Antimicrobial Effect of Malpighia Punicifolia and Extension of Water Buffalo Steak Shelf-Life.

In the present study, a multiple approach was used to characterize Malpighia punicifolia extract and to evaluate its inhibitory activity against several meat spoilage bacteria. First, volatile fraction, vitamins and phenolic compounds of the extract obtained by supercritical fluid extraction were determined by GC-MS and HPLC. Then, the antimicrobial action of the extract was in vitro evaluated against Pseudomonas putida DSMZ 291(T), Pseudomonas fluorescens DSMZ 50009(T), Pseudomonas fragi DSMZ 3456(T), and Brochothrix thermosphacta DSMZ 20171(T) by the agar well diffusion assay and by the agar dilution test. Based on the results of the minimum inhibitory concentration (MIC) against the assayed bacteria, 4 different concentrations of the extract were used in a challenge test on water buffalo steaks stored for 21 d at 4 °C. Results of chemical analyses showed that M. punicifolia extract is characterized by the presence of several compounds, already described for their antimicrobial (phenolic acids, flavonones, and furanes) and antioxidant (ascorbic acid) properties. The in vitro detection of antimicrobial activities highlighted that the extract, used at 8% concentration, was able to inhibit all the target bacteria. Moreover, very low MIC values (up to 0.025%) were detected. In situ tests, performed on water buffalo steaks treated with the extract in the concentration range 0.025% to 0.05%, showed a strong inhibition of both intentionally inoculated bacteria and naturally occurring microorganisms. Positive results, in terms of color and odor, were also observed during the entire storage of steaks preserved with the extract.

Introduction

Species belonging to Pseudomonas spp. and Brochothrix thermosphacta are considered the main contaminant bacteria of fresh meats stored under refrigerated temperature (Tremonte and others 2005; Nychas and others 2008; Nowak and others 2012b). Several studies (Nychas and others 2008; Nowak and others 2012a,b) described B. thermosphacta as an important spoiler of meat stored under both reduced and high oxygen atmospheres. On the other hand, Pseudomonas species are defined as psychrotrophic bacteria having a very fast growth rate and high affinity for the oxygen, thus able to dominate the microbiota of refrigerated meats stored aerobically or under modified atmosphere packaging (Ercolini and others 2006; Argyri and others 2011; Doulgeraki and Nychas 2013). Among Pseudomonas species, Pseudomonas fluorescens, Pseudomonas putida, and Pseudomonas fragi are capable of producing extracellular proteases and lipases, generating off‐odors and a partial or a complete meat degradation (Liao 2006).

In the last years, different bio‐protective strategies were developed to inhibit the growth of spoilage bacteria on both pork and bovine fresh meats (Tremonte and others 2005; Soultos and others 2008; Turgis and others 2008). Many studies were addressed to the search for natural alternatives to chemical additives in foods (Iturriaga and others 2012; Kim and others 2013; Casaburi and others 2014). In this field, different scientific reports highlighted the antimicrobial effect of natural compounds on pathogenic or spoilage bacteria (Holley and Patel 2005; Tipaldi and others 2011; Settanni and others 2012). Among natural substances, extracts obtained by supercritical fluid extraction (SFE) showed promising results in the control of food spoilage and foodborne pathogenic bacteria (Muñoz and others 2009; Herrero and others 2010).

Besides antimicrobial activities, natural extracts were also evaluated for their anti‐oxidant effects on stored meat. Interesting results were obtained by Naveena and others (2006), which showed the effect of clove oil plus lactic acid and vitamin C on the improvement of color in water buffalo meat. This topic is particularly interesting, taking into account that the fast meat darkening due to the formation of met‐myoglobin (Bekhit and others 2003) negatively affects the shelf‐life of meat, as well as the consumer acceptability.

The extract of Malpighia punicifolia (Acerola), among other natural compounds, shows promising results, thanks to a series of active components, which could positively affect the overall quality of fresh meat (Schreckilger and others 2010). This extract, having a long history of use in folk medicine, was screened in vitro for its antibacterial activity (Schelz and others 2010). However, the literature is poor in scientific reports on the effect of the extract in the preservation of fresh meat. In the light of previous findings, this study used a multiple approach to characterize a SFE M. punicifolia extract and to evaluate its inhibitory activity against several meat spoilage microorganisms. The effect was also tested in situ on water buffalo steaks, characterized by high perishability, during the storage at 4 °C for 21 d.

Materials and Methods

Natural extract

A fruit extract of M. punicifolia (Bioma Life Science, Quartino, Switzerland) was tested in this study. As reported by the supplier, the extract was obtained by SFE at 42.2 °C and 70 atm. In short, the extract was obtained by treatment of M. punicifolia dried powder (300 g) with liquid CO2 (1 L). The dried extract was dissolved in hydroalcoholic solution (ethanol:water 30:70, v/v) to obtain a 8% (w/v) stock solution, and it was stored at −20 °C until its use.

Chemicals

Folin–Ciocalteu phenol reagent and sodium carbonate (Na2CO3) were acquired from Merck (Darmstadt, Germany). Trifluoroacetic acid (TFA) was purchased from Scharlau Chemie (Barcelona, Spain). All vitamins, fat‐soluble (α‐tocopherol, retinol acetate, and cholecalciferol) and water‐soluble vitamins (ascorbic acid, thiamine, folic acid, pyridoxine, nicotinamide, and cobalamin) were purchased from Accustandard (New Haven, Conn., U.S.A.). All standard phenolic compounds (caffeic acid, coumaric acid, catechin, catechol, cyanidine chloride, gallic acid, genistein, hesperidin, and quercetin) were purchased from Extrasynthese (Genay, France) and chlorophylls and zeaxanthin from DHI (Hørsholm, Denmark). Milli‐Q water was obtained using a Millipore purification system (Millipore, Bedford, Mass., U.S.A.). Methanol [high‐performance liquid chromatography (HPLC)‐grade] was purchased from LabScan (Dublin, Ireland).

Standard solutions (3 mg/mL) were prepared using solutions of 0.01% TFA in water (polar vitamins, caffeic acid, and gallic acid), methanol:water (the rest of phenolic compounds) or methanol (pigments and fat‐soluble vitamins) and stored in darkness under freezing conditions (−20 °C) until use.

Total phenolic content of the extract

Total phenols were estimated as gallic acid equivalents, expressed as mg gallic acid/g extract (Kosar and other 2005). The total volume of reaction mixture was miniaturized to 2.0 mL. In detail, 1.2 mL water and 20 μL of sample were mixed, and 100 μL undiluted Folin–Ciocalteau reagent was subsequently added. After 1 min, 0.3 mL of 2% (w/v) Na2CO3 were added and the volume was made up to 2.0 mL with water. After 2 h of incubation at 25 °C, the absorbance was measured at 760 nm and compared to a gallic acid calibration curve elaborated in the same manner. Data were presented as the average of triplicate analyses.

Volatile fraction by gas chromatography‐mass spectrometry (GC‐MS)

The extract was analyzed using an Agilent‐6890N GC system with a programmed split/splitless injector coupled to an Agilent‐5973N quadruple mass spectrometer (Agilent, Palo Alto, Calif., U.S.A.). The system was controlled by means of an Agilent MSD Chemstation software. The column used in the GC was a 30 min × 0.25 mm i.d. fused silica capillary column coated with a 0.25 μm layer of SE‐54 (HP‐5MS, Agilent). The injection was carried out at 250 °C in split mode (ratio 1:20). The volume of sample injected was 1 μL. Extract was injected at a concentration of 5 mg/mL. Helium was the carrier gas (7 psi). The oven temperature was programmed as follows: 40 °C as initial temperature (maintained for 2 min) to 150 °C in 24 min, 5 °C/min, and from 150 °C to a final temperature of 300 °C at 15 °C/min. A solvent delay of 4 min was selected before analyzing the compounds reaching the MS. Compounds were tentatively identified by MS in SCAN mode using a mass interval ranging from 35 to 450 m/z. Their spectra were compared with those in a MS library (Wiley Registry of Mass Spectral Data), with data found in the literature and with standards when available.

Determination of vitamins and phenolic compounds

A modified HPLC method, previously developed in CIAL laboratory (Madrid, Spain) was used (Mendiola and others 2008). Shortly, analytes were separated on an ACE‐100 Å C18 (150 mm × 4.6 mm, 3 μm particle size; Advanced Chromatographic Technologies, Aberdeen, U.K.) in a single run, using combined isocratic and linear gradient elution with a mobile phase consisting of 0.010% TFA (solvent A, pH 3.9) and methanol (solvent B) at the flow rate of 0.7 mL/min using an Agilent 1100 HPLC apparatus. The gradient profile (A:B) started at 95:5 and was constant in the first 4 min, increased up to 0:100 in 2 min, then constant until a total analysis time of 40 min, and finally linearly increased up to 95:5 to reach initial conditions. The most suitable detection wavelength for simultaneous vitamin‐polyphenol determination was 280 nm. Quantification of the different families of compounds was performed by building a calibration curve for each of them, as described by Mendiola and others (2008).

Direct injection of pure vitamin standards was done by dissolving them in TFA 0.01% or 100% methanol depending on the polarity of the vitamin. For polyphenols, all samples were injected dissolved in hydromethanolic solutions.

In vitro detection of M. punicifolia antimicrobial activity

The antimicrobial activity of M. punicifolia extract was evaluated against 4 bacterial type strains, represented by P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and Brochothrix thermosphacta DSMZ 20171T. Strains of Pseudomonas and Brochothrix thermosphacta were revitalized at 28 °C in PAB (Pseudomonas Agar Base, Oxoid, Milan, Italy) and in STA Agar Base (Oxoid), respectively, and then stored in the same media at 4 °C. Prior to experiments, strains were overnight cultured in the proper conditions. The inhibitory action of M. punicifolia was assessed by the agar well diffusion assay (Tremonte and others 2007, 2010). Briefly, 1 mL (inoculum size of 4.0 log CFU/mL) of each bacterial suspension was inoculated into 20 mL of proper soft media (0.7% agar), gently mixed and poured into Petri plates. A well of 4.0 mm in diameter was bored into each agar plate, and 85 μL of the stock hydroalcoholic solution of M. punicifolia extract at 8% (w/v) concentration were placed into each well. Hydroalcoholic solution without extract was used as control. Streptomycin (25 μg) and tetracycline (30 μg; both from Oxoid) were used as reference controls for Pseudomonas spp. and B. thermosphacta, respectively, as well as to compare the inhibitory activity of M. punicifolia extract. After incubation at 28 °C for 24 to 48 h, inhibition halos (IH, mm) were accurately measured with a caliper and the ratio between IH (mm) of the proper reference control and IH (mm) of M. punicifolia extract was calculated. Thereafter, 4 levels of inhibition were established: very strong (ratio ≤1), strong (ratio between 1 and 1.2), moderate (ratio between 1.2 and 1.4), and low (ratio >1.4). The presence of growth around the well was considered as absence of inhibition. Data were reported as the mean value and standard deviation on 3 analyses.

Minimum inhibitory concentration (MIC) evaluation of M. punicifolia extract

The minimum inhibitory concentration (MIC) assay was carried out using the agar dilution method as described in the EUCAST definitive document 3.1 (2000) with some modifications. In detail, the extract of M. punicifolia was filter‐sterilized (Filter Unit Red rim FP 30/0.2 CA‐S, 0.22‐μm pore size; Schleicher & Schuell, Dassel, Germany) and added to sterile molten Mueller‐Hinton (MH) agar at appropriate volume to produce a concentration range between 0.0002% and 0.4% (w/v), as indicated in Table 1. The resulting MH agar solutions were poured into 90 mm Petri plates immediately after vortexing. The surface of plates was inoculated with the microbial strains at a concentration of 4.0 log CFU/mL and incubated at 28 °C for 48 h. At the end of the incubation period, plates were evaluated for the presence or absence of growth. MIC values were recorded as the lowest concentration of natural extract that completely inhibited the growth. Each test was performed in triplicate.

Table 1

Dilutions of M. punicifolia extract used in the agar dilution susceptibility test

				Final	Final
Concentration of	Volume	Distilled water	Concentration	concentration of	concentration of
extract (mg/L)	used (mL)	added (mL)	obtained (mg/L)	extract (mg/L)a	extract (% w/v)a
80000.00b	1.00	0.00	80000.00	4000.00	0.4000
80000.00	1.00	1.00	40000.00	2000.00	0.2000
80000.00	1.00	3.00	20000.00	1000.00	0.1000
20000.00	1.00	1.00	10000.00	500.00	0.0500
20000.00	1.00	3.00	5000.00	250.00	0.0250
20000.00	1.00	7.00	2500.00	125.00	0.0125
2500.00	1.00	1.00	1250.00	62.50	0.0063
2500.00	1.00	3.00	625.00	31.25	0.0031
2500.00	1.00	7.00	312.50	15.63	0.0016
312.50	1.00	1.00	156.25	7.81	0.0008
312.50	1.00	3.00	78.13	3.91	0.0004
312.50	1.00	7.00	39.06	1.95	0.0002

In medium after the addition of 19 mL of Muller Hinton agar.

8000 mg/L (stock solution) corresponding to 8% w/v of M. punicifolia extract.

Challenge test on water buffalo steaks

Water buffalo steaks (Musculus quadriceps femoris) were obtained from commercial sources and transported at 4 °C in few hours into the DiAAA laboratories (Campobasso, Italy). Steaks (each about 200 g and approximately 1 cm thick) were intentionally inoculated with a 4‐strain cocktail of P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T. The inoculation of each steak was performed by disposing 100 μL of the mixed cocktail with a micropipette on the surface and spreading over the surface with a sterile disposable plastic spatula in order to obtain an inoculum of about 4.0 log CFU/g. The meat steaks were then divided into 5 batches: batches M1, M2, M3, and M4, steaks dipped for a few seconds in a presterilized container with 0.0063%, 0.0125%, 0.025%, and 0.05% (w/v) M. punicifolia extract, respectively, and batch C (control), steaks treated with the hydroalcoholic solution without extract. Removing excess solution was done by gently pressing each piece against the inside walls of containers. Then, each steak was individually packaged in polypropylene tray wrapped in air‐permeable film, and stored at 4.0 ± 0.5 °C for 21 d. Three biological replicates were performed for each batch.

A 2nd test was performed as described previously, except that after inoculation with the 4‐strain cocktail, steaks were incubated at 4 °C for 2 h to allow the bacterial adhesion to the meat surface (Rhoades and others 2013). Then steaks were divided into 5 batches and treated as described above.

Microbiological and sensory analyses on water buffalo steaks

Microbiological analyses were performed at 0, 4, 8, 12, 15, 18, and 21 d of storage at 4 ± 0.5 °C to evaluate the microbial load in water buffalo steaks. For this purpose, about 10 g of meat from each batch were aseptically withdrawn and decimal diluted in a sterile solution of 0.1% peptone water. After homogenization in a Lab‐blender (Stomacher Seward Medical, London, SE1 1PP, U.K.), subsequent serial dilutions were inoculated in appropriate media as described below:

B. thermosphacta was enumerated on STA Agar base (Oxoid) with STA selective supplement (Oxoid) after incubation at 28 °C for 48 h.

Pseudomonas spp. was counted on PAB (Oxoid) with SR102E supplement (Oxoid) after incubation at 28 °C for 24 to 48 h.

Total and fecal coliforms were counted on Violet Red Bile Agar (Oxoid) after incubation for 48 h at 37 °C and 44 °C, respectively.

Eumycetes (yeasts and molds) were counted on Rose Bengal Agar after incubation at 25 °C for 48 h.

The results were expressed as the mean of 6 determinations performed on 2 different steaks for each biological replicate.

The growth of Pseudomonas spp. and B. thermosphacta were modeled using the DMFit v 2.0 structured dynamic model (Baranyi and Roberts 1994).

Color was determined on the surface of water buffalo steaks. Measurements were performed at 0, 4, 8, 12, 15, 18, and 21 d of storage using the CIE L*, a*, b* system (Commission Internationale de l'Eclairage 1978) with a reflectance spectrophotometer (Minolta CR300b, Suitashi, Osaka, Japan). The results were expressed as the mean of 18 determinations performed in triplicate on 2 steaks for each biological replicate.

Sensory analysis was performed on steaks from all batches at 0, 3, 7, 10, 14, and 21 d of storage. To define the sensory profile, a descriptive panel of 15 judges was employed. The judges were trained in 3 preliminary sessions to produce a list of attributes useful to define the sensory profile of water buffalo steaks. On the basis of the frequency of citation (>60%), 4 descriptors were selected: general appearance, color, flavor, and off‐flavor. Random samples were evaluated by assigning a score between 1 (poor) and 10 (excellent) for positive descriptors (general appearance, color, and flavor), and between 10 (unacceptable) and 1 (excellent) for the negative one (off‐flavor).

Statistical methods

Mean values, medians, standard deviations, and the occurrence of statistically significant differences were determined with the OriginPro 7.5 software (OriginLab Corporation, Northampton, Mass., U.S.A.).

Results and Discussion

Chemical composition of M. punicifolia extract

GC‐MS analysis of M. punicifolia extract led to the identification of 9 components in the volatile fraction (Table 2). The main relative composition of the extract, expressed as the percentage area contribution of each compound, was represented by furanes. These compounds are generally related to fruit aroma, but several authors also reported their antimicrobial activity (Baveja and others 2004; Sung and others 2007). The profile of vitamins and phenolic compounds (Table 3) was obtained using a RP‐HPLC‐DAD method previously described by Mendiola and others (2008). This method allows, in a single run, to analyze vitamins (water and fat soluble), phenolic compounds, and pigments, providing a general description of bioactive components present in a natural extract. The analysis highlighted the absence of fat‐soluble vitamins and the high content of some water soluble vitamins such as ascorbic acid and thiamine, which were observed at a concentration of 95.06 and 22.59 mg/g extract, respectively (see Table 3). Although phenolic compounds were not completely identified, it was possible, by resemblance of UV‐vis spectra, to assign them to certain families of phenolic compounds and to quantify them using the right standard. The content of phenolic compounds was correlated to the value obtained in the Folin–Ciocalteu test (121.16 ± 1.24 mg gallic acid/g extract).

Table 2

Volatile compounds in M. punicifolia extract detected by GC‐MS

		Retention index	Retention index
RT (min)	Compound	(Theorical)	(Real)	% area
6.54	2‐furancarboxaldehyde	830	867	4.31
7.58	2‐furanmethanol	885	894	1.76
9.36	2‐Propenoic acid, 2‐methyl‐ ethenyl ester	860	950	2.59
10.52	2‐Furancarboxilic acid	970	987	1.62
12.56	2‐furanpropionic	1050	1053	1.43
16.06	4H‐Pyran‐4‐one, 2,3‐dihydro‐3,5‐dihydroxy‐6‐methyl‐	1130	1170	4.03
18.30	2‐Furancarboxaldehyde, 5‐(hydroxymethyl)‐	1200	1214	43.79
21.53	5‐(hydroxymethyl)‐2‐Furancarboxaldehyde,	1250	1258	3.52
24.50	3‐hydroxy‐5‐methyl‐2(5h)‐furanone	1480	1502	1.47

Table 3

Vitamins and phenolic compounds in M. punicifolia extract detected by HPLC

Peak	Time (min)	Compound	mg/g
1	3.21	Ascorbic acid (vitamin C)	95.06
2	3.30	Thiamine (vitamin B1)	22.59
3	8.53	Phenolic acid	5.31
4	9.51	Pheno diol	2.21
5	10.02	Flavanone	1.01
6	10.43	Flavanone	0.51
7	13.37	Isoflavone	0.70
8	13.45	Isoflavone	0.26

Antimicrobial activity expressed in vitro by M. Punicifolia

The agar well diffusion showed that M. punicifolia extract, used at 8% concentration (stock solution), exerted a very strong inhibitory action (IH ratio ≤1) against B. thermosphacta DSMZ 20171T (Table 4). A strong inhibition was appreciated against P. fluorescens DSMZ 50009T (IH ratio, 1.07) and against P. fragi DSMZ 3456T (IH ratio, 1.11), while a moderate inhibition was detected against P. putida DSMZ 291T (IH ratio, 1.23). On the other hand, the hydroalcoholic solution without extract, used as control, produced no inhibition, as highlighted by the presence of microbial growth around the wells. Previous results pointed out that M. punicifolia extract has an important inhibitory action that could be associated to its composition. In fact, several compounds found in the extract, such as phenolic acids, flavonones, and furans (Table 2 and 3), were already described for their antimicrobial activity (Tiwari and others 2009; Ozen and others 2011; Ramalakshmi and Muthuchelian 2011; Chauhan and others 2012; Hyldgaard and others 2012; Nowak and others 2012a).

Table 4

Antimicrobial activity exerted by M. punicifolia extract against B. thermosphacta and Pseudomonas spp.as detected by the agar well diffusion assay

		Hydro
	M. punicifolia	alcoholic
Strains	extract (8% w/v)	solutionb	Streptomycinc	Tetracyclinec	Ratiod
B. thermosphacta DSMZ 20171T	24.02 (±0.05)a	Absent	−	23.5 (±0.04)	0.98
P. fluorescens DSMZ50009T	19.53 (±0.04)	Absent	21.01 (±0.05)	−	1.08
P. fragi DSMZ3456 T	17.49 (±0.06)	Absent	19.51 (±0.03)	−	1.11
P. putida DSMZ291T	15.01 (±0.03)	Absent	18.49 (±0.04)	−	1.23

Inhibition halos (IH, mm).

Hydroalcoholic solution was used as control.

Streptomycin and tetracycline were used as reference controls for Pseudomonas spp. and B. thermosphacta.

Ratio between IH (mm) mean value obtained with the reference control and IH (mm) mean value obtained with M. punicifolia extract.

The results of the agar dilution test revealed interesting MIC values (Table 5). In fact, a concentration up to 0.025% completely inhibited the bacterial growth of all the assayed microorganisms, and the highest activity of the extract was detected against B. thermosphacta (MIC of 0.0063%). An important antimicrobial activity was also appreciated against Pseudomonas species, which showed MIC values of 0.0125% for both P. fragi and P. fluorescens and of 0.025% for P. putida. The highest antimicrobial effect measured against B. thermosphacta confirmed that Gram‐positive bacteria are more sensitive than Gram‐negative bacteria, in agreement with results on the inhibitory action of natural extracts obtained by other authors (Burt 2004; Kim and others 2013). MIC values also indicated that the extract of M. punicifolia has an antimicrobial effect against Pseudomonas and B. thermosphacta similar or higher than that exerted by other natural extracts (Gutierrez and others 2009; Tiwari and others 2009; Nowak and others 2012a).

Table 5

MIC values determined for B. thermosphacta and Pseudomonas spp. in presence of different concentrations of M. punicifolia extract (−, absence of growth; +, presence of growth)

Concentration of	B. thermosphacta		P. fragi	P. putida
M. punicifolia extract	DSMZ20171T	P. fluorescens DSMZ50009T	DSMZ3456T	DSMZ291T
0.4000	−	−	−	−
0.2000	−	−	−	−
0.1000	−	−	−	−
0.0500	−	−	−	−
0.0250	−	−	−	−
0.0125	−	−	−	+
0.0063	−	+	+	+
0.0031	+	+	+	+
0.0016	+	+	+	+
0.0008	+	+	+	+
0.0004	+	+	+	+
0.0002	+	+	+	+

Antimicrobial activity of M. punicifolia extract in water buffalo steaks

In situ experiments were carried out to investigate the efficacy of the extract when used in steaks stored under refrigerated conditions. The first test was performed on water buffalo steaks inoculated with a 4‐strain cocktail of P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T. Steaks were then divided into 5 batches and were dipped with M. punicifolia extract at increasing concentrations (M1 to M4) or with the control solution (C). Immediately after dipping, microbiological analyses were performed on steaks from each batch to confirm the inoculum size at time zero (about 4.0 log CFU/g). Microbial counts revealed statistically significant differences (P < 0.05) between batches (data not shown), probably due to the washing out of some microbial cells during the dipping. So, the test was repeated in the same manner, except that prior to the treatment with the extract or with the control solution, inoculated steaks were incubated at 4 °C for 2 h to allow the bacterial adhesion to the meat surface (Rhoades and others 2013). This tool allowed the obtainment of constant initial microbial loads among batches, that we considered an important requisite not only to simulate the highest microbial contamination that can naturally occur in fresh meats (Ercolini and others 2010; Papadopoulou and others 2012), but also to compare the data of the antimicrobial efficacy of M. punicifolia extract used at different concentrations. The microbiological changes occurring in water buffalo steaks during 21 d of storage at 4 °C are shown in Figure 1 and 2. Results evidenced that M. punicifolia extract, when used at concentrations of 0.025% and of 0.05%, strongly affected the growth of B. thermosphacta (Figure 1A) and Pseudomonas spp. (Figure 1B) intentionally inoculated at 4.0 log CFU/g. The analysis of kinetic parameters for these bacteria (Table 6) underlined the effect of the extract used at different concentrations. In fact, B. thermosphacta increased in the control batch (C) after a lag phase of 78.14 h, reaching levels of 7.26 log CFU/g at the end of the storage with a maximum specific growth rate (μmax/h) of 0.0102. In samples from batch M1 (0.0063% concentration of extract), B. thermosphacta reached 6.56 log CFU/g with a μmax of 0.0078/h. Moreover, the lag phase of B. thermosphacta in the batch M1 was about twice as that detected in the control. A similar behavior was observed for Pseudomonas spp., but final counts were higher than those of B. thermosphacta. The in situ inhibitory activity of the extract appeared much more evident in the batch M2 (0.0125% concentration of the extract), where counts of B. thermosphacta and Pseudomonas spp. increased slightly during the entire storage period, and their lag phases were 3.7 and 2.2 times longer, respectively, than those observed in the control. A considerable inhibition of intentionally inoculated bacteria was found in batches M3 (0.025% concentration of the extract) and M4 (0.05% concentration of the extract). In fact, in the batch M3 counts of 2.11 for B. thermosphacta and 2.70 log CFU/g for Pseudomonas spp. were appreciated after 21 d under refrigerated storage. Counts lower than 1.8 log CFU/g were registered in the batch M4.

Figure 1

Kinetic data of B. thermosphacta (A) and Pseudomonas spp. (B) in water buffalo steaks stored at 4 °C for 21 d. Data were fitted with Baranyi's model (Baranyi and Roberts 1994). C, control steaks intentionally inoculated with P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T and without vegetal extract; M1, M2, M3, and M4, inoculated steaks treated with M. punicifolia extract at a concentration of 0.0063%, 0.0125%, 0.025%, and 0.05% (w/v), respectively.

Figure 2

Behavior of total coliforms (A) and Eumycetes (B) in water buffalo steaks stored at 4 °C for 21 d. C, control steaks intentionally inoculated with P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T and without vegetal extract; M1, M2, M3, and M4, inoculated steaks treated with M. punicifolia extract at a concentration of 0.0063%, 0.0125%, 0.025%, and 0.05% (w/v), respectively.

Table 6

Kinetic parameters detected for B. thermosphacta and Pseudomonas spp. in water buffalo steaks intentionally inoculated with P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T. C, inoculated control steaks without vegetal extract; M1, M2, M3, and M4, inoculated steaks treated with M. punicifolia extract at a concentration of 0.0063%, 0.0125%, 0.025%, and 0.05% (w/v), respectively

		Initial values	Lag	Shoulder	Max rate	Final values
Microorganism	Batch	(log CFU/g)	(h)	(h)	(μmax/h)	(log CFU/g)
	C	4.06	78.14	−	0.0102	7.26
	M1	4.01	138.58	−	0.0078	6.56
B. thermosphacta	M2	4.11	293.47	−	0.0027	4.62
	M3	4.10	−	76.32	−0.0049	2.11
	M4	4.13	−	18.36	−0.0059	1.33
	C	4.08	96.55	−	0.0120	8.21
	M1	4.01	108.42	−	0.0088	7.04
Pseudomonas spp.	M2	4.08	215.21	−	0.0067	6.03
	M3	3.99	−	180.23	−0.0043	2.71
	M4	4.01	−	116.31	−0.0061	1.78

In the light of previous results, it is possible to affirm that M. punicifolia extract, used in situ at concentrations of 0.025% and 0.05% (batches M3 and M4, respectively), was able not only to inhibit, but also to inactivate the assayed bacteria. Kinetic data (Table 6) showed that B. thermosphacta was more sensitive to the extract than Pseudomonas spp. In fact, the shoulders recorded in the survival curve of B. thermosphacta were 76.32 and 18.36 h in batches M3 and M4, respectively, followed by a linear decay. For Pseudomonas spp., higher lengths of the shoulders were evidenced (180 and 116 h in batches M3 and M4, respectively).

The inhibitory activity of M. punicifolia extract was also observed against undesired microorganisms other than those intentionally inoculated. In fact, a strong decrease in total coliforms was recorded in batches M3 and M4 during the storage period, in contrast with results characterizing the other batches (Figure 2A). The effect of the extract used at 0.025% concentration (batch M3) was lower against Eumycetes than that against total coliforms, whereas the concentration of 0.05% (batch M4) caused an important decrease of Eumycetes during time, similar to that characterizing total coliforms (Figure 2B).

Previous results revealed that higher concentrations of M. punicifolia extract were required in situ than those required in vitro, and this fact was particularly evident against B. thermosphacta. In fact, the MIC value for B. thermosphacta was 0.0063%, but a concentration at least of 0.025% was required to assure the inhibitory effect in the steaks. These data are in agreement with other studies, which highlighted that the application in situ of natural extracts may be affected by the food composition (Tiwari and others 2009; Hyldgaard and others 2012). In this context, results obtained in our study are actually promising, considering that the difference between inhibitory concentrations of M. punicifolia extract in vitro and in situ was considerably lower than that reported in other studies on natural extracts (Cava‐Roda and others 2012; Kyung 2012). For instance, Gill and others (2002) observed that cilantro oil had a significant antimicrobial activity in vitro at 0.018% concentration, while no antimicrobial activity was recorded in a food model using a concentration of 6% (over 300 times higher). In our study, the antimicrobial activity against B. thermosphacta was obtained by using M. punicifolia extract at concentrations of 0.0063% in vitro and 0.025% in situ, that is, only 4 times higher. This datum is particularly interesting, taking into account that the use of natural preservatives can alter the taste of food and exceed the flavor threshold acceptable to consumers (Lv and others 2011).

Among other factors able to influence the consumer choice, the color represents one of the most important visual clues for red fresh meat (Tremonte and others 2005). In our study, the color analysis performed on water buffalo steaks treated with M. punicifolia extract showed that the red color was preserved in samples from batches M3 and M4 (a* > 12.0) until the end of the storage (Table 7), whereas a marked decrease in this value was observed in the samples from the other batches. The loss of redness in meat is generally related to the accumulation of met‐myoglobin, caused to primary lipid oxidation products such as hydroperoxides and other free radicals, which oxidize the ferrous ion (Fe2+) from oxymyoglobin into the ferric form (Fe3+) present in metmyoglobin. Moreover, the accumulation of metmyoglobin in meat products is accelerated by secondary lipid oxidation products, such as unsaturated aldehydes (Faustman and others 2010). Phenolic acids, allowing the direct interaction with myoglobin, may retard myoglobin oxidation and discoloration (Kroll and Rawel 2001), as also reported by other authors with regard to the use of vegetable extracts in meat products (Gil‐Chavez and others 2013; Kim and others 2013).

Table 7

Lightness (L*), redness (a*), and yellowness (b*) values recorded after 21 d of refrigerated storage on the surface of water buffalo steaks intentionally inoculated with P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T. C, inoculated control steaks without vegetal extract; M1, M2, M3, and M4, inoculated steaks treated with M. punicifolia extract at a concentration of 0.0063%, 0.0125%, 0.025%, and 0.05% (w/v), respectively

	Time (d)	C	M1	M2	M3	M4
	0	45.02 ± 0.89a	45.56 ± 0.54a	45.93 ± 0.63a	45.74 ± 0.54a	45.89 ± 0.71a
	4	44.99 ± 0.73a	45.07 ± 0.69a	45.07 ± 0.47a	46.02 ± 0.42a	45.94 ± 0.63a
	8	42.52 ± 0.66b	44.92 ± 0.73a	45.35 ± 0.42a	46.21 ± 0.71a	46.09 ± 0.38a
L*	12	41.14 ±1.22b,c	45.66 ± 0.27a	45.09 ± 0.67a	46.11 ± 0.47a	46.01 ± 0.53a
	15	40.52 ± 0.71c	45.66 ± 0.56a	45.12 ± 0.88a	46.01 ± 0.52a	45.92 ± 0.66a
	18	38.55 ± 0.53d	44.83 ± 0.91a	45.15 ± 0.92a	45.95 ± 0.74a	45.95 ± 0.54a
	21	36.05 ± 0.54e	42.73 ± 0.56b	45.77 ± 0.61a	46.11 ± 0.55a	46.29 ± 0.48a
	0	13.85 ± 0.45a	13.75 ± 0.53a	13.70 ± 0.23a	13.95 ± 0.23a	13.88 ± 0.24a
	4	10.54 ± 0.47b	11.42 ± 0.58b	11.96 ± 0.42b	13.72 ± 0.41a	13.68 ± 0.42a
	8	8.62 ± 0.71c	10.43 ± 0.61b	11.15 ± 0.51b	13.32 ± 0.45a	13.45 ± 0.38a
a*	12	7.53 ± 0.71c,d	10.01 ± 0.73b,c	10.75 ± 0.45b	13.15 ± 0.51a	13.32 ± 0.27a
	15	7.15 ± 0.49d	9.10 ± 0.72c	10.68 ± 0.57b	12.85 ± 0.42a	13.28 ± 0.31a
	18	6.75 ± 0.56d	8.82 ± 0.47c	9.35 ± 0.22c	12.95 ± 0.36a	13.33 ± 0.29a
	21	5.34 ± 0.52e	7.61 ± 0.58d	8.75 ± 0.29c	12.45 ± 0.43a,b	12.98 ± 0.30a
	0	10.20 ± 0.63a	10.85 ± 0.37a	10.44 ± 0.41a	10.75 ± 0.51a	10.88 ± 0.31a
	4	10.83 ± 0.67a	11.03 ± 0.48a	10.43 ± 0.57a	10.83 ± 0.63a	11.08 ± 0.42a
	8	11.22 ± 0.48a	11.42 ± 0.57a	10.72 ± 0.51a	11.32 ± 0.42a	10.97 ± 0.33a
b*	12	13.16 ± 0.59b	12.33 ± 0.32b	12.53 ± 0.47b	10.70 ± 0.63a	11.38 ± 0.41a
	15	14.85 ± 0.57c	13.22 ± 0.68b	13.75 ± 0.51b	10.96 ± 0.49a	10.89 ± 0.65a
	18	15.14 ± 0.44c	14.95 ± 0.47c	14.99 ± 0.42c	11.40 ± 0.71a	11.29 ± 0.46a
	21	16.85 ± 0.38d	15.12±0.52c	15.17±0.61c	11.10 ± 0.56a	11.42 ± 0.52a

Same letters in rows and columns indicate the absence of significant differences (P > 0.05).

The positive effect expressed by M. punicifolia extract on the color of water buffalo steaks was also supported by the lightness (L*), and the yellowness (b*) measurements. In fact, these parameters revealed slight changes in batches M3 and M4, whereas a marked change was found in batches C and M1, where L* decreased and b* increased during the time of storage. Intermediate values were recorded in steaks from the batch M2.

The sensory analysis performed on water buffalo steaks (Table 8) confirmed previous results. In fact, after 21 d under refrigerated storage, steaks from batches M3 and M4 (0.025% and 0.05% concentration of the extract, respectively) were highly appreciated by the panelists, in terms of maintenance of red color, flavor, and general appearance, and low presence of off‐flavors.

Table 8

Mean values of sensory attributes evaluated after 21 d of refrigerated storage on water buffalo steaks inoculated with P. putida DSMZ 291T, P. fluorescens DSMZ 50009T, P. fragi DSMZ 3456T, and B. thermosphacta DSMZ 20171T. C, inoculated control steaks without vegetal extract; M1, M2, M3, and M4, inoculated steaks treated with M. punicifolia extract at a concentration of 0.0063%, 0.0125%, 0.025%, and 0.05% (w/v), respectively

				General
Batch	Flavor	Off‐flavor	Color	appearance
C	1.11 ± 0.35a	9.80 ± 0.41a	2.14 ± 0.63a	1.32 ± 0.59a
M1	1.63 ± 0.51a	8.70 ± 0.45a	4.12 ± 0.51b	2.45 ± 0.63a
M2	3.92 ± 0.25b	5.83 ± 0.41b	5.44 ± 0.50b	4.91 ± 0.59b
M3	8.12 ± 0.63c	2.23 ± 0.56c	8.56 ± 0.51c	8.66 ± 0.50c
M4	9.21 ± 0.51c	1.31 ± 0.43c	9.15 ± 0.35c	9.42 ± 0.51c

Same letters indicate the absence of significant differences (P > 0.05).

As expected, control steaks showed unacceptable sensory parameters at the end of the storage period, and similar values were attributed to steaks from the batch M1 (P > 0.05). Steaks from the batch M2 presented intermediate values between those appreciated in batches C/M1 and M3/M4. Generally, significant differences were not appreciated by panelists between batches M3 and M4 (P > 0.05).

Conclusions

Results reported in this study highlight that M. punicifolia extract is able to produce a strong antagonistic effect against Pseudomonas species and B. thermosphacta, involved in the spoilage of fresh meat products. In vitro assays evidenced low MIC values against the assayed spoilage bacteria, and promising results were also obtained in situ. The strong antimicrobial activity of M. punicifolia extract could be due to the presence of several compounds belonging to different bioactive groups, which represent a series of hurdles (more or less strong) for the microbial growth. Moreover, the presence of bioactive compounds also allowed the preservation of sensorial properties in water buffalo steaks.