Raw milk cheeses from Beira Baixa, Portugal—A contributive study for the microbiological hygiene and safety assessment

Due to specific bacterial microbiota, raw milk cheeses have appreciated sensory properties. However, they may pose a threat to consumer safety due to potential pathogens presence. This study evaluated the microbiological contamination of 98 raw milk cheeses from Beira Baixa, Portugal. Presence and enumeration of Coagulase Positive Staphylococci (CPS), Listeria monocytogenes, Salmonella spp., pathogenic Escherichia coli, and indicator microorganisms (non-pathogenic E. coli and Listeria spp.) was attained. E. coli antimicrobial resistance (AMR) was also evaluated. PCR and/or Whole genome sequencing (WGS) was used to characterize E. coli, Salmonella spp. and L. monocytogenes isolates. Sixteen cheeses (16.3%) were classified as Satisfactory, 59 (60.2%) as Borderline and 23 (23.5%) as Unsatisfactory/Potential Injurious to Health. L. monocytogenes, CPS > 104 cfu g−1, Extraintestinal pathogenic E. coli (ExPEC) and Salmonella spp. were detected in 4.1%, 6.1%, 3.1% and 1.0% of the samples, respectively. Listeria innocua (4.1%) and E. coli > 104 cfu g−1 (16.3%) were also detected. AMR E. coli was detected in 23/98 (23.5%) of the cheese samples, of which two were multidrug resistant. WGS identified genotypes already associated to human disease and Listeria spp. cluster analysis indicated that cheese contamination might be related with noncompliance with Good Hygiene Practices during cheese production. Supplementary Information The online version contains supplementary material available at 10.1007/s42770-024-01332-y.


Introduction
Traditional cheeses are dairy products characteristic of a certain geographic region.These cheeses may be made from raw milk and the method of production is passed down from generation to generation.The microbiota of raw milk is complex, derived from many sources (e.g.microorganisms from teats, the farm environment, feedstuffs, as well as milking and processing equipment) and can have both positive and negative impacts on the cheese quality and shelf life and, consequently, on its economic potential.The diverse indigenous microbiota of raw milk cheeses provide fermentation, contribute to ripening and is responsible for the specific sensory properties of raw milk cheeses, namely a more intense and stronger flavor [1][2][3].Also, some authors suggest a beneficial link between the consumption of raw milk microbes, and protection against the development of asthma and atopy, later in life [4,5] and the reduction of the blood pressure in people with mild to moderate hypertension [6].However, raw dairy products can be contaminated with pathogens and consequently may have health-related implications leading to severe illnesses [7].In fact, estimates by the Center for Disease Control and Prevention, in the USA, based on data collected between 1993 and 2006, suggest that non pasteurized milk and milk products caused a disproportionate number of outbreaks and outbreak-associated illnesses relative to consumption of pasteurized products (≈150 × greater/unit of product consumed) [8].Furthermore, Costard et al., made a study on raw milk and cheese data from 2009 to 2014 and estimated that consumption of these raw products causes 840 times more illnesses and 45 times more hospitalizations than pasteurized products [9].
Microbial contamination during cheese-making, ripening and storage may occur directly or through cross contamination events, during processing, in retail or in domestic environments [10].Although the microbiological quality and safety of raw milk cheeses begin with milk, which may be the primary source of contamination [11], other sources of contamination can be present.Indeed, in the production environment, in food-contact surfaces and utensils that are not properly clean and sanitized some pathogenic microorganisms are able to adhere and persist forming biofilms [12].The hands of the workers, and the unsafe handling and inadequate storage practices in domestic environments [13] are also potential sources of cheese contamination.
According to the European Food Safety Authority (EFSA) [14], between 2018 and 2022, milk and milk products caused 149 strong evidence outbreaks, with 1754 human cases, 298 hospitalizations and 22 deaths, on Europe [14].Salmonella spp. was the leading causative agent being responsible for 27.52% of the outbreaks, Staphylococcus aureus toxins were associated with 15.44% and Shiga-toxin producing E. coli (STEC) was detected in 6.04% of these 149 outbreaks.L. monocytogenes, although only associated with 4.03% of the outbreaks, was responsible for a high number of deaths, 14 of the 22 reported [14].
Although raw milk cheeses are niche products on the global market, they constitute an important fraction of the cheese production in Portugal and are part of Portuguese cultural and gastronomic identity [15].The protection and promotion of these products is essential as they help to reduce rural depopulation, develop existing resources and generate employment opportunities, among other benefits.Some of the Portuguese raw milk cheese brands (n = 11) have a registration of Protected Designation of Origin (PDO), which means that they have special characteristics related to the geographical place in which they were produced/ processed [16].Beira Baixa is a historically important geographical area in terms of cheese production, and is one of the Portuguese demarcated regions for this artisanal activity.The geographical area of Beira Baixa cheeses production covers the municipalities of Castelo Branco district and part of the Santarém district, in the central region of Portugal.The type of rennet (of animal or of vegetal origin-Cynara Cardunculus) and the temperature, relative humidity and duration of the maturation period vary between brands.Beira Baixa cheeses with PDO label are "Castelo Branco", "Amarelo da Beira Baixa" and "Picante da Beira Baixa"."Castelo Branco" PDO labeled cheese is a semi-hard or semi-soft paste cured cheese with a yellowish color produced with ewe's raw milk; "Amarelo da Beira Baixa" is a semi-hard or semi-soft paste cured cheese with a yellowish color produced with ewe's or ewe's and goat's raw milk and "Picante da Beira Baixa" is a semi-hard to hard paste cured cheese with a greyish white color, produced with ewe's and goat's raw milk.Beira Baixa PDO labeled cheeses have a ripening period of at least 45 days.
The goal of this study was to contribute to the microbiological contamination assessment of cured raw milk cheeses produced in Beira Baixa region, Portugal.This assessment could potentially contribute to improve the implemented safety systems and consequently the quality of the cheeses in the studied region.

Sampling
In this study, a total of 98 cured raw milk cheeses from Beira Baixa region, Portugal, were analyzed, corresponding to 32 different brands produced by 9 producers (Supplementary Table 1).All samples came from different batches, and were purchased in several retail establishments, nearby Lisbon area, from October 2022 to June 2023.Samples were kept under refrigeration conditions (2 °C to 4 °C) from purchase until analysis, which occurred within a maximum of 24 h after purchase and within their assigned shelf-life period.

Microbiological analysis
ISO 7218: 2007 general requirements and guidance for microbiological examinations [17] were followed for all microbiological analysis.
The 98 cheese samples were analyzed for the presence and enumeration of some common foodborne pathogens such as Coagulase Positive Staphylococci (CPS), Listeria monocytogenes, Salmonella spp.and pathogenic Escherichia coli, as well as of indicator microorganisms (non-pathogenic E. coli and Listeria spp.other than L. monocytogenes).Staphylococcal enterotoxins (SE) detection was also assessed for those samples with a CPS concentration ≥ 4.9 × 10 4 cfu/g.

Salmonella spp. detection and E. coli and Coagulase Positive Staphylococci (CPS) detection and enumeration
Sample preparation was performed as followed: 25 g of each cheese sample was homogenized, at 230 rpm for 1 min using a stomacher (Stomacher, 400 Circulator, London, UK), in a sterile bag with 225 mL of Buffered peptone water (BPW-Oxoid, Basingstoke, Hampshire, UK), as described on ISO 6579-1: 2017 protocol [18].
For E. coli plating out, a loopful of the incubated nonselective pre-enrichment was streaked on the surface of Chromogenic Coliform Agar (CCA, Biokar Diagnostics) plates, and incubated at 37 °C during 24 h ± 2 h.Hemolytic activity of presumptive E. coli colonies was tested by sub-culture on Columbia Agar + 5% Sheep Blood (COS; bioMérieux) and incubation at 37 °C during 24 h ± 2 h.The presumptive E. coli and Salmonella spp.isolates were confirmed on VITEK®2 compact system (bioMérieux) and all positive isolates stored at -80 °C in broth with 20% glycerol.

Detection and enumeration of Listeria spp.
For L. monocytogenes detection, 25 g of each cheese sample was added to 225 ml of half Fraser broth (bioMérieux), as the primary enrichment culture, in a stomacher bag and were homogenized in a stomacher for 1 min and incubated for 24 ± 1 h at 30 °C.One hundred microliters of the primary enrichment culture, was then added to 10 mL of Fraser broth (bioMérieux), as a second enrichment culture, and incubated at 37 °C for 24 h ± 2 h [19].After incubation, 0.5 mL of the second enrichment culture was tested using the AFNOR validated VIDAS® LMO2 automated method (BIO12/11-03/04), according to the manufacturer's instructions.
L. monocytogenes enumeration was performed according to ISO/11290-2 horizontal method [20].Briefly, 10 g of each VIDAS® LMO2 positive sample was added to 90 mL of BPW and homogenised (Initial Suspension).L. monocytogenes presumptive colonies were counted after the spread of 1 mL of the Initial Suspension on the surface of Microinstant® Listeria Agar (Ottaviani e Agosti) (Biokar Diagnostics) plates and incubation at 37 °C for 48 h ± 2 h.L. monocytogenes presumptive colonies (blue colored surrounded by an opaque halo) as well as Listeria spp., other than L. monocytogenes (blue colonies without an opaque halo) were isolated on Columbia Agar + 5% Sheep Blood (COS; bioMérieux) at 37 °C for 24 h ± 2 h, where hemolytic activity was assessed.Confirmation of the identification of the isolates was attained on VITEK®2 compact system (bioMérieux), following the manufacturer's instructions.All positive isolates were stored in Tryptone Soy Broth (TSB; Biokar Diagnostics) with 20% glycerol, at -80 °C.

Detection of Staphylococcal Enterotoxins (SE)
All cheese samples presenting Coagulase Positive Staphylococci levels ≥ 4.9 × 10 4 cfu/g were tested for the presence of staphylococcal enterotoxins (SE), as described on ISO 19020:2017 [21].In summary, the first step consisted in an extraction where toxin diffusion was attained by adding 25 g of each tested cheese sample to 40 mL of distilled water at 37 °C ± 1 °C and, after homogenization, the mixture was shaken for 30-60 min at room temperature in an VXR basic Vibrax orbital shaker (Ika®,Staufen, Germany).The pH was adjusted, the mixture centrifuged and then the resulting supernatant was concentrated by dialysis with a 6000-8000 Da molecular cut-off membrane (Spectrum Laboratories,Rancho Dominguez, CA, USA) against 30% (w/v) of polyethylene glycol 20,000 (Merck,Darmstadt, Germany), overnight, at 4 °C.Finally, immunoenzimatic detection was performed using the automated method VIDAS® Staph enterotoxin II (SET 2) (bioMérieux).

Microbiological results interpretation and statistical analysis
Microbiological results were interpreted according to the criteria showed on Table 1.The used criteria was based on the National Institute of Health Doutor Ricardo Jorge (INSA) guidelines for the interpretation of microbiological assays [22], the Commission Regulation (EC) Nº 2073/2005 [23], the Health Protection Agency-Guidelines for assessing the Microbiological Safety of Ready-to-eat Foods Placed on the market [24] and on the Luxembourg Microbiological criteria applicable to foodstuffs [25].
A cheese brand obtained a Satisfactory microbiological quality classification when all the samples of that brand were classified as Satisfactory for all the tested parameters.Borderline classification was attained when at least one of the tested samples was classified as Borderline, in at least one of the tested parameters, and none of them was classified as Unsatisfactory, in any of the tested parameters.Unsatisfactory/Potential Injurious to Health (U/PIH) classification was attributed when at least one sample was classified as Unsatisfactory/Potentially Injurious to Health, in at least one of the tested parameters.
Fisher-Freeman-Halton's exact test of independence, with a 99.0% degree of confidence, was performed, using IBM SPSS Statistics 27.0.1 software, to determine if there was a significant relationship between the independent variables "microbiological safety classification" and the type of milk used in cheese production ("presence of cow's milk", "presence of ewe's milk", "presence of goat's milk") and between "microbiological safety classification" and registration of "Protected Designation of Origin (PDO)".

Antimicrobial susceptibility testing of E. coli and Salmonella spp. isolates, Salmonella spp. serotyping and pathogenic E. coli identification
Since the use of antimicrobial agents in animal farming is considered as one of the most critical factors that contribute to the emergence and dissemination of antibiotic resistant bacteria, and because E. coli is considered a potential indicator of antimicrobial resistance (AMR), E. coli isolates (pathogenic and non-pathogenic) AMR was evaluated.
An isolate was considered as multidrug resistant (MDR) when presenting resistance to three or more antimicrobial classes [27].
Salmonella spp. was serotyped by the slide agglutination method for O and H antigens (SSI, Copenhagen, Denmark), according to the Kauffmann-White-Le Minor scheme [28].
The detection of at least one of the pathotype-specific genes in each isolate allowed the classification of potentially pathogenic (STEC, Shiga Toxin-producing E. coli; EAEC, Enteroaggregative E. coli; EPEC, Enteropathogenic E. coli; ETEC, Enterotoxigenic E. coli; EIEC, Enteroinvasive E. coli).Furthermore, E. coli pathogenicity was also inferred by Whole Genome Sequencing (WGS), since some of the likely nonpathogenic E. coli isolates were MDR and/or hemolytic, and were classified as Extraintestinal pathogenic E. coli (ExPEC).In these cases, the presence of two or more typical virulence genes allowed the ExPEC classification [30].

Listeria spp., Salmonella spp. and E. coli whole-genome sequencing, in Silico typing and screening of E. coli virulence/AMR genes
Extraction of genomic DNA from all MDR and hemolytic E. coli as well as from all Listeria innocua, L. monocytogenes and Salmonella spp.isolates was achieved using the ISO-LATE II Genomic DNA Kit (Bioline, London, England, UK).Extracted DNA was quantified, with the dsDNA HS Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA), in the Qubit fluorometer (Invitrogen, Waltham, MA, USA), according to the manufacturer's instructions.NexteraXT library preparation protocol (Illumina, San Diego, CA, USA) was used for DNA preparation for sequencing.Cluster generation and sequencing (2 × 150 bp) were performed on either a MiSeq, a NextSeq 550 or NextSeq 2000 instrument (Illumina).
Sequencing reads were deposited on the European Nucleotide Archive (ENA) under the bioprojects PRJEB31216 (Listeria spp.), PRJEB54735 (E.coli) and PRJEB32515 (Salmonella spp.), as well as on EFSA WGS portal.Supplementary Table 1 presents the accession numbers for each isolate.
For L. innocua, in the absence of a cgMLST schema, a core-genome alignment of the INNUca polished assemblies was constructed with Parsnp v.1.7.4 implemented on Harvest suite [43], using the default parameters, with exception of parameter -C, which was adjusted to 2000 in order to maximize the resolution.The core-genome SNP-based clustering analysis was performed with ReporTree v.2.0.3 (https:// github.com/ insap athog enomi cs/ Repor Tree) [40] using Gra-peTree (MSTreeV2 method) [41], with clusters of closely related isolates being determined and characterized at a SNP thresholds of 1, 4, 7 and 15 SNPs.This core-genome SNPbased clustering analysis relied on a core-genome alignment (comprising 95% of the L. innocua genome size) involving a total 33,081 variant sites.Interactive phylogenetic tree visualization was conducted with GrapeTree [41].
When looking at the obtained results considering the type of milk used in cheese production, it was possible to find a statistical significant association between microbiological safety classification and the presence of cow's milk (Fisher-Freeman-Halton = 11.785,p = 0.001) (Table 2).There was a clear decrease in the number of Satisfactory samples and an increase of Unsatisfactory/Potentially Injurious to Health samples, when cheese samples contained cow's milk (Table 2).This association occurred also for microbiological safety classification regarding Coagulase Positive Staphylococci and E. coli enumeration and the presence of cow's milk in the cheese samples (Table 2).There was no statistical significant association between microbiological safety classification and the presence of ewe's or goat's milk, nor statistical significant association between microbiological safety classification and registration of Protected Designation of Origin (Table 2).There was also no statistical significant association between microbiological safety classification of the cheese samples and the number of types of milk used (Fisher-Freeman-Halton = 10.250,p = 0.025) (data not shown).

Genotypic and phenotypic characterization of E. coli, Salmonella spp. and Listeria spp. isolates
E. coli isolates were recovered from 91 out of the 98 samples.The remaining seven samples either did not contain E. coli or it was not possible to isolate E. coli, due to high levels of contamination with other microorganisms.
Salmonella spp.strain, isolated from one of the cheeses, belongs to serotype Duisburg, and ST4046.Although containing one gene and one point mutation associated with antimicrobial resistance (aac(6')-Iaa; parC:p.T57S), this isolate was phenotypically susceptible to all the tested antimicrobials.
The four Listeria monocytogenes isolates belong to the following STs: ST1 (isolated from one of the 6B samples), ST5 (isolated from two samples of brand 1B) and ST7 (isolated from one of the 9A samples).The Listeria innocua isolates belong to ST492 (two of the 2G and one of the 2F samples) and ST603 (isolated from one of the 1C samples).

Core-genome clustering analysis of Listeria spp. isolates
Despite the low number of isolates (four per species), coregenome clustering analysis of L. monocytogenes and L. innocua isolates was performed in order to assess the genetic relatedness among them and its potential correlation with cheese producer/brands (Fig. 2).
Fig. 1 Microbiological results of the 98 tested cheeses, by producer (N = 9; 1-9) and brand (N = 32; 1A-9A).In the three first columns of the heatmap-like figure, the code Blue for Yes and Grey for No was used to inform about the existence of a Protected Designation of Origin (PDO) for each sample, the isolation of an E. coli with antimicrobial resistance (AMR) to at least one of the tested antimicrobials, and the detection of an hemolytic (Hemo) E. coli.The Green for Satisfactory; Yellow for Borderline and Red for Unsatisfactory/ Potential Injurious to Health code was used to characterize the Microbiological Safety classification of each of the 98 samples regarding the tested parameters (E.coli, Coagulase Positive Staphylococci (CPS), pathogenic (pathog) E. coli; Salmonella spp.(Salm) and Listeria spp.(List).In the last two columns of the heatmap, the global Microbiological Safety classification of each sample and brand are presented ◂

Discussion
Cheeses may have physicochemical characteristics and shelf life favorable to the growth or survival of pathogenic bacteria [44,45].For instance, a pooled prevalence of Listeria monocytogenes (3.6%-sheep; 12.8%-goat), Salmonella spp.(5.9%-goat), Staphylococcus aureus (16.0%-goat) and STEC (2.8%-sheep; 4.3%-goat) in sheep and goat milk cheeses has been reported in a recent meta-analysis approach [46], which clearly demonstrates that these microorganisms are frequently detected in this type of foodstuffs.In particular, the production of cheeses free from L. monocytogenes may be a challenge for producers, since this pathogen is ubiquitous in nature, can grow at refrigeration temperatures, and forms biofilms, being difficult to eradicate from food contact surfaces in  the production environment.In fact, many outbreaks associated with the consumption of cheeses have been caused by L. monocytogenes [47][48][49][50][51].
In this study, we have detected L. monocytogenes in four samples (4/98 = 4.1%) (3 with concentrations > 10 2 cfu/g), and further found the presence of the indicator microorganism L. innocua in another four (4/98 = 4.1%).These results emphasize the need for monitoring good hygiene and manufacturing practices, as well as raw milk microbiological quality, in the context of Regulation (EC) N°852/2004 on the hygiene of foodstuffs [52], in the sector of Portuguese artisanal cheese industry.
The three identified Listeria monocytogenes sequence types (ST1, ST5, ST7) were already isolated, by others, in foodstuffs and associated to human disease [38,53,54].All the Listeria monocytogenes detected STs were also already associated to Dairy Processing Facilities [55][56][57], with ST5 being described as a particularly persistent ST that might harbor specific efflux pump systems and heavy metal resistance genes that may possibly provide a higher tolerance to disinfectants [55].
The cgMLST analysis of L. monocytogenes isolates and SNP-analysis of L. innocua isolates suggest that, in this study, cheese contamination is potentially related with noncompliance with Good Hygiene Practices at producer level, since the two observed clusters (one for L. monocytogenes and one for L. innocua) are producer specific.
S. aureus is a common cause of bovine mastitis and is frequently detected in raw milk [58,59].In fact, in this study, we have found a statistically significant association between the microbiological safety classification of the samples regarding CPS enumeration and the use of cow's milk in cheese production (Fisher-Freeman-Halton = 10.624,p = 0.004) (Table 2).Furthermore, enterotoxin-producing S. aureus exhibits a high osmotolerance, growing or surviving in cheeses with a w levels as low as 0.86 (equivalent to about 20% NaCl), provided that all other conditions are optimal [60].Thus, this bacterium may be a safety problem in raw milk cheeses, being already described as implicated in several outbreaks linked with milk and milk products [61].In fact, in the present study, the detection of coagulase-positive Staphylococci > 10 4 cfu/g in 6.1% (6/98) of the samples highlights for the importance of implementation of more efficient hygiene control measures during the process of cheese production and adequate selection of the raw materials.High counts of Staphylococci are also the main nonconformities found in Brazilian artisanal cheeses [62].
In this research, Salmonella spp. was detected in one sample (1/98 = 1.0%).Serotype Duisburg, identified in the Salmonella spp.isolate, was recently associated to a foodborne outbreak that occurred in the United States of America in 2021, involving the consumption of cashew brie, a vegan alternative to Brie cheese [70].
Although no STEC isolates were found in the analyzed cheeses, we have detected the presence of Extraintestinal pathogenic E. coli (ExPEC) in three samples (3/98 = 3.1%), in accordance with our previous study performed on Alentejo raw milk cheeses [71].ExPEC is recognized as the most common gram-negative pathogen in humans, causes most urinary tract infections (UTIs) in young healthy women, is the leading cause of bacteremia in adults, and is the second most common cause of neonatal meningitis [72].This pathotype was also already detected in raw milk cheeses from various Latin America countries [73][74][75], and may potentially be transmitted to humans via food [76].We have further found the presence of non-pathogenic E. coli > 10 4 cfu/g in 16.3% of the samples (16/98), highlighting for potential non-conformities during processing along cheese production chain.ExPEC isolates were characterized as belonging to STs 14755, 155, and 362.E. coli ST155 and ST362 are highly widespread in nature, and have been isolated from wildlife, livestock, water, foodstuffs and humans, among other sources (Enterobase, https:// enter obase.warwi ck.ac.uk/ speci es/ index/ ecoli).E. coli ST155 has been described as a potential foodborne pathogen [77,78].E. coli ST 362 is known as a biofilm-producer [79] and was the dominant ST found in the calf ESBL-producing E. coli isolates from a dairy farm in Germany [80].The ST14755, a novel sequence-type, was detected in a susceptible ExPEC hemolytic isolate, and belongs to serotype O175:H16.
Drug resistant microorganisms have become a threat to public health, with acute and chronic infections increasingly failing to respond to antimicrobial drugs.It is estimated that 700,000 people die because of resistant infections every year, and that by the year of 2050, 10 million deaths/ year can occur.This will result in a cumulative global economic loss of USD 100 trillion, due to the antimicrobial resistance effect [81].It is well known that the unrestrained use of antibiotics to control disease in farm animals and in humans increases the number and frequency of antibiotic resistant bacterial isolates [82,83].In fact, in this study, 23.5% of the tested cheese samples presented E. coli isolates resistant to antimicrobials, with two of these strains being MDR.Antimicrobial resistance gene content of unprocessed animal products may potentially play a role in the acquisition of antimicrobial resistance of human pathogens [84].
The link of the above-mentioned pathogens with foodborne outbreaks related to cheese consumption, as well as the association between enhanced virulence and bacteria stress responses to cheese manufacturing, cheese matrix itself and storage conditions [85], highlights for the importance of the development of new instruments to control the presence of these bacteria during the process and, consequently, in the ready-to-eat end-product.Microbiological safety assessment of raw milk cheeses placed on the market is one of these instruments and may be essential to identify critical points along cheese production and to implement timely corrective measures, that may allow artisan cheesemakers to continue to produce their appreciated artisanal cheeses but with improved safety standards.
The results of this study clearly demonstrate that raw milk cheeses from Beira Baixa region, Portugal, may contain pathogenic and/or indicator microorganisms above the stipulated guideline limits, being some resistant to antimicrobials.It is essential not only to monitor the hygiene at the primary production (farm level), but also to implement corrective measures during cheese processing, to ensure control at the established critical points.In this study, we have found a statistical significant association between the microbiological safety classification of the samples and the use of cow's milk in cheese production (Fisher-Freeman-Halton = 11.785,p = 0.001) (Table 2).This association also occurs when considering the microbiological safety classification of the samples regarding CPS and E. coli enumeration (Table 2), the two parameters that mostly contributed for the U/PIH classification of the samples.This result may point out for a potential safety critical point in Beira Baixa cheese production related with the hygiene at the cow's milk production farms, and is a good example of how the microbiological safety assessment of the final product may be important to identify potential critical points along production.
The implementation and maintenance of training for the application of good manufacturing practices, and the constant monitoring of the quality of the production process in the Portuguese artisanal cheese industry, is of critical importance.To highlight also the importance of controlling the quality of the milk and to test environmental surfaces like food contact surfaces (e. g. cheese-making equipment, utensils, shelves for maturation/ripening) or non-food contact surfaces for preventing cross contamination with microbiological hazard thus improving the quality and safety of the final product.

Fig. 2
Fig.2Core-genome clustering analysis of Listeria spp.isolates.(A) For L. monocytogenes (four isolates), the Minimum Spanning Tree (MST) was constructed based on the cgMLST 1748-loci Pasteur schema[38].Each circle (node) contains the brand code and represents a unique allelic profile, with numbers on the connecting lines representing allelic distances (AD) between nodes.Straight and dotted lines reflect nodes linked with ADs below and above a threshold of seven ADs, which can provide a proxy to the identification of genetic clusters with potential epidemiological concordance[42].(B) For L. innocua (four isolates), the MST was constructed based