The use and conservation of fish gonads as foodstuff dates back more than three thousand years ago in Sardinia as gift of the Phoenicians. Today, they are considered a valuable material and sometimes, as regards of some fishes, even difficult to find. Both female and male gonads are usually considered edible foods. Male gonads are constituted by fish testes and are commonly known as latti or lattumi (Palese and Palese, 1992). Female gonads have a wider distribution and are commercialized as fresh products but also salted, dried, smoked and marinated. Depending on the source, the local traditions and the preservation technologies, gonads are processed and then commercialized like ovarian sack (i.e. Bottarga) or as eggs, without the serous coating (caviar, salmon eggs, lumpfish roes, etc.) (Giuffrida and Panebianco, 2008). In the last few years, the consumption of fish eggs has increased rapidly, finding widespread use also in mass catering (EUROSTAT, 2017).
This increase has involved also those of the Peter’s fish (Zeus faber). Female of this species, thanks to their reproductive characteristics, show highly developed gonads in different periods of the year (Fulton, 1898). Moreover, this species is widespread all over Mediterranean areas, making the raw material easy to find. Several studies reported microbiological characterization of different fish roe product, revealing the presence, in some occasion, of food borne pathogens (Altug and Bayrak 2003; Boĭko et al., 2004; Oeleker et al., 2015; Razavilar and Rezvani, 2004; Voidarou et al., 2011). No information on Peter’s fish roe microbiology are, instead, available. Furthermore, Zeus faber is often subjected to Anisakidae colonization that can be found frequently in ovaries (Pekmezci et al., 2014; Yardimci et al., 2014). For all these reasons, the aim of the present study was to perform a quality assessment of Zeus faber ovaries regularly commercialized for human consumption.
Materials and Methods
Sample collection and macroscopic observations
A total number of 34 Zeus faber ovaries (11 fresh and 23 frozen samples) were collected from local market of Mazara del Vallo in a period from October to May. Fishes were caught in FAO area 37.1.2 and 37.2.2, by the trawl fleet of Mazara del Vallo (Sicily, Italy). Mazara del Vallo is widely considered to be the most important fishing center in Italy and contributes for more than 3/4 to the production and turnover of the national trawl fleet (ISTAT data).
Fresh ovaries were transported to our laboratory under refrigerated conditions (4°C) and processed within 24h from their arrival. Frozen samples were, instead, primarily thawed at 4°C for 24h. Each sample was carefully examined for the presumptive presence of parasites. All parasites were examined under stereoscopic microscope (Leica M 205C), and the belonging to the Anisakis genus was made according to guidelines proposed by Murata et al. (2011).
Portions of each sample were fixed in buffered formalin (10%), embedded in paraffin, and sliced into 5-μm sections. Sections were stained using hematoxylineosin (HE) and trichrome Masson. Stained sections were examined using light microscopy (Leica DM 4000B).
Each sample was processed for the count of: i) Aerobic mesophilic bacteria (AMB) according to UNI EN ISO 4833:2004; ii) Enterobacteriaceae according to UNI EN ISO 21528-2:2004; iii) Vibrio spp. on Thiosulphate Citrate Bile Salt Sucrose Agar (TCBS Oxoid, Italy), with 3% NaCl, incubated at 37°C for 24h; iv) Aeromonas spp. in Glutamate Starch Phenol Red Agar (GSP) with Ampicillin Selective Supplement (Merk, Italy), incubated at 30°C for 24h.
The detection of the following parameter was also conducted: i) Vibrio spp. with a preliminary enrichment on Phosphate Buffer Saline (PBS) at 30°C for 24h, then spread on TCBS, with 3% NaCl and incubated at 37°C for 24h; ii) Aeromonas spp. with a preliminary enrichment on PBS at 30°C for 24h, then spread on in GSP with Ampicillin and incubated at 30°C for 24h; iii) Listeria monocytogenes detection according to UNI EN ISO 11290-1: 2005; iv) Salmonella spp. detection according to UNI EN ISO 6579: 2002.
Vibrio identification: biochemical protocol
Suspected alophilic vibrios were confirmed to genus level to according to biochemical protocol suggested by Ottaviani et al. (2003). Isolated colonies growth on TCBS, were tested for Gram stains, Ossidase test, vibriostatic factor O/129 (150 mg, Oxoid) and growth on Kliger Iron Agar (KIA, Biolife, Italy). All the strains that resulted belonging to Vibrio spp. were, then, confirmed by molecular methods.
Vibrio identification: multiplex PCR assay
For DNA extraction 1 mL of broth culture was centrifuged at 12.000 rpm for 5 min; the pellet was re-suspended in 1 mL of sterile distilled water, boiled for 5 min, and centrifuged again. The supernatant was stored at 20°C until use. DNA was quantified by means of a spectrophotometer (SmartSpec Plus, Bio-Rad, Milan, Italy). For Vibrio genus identification specific gene rpoA primers were employed, according to La Neve et al. (2006) and Dalmasso et al. (2009). Confirmed strains belonging to Vibrio genus were selected for further identification at species level. For V. alginolyticus, and V. parahaemolyticus identification, were employed specific primers encoding a collagenase gene portion, as suggested by Di Pinto et al. (2005). Primers for the screening of pathogenic factors toxR and toxRS were also added to the PCR mix as reported by Xie et al. (2005).
Multiplex PCR mix consisted in a total volume of 50 mL containing 5 mL of Buffer 10X, 200 mmol of each dNTPs, 1.5 mmol of MgCl2, 100 nmol of each primer, 1 mL of DNA template and 1U of Platinum Taq DNA Polymerase (Invitrogen).
The pH of each sample was measured at room temperature with a pH meter HI90023CW (Hanna Instruments, Italy) equipped with a Mettler Toledo electrode (InLab 427) after homogenizing 3 g of sample in 27 mL distilled water for 10 s at 1300 rpm with an Ultra-Turrax T25 macevator (Janke & Kunkel, Staufen, Germany). Each value was the mean of three replicates.
A t-Test was performed in order to evaluate significative differences in our results among the samples and between fresh and frozen products (XLSTAT, Microsoft Excel, Addinsoft). Significance level was assumed as P<0.01. Microbiological charges were converted in Log CFU/g for a better expression of results.
Results and Discussion
Table 1 shows mean values for each microbiological parameter. The mean values for AMC and Enterobacteriaceae were 3.94±0.5 Log CFU/g and 1.30±1.2 Log CFU/g respectively. Fresh and frozen samples have significant (P<0.01) differences in AMC, with mean values of 4.75±0.5 log CFU/g and 3.65±0.7 log CFU/g respectively. In fresh ovaries, AMC oscillated from 3.65 to 5.17 Log CFU/g, while, in frozen samples ranged from 1.74 to 4.60 Log CFU/g. Also for Enterobacteriaceae loads, significant differences (P<0.01) were observed between fresh and frozen products. In fresh products Enterobacteriaceae charges ranged from 1.17 to 3.43 Log CFU/g, while 52.17% (12) of frozen samples reported Enterobacteriaceae loads <1 Log CFU/g.
Considering bacteriological hygiene indicators (AMC and Enterobacteriaceae), Zeus faber ovaries presented an overall satisfactory condition even as fresh product stored at refrigeration temperatures (Altug and Bayrak 2003; Boĭko et al., 2004; Oeleker et al., 2015; Razavilar and Rezvani, 2004; Voidarou et al., 2011).
The only microbiological concern may be related to pathogen detection. No Salmonella spp. and Listeria monocytogenes were found. Otherwise, Aeromonas spp. and Vibrio spp. were detected in several samples of Zeus faber ovaries. Aeromonas strains have a significative incidence in marine environment all over Europe and were frequently detected in fishery product, including fish eggs (Davies et al., 2001; Hänninen et al., 1997; Hansen and Olafsen, 1989; Muscolino et al.; 2014). Bacteria belonging to Aeromonas genus may be, also, responsible of gastro-enteric diseases outbreaks in humans, thanks to the production of enterotoxins, cytotoxins and haemolysins (Deodhar et al., 1991; Kirov, 1993). For all these reasons, they could represent a microbiological hazard for the consumer.
In our study, Aeromonas spp. was found in seven samples: two in frozen and five in fresh samples. The two positive frozen samples, were characterized by loads of 2.2 Log CFU/g and <100 CFU/g respectively. In fresh ovaries, instead, Aeromonas spp was reported in 45.45% (5) of samples with loads ranging from 2.0 to 3.0 Log CFU/g. However, Aeromonas spp. mean loads revealed no significative difference (P>0.01) between fresh and frozen samples (Table 1). Considering its psychrophilic behavior, its tolerance and survival at frozen temperatures as well as its stress resistant variants, this study confirms Aeromonas spp. as potential microbiological hazard in Zeus faber ovaries intended for consumption (Castro-Escarpulli et al., 2003; Giuffrida et al., 2010).
Vibrio spp. was detected in the 36.36% (4) and 13.04% (3) of fresh and frozen samples respectively, but always with loads <1 Log CFU/g.
A total number of 31 suspected Vibrio spp. strains (14 from fresh and 17 from frozen samples) were isolated from TCBS with 3% NaCl. All suspected strains were confirmed as belonging to Vibrio spp., resulted positive for biochemical test and for molecular identification at genus level (rpoA +). Their presence in frozen samples, confirmed the remarkable resistance of halophilic Vibrio at freezing temperatures (Bang and Drake, 2002; Johnston and Brown, 2002). Multiplex PCR analysis revealed that all 31 strains isolated were V. alginolyticus.
This bacteria is often found in Italian marine coastal environments and sea products (Dumontet et al., 2000; Panebianco et al., 2011; Narracci et al., 2013; Ziino et al., 2010; Ziino et al., 2014). It is also, widely spread all over the world, and usually has the highest prevalence in Vibrio communities (Schets et al., 2010; Jones et al., 2013). V. alginolyticus pathogenic potential in humans is well recognized, causing various infections and inflammation patterns (Schmidt et al., 1979). It raised great importance as emergent foodborne pathogen, responsible for several cases of gastroenteritis linked to the consumption of fish products (Mustapha et al., 2013). Moreover, in immune-compromised subjects, serious extra-intestinal infections, such as septic shock, have been reported as consequence of consumption of seafood contaminated by V. alginolyticus (Dong-Young et al., 2008). ToxRS and ToxR pathogenic factors encode trans-membrane proteins that mediate the regulation of the virulence gene expression in several pathogenic Vibrio strains, including V. alginolyticus (Das et al., 2016). These factors are frequently investigated for a better characterization of bacterial pathogenic potential (Fu et al., 2016). Moreover, ToxR is widely proposed for the specie identification and phylogenetic analyses (Montieri et al., 2010). Considering all the 31 V. alginolyticus strains, 61.5% carried the ToxRS factor and in lower percentage (6.5%) the ToxR gene. In particular, among the 14 strains isolated from fresh samples, 12 (85.7%) were ToxR positive and 2 (14.3%) carried the ToxRS pathogenic factor. Otherwise, among the 17 strains isolated from frozen samples, 7 (41.2%) were ToxR positive but no ToxRS positive strains were found. It was asserted that ToxR may assist V. alginolyticus in host target cells adhesion, which is crucial step for the initiation of the infection (Chang et al., 2002). As reported by previous studies, fish roe represent an ideal setting for Vibrio colonization (Voidarou et al., 2011). Our results highlight the role of Zeus faber ovaries as source of V. alginolyticus pathogenic strains, which could be potentially implicated in foodborne disease episodes.
The pH mean values was 6.32±0.2 for fresh ovaries and 6.22±0.3 for frozen products with no significant differences (P>0.01) between the two kind of samples.
Macroscopic and microscopic analysis results
The 47.05% (16) of total samples revealed parasitic infestation by presumptive nematode, confirmed by stereomicroscopic observation as Anisakis larvae type I. Specifically, larvae were found in 16 ovaries: 45.45% (5) and (11) 47.82% of fresh and frozen samples respectively. Various levels of infestation were observed: from isolated larvae to a massive infestation, in which nematodes were bunchshaped aggregated (Figure 1).
The histological investigation showed the presence of various patterns, ascribable to different gonadic development stages (Abou-Seedo et al., 2003; Macrì et al., 2011). Specifically, in some analyzed ovaries, a primary stage of oocyte growth (Previtellogenic stage) was observed (Figure 2), while in other samples a secondary stage of oocyte maturation (Vitellogenic stage) was predominant (Figure 3), with no relation to season or area of fishing.
In samples infested by Anisakis larvae, parasites were found encysted on the serosa, and were not observed tissue and cellular reactions (Figure 4). Only in two cases, however, it was possible to observe parasites inside the gonads, not surrounded by any inflammatory reaction (Figure 5). Our results confirm the parasitological risk due to the consumption of Zeus faber ovaries. Anisakis larvae detection deserve great attention, as the nematode have been also observed in the depth of the gonads. The lack of connective capsule and inflammatory response around the parasites, are consequences of larvae tissue mobilization, searching for favorable conditions after the fish death. In these cases, Anisakis could be unnoticed to a superficial inspection, especially if a moderate infestation occur. However, it would, still, represent a risk for the consumer as thermal shock is not sufficient to avoid possible allergic reaction in sensitive subjects, caused by allergens resistant to high temperatures and freezing conditions (Audicana and Kennedy, 2008; Speciale et al., 2017).
In frozen samples, other microscopic findings are the damage related to the presence of intra and extra-cytoplasmic ice crystals. During thawing process, ice particles damaged the oocyte cell wall, causing the loss of cellular content. In particular, oocytes appeared coerced, with irregular contours, fragmented and well spaced. The cytoplasm lost its characteristic granular appearance, becoming uniformly acidophilus and, sometimes, voluminous intracytoplasmic vacuoles were observed (Figure 5). As reported in previous study microscopic analysis resulted a valid means to discriminate fresh products from those thawed (Meistro et al., 2016; Muscolino et al., 2012).
In conclusion, Zeus faber fish ovaries showed more than satisfactory hygiene conditions. The only microbiological concern was related to the presence of potentially pathogenic strains of V. alginolyticus and Aeromonas spp.. However, microbiological hazard can be reduced by adequate cooking and the prevention of cross-contamination. In relation to the presence of Anisakis larvae, only an appropriated consumer information seems the most important measure to prevent allergic manifestations, beside an efficient thermal treatment or similar strategies in order to inactivate nematodes (Anastasio et al., 2015; Giarratana et al., 2012; Giarratana et al., 2014; Giarratana et al., 2015; Giarratana et al., 2017a; Giarratana et al., 2017b; Valero et al., 2015).