Recent developments in the use of viability dyes and quantitative PCR in the food microbiology field
The increase in foodborne outbreaks highlights the need for rapid, sensitive and specific methods for food safety monitoring, enabling specific detection and quantification of viable foodborne pathogens. Real-time PCR (qPCR) combined with the use of viability dyes, recently introduced, fulfils all these requirements. The strategy relies on the use of DNA-binding molecules such as propidium monoazide (PMA) or ethidium monoazide (EMA) as sample pretreatment previous to the qPCR. These molecules permeate only membrane-compromised cells and have successfully been applied for different types of foodborne pathogens, including bacteria and viruses. Moreover, those dyes have been explored to monitor different food manufacturing processes as an alternative to classical cultural methods. In this review, state-of-the-art information regarding viability PCR (v-PCR) is compiled.
Provide feedback or get help You are viewing our new enhanced HTML article.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Foodborne diseases are a significant and widespread global public health threat. World Health Organization (WHO) cites food safety as one of the top 11 priorities and challenges of this century estimating that foodborne and waterborne diseases taken together kill about 2·2 million people annually, 1·9 million of them children. In the EU, more than 320 000 human cases are reported each year, but the real number is likely to be much higher. According to the European Food Safety Authority (EFSA) in 2011 (EFSA 2013), most of the 5648 reported foodborne outbreaks were caused by Salmonella, Campylobacter, bacterial toxins and enteric viruses, and the main food sources were eggs, mixed or buffet meals and vegetables. In 2011, the Centers for Disease Control and Prevention (CDC) issued new figures for the incidence of foodborne illness, estimating that about 48 million people in the USA suffer from foodborne illnesses each year, resulting in 128 000 hospitalizations and 3000 deaths. Moreover, the cost of foodborne illness in this country is now estimated to be up to $77·7 billion a year, according to the analysis of the Ohio State University (Scharff 2012).
At present, not only bacterial pathogens are relevant in food safety. The importance of foodborne diseases caused by enteric viruses is increasingly being recognized, and the World Health Organization has found that there is an upward trend in their incidence. This recognition is reflected by the attention that national and international organizations give to consider the control of foodborne viral infection in the report of the Advisory Committee on the Microbiological Safety of Food (Food 1998), the recent proposed guidelines on the application of food hygiene to the control of viruses for Codex Alimentarius (CX/FH/10/42/5), the scientific opinion of the EFSA (Hazards 2011) and the expert advice on foodborne viruses for Codex Alimentarius (www.who.int/foodsafety/publications/micro/mra13/en/index.html).
With all these data, the increase in public attention to food safety has led to an increase in research into new and rapid methods for pathogen detection in food products. A wide range of methods are available that differ enormously according to the type of micro-organism and food matrix. Traditionally, foodborne pathogenic bacteria are monitored and characterized by food laboratories using standard culture methods. They are based on traditional microbiological methods described in the Bacteriological Analytical Manual (BAM; http://www.cfsan.fda.gov/~ebam) which include the use of selective media, biochemical confirmation and other parameters for bacterial identification (e.g. ISO 16654:2001, ISO 11290 and ISO 6579). Culture methods require 4–5 days to obtain presumptive positive or negative results and can take up to 7 days, depending on the biochemical and serological confirmations. Moreover, these tests are based on an enrichment step, which complicates experimental techniques for obtaining quantitative data (Brehm-Stecher et al. 2009). Altogether, the culture-based approach makes the routine analyses of food samples or outbreak investigations very expensive and time-consuming. Therefore, rapid and accurate methods to detect and identify bacterial pathogens in food products are of great value to the food industry (Perelle et al. 2007). As an alternative to culture methods, molecular techniques have revolutionized microbial detection. DNA-based methods such as polymerase chain reaction (PCR) and real-time PCR (qPCR) are rapid, versatile, sensitive, precise and allow both specific and quantitative detection of micro-organisms from a variety of origins (Malorny et al. 2003). Moreover, these techniques enable detection of subdominant bacterial populations, even in the absence of selective enrichment medium and in the presence of other dominant populations (Postollec et al. 2011).
However, PCR methods still face up some drawbacks such as the inhibition of amplification by substances naturally found in the food matrix or the inability of the PCR itself to discriminate between viable and dead cells, resulting in an overestimation of the target micro-organism. Inhibition problems may be solved or improved by the optimization of sample preparation (Brehm-Stecher et al. 2009) and by incorporation of internal amplification controls (Rodríguez-Lázaro et al. 2007). Nevertheless, detection of only viable micro-organisms is still considered a major disadvantage when using PCR or qPCR (Rudi et al. 2002; Wang and Levin 2006). This drawback is especially relevant when the aim of the analysis is the quantification of the micro-organisms. In these cases, samples are processed and analysed without a prior enrichment step, and DNA from dead cells, killed by processing procedures or other factors, can interfere in the PCR amplification (Nogva et al. 2003). It is particularly significant for processed food or foods subjected to long-time storage due to the relatively long persistence of DNA after cell death resulting in false-positive results. In the past, the most commonly used strategy to avoid this overestimation relied on the detection of mRNA, which is a direct indicator of active bacterial metabolism. Thus, reverse transcription (RT)-PCR and nucleic acid based sequence amplification (NASBA) were applied on different micro-organisms and food matrices (Morin et al. 2004). However, due to RNA instability, degradation can occur during sample handling and storage giving false-negative results.
During the last decade, another approach for the detection of viable cells has been introduced consisting in coupling PCR or qPCR with nucleic acid intercalating dyes such as propidium monoazide (PMA) or ethidium monoazide (EMA) (Nogva et al. 2003; Nocker et al. 2006), namely viability PCR (v-PCR). This procedure is based on the integrity of bacterial cells as viability dyes penetrate only into compromised membrane cells. Once inside the cell, the dye intercalates covalently into the DNA after exposure to strong visible light, interfering DNA amplification. The principle is not only limited to bacteria, and it has also successfully been applied to viruses, fungi and protozoa. Moreover, there are other nucleic acid intercalating dyes such as reagent D (commercially available reagent from Biotecon) which has been applied for the detection of Enterobacteria in infant formula. This reagent contains a light sensitive substance which can only penetrate the cell membranes of dead cells. Exposure to visible light leads to covalent binding of this substance to DNA and prevents the DNA from being amplified via PCR. Although reagent D successfully proved to inhibit signal from dead Enterobacteria, other studies have demonstrated that it bounds to DNA from viable cells as well as dead cells in Gram-positive bacteria such as Listeria monocytogenes (Elizaquível et al. 2012a; Martinon et al. 2012).
Recently, Fittipaldi et al. (2012) reviewed the use of PMA and EMA as intercalating dyes. Although the two dyes behave nearly identically as intercalating stains, they differ in regard to their permeation through cell membranes, concluding that EMA, due to its chemical composition, is slightly more efficient in signal suppression than PMA, but PMA is more effective than EMA in terms of live–dead discrimination. These authors have extensively reviewed different technical aspects when employing viability dyes, such as dye concentration, type of micro-organism, concentration of the micro-organism, the ratio between live and dead cells, the length of PCR product, the turbidity, pH and salt concentration of the sample, type of the light source, and time and temperature of the activation process with visible light.
The purpose of the present review is to consolidate information on various aspects of the application of these viability dyes for the selective detection and quantification of foodborne pathogens and other relevant micro-organisms in the food industry. Moreover, it also summarizes the challenges that can be found when applying v-PCR to real food samples submitted to different food manufacturing processes.
Detection of relevant food bacteria using viability dyes and molecular techniques
Detection and quantification of viable bacterial foodborne pathogens is essential to identify their contribution to public health. Classical growth-based methods can underestimate the actual counts due to the presence of viable but nonculturable cells (VBNC) which can maintain metabolic and virulent activity. This drawback is also present when applying alternative culture-independent molecular methods, such as PCR, which may detect not only VBNC bacteria but also DNA from dead cells. To overcome this drawback, PCR has been combined with viability dyes to selectively detect and quantify viable foodborne pathogens. The selective analysis of DNA from live foodborne bacterial pathogens by v-PCR was first described for the detection of Escherichia coli O157:H7, Listeria monocytogenes and Salmonella using EMA (Nogva et al. 2003). After heat and chemical treatments, EMA-PCR was able to discriminate between live and dead cells which made the use of EMA a promising viability detection method. Since then, EMA-PCR has been extensively applied for the detection of viable bacteria both in pure cultures (Nogva et al. 2003; Chang et al. 2009; He and Chen 2010) and in complex food matrices (Rudi et al. 2005a,b; Wang and Mustapha 2010; Soejima et al. 2012a,b) (Table 1). However, latter studies demonstrated that EMA penetrates viable cells when applied for E. coli O157:H7 detection (Nocker and Camper 2006) and also for other bacterial species (Nocker et al. 2006) resulting in partial DNA loss and causing underestimation of live microbial counts (Nocker et al. 2006b; Cawthorn and Witthuhn 2008; Levin and Lee 2009). As an alternative, PMA was evaluated to substitute EMA. PMA proved to be more selective due to the higher charge of the molecule (Nocker et al. 2006) which penetrates more specifically into membrane-compromised dead cells. PMA treatment combined with qPCR has been successfully tested on several bacterial foodborne pathogens such as L. monocytogenes (Pan and Breidt 2007), E. coli O157:H7 (Nocker et al. 2009; Elizaquível et al. 2012a), Campylobacter jejuni (Josefsen et al. 2010), Bacillus subtilis (Kim and Ko 2012) or B. subtilis spores (Rawsthorne et al. 2009) and even prior to a multiplex PCR method that successfully detects three different Salmonella serotypes (Yang et al. 2012).
Table 1. Application of PMA/EMA for the detection of viable bacteria and yeast in different types of food matrices
|Chicken breasts and legs|| Campylobacter jejuni ||EMA||Rudi et al. (2005a,b)|
|Chicken carcasses|| Campylobacter ||PMA||Josefsen et al. (2010)|
|Chicken and eggs|| Salmonella ||EMA||Wang and Mustapha (2010)|
|Chicken, beef and ham||Salmonella enterica serovars||PMA||Yang et al. (2012)|
|Cooked ham|| Salmonella ||PMA||Martin et al. (2013)|
|Vegetables and juices|
|Spinach and mixed salad||Escherichia coli O157:H7, Salmonella and Listeria monocytogenes||PMA||Elizaquível et al. (2012a)|
|Lettuce|| Salmonella ||PMA||Liang et al. (2011)|
|Lettuce, tomato and ground beef||Salmonella typhimurium, E. coli O157:H7 and L. monocytogenes||PMA||Yang et al. (2013)|
|Lettuce and spinach||E. coli O157:H7||PMA||Dinu and Bach (2013)|
|Iceberg lettuce and soya sprouts||E. coli O157:H7||PMA||Elizaquível et al. (2012a)|
|Cantaloupe, spinach, and tomato|| Salmonella ||PMA||Chen et al. (2011)|
|Tomato||Salm. enterica serovars||PMA||Yang et al. (2012)|
|Tomato|| Alternaria ||PMA||Crespo-Sempere et al. (2013)|
|Apple, orange or grape juices|| Zygosaccharomyces bailii ||EMA||Rawsthorne et al. (2009)|
|Yoghurt|| Bifidobacterium ||EMA||Meng et al. (2010)|
|Pasteurized milk||Total coliforms||EMA||Soejima et al. (2012a)|
|Pasteurized milk|| Enterobacteriaceae ||PMA||Soejima et al. (2012b)|
|Gouda cheese|| L. monocytogenes ||EMA||Rudi et al. (2005a,b)|
|Powdered infant formula|| Cronobacter sakazakii ||EMA||Minami et al. (2012)|
|Salmon|| Photobacterium phosphoreum ||PMA||Macé et al. (2013)|
|Shrimp and salmon|| Brochotrix thermosphacta ||PMA||Mamlouk et al. (2012)|
|Must and wine|| Oenococcus oeni ||PMA||Vendrame et al. (2013)|
|Food processing surfaces|| E. coli, Staphylococcus aureus, L. monocytogenes ||PMA||Martinon et al. (2012)|
However, the difference between bacterial genera may have a great impact on the effect of viability dyes, especially the cell envelope structure differences between Gram-negative and Gram-positive bacteria. In this sense, Elizaquível et al. (2012a) found that PMA-qPCR successfully quantified viable Salmonella and E. coli O157:H7. However, residual detection signals of dead cells were obtained when quantifying L. monocytogenes which led to an overestimation of viable L. monocytogenes population. Similar results were obtained for heat-treated Listeria innocua by Lovdal et al. (2011). In Gram-negative bacteria, the complex structure of the outer membrane represents the major permeability barrier, while the corresponding barrier in Gram-positive is the peptidoglycan layer (Nogva et al. 2003). This difference facilitates the PMA penetration in membrane-compromised dead Gram-negative bacteria and could explain the better performance of PMA in the differentiation of live and membrane-compromised dead Gram-negative bacteria.
The use of viability dyes has been extensively applied for pathogen detection in pure cultures, and the method has already been evaluated in several food matrices (Table 2).
Table 2. Application of viability dyes for the detection of infectious viruses
|Coxsakievirus, Echovirus||Heat treatment||PMA||Yes||Parshionikar et al. (2010)|
|HAV||Heat treatment||PMA||Yes||Sanchez et al. (2012)|
|High pressure processing||Partially|| |
|Norovirus||Heat treatment||PMA||Yes by RT-PCR but not by RT-qPCR||Parshionikar et al. (2010)|
|Poliovirus||Heat treatment||PMA/EMA||Yes||Parshionikar et al. (2010) and Kim et al. (2011)|
|Avian influenza||Stability in water||PMA||No||Graiver et al. (2010)|
|MS2 phage||Heat treatment||PMA||Yes||Kim and Ko (2012)|
|Murine norovirus||Heat treatment||PMA||No||Kim and Ko (2012)|
|EMA||Yes||Kim et al. (2011)|
|T4 phages||Heat treatment||PMA||No||Fittipaldi et al. (2010)|
Among food products, milk, dairy products, eggs and meat are the more habitually related to foodborne outbreaks (EFSA 2013). In this sense, Soejima et al. (2012a) successfully applied PMA-PCR in milk allowing the detection of low levels (≥1 CFU ml−1) of viable total coliforms. In a latter study, Soejima et al. (2012b) developed a PMA-qPCR assay that successfully detected low numbers of viable Enterobacteriaceae (5–10 CFU ml−1) in pasteurized milk. Detection of Camp. jejuni and Salmonella has been tested in chicken and beef by using both PMA (Josefsen et al. 2010; Yang et al. 2012) and EMA (Rudi et al. 2005a,b; Wang and Mustapha 2010), and in all these studies, the use of viability dyes allowed the detection of viable pathogens without any interference from the natural microbiota of these complex food products. More recently, Martin et al. (2013) approached the detection of Salmonella in cooked ham by testing three different specific PCR targets differing in length (95, 285 and 417 bp). Results showed that the inhibition effect was dependent on the PCR amplification product length, and only the longer product achieved the suppression of dead cells PCR signals.
Fruit and vegetables, in particular leafy green vegetables, are increasingly being recognized as important vehicles for the transmission of human pathogens (Lynch et al. 2009). Thus, in the last years, there has been an increasing number of papers related to the use of viability dyes for pathogen detection in this type of food products (Chen et al. 2011; Elizaquível et al. 2012a; Yang et al. 2012, 2013). The application of viability dyes on vegetables for the selective detection of live foodborne bacterial pathogens by qPCR was first described for E. coli O157:H7, L. monocytogenes and Salmonella using PMA in spinach and mixed salad (Elizaquível et al. 2012a). In this study, only residual detection signals were obtained in mixed salad when inoculated with high amounts of dead L. monocytogenes cells, above 105 CFU g−1. Liang et al. (2011) developed an approach to detect viable Salmonella in lettuce by using PMA-qPCR after a heat treatment. They observed that PMA treatment can effectively prevent PCR amplification from dead Salmonella cells in lettuce at 103 CFU g−1. Moreover, they combined the PMA-qPCR with a prior 12-h enrichment allowing the detection of as low as 10 CFU g−1. In another study, Chen et al. (2011) successfully detected and quantified viable Salmonella cells in produce by coupling PMA treatment with loop-mediated isothermal amplification (PMA-LAMP). All the above-mentioned studies were performed inoculating dead pathogens into the food matrices. In a more recent study, Dinu and Bach (2013) applied a PMA-qPCR assay in lettuce and spinach with a background of added dead pathogens and also VBNC E. coli O157:H7 with a detection limit of 103 CFU g−1 and a linear quantitative detection range of 5 log.
Spoilage bacteria include a variety of species that cause food deterioration and develop unpleasant odours, tastes and textures. Food spoilage is a complex process which can be produced by a variety of species that cause food deterioration and develop unpleasant odours, tastes and textures in food products. The heterogeneity of raw materials, processing conditions and natural microbiota present in different food matrices complicate the determination of the actual micro-organism which causes the spoilage. In milk, thermophilic bacilli have significant economic consequences when they exceed specification limits and may result in downgrading of the product. Rueckert et al. (2005) evaluated the use of EMA-qPCR to detect and enumerate viable vegetative cells and spores of thermophilic bacilli in milk powder. Very recently, Cattani et al. (2013) have developed a seminested touchdown PCR assay combined with PMA treatment for the detection of viable Bacillus sporothermodurans vegetative cells. B. sporothermodurans is a Gram-positive, aerobic and mesophilic bacterium, which is characterized by the production of spores that are highly heat-resistant and capable of surviving industrial ultra-high temperature (UHT) milk processing (140°C for 4 s). The described method was shown to amplify DNA specifically from viable cells and presented a detection limit of 102 CFU ml−1 in UHT milk. This detection limit meets the criteria of the European Union (EU) and Brazilian legislation for the maximum count of mesophilic micro-organisms in UHT milk.
Brochothrix thermosphacta is a Gram-positive bacterium that predominates the spoilage microbiota of modified atmosphere packaging (MAP) meat and fish. As an alternative to the traditional methods currently used to detect B. thermosphacta in foods, Mamlouk et al. (2012a) have successfully combined PMA and SYBR Green-based qPCR which allows the selective detection and quantification of live cells on naturally contaminated shrimp and salmon. Photobacterium phosphoreum has been identified as the main spoilage bacterial species of several chilled marine fish, and very recently Macé et al. (2013) have develop a PMA-qPCR assay for its detection in salmon steaks.
Detection and quantification of viable bacterial cells is essential, as well, in some food products or food processes as a food quality parameter. That is the case of probiotic food. Probiotics are defined as ‘living micro-organisms which, when administered in adequate amounts, confer health benefits on the host’ (FAO/WHO 2002). They can be incorporated into a food matrix which becomes a functional food. The basic requirement for probiotic bacteria to be able to exert expected positive effects is to be alive; therefore, appropriate quantification methods are crucial. Probiotic viability is usually monitored by traditional culture-based methods. However, as stated above, these methods are time-consuming and can underestimate microbial counts of live cells. Moreover, they lack of specificity, especially for closely related bacteria found in similar environments (Ndoye et al. 2011). In this sense, PMA-qPCR assays have been progressively incorporated to the determination of probiotic viability in different food matrices. Thus, Kramer et al. (2009) evaluated the possibility of using PMA-qPCR for enumeration of viable lactic acid bacteria (LAB) in a commercial lyophilized product. In this study, a great correlation was found between plate count values and PMA-qPCR quantification which demonstrated that VBNC cells were not present in the sample at any storage time. Both PMA and EMA have also been applied for viable probiotic detection in yoghurt (García-Cayuela et al. 2009; Meng et al. 2010). In both cases, the use of viability dyes in combination with qPCR showed a good correlation with plate counts and allowed the specific quantification of viable probiotics in yoghurt. PMA-qPCR has also been applied for probiotic detection in other matrices such as Cheddar cheese (Desfosés-Foucault et al. 2012), demonstrating to be a powerful tool for the quantification of viable cells compared with traditional methods for three different LAB species.
Whole bacterial populations
Differentiation of live bacteria is as well of interest in the analysis of the whole bacterial community when working with food and other environmental samples. Nocker et al. (2007a,b) applied both, end-point PCR and denaturing gradient gel electrophoresis (DGGE) in combination with PMA for the analysis of microbial communities present in water. Their results showed that, although end-point PCR did not show any difference between PMA treated and nontreated samples, when using DGGE, banding patterns were modified after PMA treatment. This indicates that PMA successfully suppressed DGGE bands corresponding to dead cells. In a latter study, Nocker et al. (2010) applied PMA in combination with pyrosequencing showing promising results for the monitoring of live bacterial populations. The analysis of mixed populations has also been approached in food matrices. Thus, Lee and Levin (2006) approached the analysis of mixed bacterial cultures of fish origin by amplifying a 16S rDNA region after mixing live and heat-killed bacteria. In this study, they observed that both PMA and EMA were effective after high temperature treatments (>60°C), but they are ineffective at marginally lethal temperatures due to the insufficient cell membrane damage.
Detection of enteric virus using viability dyes and RT-qPCR
Enteric viruses are those human viruses that are primarily transmitted by the faecal–oral route, either by person-to-person contact or by ingestion of contaminated food or water. There is a growing concern over human exposure to enteric viruses through contaminated food products or water. Data on viral foodborne diseases are still fragmented, focusing on either particular countries or particular pathogens only. Nonetheless, epidemiological evidence indicates that enteric viruses, particularly human norovirus (NoV) and also hepatitis A virus (HAV), are the leading cause of foodborne illness in developed countries (Koopmans and Duizer 2004; Authority 2012; Scharff 2012). To date, foods are rarely tested for enteric virus contamination and, when done, it is always performed by molecular techniques (Bosch et al. 2011). Over the last decade, several methods for enteric virus detection have been developed. In Europe, the CEN/TC25/WG6/TAG4 working group was entrusted by the European Committee for Standardization to establish a horizontal method for detecting norovirus and HAV in foods and bottled water. This method has recently been issued as an ISO norm (ISO/TS 15216-1:2013), and it is based on RT-qPCR. However, molecular detection of enteric viruses does not directly imply infectivity. Generally, enteric virus genome is a single-stranded RNA protected by a viral protein capsid. Although the use of viability dyes has been investigated for discrimination of infectious enteric viruses, much less information is available if compared with other relevant foodborne pathogens. So far, PMA combined with RT-qPCR has successfully been applied to discriminate between infectious and noninfectious poliovirus, coxsackievirus, echovirus and HAV that were inactivated by heat treatment (Parshionikar et al. 2010; Sanchez et al. 2012) (Table 2). Moreover, EMA has also been used to distinguish between thermally inactivated murine norovirus (MNV) and poliovirus suspensions (Kim et al. 2011). All the afore-mentioned viruses are RNA viruses, and as suggested by Parshionikar and collaborators (Parshionikar et al. 2010), stable secondary structures of the RNA genome may facilitate covalent binding of PMA/EMA to viral RNA. This is in agreement with the successful application of PMA–RT-qPCR for enterovirus (Parshionikar et al. 2010) and HAV where primer sets targeted the 5′ noncoding region (NCR), which has a strong secondary structure.
Human noroviruses are the leading cause of foodborne illness in developed countries. Although attempts to culture human NVs have been made, research with human NV has been hampered by the lack of suitable laboratory animals and the inability to propagate the virus in vitro. Consequently, the use of surrogates including feline calicivirus (FCV) and MNV is being extensively applied. Moreover, some attempts to distinguish between infectious and noninfectious noroviruses by PCR have been made. These methods include pretreatments by RNase for digesting viral RNAs from nonintact virus particles (Topping et al. 2009) or the use of viability dyes to assess infectivity. Despite the successful use of PMA to monitor human norovirus infectivity by conventional RT-PCR, results were not consistent when using real-time RT-PCR (Parshionikar et al. 2010).
Other studies have also demonstrated that the effects of viability dyes differ depending on the type of virus. For instance, Kim and Ko (2012) reported that PMA–RT-qPCR was suitable to discriminate between infectious and noninfectious bacteriophage MS2, but not MNV. In the same line, EMA did not inhibit RT-PCR amplification of nonviable avian influenza RNA (Graiver et al. 2010), while Fittipaldi et al. (2010) reported the discrimination of infectious bacteriophage T4 virus by PMA–qPCR. All these results point out the importance of the target region in terms of length and secondary structures.
Propidium monoazide combined with RT-qPCR has been also evaluated on HAV suspensions inactivated with high hydrostatic pressure (HPP) (Sanchez et al. 2012). In this study, a fraction of HPP-inactivated viruses were still detected by the PMA–RT-qPCR assay, indicating that RNA of HPP-inactivated HAV was not completely accessible to PMA. Similar results were obtained by Kingsley et al. (2002) when using RNase treatment, suggesting that HPP inactivation is mainly due to subtle alterations of viral capsid proteins (Kingsley et al. 2002). In line with these results, inactivation of MNV by HPP and UV was not directly correlated with effects on the integrity of virus genome or capsid protein. Viral inactivation was most likely due to effects on proteins involved in adhesion or invasion stages (Díez-Valcarce et al. 2011).
Finally, further studies need to be undertaken to prove this concept as a method to detect infectious enteric viruses in food samples.
Detection of yeast and moulds using viability dyes and qPCR
The large and diverse group of foodborne yeasts and moulds (fungi) includes several hundred species. Both yeasts and moulds cause various degrees of deterioration and decomposition of foods. They can invade and grow any type of food at any time, and they also grow on processed foods and food mixtures. Several foodborne moulds, and possibly yeasts, may also be hazardous to human health because of their ability to produce toxic metabolites known as mycotoxins. Moreover, certain foodborne moulds and yeasts may also elicit allergic reactions. On a positive way, yeasts are widely used in the food industry, and for instance, they are involved in the wine production process.
The detection of yeasts and moulds in food products by culture methods is frequently complicated due to the presence of injured or stressed cell populations. Therefore, attempts have been performed to combine molecular detection methods and viability dyes. EMA combined with qPCR has been successfully employed to detect viable Zygosaccharomyces bailii, a notorious food spoilage yeast, in fruit juices (Rawsthorne and Phister 2009). In this study, the type of juice, apple, grape and orange had minimal effect on the EMA-qPCR assay (Table 2). Andorrà et al. (2010) evaluated the use of viability dyes, PMA and EMA, to differentiate between live and dead yeast in alcoholic fermentation. Results showed that both dyes presented similar results on yeast monitoring, namely total yeast, Saccharomyces cerevisiae, Candida zemplinina, Hanseniaspora sp., Dekkera bruxellensis and Z. bailii, in grape must fermentation and ageing wines.
Application of DNA-binding dyes and qPCR to evaluate the efficacy of decontamination food processes
The safety of food products cannot be assured by testing only for specific foodborne pathogens and this should be supplemented by the prevention of contamination and the implementation of manufacturing processes that inactivate or eliminate them. For the food industry as well as for food authorities, it is important to develop effective and rapid techniques to monitor the effects of food manufacturing processes. Some rapid methods such as flow cytometry-based viable/dead methods (e.g. LIVE/DEAD BacLight kit) do not provide sufficient specificity or sensitivity. Thus, the molecular monitoring of disinfection efficacy using viability dyes in combination with qPCR has been investigated, mainly in pure culture suspensions (Table 3). Using S. typhimurium, L. monocytogenes, E. coli 0157:H7 and Mycobacterium avium, Nocker et al. (2007a) reported that PMA-qPCR was suitable to monitor the disinfection efficacy of hypochlorite, benzalkonium and heat, whereas it fails to assess loss of viability after UV treatment. In agreement with this study are the results from Rudi et al. (2005a,b), where EMA-qPCR was successfully used to monitor Campylobacter jejuni inactivation after boiling, heating at 72°C, disinfection with 70% ethanol and 500 ppm benzalkonium chloride. Pasteurization has been extensively evaluated by using viability dyes by Soejima et al. (2012a,b). They obtained successful results for the detection of Enterobacteriaceae and total coliforms using both EMA and PMA in milk. However, although being successful, dye performance was strongly dependent on the pasteurization treatment of the milk. Thus, after heat treatments below 72°C, cells were not permeable to PMA. To overcome this issue, Yang et al. (2011) have proposed the addition of deoxycholate previous to PMA treatment to increase dye penetration in dead cells. Deoxycholate disrupts the membranes of dead cells while does not affect live cells. In line with this study, Nkuipou-Kenfack et al. (2013) recently showed that the addition of deoxycholate improved the exclusion of dead cell signals for Salmonella but not for L. monocytogenes.
As a safe and effective alternative to chemical preservatives, nowadays particular interest has been focused on the application of natural ingredients. In this sense, essential oils (EOs) possess strong antimicrobial properties against foodborne pathogens, and their effects on bacterial growth and survival have extensively been studied for many years. To evaluate the effectiveness of EOs, PMA-qPCR has recently been applied for the detection of viable foodborne pathogens in fresh cut salads treated with cumin, clove, oregano, cinnamon and zataria EOs, and results corroborated the suitability of this procedure (Elizaquível et al. 2012b,c; Elizaquível et al. 2013).
The use of viability dyes has also been assayed after disinfection processes not directly applied to the food matrices. For instance, they have been tested in washing water treated by ultrasound, and good correlation between PMA-qPCR and plate counts was observed for E. coli and Bacillus subtilis by Zhang and Yu (2010) after 100 s of ultrasounds treatment in tap water. These results were corroborated by Elizaquível et al. (2012d) where PMA-qPCR was applied to detect viable cells of E. coli O157:H7 after the ultrasound treatment not only in tap water but also in two batches of vegetable industry lettuce wash water. Moreover, disinfection efficacy has been evaluated in food-contact surfaces at a beef processing plant showing that viable cells detected by EMA-qPCR are two log greater that the plate count results demonstrating the presence of VBNC cells in the surfaces (Khamisse et al. 2012).
These results demonstrate the potential of PMA/EMA for the monitoring disinfection methods in food industry. Application of this technique for the evaluation of other disinfection methods in food including treatment with hydrogen peroxide, phenol, aldehydes, surfactants, high pressure or heavy metals like silver and copper, and the correlation with traditional cultivation-based screening still remains to be tested.
Table 3. Application of Pv-PCR to monitor disinfection treatments and food manufacturing processes
|Benzalkonium chloride|| Campylobacter jejuni, Listeria monocytogenes ||Pure culture||EMA/PMA||Yes||Rudi et al. (2005a,b) and Nocker et al. (2007a)|
|Hypochlorite|| Salmonella enterica ||Pure culture||PMA||Yes||Nocker et al. (2007a)|
|UV light||Escherichia coli O157:H7||Pure culture||PMA||No||Nocker et al. (2007b)|
|Heat||Mycobacterium avium, Camp. jejuni, Zygosaccharomyces bailii||Pure culture||PMA/EMA||Yes||Rudi et al. (2005a,b) and Rawsthorne and Phister (2009)|
|Ethanol|| Camp. jejuni, Z. bailii ||Pure culture||EMA||Yes||Rawsthorne and Phister (2009)|
|Chlorinated alkaline cleaner||E. coli O157:H7||Swabbing suspensions||EMA||Yes||Marouani-Gadri et al. (2010)|
|Oregano, clove EOs|| E. coli O157:H7 ||Mixed salad||PMA||Yes||Elizaquível et al. (2012c)|
|Ultrasounds||E. coli O157:H7||Lettuce wash water||PMA||Yes||Elizaquível et al. (2012d)|
|Ultrahigh pressure, ultrasound and high-pulsed electric field||E. coli O157:H7||Pure culture||PMA||Yes||Xing-Long et al. (2013)|
|Zinc oxide nanoparticles|| Camp. jejuni ||Pure culture||EMA||Partial||Xie et al. (2011)|
|Pasteurization||Total coliforms, Enterobacteriaceae||Milk||EMA||Yes||Soejima et al. (2012a,b)|
|Pasteurization|| E. coli ||Milk/ground beef||PMA/deoxycholate||Yes||Yang et al. (2011)|
|Heat treatment|| Salmonella ||Cooked ham||PMA||Yes||Martin et al. (2013)|
Limitations and practical solutions
Although the application of v-PCR for the detection of pathogens in food products greatly helps to generate more meaningful data, this technique faces some challenges that can affect their sensibility. On the one hand, PCR-based procedures have a recognized detection limit which is, at best, 1 cell per reaction (103 cells per g) (Maurer 2006). However, bacterial pathogens are usually found at very low levels in naturally contaminated food products, mostly between 102 and 104 cells per g (Elizaquível et al. 2011). On the other hand, food products and other environmental samples are complex matrices which can affect the efficiency of v-PCR (Kramer et al. 2009) due to the presence of inhibitors that can act in different points of the detection process: lowering the effective dye concentration by chemical adsorption, preventing photo-activation due to the presence of organic compounds, interfering in the cell lysis, degrading nucleic acids or inhibiting polymerase activity (Wilson 1997; Fittipaldi et al. 2011a,b). Other factors such as dead cell concentration, turbidity, pH or salt concentration also have the potential to interfere with v-PCR results (Fittipaldi et al. 2012).
Although PMA, and in some instances EMA, has been successfully used for the detection of viable bacterial cells, some studies have found that the use of these viability dyes was not fully effective to remove the signal from membrane-compromised dead cells. This issue needs to be solved as, for some pathogens, a policy of zero tolerance is established, and therefore, false-positive results will implicate serious consequences. Thus, for PMA-PCR, Kralik et al. (2010) reported only 2 log reduction in PCR signal using permeable cells of Mycobacterium paratuberculosis. Other studies have found difficulties to eliminate the signal of dead Listeria monocytogenes cells (Pan and Breidt 2007; Elizaquível et al. 2012a). Also for EMA-PCR, some studies have shown no reduction in PCR signals in dead cells for Campylobacter (Flekna et al. 2007), Staphylococcus (Kobayashi et al. 2009) or environmental samples (Wagner et al. 2008).
To overcome this issue, Nkuipou-Kenfack et al. (2013) have recently evaluated different conditions to increase the penetration of PMA into dead cells of Salmonella typhimurium and L. monocytogenes. This study shows that exposing cells to PMA at a temperature exceeding their optimal growth temperature by 10°C for 30 min enhanced the performance of PMA treatment.
Another point to be considered is that the use of viability dyes does not directly measure live or dead but determine a particular damage of the cells (Hammes et al. 2011). Then, the successful application of viability dyes is based on membrane integrity as viability criterion. However, in the food industry and many other environments, bacteria are killed by processes that do not directly cause membrane damage and thus do not allow PMA or EMA to penetrate. PMA, for instance, was not able to differentiate between live and dead Escherichia coli after a UV treatment (Nocker et al. 2007a). In this sense, some authors have postulated that more than one viability criterion must be considered: culturability, metabolic activity and membrane integrity (Nocker and Camper 2009). Therefore, the use of different parameters to determine cell viability is recommended, and multiple cellular criteria must be assessed to determine cell's viability (Nocker et al. 2011).
As stated previously, viability dyes enter membrane permeable cells, and, once inside the cell, they intercalate into their nucleic acid. Taking this into account, some studies have found that the longer the sequence targeted for the detection, the higher the probability for the dye to intercalate (Contreras et al. 2011). For heat-killed L. monocytogenes, Soejima et al. (2011) determined that EMA was more efficient at suppressing signal from dead cells when targeting a 894-bp fragment comparing to a 113-bp fragment. Accordingly, Martin et al. (2013) found an inhibition effect dependent on the PCR amplification product length. They tested three PCR targets of 95, 285 and 417 bp combined with a PMA pretreatment to enumerate viable Salmonella cells in cooked ham and only the longer product achieved suppression of 108 CFU g−1 of heat-killed cell. Similar results were obtained for PMA when applied to denaturing gradient gel electrophoresis (DGGE) (Luo et al. 2010; Contreras et al. 2011). Moreover, Banihashemi et al. (2012) reported that the use of longer amplicon-sizes allows the differentiation of UV-dead cells and determined that this method may be able to target viable cells regardless of the mechanism causing cell inactivation.
Another limitation for the successful application of v-PCR in food samples is the micro-organism concentration. Several authors have reported that bacterial detection and quantification using PMA-qPCR are limited when high bacterial concentrations are present (Elizaquível et al. 2012a; Slimani et al. 2012; Zhu et al. 2012). In these cases, PMA enters the membrane of compromised cells, but the high concentration of cells might interfere the cross-linking step where PMA is activated, either by shadowing the light or competing for the PMA. In these cases, optimization of the concentration of viability dye, incubation temperature and time as well as size of the target sequence is crucial to overcome this limitation.
Conclusions and future perspectives
Overall, data provided in this review show the potential of viability dyes to be used with molecular techniques for the selective detection of viable micro-organisms in food samples. This review reports that micro-organisms exposed to agents that induce membrane damage followed by PMA/EMA treatment had reduced PCR signals. A linear correlation is observed between loss of culturability and PCR signal reduction for heat, ultrasounds, some sanitizers and EOs. Moreover, PMA has successfully been applied for the detection of some other relevant food micro-organisms, such as enteric viruses, yeast and moulds.
At the moment, the bottleneck in the detection of pathogenic bacteria in food by v-PCR relies on improving sample preparation in order to separate bacterial cells from the food matrix as well as to concentrate them above the detection level, at best 103-104 CFU g−1, without an enrichment step.
An intrinsic drawback of v-PCR is that the exclusion of dead micro-organisms can be incomplete leading to false-positive signals, and therefore, efforts need to be carried out to improve the efficacy of this approach or complement with other strategies. For instance, novel approaches, for example, assessing metabolic activity towards preferential detection of viable cells could complement the use of viability dyes. The effective implementation of v-PCR for food safety purposes will mostly depend on the improvements of the technology.
Conflict of interest
No conflict of interest declared.