and 0.53 ± 0.02 mg MDA/kg of the sample, respectively, as shown in Figure 4. Similar results were obtained at the other two storage temperatures (10 and 15 °C), with low TBAR values and insignificant differences among the samples. Similar results were reported by other researchers, where TBAR values did not exceed 1.0 mg of malonaldehyde/kg of the sample until day 126 of storage [43]. If the storage period was extended, an increase in TBA values would be expected, as reported by previous studies on vacuum-packaged bacon [44]; however, the microbiological spoilage of the samples was much more rapid in comparison to autooxidation, rendering the samples rejectable before their deterioration due to oxidation.

3.1.4. Other Indices—Color/pH/Sensory Attributes

Vacuum-packaged bacon slices, during their storage under chilled conditions (5, 10 and 15 °C), showed a slight decrease in pH values. The differences observed in pH values between the different types of samples (CNT, RN, RSE and RSE–RSME) were only marginal, and thus are considered as insignificant. At 5 °C, the average initial values for the CNT, RN, RSE and RSE–RSME samples were 5.93 ± 0.06, 5.61 ± 0.05, 5.51 ± 0.06 and 5.64 ± 0.08, respectively, while after 49 days of storage, the pH values decreased to 5.26 ± 0.12, 5.12 ± 0.04, 5.08 ± 0.15, and 5.01 ± 0.08, respectively. This reduction in pH values is attributed to the production of lactic acid due to the growth of lactic acid bacteria (LAB) during the storage of the samples under anaerobic conditions. Similar results were also obtained at the other storage temperatures (10 and 15 °C).

The color changes of bacon samples during chilled storage (at 5 °C) are presented in Figure 5. The L* value, representing color brightness or lightness on a scale from 0 to 100 (where 0 stands for white and 100 for black), remained consistent for both treated and untreated samples throughout the 49-day storage period at 5 °C, showing no significant changes from day 0. The control samples seem to be the darkest samples and the RSE–RSME samples seem to be the brightest ones. The control samples (CNT) had an average L* value of 63.19 ± 0.76 during the storage period and the RN, RSE and RSE–RSME samples had values of 67.22 ± 0.63, 65.21 ± 0.44 and 68.31 ± 0.63, respectively. The a* axis represents the position of the color on the green to red spectrum, with positive a* values indicating red color, while negative values indicate green color. The a* values for all samples also remained stable during storage, with the RN and RSE–RSME samples displaying lower values and the RSE samples showing the highest values displaying more red tones. The average a* values of the samples during storage, depicted in Figure 5a (where a* values are presented cumulatively for all storage times, as distributions, for the different sample categories), were 8.85 ± 0.12, 8.10 ± 0.16, 9.72 ± 0.27 and 7.85 ± 0.34 for the CNT, RN, RSE and RSE–RSME samples, respectively. Hence, the treated samples using only rose by-products seem to present an enhanced red color, while the treated samples with rose and rosemary extract managed to maintain the red color of meat in comparison to the control (CNT) samples, both with half the amount of nitrites. It is worth noting that increased values of the color parameter a* have also been observed in bacon enriched with tea, apple, and cinnamon polyphenols [45]. Polyphenols are known antioxidants, meaning they have reducing properties. Their presence, as a reducing agent, favors the conversion of nitrites into nitric oxide, which reacts with the main protein of the meat and forms nitrosomyoglobin, which has a bright red color that is stabilized after the mild heat treatment/smoking of the bacon. The b* axis represents the position of the color on the blue to yellow spectrum, with positive b* values indicating yellowness, while negative values indicate blue tones. Variations in b* values were observed between the samples. The control samples had the lowest b* values, the RSE and RSE–RSME samples had intermediate values, while the RN samples showed the highest values. The average b* values of the samples during storage were 2.57 ± 0.27, 5.37 ± 0.27, 4.32 ± 0.22 and 4.64 ± 0.31 for the CNT, RN, RSE and RSE–RSME samples, respectively. This ‘clustering’ of samples, regarding their yellowish tone, is also indicated in Figure 5b. Overall, the color results indicate that while the treated samples exhibit similar lightness and redness to the control samples, despite containing half the amount of nitrites, they are slightly more yellowish than the control (CNT) samples. Nevertheless, the RSE samples seem to simulate control samples’ color in a satisfactory way. Similar results were also obtained at the other storage temperatures (10 and 15 °C).

The sensory characteristics (appearance and odor) of the control and extract-enriched bacon samples were evaluated, and the representative results are presented in Figure 6. A slight decrease in the sensory scores of appearance was observed in all samples during the storage time, due to the slight yellowish tone (Figure 5b and Figure 7). However, all samples remained acceptable after 49 days of storage under chilled conditions in vacuum-sealed packaging, as none of them received a score below 5, which was considered the minimum threshold for acceptability. The final scores for appearance at 49 days of storage were 7.5 ± 0.2, 6.4 ± 0.4, 7.0 ± 0.4 and 6.2 ± 0.3 for the CNT, RN, RSE and RSE–RSME samples, respectively, indicating that the control samples (CNT) received the highest score, followed closely by the RSE samples, while the RSE–RSME samples received the lowest score. The sensory scores for odor also decreased with storage time in all samples, with the control samples retaining their odor for longer, followed closely by the RSE samples. Over time, the smell without being unpleasant was more intense compared to the samples measured during the first days of chilled storage. This should be attributed to the proteolysis taking place [46], as well as to the volatile metabolic products of the bacteria, which, due to the low permeability of the packaging material, do not escape into the atmosphere and remain inside the food. As described in [47], using metabolomic studies, the action of endogenous enzymes (e.g., proteases) and spoilage microorganisms during the chilled storage of bacon leads to the decomposition of proteins and fats, gradually forming aldehydes, ketones, alcohols, esters, acids, and other off-flavor compounds. More specifically, the CNT samples remained acceptable in terms of their sensory attributes until the end of the 49-day storage period, reaching a minimum score of 5.0 ± 0.2. The RSE samples reached a minimum score of 4.3 ± 0.4 at the end of the storage period and remained sensorily acceptable until day 41 of storage. On the other hand, the RSE–RSME and RN samples’ odor deteriorated, reaching a final score of 2.2 ± 0.2 and 2.0 ± 0.4, respectively, surpassing the limit score of 5 by day 29 of storage. Overall, the bacon samples enriched with rose extract (RSE) retained their sensory attributes as effectively as the control samples. However, the RSE–RSME samples, enriched with both rose and rosemary extract, maintained their sensory acceptability for a shorter time, exhibiting a behavior similar to samples containing half the nitrite amount (RN). Similar results were also observed at the other storage temperatures (10 and 15 °C), with the sensory characteristics of the samples exceeding their limits in a shorter period of time due to more rapid microbial spoilage.

4. Conclusions

The bacon samples enriched with Rosa damascena distillation by-product extract (RSE) presented the lowest microbial growth rate for all microorganisms examined, followed by the samples representing the conventional manufacture procedure (CNT, containing the maximum nitrite content permitted). On the other hand, rose–rosemary extract did not show any particular antimicrobial effect in this study, as would be anticipated based on the available literature. At the lowest temperature, 5 °C (~refrigerated conditions), the lowest microbial growth rates and the maximum lag phase (λ) were observed for all of the examined bacon samples, as expected. An important observation in our study is that at that at this temperature, simulating the proper refrigerated conditions, the microbial stability was significantly improved when the rose extract replaced half of the nitrites.

On the other hand, at the highest tested temperatures (10, and 15 °C), natural antimicrobials appear to have only a mild positive effect on microbial growth.

The color of the bacon samples, a crucial factor influencing consumer acceptability and a primary reason for the amount of nitrite additives used, appeared to be influenced by the reduced nitrite content in the RN and RSE–RSME samples. On the contrary, the RSE samples exhibited a color similar to the CNT samples, despite containing only half the amount of nitrites. This suggests that the rose extract used did not deteriorate color stability, offering a viable option for the reduction in the content of nitrites. Similar observations can be made for the overall sensory characteristics of the samples, with the RSE samples showing similar acceptability to the CNT samples. Overall, the results of this study provide preliminary proof on the potential use of rose extract (derived out of side streams of rose steam distillation) as an effective natural antimicrobial agent, that could potentially reduce or partially substitute nitrites in cured meat products. Given the current tendency of law and control authorities to reduce the nitrite/nitrate limit in food applications, it is important to investigate other sustainable sources of antimicrobial properties. As such, further studies on rose extract use as a natural preservative are essential, aiming mostly at optimizing the concentration applied in the infusion brine of bacon samples. This optimization should result in a novel meat product, with the minimum nitrite content possible (without jeopardizing product safety), the maximum delay in spoilage microorganisms, while obtaining the maximum retention of bacon’s sensory attributes.