Microbiological characterization of different formulations of alheiras (fermented sausages)

: Different ingredients in old recipes are becoming popular and the traditional alheira did not escape to this new trend. The objective of this preliminary study was to characterize microbiologically nine different formulations of alheira from five producers. In this sense, isolates obtained were characterized through different phenotypic and biochemical tests. Their susceptibility to different antimicrobials and the presence of virulence factors was also investigated. Lactic acid bacteria were the predominant microbiota, but pathogenic bacteria as coagulase-positive staphylococci, Listeria monocytogenes and Salmonella spp. as well as indicator organisms were also found. Several virulence factors were produced among the different groups of isolates, with a high incidence of isolates producing β-haemolysis. Along with their potential pathogenic activity, also several antimicrobial resistances were found being the majority of isolates classified as multi-resistant. At our knowledge, this is the first study with these new formulations of alheira. A higher number of products must be analyzed, but we believe that results obtained in this study should help to alert consumers for the need of safe cooking time/temperatures of these products.


Introduction
Fermented meat products are important elements in the economy of certain regions, not only because of their culinary heritage, but also due to their unique flavors and textures [1]. In Portugal, there are a wide variety of fermented meat products and their manufacture represents an important income in specific regions, predominantly in the North and the Southern regions [2]. Many of these products are fashionable food products that command high prices in urban centers and in export markets.
A traditional and naturally fermented meat sausage, typical from Trá s-os-Montes region of northern Portugal is alheira. Traditionally, alheira is produced from chopped pork and poultry meat, lard, wheat bread, olive oil and pork fat, which are mixed with salt, garlic and spices. The meat, lard, olive oil and spices are boiled together with water and then the bread is added and the mass mixed. When everything is completely mixed the paste is stuffed into cellulose or natural pig casings and submitted to a smoking process for no longer than 8 days [3]. Nowadays, due to consumers and market demands, an increase of new products has been observed. Since 2015 that several formulations of alheira have been produced, both with different meats as turkey, piglet, lamb or veal, as well as with completely different ingredients like tuna, olives, between others.
Different formulations of traditional products, new ingredients in old recipes are trending. It is therefore necessary to characterize these new products in order to uphold their food safety and create the science base on which strategies for competitiveness and sustainability can be built. Such knowledge would allow the development of new methods of production and quality control, simultaneously compatible with modern retail channels and respectful of the unique characteristics of these products.
This preliminary study aimed to analyze the microbiota of nine different formulations of alheiras. The characterization of isolates through different phenotypic and biochemical tests as well as evaluation of their susceptibility to different antimicrobials and the presence of virulence factors was also performed.

Sampling
Nine alheiras with different compositions and from five different producers were purchased from retail stores in two different occasions during October 2015 (referred as batch 1) and April 2016 (referred as batch 2). The innovative ingredients, pH and water activity of each alheira are presented in Table 1. Samples were transported to the laboratory in insulated bags and stored at 4 ℃ until they were analyzed.
After appropriate incubation, colonies were counted and/or confirmatory tests performed and the colony forming units (CFU)/g calculated. Microbial counts were transformed to log CFU/g.

Statistical analysis
An analysis of variance was carried out to test the differences between each formulation of alheira as well as between different producers. Multiple comparisons were evaluated by Tukey's post-hoc test and all analyses were performed using IBM SPSS Statistics, 24 (IBM Corporation, USA). The mean difference was considered significant at the 0.05 level.

Origin of isolates
Colonies grown on each selective media were randomly selected and purified by repeated streaking onto the respective growth media. All isolates (10%) recovered were stored at −80 ℃ in Tryptic Soy Broth (TSB, Pronadisa) or MRS broth (in the case of LAB) with 30% (v/v) of glycerol (Sigma, Steinheim, Germany) and sub-cultured twice before use.

Identification by phenotypic and biochemical tests
Isolates were characterized using their colonial and cellular morphology, Gram reaction and conventional biochemical tests: enterococci (n = 38) were tested for different growth conditions and acid production from several sugars [14]; Listeria spp. isolates (n = 7) were identified according to the ISO 11290-1: 1996 [11]; staphylococci (n = 49) were tested for the presence of several enzymes, acid production from several sugars [15] and susceptibility to novobiocin (5 µg) and polymyxin B (300 IU) according to Iorio et al. [16]. Enterobacteriaceae (n = 33) were tested for motility, production of hydrogen sulfide, presence of several enzymes and acid production from several sugars [15]. Lactic acid bacteria (LAB, n = 111) were tested by Gram staining and for the presence of enzymes catalase and oxidase [5].
Identification of each group of isolates was based on Bergey's Manual of Determinative Bacteriology [15].

Screening for botulinum neurotoxin producing strains
Samples which showed the presence of sulfite-reducing Clostridium were screened for botulinum neurotoxin producing strains by mouse bioassay [17] following enrichment in TPGY (Trypticase-Peptone-Glucose-Yeast Extract Broth).

Antimicrobial susceptibility
The minimum inhibitory concentrations (MIC; µg/ml) were determined by ε-test for trimethoprim/sulfamethoxazole (SXT, AB Biodisk, Solna, Sweden) and by the agar dilution method for fifteen other antimicrobials, according to the Clinical Laboratory Standards International [18]. Each test was carried out on Muller-Hinton Agar (MHA, BioMé rieux) with cations adjusted for penicillin G (Sigma) and ampicillin (Fluka, Steinheim, Germany) and on MHA for vancomycin (Fluka), ceftazidime, chloramphenicol, kanamycin, nalidixic acid, nitrofurantoin, oxacillin, streptomycin (Sigma), ciprofloxacin, erythromycin, gentamicin, tetracycline and rifampicin (all kindly supplied by the company Labesfal, Portugal). Each test was carried out on Muller-Hinton Agar, using a different set of antimicrobials for each group of microorganisms [18,19]. All the isolates were grown on plates of MHA and MHA with cations adjusted with no antimicrobial. Each experiment was performed in duplicate and the quality control strains Enterococcus faecalis ATCC 29212 and Staphylococcus aureus ATCC 29213 were used to monitor the accuracy of MICs [18]. Plates were incubated at 37 ℃ for 24 hours.
Enterococcus spp., Staphylococcus spp. and Enterobacteriaceae isolates were classified as sensitive, intermediate or resistant to each antimicrobial according to the Clinical and Laboratory Standards Institute [18]. Lactic acid bacteria (LAB) and Listeria spp. isolates were classified as described by EFSA [19] and Barbosa et al. [20], respectively.
Isolates exhibiting resistance to, at least, two of the antimicrobial agents of different classes were considered to be multi-resistant strains.

Presence of virulence factors
Presence of different virulence factors was only determined for Staphylococcus spp., LAB isolates, Enterococcus spp. and Enterobacteriaceae.

1) Production of gelatinase, DNase and hemolytic activity
Presence of hydrolytic enzymes gelatinase [21] and DNase [22] was determined using the modified Luria-Bertani (MLB) broth supplemented with 50.0 g/L of gelatin and DNase agar (Pronadisa) with 0.05 g/L of methyl green (Sigma), respectively. Presence of haemolysin was assessed using 5% v/v Sheep blood agar [23]. Each experiment was performed in duplicate and S. aureus ATCC 25213 was used as positive control.

2) Detection of decarboxylase activity
Only isolates of LAB, Enterococcus spp. and Enterobacteriaceae were screened for the production of histamine, tyramine, putrescine and cadaverine, according to the method described by Bover-Cid and Holzapfel [24]. Briefly, each isolate was sub-cultured seven times in MRS (LAB) or Brain Heart Infusion broth (BHI; Enterobacteriaceae and Enterococcus spp.) with 0.1% of each precursor amino-acid (all from Sigma), in order to promote enzyme induction. Then, all isolates were spotted in duplicate on the Bover-Cid medium plates with and without (as control) each amino acid and incubated at 37 ℃ for 4 days under aerobic conditions. Positive reaction was confirmed when a purple color occurred or tyrosine precipitate disappeared around the colonies [24,25].

Results and discussion
Microbiological characteristics of nine alheiras with different formulations were studied in two different time periods. The innovative ingredients of each fermented sausage and their respective values of pH and water activity (a w ) are presented in Table 1. Values of pH varied between 3.84 and 5.60 and values of a w between 0.968 and 0.998. Traditionally, besides displaying distinctive organoleptic and sensory characteristics in these products, low values of pH and a w are also important parameters to prevent microbial spoilage and growth of pathogenic bacteria [26]. Combination of different hurdles is important to ensure the safety of these products. The use of spices such as thyme or oregano (producer A) may contribute as additional hurdles, due to their recognized antimicrobial activities [27].
Regarding composition of the selected fermented sausages, only one did not include meat (alheira 1, tuna fish), four were produced with different meats, such as turkey (alheira 5), lamb (alheira 6), veal (alheira 7) and piglet (alheira 9), two were produced with chicken meat and uncommon ingredients such as brown bread instead of wheat bread (alheira 4) and pitted olive and oregano (alheira 2) and, finally, despite their traditional composition corn bread and red wine were added to alheiras 8 and 3, respectively.
Results for the enumeration and detection of different microorganisms are presented in Table 2 and statistical differences obtained between formulations of alheiras and producers are shown in supplementary Tables S1 and S2. Differences in manufacturing processes and/or composition of alheiras could be the reason of the great variability between different producers as well as different products and different time of production from the same producer. With the exception of alheiras from producer A, no significant differences were obtained between the other producers (p > 0.05), since all presented high values of total microorganisms at 30 ℃. Counts in MRS were also high for the majority of the samples, which was expectable since alheiras are fermented products and LAB play an important role in fermentation processes [28]. Also Enterococcus spp. were present in most of the samples and, when present, their values varied between 2.8 and 8.7 log CFU/g. Although a lot of benefits have been attributed to their presence in fermented products, such as contribution on ripening and aroma development [29] and also the production of antimicrobial substances [25], the existence of many strains possessing virulence factors and which are becoming increasingly resistant to antimicrobials is a reason of concern [23, 30,31]. Apart from producer E, in which none sample presented yeasts or moulds, in the other samples, higher growth was observed for yeasts compared to moulds. It is described that these microorganisms also play an important role in the development of the organoleptic characteristics of products [32].
The presence of indicator microorganisms was also found in a few samples, which might be result of poor hygiene or poor process control [26]. Enterobacteriaceae were found in alheiras from three producers (A, D and E), but counts were lower than 4 log CFU/g in all samples and in alheira 8 (batch 2) of producer E, E. coli was found in numbers of 2.3 log CFU/g. Also three samples from batch 1 (alheiras 1, 4 and 7) and three samples from batch 2 (alheiras 2, 5 and 8) were positive for indicator organism Listeria spp. and, in addition, the pathogen L. monocytogenes was present in three samples (from batch 1: alheira 4 and from batch 2: alheiras 2 and 5). Although this pathogen was present in only one alheira of producer A, the remaining two positive alheiras, although having different compositions, belong to producer B. This may be indicative of crosscontamination during their manufacture [33]. Although alheiras are cooked before consumption, it is important to highlight that cooking methods might not be sufficient to inactivate this foodborne pathogen [34]. Regarding pathogens, significant differences were obtained between formulations (p < 0.05) with three samples: one from batch 1 (alheira 7) and two from batch 2 (alheiras 2 and 7) presenting more than 2 log CFU/g of coagulase-positive Staphylococcus and other three: one from batch 1 (alheira 6) and two from batch 2 (alheiras 6 and 8) presenting more than 4 log CFU/g. These high concentrations are probably the result of the considerable levels of handling product-in-process or even crosscontamination [35]. Salmonella spp. was present in 25g of alheira 2 (batch 1) and sulfite-reducing Clostridium spores in 1 g of three samples (alheira 8 (batch 1) and alheira 9 (batches 1 and 2)). As in September 2015 there were cases of botulism associated with the consumption of alheiras in Portugal, positive samples for sulphite-reducing Clostridium spores were sent to Instituto Ricardo Jorge to evaluate their possible contamination by Clostridium botulinum. Neurotoxigenic strains of C. botulinum were not detected in the three positive samples. Although different, all samples belong to the same producer, which indicates that cross-contamination could occur during the production processes.
Only samples from producers A and B -alheira 1 (batch 2), alheira 3 (batches 1 and 2), alheira 4 (batch 2) and alheira 5 (batch 1) did not show the presence of any pathogen. It is interesting to note that from those, only alheira 3 had no pathogens in the two batches tested. Although very similar to traditional alheiras, the huge difference is that these samples were produced with red wine (wine marinade with added garlic), which may contribute to their good microbiological quality. Antimicrobial activities of wine marinades are already reported [36].
Results obtained by different phenotypic and/or biochemical tests (data not shown) only allowed the characterization and identification of LAB to group level, Enterobacteriaceae to family level, enterococci and staphylococci to the genus level, and Listeria spp. to species level (L. monocytogenes and L. innocua).
The percentage of isolates (belonging to each group of bacteria) that were sensitive, intermediate and resistant to each tested antimicrobial is presented in Figure 1.
All isolates of enterococci (n = 38) were susceptible to ampicillin (100%) and a large number to penicillin (97.37%), chloramphenicol (86.84%), erythromycin (84.21%) and rifampicin (71.05%). Those results are in agreement with the study of Barbosa et al. [30] for enterococci isolated from traditional fermented products, with the exception of erythromycin and rifampicin, since the authors found a high percentage of resistant isolates. The same authors also reported intermediate resistance of 0.5% to vancomycin, instead of 71.05% of resistance found in the current study. Apart from the high percentage of enterococci isolates resistant to vancomycin, also high percentages were found to ciprofloxacin ( [30,37]. One of the most important concerns regarding the presence of antimicrobial resistant enterococci is their remarkable ability to acquire new mechanisms of resistance and also to transfer resistance genes to other pathogens [38]. Regarding other isolates belonging to LAB group (n = 111), none of the isolates showed intermediate resistances, more than 90% of the isolates were sensitive to all antimicrobials, except to vancomycin (63.96% resistant) and tetracycline (63.06% resistant) and 31 out of 111 isolates were multi-resistant. Several species of Lactobacillus, Leuconostoc and Pediococcus are intrinsically resistant to vancomycin, which could explain the high percentages found [39]. In the study of Federici et al. [40], the authors also found resistances to tetracycline among isolates from different genera of LAB isolated from smoked and dry-cured meat sausages. Unlike resistance to vancomycin, resistance to tetracycline is acquired and there is evidence of conjugative transfer of the gene tet(M) from lactobacilli to other microorganisms in vitro [41], which is a matter of concern.
Legend: PEN-penicillin G, AMP-ampicillin, VAN-vancomycin, ERY-erythromycin, TETtetracycline, CIP-ciprofloxacin, NIT-nitrofurantoin, RIF-rifampicin, CHL-chloramphenicol, GEN-gentamicin, KAN-kanamycin, STR-streptomycin, OXA-oxacillin, CEF-ceftazidime, NAL-nalidixic acid, SXT-trimethoprim/sulfamethoxazole.   [42], less than 50% of coagulase-negative staphylococci isolated from spontaneously fermented meat products were resistant to the antimicrobials tested. Rebecchi et al. [43] found a high frequency of coagulase-negative staphylococci isolated from a typical Italian salami resistant to tetracycline and erythromycin. Pereira et al. [44] also studied 65 coagulasepositive staphylococci isolated from traditional Portuguese fermented meat products and despite the authors having found a higher percentage of penicillin and ampicillin resistant isolates, they also found percentages of isolates resistant to oxacillin, ciprofloxacin, erythromycin and vancomycin similar to those found in this study. Concerning Enterobacteriaceae (n = 33), high percentages of sensitive isolates were found to gentamicin (93.94%) and nalidixic acid (84.85%), but more than 30% of the isolates were resistant to ampicillin (36.36%) and tetracycline (33.33%) and intermediate resistant to chloramphenicol (33.33%). In addition to all the resistances observed, 21 out of 33 isolates were multi-resistant. During recent years several studies have reported the antimicrobial resistance of some genera of Enterobacteriaceae isolated from meat, such as Escherichia spp. and Salmonella spp. [45]. The scarcity of studies with antimicrobial resistances of Enterobacteriaceae isolated from fermented products turns difficult the comparison of the results obtained in the present study. However, despite the large percentage of susceptible isolates, the presence of antimicrobial resistance genes that can be transmitted to other microorganisms and the ability of Enterobacteriaceae to acquire new resistances must be taken into account.
It is possible to observe that isolates of Listeria spp. (n = 7) were sensitive to most of the antimicrobials tested. Only resistance to ampicillin (28.6% resistant), ciprofloxacin (14.3% intermediate resistant) and chloramphenicol (14.3% intermediate resistant) were observed. Also none of the isolates were multi-resistant. It is important to highlight that only isolates of L. innocua were resistant to ampicillin. In a study with 121 L. monocytogenes isolated from alheiras and/or their raw materials, the authors also found a low incidence of antimicrobial-resistant isolates [20].
Phenotypic expression of virulence factors gelatinase, DNase, haemolysis and biogenic amines of each group of isolates are presented in Table 3.
Only 10.2% of staphylococci isolates produced gelatinase, but DNase production was detected in 36.7% of isolates and also 57.1% of the isolates showed β-haemolytic activity. In the study of Pereira et al. [44], the authors found higher incidence of staphylococci isolates producers of gelatinase and β-haemolysis. This could be explained due to the fact that isolates in this study whereas in the study of Pereira et al. [44] only S. aureus were investigated.
Incidence of virulence factors tested among LAB and enterococci isolates was similar. About 10% produced gelatinase and only 4 LAB isolates (3.6%) produced DNase, but higher incidences were observed for haemolysin production with almost 50% of isolates showing β-haemolysis (48.6% of LAB and 52.6% of enterococci). Some studies with fermented products have been demonstrating similar results in terms of production of gelatinase [23,31,46] and DNase [14,23] for enterococci and other LAB. However, the same is not valid for haemolysin production, since the same authors reported very low incidences (nearly 0%) of β-haemolytic activities. In fact, the protein toxin cytolysin, which is responsible to β-haemolysis is the most important virulence factor recognized to enterococci genus [31] and the high incidence of β-haemolytic isolates found is a matter of concern. In contrast, low incidences of haemolysin production were found among Enterobacteriaceae isolates (6.1% β-haemolysis and 3.0% α-haemolysis), but 27.3% and 15.2% of isolates produced gelatinase and DNase, respectively. Although being virulence factors, gelatinase and DNase are tests used in identification of some genera/species of Enterobacteriaceae. Therefore, to our knowledge, does not seem to exist data about the presence of these factors in Enterobacteriaceae isolated from fermented sausages. However, in a study with Enterobacteriaceae associated with ready-to-eat fruits, the authors found high incidences of gelatinase (81.7%) and α-haemolysin (96.6%) producers [47].
Positive reactions were also found for all biogenic amines in the screening medium for LAB and enterococci isolates, except for tyramine in the case of LAB isolates. It has been reported that tyramine is frequently the most abundant biogenic amine found in fermented sausages [48] and produced mainly by LAB, including enterococci [49]. Although the low incidence, 7.9% of enterococci isolates produced tyramine. For isolates of Enterobacteriaceae only positive reactions for tyramine were observed (24.2%). Durlu-Özkaya et al. [50] also reported the ability of Enterobacteriaceae isolated from meat to produce tyramine, as well as putrescine, cadaverine and histamine.
Usually, putrescine and cadaverine are found in fermented products in lower levels than tyramine but some authors have been demonstrating the ability of several LAB in producing those two biogenic amines [46,49,51]. Histamine is rarely found in fermented sausages and its production appears to be restricted to some strains of a small number of Enterobacteriaceae or LAB isolates (reviewed by [52]).

Conclusions
New formulations of alheira were microbiologically characterized. Although lactic acid bacteria were the predominant microbiota, pathogenic bacteria were also found. This means that the same microbiological hazards found in the traditional alheira were also found in the new formulations. This is not surprising, since the same producers are producing the new products and the possibility of cross-contamination is high. It is still necessary to alert to safe cooking time/temperatures of these products, since some common cooking practices of alheira, may not guarantee a sufficient temperature in its center to eliminate the pathogens potentially present. Also several virulence factors were found among the different group of isolates tested, with a high incidence of isolates producing β-haemolysis. Along with their potential pathogenic activity, also several antimicrobial resistances were found being the majority of isolates classified as multi-resistant. Given that these products provide a perfect environment for contact between bacteria, the easiness of horizontally transmission of antimicrobial resistance genes, or virulence determinants, is a matter of concern.
Even taking into account that a larger number of products must by analyzed, the results obtained in this study should not be ignored and, moreover, should serve to alert consumers for a correct preparation of these products.