1 Introduction

Fresh fruits and vegetables are essential parts of the world’s diet, contributing necessary vitamins and minerals, often eaten raw or minimally processed, and grown using conventional agricultural methods. Further, because of the consumer perception that these products are healthy, increased consumption has been recorded through time. However, as these produce are usually eaten raw, they are at risk to microbiological contamination, hence causing foodborne illnesses. There are several reports of diseases related to the consumption of produce throughout the world. For instance, there are at least 713 produce-related outbreaks between 1990 and 2005 in the USA and 88 outbreaks between 1996 and 2006 in the UK (Goodburn and Wallace 2013). Fresh produce have become one of the vehicles of transmission of various opportunistic pathogens such as enteric microorganisms (Rangel et al. 2005) which are typically linked to fecal contamination from different farm animals (Aruscavage et al. 2006; Chapman et al. 1997).

There are many questions about the transmission of microorganisms from their potential reservoirs to fruits and vegetables, and the vectors which may be involved in this process. While all produce items have factors in common, it is important to recognize that each fruit and vegetable has a unique combination of composition and physical characteristics, growing and harvesting practices, cooling techniques, and optimum storage temperatures and environment (De Roever 1998). The increase in the risk of foodborne outbreaks may also be caused by the lack of many developing countries in making attempts to accurately obtain proper risk assessment over the different agricultural produce.

In the Philippines, only few studies and publication can be linked to scientific evidences and accounts of foodborne outbreaks, and some cases were even due to the contamination of enteric microorganisms to food products other than fresh produce (Azanza 2006). In the study of Vital et al. (2014), the contamination of Escherichia coli and Salmonella including bacteriophages was found to be significant in various retail fresh produce that are commonly used in salad preparation such as bell pepper, carrot, lettuce, cabbage, and tomato. Thus, this study comprehensively surveyed the relative levels of microbial contamination of agricultural produce collected from major open air markets and supermarkets in Luzon, Philippines using culture and molecular methods. Moreover, the microbiological quality of fresh produce prior to consumption was assessed so as to determine the safety when served on the dining table.

2 Methodology

2.1 Fresh produce sampling

Fresh produce typical for raw consumption was comprehensively surveyed. Sampling areas encompassed all the geographic regions of Luzon, Philippines, with the following sites: open air markets in the National Capital Region (Balintawak, Divisoria, Pasig, Pateros), North Luzon (Baguio City), South Luzon (Laguna), and Central Luzon (Pampanga); supermarkets in the National Capital Region (Quezon City), North Luzon (Baguio City), South Luzon (Laguna), and Central Luzon (Pampanga). Five types of vegetable produce that were consumed uncooked or minimally processed, namely bell pepper, carrot, lettuce, mung bean sprout, and tomato, were collected from a variety of sources to encompass the different produce handling and distribution practices. These produce were selected based on the results of previous studies (Vital et al. 2014). A total of 410 fresh produce were collected, i.e., 50 vegetables from open air markets per site and 25 vegetables from supermarkets from each site. The samples were transported to the laboratory in an improvised ice box (kept under 10 °C) and processed within 3–8 h after collection.

2.2 Culture-dependent analysis

The fresh produce samples underwent washing steps, and the wash buffer was examined for the presence of test microbial organisms (E. coli, Salmonella, somatic phages) using a method similar to EPA 1602 (US-EPA 2001) and Vital et al. (2014). Thermotolerant E. coli and somatic bacteriophages are classic indicator organisms for fecal contamination, while Salmonella spp. is a recognized foodborne pathogen (Mead et al. 1999).

For the detection and enumeration of thermotolerant E. coli, tenfold serial dilution was performed from the wash solution. They were filtered through 0.45-μm filters, plated on modified membrane thermotolerant E. coli (mTEC) agar (BD Difco Laboratories, MD, USA), and incubated at 44.5 °C for 24 h. Blue to deep blue colonies were considered presumptive for E. coli, and at least three isolated colonies were streak-plated on eosin methylene blue (EMB) agar plates (BD Difco Laboratories, MD, USA) and incubated at 35 °C for 24 h. Colonies with dark center and green metallic sheen were considered as thermotolerant E. coli (Vital et al. 2014).

For the isolation of Salmonella spp., the wash solution was mixed with double strength buffered peptone water (BPW) (BD Difco Laboratories, MD, USA), and three further tenfold serial dilutions were performed in BPW as an enrichment step. Afterward, a volume per dilution series was added to Rappaport–Vassiliadis (RV) enrichment broth (BD Difco Laboratories, MD, USA) using a sterile multi-well plate, and incubated at 42.5 °C for 24 h. Results were recorded as positive (+) or negative (−) and compared to an MPN table. Positive wells were streak-plated in triplicate onto xylose lysine deoxycholate (XLD) agar (BD Difco Laboratories, MD, USA) and incubated at 35 °C for 18–24 h. Salmonella spp. isolates were indicated by red colonies with a dark center (Vital et al. 2014).

For the detection of somatic bacteriophages, the double agar assay was performed (Lasobras et al. 1999). Tryptic soy agar (TSA) (BD Difco Laboratories, MD, USA), 0.7% soft agar layers, E. coli CN-13 (nalidixic acid resistant strain), and antibiotic nalidixic acid (Sigma-Aldrich, USA) were used. After 24 h of incubation at 35 °C, the presence of somatic phages was indicated by clearing zones in the plates. Positive control included Phi X-174 (DSM-4497) in the mixture. Additionally, negative control (without any phage) was also considered as another setup to ensure the specificity of the results.

2.3 Molecular analysis

Produce samples underwent molecular analysis to confirm the presence of the enteric microorganisms. Isolates were obtained from the confirmed positive from the culture methods. For the DNA extraction, 1 mL of bacterial culture was prepared in a tryptic soy broth (TSB) (BD Difco Laboratories, MD, USA) grown at 35 °C for 24 h. After overnight incubation, the culture was centrifuged at 10,000×g for 10 min and the supernatant was discarded. One milliliter of sterile distilled water was added to the pellet, mixed by vortexing, centrifuged at 10,000×g for 10 min, and the supernatant was discarded. One hundred microliters of sterile distilled water was added to the pellet and mixed by vortexing. The suspension was placed in a dry bath at 95–100 °C for 15 min. After boiling, the sample was centrifuged at 10,000×g for 10 min and the supernatant containing the DNA was collected.

Confirmation of the isolates was done with polymerase chain reaction (PCR) amplification using specific primers for E. coli (Takahashi et al. 2009) and Salmonella spp. (Malorny et al. 2003). Amplicons were characterized using agarose gel electrophoresis (AGE). Together with a KAPA 1000 bp DNA ladder (Kapa Biosystems, MA, USA) and a negative control, the amplicons were loaded into 1.5% (w/v) agarose gels with 5000× SYBR Safe DNA Gel (Thermo Fisher Scientific, MA, USA). The loaded gel was run in a Mupid-exU submarine electrophoresis system containing 1X TAE buffer at 100 V for 30 min. The resulting DNA bands were then viewed under UV illumination using Bio-Print ST4 (Vilber-Lourmat, Germany) and Vision-Capt version 16.08 (Vilber-Lourmat, Germany).

2.4 Data and statistical analysis

All values were log-transformed to improve normality and homoscedasticity. Two-way analysis of variance and Tukey’s HSD post hoc test were used to compare means of foodborne microorganism counts among the five vegetable samples and among the sampling sites. Data were analyzed using the software SPSS 15.0 (SPSS Inc., Chicago, Il, USA). The values were considered statistically significant based on a p value <0.05.

3 Results

A total of 410 raw vegetable samples were collected and processed from various open air markets and supermarkets in Luzon, Philippines. Out of these collected produce, 23 (5.61%) were positive for E. coli, 21 (5.12%) for Salmonella spp., and 21 (5.12%) for somatic coliphages (Table 1).

Table 1 Prevalence of enteric microorganisms in fresh produce collected in Luzon, Philippines
Full size table

Based on the culture-dependent and molecular methods employed in this study, E. coli, Salmonella spp., and somatic coliphages were found to be of significant numbers in selected vegetable produce (bell pepper, carrot, lettuce, mung bean sprouts, and tomatoes) sampled in different areas (Table 1; Fig. 1).

Fig. 1
figure 1

Prevalence of microbial contamination among fresh produce (per sample type) collected in Luzon, Philippines: mean log CFU/g (for E. coli), log MPN/g (for Salmonella), log PFU/g (for coliphage) with standard deviation of mean (SD); BP bell pepper, CR carrot, LT lettuce, MS mung bean sprouts, TM tomato. Significant values indicated by asterisk (p < 0.05). Rankings: log CFU/g (MS > BP = TM); log MPN/g (TM > (LT = CR = BP); log PFU/g (BP > (TM = CR)

Full size image

Microbial loads among the raw vegetables differed in the amount of thermotolerant E. coli, Salmonella spp., and coliphage per produce type (Fig. 1). Mung bean sprouts were contaminated with E. coli significantly (p < 0.05). Also, tomatoes had the highest counts of Salmonella spp. and bell pepper with phages (p < 0.05). Thermotolerant E. coli had values as high as 2.74 log CFU/g, Salmonella as high as 4.93 log MPN/g, and phages with 1.14 log PFU/g microbial content.

Because of this highly significant E. coli contamination in mung bean sprouts, based on Codex Alimentarius, the vegetables are included in Class C category (unsatisfactory level, i.e., potentially injurious to human health and not advisable for human consumption) in terms of food safety (Table 2). Lettuce also had high numbers of E. coli but not that significantly high as compared to mung bean sprouts, but still belonged to the Class C (unsatisfactory level) (Table 2).

Table 2 Microbiological quality of fresh produce in selected sampling sites of Luzon, Philippines.
Full size table

Meanwhile, tomatoes and bell peppers were found to be less contaminated by E. coli and, hence, were classified as Class A (satisfactory level, i.e., poses no potential health risks to humans) (Table 2). Though carrots were contaminated with E. coli, they were still under acceptable category (Class B) for human consumption as the coliform counts are insignificant to pose health dilemma (Table 2).

In this study, Salmonella spp. contamination was highly significant in numbers in tomatoes as compared to lettuce, carrots, and bell pepper (Fig. 1). Obviously, Salmonella spp. presence in vegetable produce may directly cause serious food poisoning and these vegetables are automatically considered as Class D (unacceptable level, i.e., unfit for human consumption and can pose threat to human health) (Table 2).

The coliphage presence was highest in bell peppers (Fig. 1). The detection of coliphage from the wash solution indicated the presence of host coliform. This was confirmed by the positive (by adding Phi X-174 to the lawn) and negative (without any virus) control. In addition, the presence of bacteriophages in the sample may indicate contamination through water resources and thus can serve as an implication of virus contamination in agricultural fresh produce. Moreover, this bacteriophage can pose health risk as it may indicate the presence of other pathogenic enteric viruses having a low infective dose and by surviving harsh environments such as extreme pH and temperature.

Microbial contamination was also compared per sampling site. Results show that the sites differed significantly (p < 0.05). Baguio City had the highest number of samples with E. coli contamination, followed by Laguna and Pampanga (Fig. 2). Salmonella spp. contamination was highest in Divisoria as compared to the remaining sites (Fig. 2). Meanwhile, there were no significant differences in coliphage presence among vegetable samples in the three sampling sites (Fig. 2). The results suggest the prevalence of enteric microorganism in various minimally processed fresh produce in different open air markets and supermarkets in the Philippines.

Fig. 2
figure 2

Prevalence of microbial contamination of fresh produce among open air markets and supermarkets (per sampling area) collected in Luzon, Philippines: mean log CFU/g (for E. coli), log MPN/g (for Salmonella), log PFU/g (for coliphage), with standard deviation of mean (SD); DIV Divisoria, PAT Pateros, BAG Baguio, LAG Laguna, PAM Pampanga, QUE Quezon City, PAS Pasig. Significant values indicated by asterisk (p < 0.05). Rankings: log CFU/g (BAG > DIV), log MPN/g (DIV > (PAT = BAG = LAG = PAM = PAS = QUE), log PFU/g

Full size image

4 Discussion

Fresh produce has earned great popularity in the past years due to the awareness of consumers to a healthy lifestyle. However, there had been an increase in incident reports related to the fresh produce-associated disease outbreaks in humans which may be accounted to food preference, contamination, and emergence of new pathogenic or antibiotic resistant strains. Bacteria such as E. coli O157:H7 and Salmonella are among the most relevant foodborne pathogens that are strongly associated in outbreaks (Ackers et al. 1998; Isaacs et al. 2005; Singh et al. 2006).

Foodborne pathogens in contaminated food can cause serious human diseases. Specific detection of pathogens in food is quite expensive, time-consuming, complex, and impractical due to the presence of several pathogens. Hence, indicator microorganisms are routinely used instead to verify effective implementation of good agricultural practices and good manufacturing practices (FAO/WHO 2008; Sagoo et al. 2003).

In this study, coliforms considered as indicator organisms were used and considered as Gram-negative, non-spore-forming rods that ferment lactose within 24–48 h and produce dark colonies with a metallic green sheen on Endo-type agar. Coliforms are represented by four genera of the family Enterobacteriaceae: Citrobacter, Enterobacter, Escherichia, and Klebsiella. E. coli in foods in sufficient numbers is taken to indicate the possibility of fecal contamination and the possible presence of other enteropathogens such as Salmonella. On the other hand, enteropathogenic Salmonella are Gram-negative, non-spore-forming rods that are widely distributed in nature, with humans and animals being their primary reservoirs. Salmonella consists of around 2200 serotypes (serovars) based on 67 O-antigen groups and various H-antigens (FAO/WHO 2008). Salmonella species are the common contaminant in the meat products; however, they also persisted in various vegetables (Vital et al. 2014).

The results of this study are supported by an observed microbial contamination of Vital et al. (2014) and Nguyen-the and Carlin (1994) in minimally processed fresh vegetables. These imply that certainly there is widespread contamination of fecal microorganisms in normally eaten raw fresh produce in the Philippines. Although E. coli and Salmonella spp. were detected in this study, it should be emphasized that the vegetables did not show any visible signs of spoilage. In fact, vegetables (or other agricultural crops) cannot be considered microbe-free because they have normal microflora present and also the given processing stages (e.g., crop planting, harvesting, handling, cutting) may be potential sources of contamination (Sagoo et al. 2003).

Among the fresh produce, lettuce is regarded as the most susceptible to contamination by pathogens because the fluid leakage from lettuce tissue provides sufficient nutrients to support the growth of diverse food pathogens (Beuchat and Ryu 1997), but in this study, mung bean sprouts tended to be highly contaminated by E. coli (Fig. 1). Mung bean sprouts may be possibly initially contaminated due to the use of contaminated water from sprouting (aside from the fact that sprouts have high water content that makes them at high risk of spoilage or contamination). Research has demonstrated that when the roots of fully developed sprouts are immersed in water containing E. coli, the pathogen is found throughout the edible part of the sprout even after surface sterilization. This means that the inner tissues of sprout can become endogenously contaminated by bacteria not solely by external contamination (Itoh et al. 1998).

Meanwhile, Salmonella spp. are the most frequently identified pathogenic agent associated with fresh produce-related infection. A range of fresh fruit and vegetable products have been implicated in Salmonella infection, most commonly lettuce, sprouted seeds, melon, and tomatoes (Doran et al. 2005). In the study of Vital et al. (2014), it was found out that tomatoes were least infected with these bacteria. However, this study revealed that tomatoes are contaminated with Salmonella which may be alarming and probably accounted to pathogen internationalization (Zhuang et al. 1995; Ingham et al. 2005).

Bacteriophages were also detected in this study and could act as an indicator of fecal contamination in produce. There are very limited studies on the use of phages and viruses as pathogenic viruses are difficult to handle. These phages may serve as model organisms for virus contamination of fresh produce (Lasobras et al. 1999). Also, since the assay performed was simple, accurate, and fast, this can be recommended as a test to detect fecal contamination in agricultural products.

High E. coli and Salmonella contamination in Baguio City and Divisoria, respectively, may be explained by the degree of food handling by the vendors and the nature of the food storage and transportation. Actually it would be very difficult to deduce the levels of vegetable contamination in these sampling sites due to lack of references and/or sources of information regarding the manner of vegetable cultivation, harvesting, storage, transportation, and safe product handling. It was keenly observed that all the open air markets failed to meet the good microbiological safety of food standards due to the unhygienic food handling of the vendors. In the Philippines, protection of consumers against contamination is part of the newly approved Food Safety Act in 2013. However, the implementing rules and regulations are still being formulated and this study aims to help in terms of microbial food safety in produce.

In light of these results, it is alarming to know that E. coli and Salmonella can both survive in soil and in plants. Usually environmental parameters play the major role in food produce contamination, but several studies revealed that pathogen survival is also controlled or defined by the complex interactions with plants. Understanding the interactions of pathogens with their external environment and plant would assist in the development of new strategies to improve food safety. Other factors that can also contribute to microbial contamination may include contaminated irrigation waters (Garcia et al. 2015), use of manure fertilizers, poor hygienic practices and poor equipment sanitation.

5 Conclusions

This study involved a widespread survey of the microbiological quality and safety of fresh produce from various markets in the Philippines. The results of the study can affect the agricultural sector of the country in terms of food supply (Jacxsens et al. 2010) or exportation (Chaidez et al. 2005). With the rising demand for convenient and fresh foods for a healthier lifestyle, the fresh produce industry has a great advantage for consumers, thus microbiological quality must be maintained, and sustainable sanitation strategies should be implemented for keeping the quality and safety (Wiley 1994; Artés et al. 2009). Further, this study opens the door for more scientific researches and development of new and/or improved tools and techniques to identify microbiological risks (Jacxsens et al. 2010). The data may be used to establish guidelines in the Philippines and other Southeast Asian countries for microbial food safety and quality, as well as prevention and detection of illnesses associated with fresh produce consumption.