Abstract
The zoonotic protozoan parasites Toxoplasma gondii, Cryptosporidium parvum and Giardia duodenalis have been recorded worldwide in economically important edible shellfish, and are thus likely to represent a significant public health risk. Therefore, an innovative, user-friendly diagnostic tool is required in order to improve food safety control. The Q3 system is a miniaturized platform whose efficiency and applicability were investigated and compared with results obtained using standard Real-Time PCR. Tanks of saltwater containing acclimated Mytilus galloprovincialis, Ruditapes philippinarum and Ostrea edulis specimens were spiked with purified Cryptosporidium, Giardia and Toxoplasma cysts/oocysts at different concentrations (i.e., 103, 104 and 105). We then collected 30 specimens for each shellfish species from each group at 24 h and 72 h post-contamination. After DNA extraction, we tested all samples by Real-Time-PCR and Q3, and evaluated the sensitivity, specificity, predictive values, repeatability and concordance between the two systems. Concordance between Real-Time-PCR and Q3 was very good (p < 0.01), especially for Toxoplasma in M. galloprovincialis at both 24 h and 72 h after contamination, and in O. edulis at 72 h. The ability of Q3 to detect all the investigated pathogens was similar to that of Real-Time-PCR, and Q3 was efficient in detecting Toxoplasma in both M. galloprovincialis and O. edulis. This is the first study concerning the use of lab-on-chip technology in a food matrix, and in edible marine mollusks in particular.
Electronic supplementary material
The online version of this article (10.1007/s13197-019-03972-7) contains supplementary material, which is available to authorized users.
Keywords: Protozoans, Shellfish, Real-Time PCR, Lab-on-chip efficiency, Food safety
Introduction
Bivalve mollusks are an important food resource worldwide (FAO-WHO 2014a), and the annual world production of farmed food mollusks is 16.4 million tonnes, i.e. 21% of the entire aquaculture sector (FAO 2015). The main species farmed in the EU are mussels (mostly Mytilus galloprovincialis), at 328,000 tonnes (26% of total EU production), followed by oysters (Ostrea edulis) at 138,000 tonnes (11% of total EU production), and clams (Ruditapes spp.), at over 60,000 tonnes (Eurostat 2015).
In addition to the pathogens and contaminants defined by the current EU legislation regarding the sanitary quality and control of mollusks (Reg. EC 853/2004; Reg. EC 854/2004; Reg. EC 882/2004; Reg EC 1881/2006; Reg. EC 2285/2015), the protozoans Giardia, Cryptosporidium and Toxoplasma are important agents whose environmental dissemination is closely associated with fecal matter of human and/or animal origin (Thompson et al. 2005; Dubey and Jones 2008; Robert-Gangneux and Dardé 2012; Mmbaga and Houpt 2017). When these protozoans are discharged into rivers via farming and urban wastewater or run-off water, they subsequently contaminate coastal waters and are filtered and concentrated by mollusks. It has been shown that Giardia, Cryptosporidium, and Toxoplasma oo/cysts are present in farmed mollusks (including M. galloprovincialis, R. philippinarum, and O. edulis), or wild mollusks in lagoons and other marine environments (Robertson 2007; Aksoy et al. 2014; Coupe et al. 2018), and even those cultivated in Class A waters, for which no depuration treatment is required (Reg. EC 2285/2015; Giangaspero et al. 2009; Putignani et al. 2012).
Detection of these protozoan parasites in mollusks is of interest for public health, especially if mollusks are eaten raw or lightly cooked (Robertson 2007; Aksoy et al. 2014; Coupe et al. 2018) and is also important because they act as bio-indicators of water contamination and are thus of interest in terms of environmental health (Ghozzi et al. 2017).
Due to their high specificity and sensitivity (one oocyst per ml of hemolymph), PCR techniques have replaced microscopy tools for detecting parasitic pathogens in a complex matrix such as mollusk tissue (Miller et al. 2006). Several molecular methods, such as nested-PCR, PCR, and PCR-RFLP, have been developed to identify these protozoans (reviewed by Hohweyer et al. 2013). Other PCR methods (i.e. PCR-DGGE) have been used to detect DNA fragments of pathogens in fish and shellfish samples (El Sheikha and Montet 2016).
More recently, qPCR has been used to detect pathogens in mollusks (Aksoy et al. 2014; Giangaspero et al. 2014; Marangi et al. 2015), demonstrating an extremely high degree of sensitivity and specificity (up to 100%) (Hohweyer et al. 2013). However, molecular tools are currently applied to scientific research and not used in routine diagnosis, although Food and Agriculture Organization and World Health Organization guidelines support the need for tracking, monitoring and surveillance of these protozoan parasites (FAO-WHO 2014b).
At present, the molecular diagnostic industry is actively involved in developing simple and cost-effective platforms to use in pathogen detection (Biava et al. 2018). Driven by the demand for rapid and easy detection, lab-on-chip (LOC) technology has developed rapidly because multiple laboratory processes can be integrated in a miniaturized device (Yoon and Kim 2012; Marziliano et al. 2015). In the EU (Reg. EC 178/2002), traceability of all foods, including bivalve mollusks, is mandatory due to its importance in achieving food safety objectives, and a rapid tool for the detection of hazards in foods would help to prevent contaminated foods reaching consumers.
The Q3 system is a small, compact LOC platform (14 × 7 × 8.5 cm, 300 g), which was developed by STMicroelectronics. Rapid and easy-to-use, it enables the amplification and detection of nucleic acids by Real-Time PCR (RT-PCR) in under 30 min (Marziliano et al. 2015; Biava et al. 2018). It has three main components (Fig. 1): the Q3 LOC Disposable Unit (DU or cartridge) where the RT-PCR reactions take place, the Q3 Reader, and Q3 software. The Q3 Reader is an electrical optical instrument, which acts as a mechanical holder of the DU, controls the thermal process on the DU, and excites and reads the optical fluorescent signals.
This study aimed to evaluate the efficiency of the Q3 system on RT-PCR protocols developed in our laboratories (Marangi et al. 2015) for detecting Giardia, Cryptosporidium and Toxoplasma in bivalve mollusks cultivated for human consumption.
Materials and methods
Aquarium set up and efficiency assessment
An aquarium system (Acqua and Co SRL, Cadelbosco Sopra, Ravenna, Italy)1 of nine tanks was installed at the Bonassisa Lab, Foggia, Italy. Each tank contained 20 L of water with NaCl and water conditions (salinity 34 g/L; water temperature 21 °C; oxygenation 7–8 mg/L) and room temperature (15 °C) monitored daily.
Protozoan strains
Giardia duodenalis trophozoites (strain ATCC 30957) were purchased from LGC Standards S.r.l.—Italy Office, Milan, Italy and cultured in vitro to obtain cysts, using a medium of TYI-S-33 plus 10% bovine serum and 10 mg/ml bovine bile at pH 7.8 (Keister 1983).
The maximal number of trophozoites was obtained by growing them to confluence. The culture medium containing dead trophozoites unattached to the walls was removed, and the encystation medium was added. The tube was then inserted into a CO2 incubator at 37 °C, and after 24 h the culture tube was placed in ice for 30 min and then centrifuged at 500×g for 10 min. The encystation medium was subsequently removed, the pellet was resuspended in the growth medium (Keister’s Modified TYI-S-33 Giardia Medium [ATCC Medium 2695]), and the culture tube was inserted again into the CO2 incubator at 37 °C. After 24 h, the culture tube was placed in ice for 30 min and then centrifuged at 500×g for 10 min. The growth medium was subsequently removed, and the pellet containing the cysts was washed twice with PBS. To obtain a pure cyst suspension, the dead trophozoites were removed through hypotonic lysis in distilled water for 12–24 h at 4 °C, followed by subsequent washings in distilled water to remove debris (Kane et al. 1991). The hypotonic lysis cysts were then subjected to a first quality control, either by optical microscopy or direct immunofluorescence, using the Merifluor Cryptosporidium/Giardia (Meridian Bioscience) kit, according to the manufacturer’s instructions.
Giardia (8–12 μm × 7–10 μm) fluorescent green apple structures were identified, and the number of cysts was recorded. Approximately 5 × 106 cysts were obtained from one single in vitro culture step.
Cryptosporidium parvum oocysts (at a dose of 1 × 108) were provided by the laboratory of Creative Science Company, Pentlands Science Park, Edinburgh (UK).
Toxoplasma gondii oocysts (at a dose of 6 × 106) were provided by the Department of Tropical Parasitology Medical University of Gdansk (Poland) and by the Institute of Parasitology, University of Wien (Austria), and held in a 2% H2SO4 solution at 4 °C.
Mollusk collection, pre-evaluation of protozoan contamination, acclimatization and spiking
Before each experiment, 2000 specimens of M. galloprovincialis, 2000 of R. philippinarum and 1000 of O. edulis were purchased from a depuration plant in the province of Bari, and underwent depuration for 3 weeks.
For each species, a pool was randomly selected and then subjected to RT-PCRs following the procedures described in Marangi et al. (2015) in order to exclude possible natural contamination by the three pathogens investigated.
Then 500 specimens of M. galloprovincialis, 500 of R. philippinarum and 300 of O. edulis were placed in the tanks and acclimatized for 10 days before processing. Salinity, nitrites, oxygenation, water temperature, room temperature and mortality parameters were constantly monitored.
Each tank and each mollusk species was spiked with each quantity (e.g., 1000, 10,000, 100,000) of each pathogen (Toxoplasma, Cryptosporidium and Giardia) separately and one at a time. An unspiked control tank was included for each mollusk species and each pathogen.
For each mollusk species, 30 live and non-injured specimens for each concentration and for each pathogen, were taken from the aquarium at 24 h and 72 h post-contamination.
The 30 specimens of each mollusk species were then divided into two subpools of 15. Hemolymph (H) and intervalvular liquid (IL) were aspirated individually from each subset of mollusks, using a needle inserted into the lateral adductor muscle; the gills (G) and digestive glands (DG) were removed following the procedures previously described (Graczyk et al. 1999) and then pooled. Four aliquots (H, IL, G, DG) were created for each pool, corresponding to the four anatomic sites of each mollusk species. All aliquots were stored at − 20 °C pending molecular analysis.
The experiments were conducted sequentially, each time using one mollusk species and one protozoan parasite species. The tanks were washed and refilled with clean water for each experiment.
DNA extraction and Real-Time PCR protocols
The Nucleospin Tissue Kit (Macherey-Nagel, Germany) was used according to the manufacturer’s instructions, to extract DNA from all aliquots according to anatomic sites, mollusk species, pathogen (at all concentrations), and at 24 h and 72 h post-contamination.
RT-PCR protocols were set up based on EvaGreen® for Toxoplasma and Cryptosporidium, and TaqMan for Giardia. The specificity of each primer-pair was first tested with DNA of all positive controls for Toxoplasma, Giardia and Cryptosporidium and then with DNA of mussels used as a negative control. Each primer-pair released a fluorescent signal specific for the investigated species and genes.
For G. duodenalis, standard RT-PCR was performed using a CFX-96 Real-Time PCR system (BioRad Laboratories, Hercules CA, USA). Briefly, the reaction mixture (25 μl) contained 10 μL of iQ™ Supermix (Bio-Rad Laboratories, Hercules CA, USA), 0.24 μM of TaqMan® TAMRA™ probe (FAM-5′-CCCGCGGCGGTCCCTGCTAG-3′-TAMRA) (Applied Biosystems, UK) and 3.12 μM of species-specific primers targeting SSU Rrna gene (Giardia-80F 5′-GACGGCTCAGGACAACGGTT-3′ and Giardia-127R 5′-TTGCCAGCGGTGTCCG-3′) to amplify a 62-bp fragment (Nazeer et al. 2013). The PCR cycling conditions were as follows: initial denaturation at 95 °C for 3 min, followed by 42 cycles at 95 °C for 30 s, and 1 min at 62 °C. Extracted genomic DNA, the positive control for Giardia, and the no template (negative control, NTC, ultrapure Millipore water) in 5 μL were added to each reaction mix. Each sample was run in duplicate, and the amplification cycle threshold (Ct) mean value was recorded.
For Cryptosporidium and Toxoplasma, RT-PCR was performed using a CFX-96 Real-Time PCR system (BioRad Laboratories, Hercules CA, USA). Briefly, PCRs were carried out in a final volume of 20 μL, using 10 μL of SsoFast™EvaGreen® Supermix (cat. no. 172-5201; Bio-Rad Laboratories, Hercules CA, USA) and 0.5 μM of species-specific primers for COWP C. parvum gene (CRYINT2D-F: 5′-TTTGTTGAAGARGGAAATAGATGTG-3′ and CRY2D-R: 5′-GGACKGAAATRCAGGCATTATCYTG-3′) or B1 T. gondii locus (TOXB41-F: 5′-CGAAGCTGAGATGCTCAAAGTC-3′ and TOXB169-R: 5′-AATCCACGTCTGGGAAGAACTC-3′) (see Aksoy et al. 2014; Marangi et al. 2015). Extracted genomic DNA, the positive control for Cryptosporidium and Toxoplasma, and the no template (negative control, NTC, ultrapure Millipore water) in 5 μL were added to each reaction mix. PCR cycling conditions were as follows: initial denaturation at 98 °C for 2 min, followed by 40 cycles at 98 °C for 5 s, and 15 s at 50 °C (Cryptosporidium) or 62 °C (T. gondii). Melting analysis was performed at the end of each PCR run (70 °C to 95 °C at 0.5 °C/5 s). Each sample was run in duplicate, and the amplification cycle threshold (Ct) and melting temperature (Tm) mean values were calculated. The criteria used to define a positive sample were (a) a detectable amplification curve, (b) a Tm value equal to the Tm value of the specific positive control, and (c) a–dF/dT fluorescence value > 2.
Sample analysis by Q3 platform
All samples tested by standard RT-PCR were also tested by Q3. For Q3 assay, the reaction mixture consisted of 5 μL of mix (containing 2.5 μL of SsoFast™EvaGreen® Supermix (2X), 1 μL of species-specific primers (0.5 μM) for Toxoplasma and Cryptosporidium and 0.5 μL of DEPC water, or 2.5 μL of iQ™ Supermix (2X), 1 μL of TaqMan probe (3.12 μM), 0.5 μL of primers for Giardia and 0.5 μL of DEPC water), + 2 μL of extracted DNA sample added in each cartridge well.
A positive control (DNA extracted from T. gondii, Cryptosporidium and Giardia culture) and a negative control (water) were also included in each cartridge. The reaction parameters were as follows: initial denaturation at 98 °C for 120 s, followed by 40 cycles at 98 °C for 5 s, and 62 °C for 15 s (Toxoplasma); 97 °C for 180 s followed by 42 cycles at 97 ° C for 30 s and 62 °C for 60 s (Giardia); 98 °C for 120 s followed by 40 cycles at 98 °C for 5 s and 55 °C for 15 s (Cryptosporidium). Each sample was tested in triplicate.
Statistical analysis
Sensitivity (S), specificity (SP), prevalence (P), positive and negative predictive values (PV +/−) were calculated for each PCR (RT-PCR and Q3) assay, according to mollusk species, pathogen, and at 24 h and 72 h post-contamination.
The concordance between the two molecular tests used was evaluated using the kappa (K) coefficient (Landis and Koch 1977). K value varies from 0 (perfect discordance) to 1 (perfect concordance) and its interpretation was given as follows: 0.81–1.00, very good concordance; 0.61–0.80, substantial concordance; 0.41–0.60, discrete concordance; 0.20–0.40, moderate concordance; 0.10–0.20, weak concordance (Landis and Koch 1977).
The differences in the performance of RT-PCR and Q3 for the various anatomic sites (hemolymph, intervalvular liquid, gills, and digestive gland), compared to the initial cyst/oocyst number and in relation to the two groups of mussels/clams (repeatability) were tested by Chi Square or by Fisher’s exact test, as appropriate. The software programs used were SPSS for Windows version 13.0 and WinEpi 2.0 (available online at: http://www.winepi.net/uk/index.htm).
Results
All the randomly pooled samples tested for possible natural contamination by Toxoplasma, Cryptosporidium and Giardia tested negative for the three investigated pathogens (data not shown).
Both RT-PCR and the Q3 platform were used to test a total of 552 pooled samples from the four anatomic sites of each mollusk species (n = 144 for M. galloprovincialis, n = 144 for R. philippinarum and n = 120 for O. edulis) spiked with the three pathogens at all concentrations (1000, 10,000 and 100,000 oo/cysts) at 24 h and 72 h post-spiking, together with 144 negative controls. The raw number of positive (with the corresponding Ct values) and negative pool samples obtained is shown in Supplementary Table 1, and the results are summarized in Table 1. The negative controls were always negative, with both standard RT-PCR and Q3. Thus, in the absence of false positive results, the specificity and positive predictive values were 100% in all cases and are not further reported in the text and tables. In general, both RT-PCR and the Q3 system detected the three pathogens in all the mollusks species, although RT-PCR performance was better than the Q3 system, both at 24 h (20.2% vs 10.5%, respectively; p < 0.0001) and at 72 h (12.3% vs 5%, respectively; p < 0.0001) (Table 1). However, performances differed according to the protozoan species.
Table 1.
Mollusk species and Protozoans | N. of positive pools (%) | |||||
---|---|---|---|---|---|---|
24 h | 72 h | Total (24 h + 72 h) | ||||
RT (n = 24) |
Q3 (n = 24) |
RT (n = 24) |
Q3 (n = 24) |
RT (n = 48) |
Q3 (n = 48) |
|
Mytilus galloprovincialis | ||||||
Toxoplasma | 8 (33.3) | 10 (41.7) | 4 (16.7) | 4 (16.7) | 12 (25.0) | 14 (29.2) |
Cryptosporidium | 0 | 6 (25) | 0 | 0 | 0 | 6 (12.5) |
Giardia | 8 (33.3) | 2 (8.3) | 3 (12.5) | 0 | 11 (22.9) | 2 (4.2) |
Negative control | – | – | – | – | ||
Ruditapes philippinarum | ||||||
Toxoplasma | 19 (79.2)a | 6 (25)a | 8 (33.3) | 5 (20.8) | 27 (56.3)b | 11 (22.9)b |
Cryptosporidium | 8 (33.3) | 2 (8.3) | 3 (12.5) | 0 | 11 (22.9) | 2 (4.2) |
Giardia | 12 (50) | 10 (41.7) | 3 (12.5) | 0 | 15 (31.3) | 10 (20.8) |
Negative control | – | – | – | – | ||
Ostrea edulis | ||||||
Toxoplasma | 24 (100)d | 14 (58.3)d | 4 (16.7) | 4 (16.7) | 28 (58.3) | 18 (37.5) |
Cryptosporidium | 8 (33.3) | 0 | 9 (37.5) | 2 (8.3) | 17 (35.4)e | 2 (4.2)e |
Giardia | NP | NP | NP | NP | NP | NP |
Negative control | – | – | – | – | ||
Total |
87/192f (45.3) |
50/192f (26.0) |
34/192 (17.7) |
15/192 (7.8) |
121/384c (31.5) |
65/384c (16.9) |
NP not performed
a–fStatistical significance is marked with the same letters for p < 0.01
For M. galloprovincialis, the prevalence, sensitivity and negative predictive values of standard RT-PCR and the Q3 platform at 24 h and 72 h for Toxoplasma, Cryptosporidium and Giardia are shown in Table 2. The performance of the Q3 platform was comparable to RT-PCR in detecting Toxoplasma oocysts at 24 h (41.7% vs 33.3%, respectively) and 72 h (both 16.7%), while only Q3 was able to detect Cryptosporidium oocysts (25% vs 0%; p = 0.0219) at 24 h (Table 2). RT-PCR was superior to Q3 in detecting Giardia cysts both at 24 h (33.3% vs 8.3%) and at 72 h (12.5% vs 0%), although these differences were not statistically significant (Table 2). The two techniques performed equally well in relation to the different number of parasites spiked, detecting Toxoplasma at the three concentrations at 24 h, but only at the highest concentration at 72 h. For Giardia only RT-PCR was able to detect cysts at all three concentrations at 24 h (Supplementary Table 1). The highest positivity for the three protozoa was detected in the digestive glands (Supplementary Table 1).
Table 2.
PCR | Toxoplasma | Cryptosporidium | Giardia | |||
---|---|---|---|---|---|---|
P/Sa | PV− | P/S | PV− | P/S | PV− | |
RT 24 h |
33.3% (14.5–52.2) |
60.0% (44.8–75.2) |
0% |
50.0% (35.9–64.1) |
33.3% (14.5–52.2) |
60.0% (44.8–75.2) |
Q3 24 h |
41.7% (21.9–61.4) |
63.2% (47.8–78.5) |
25.0% (7.7–42.3) |
57.1% (42.2–72.1) |
8.3% (− 2.7 to 19.4) |
52.2% (37.7–66.6) |
RT 72 h |
16.7% (1.8–31.6) |
54.5% (39.8–69.3) |
0% |
50.0% (35.9–64.1) |
12.5% (− 0.7 to 25.7) |
53.3% (38.8–67.9) |
Q3 72 h |
16.7% (1.8–31.6) |
54.5% (39.8–69.3) |
0% |
50.0% (35.9–64.1) |
0% |
50.0% (35.9–64.1) |
RT 24 h + 72 h |
25.0% (7.7–42.3) |
57.1% (42.2–72.1) |
0% |
50.0% (35.9–64.1) |
22.9% (11.0–34.8) |
56.5% (45.9–67.0) |
Q3 24 h + 72 h |
29.2% (16.3–42.0) |
58.5% (47.9–69.2) |
12.5% (− 0.7–25.7) |
53.3% (38.8–67.9) |
4.2% (− 1.5 to 9.8) |
51.1% (41.0–61.2%) |
aPrevalence coincides with sensitivity because no false positives were detected
RT-PCR and Q3 amplification curves of digestive glands at the concentration of 100,000, digestive glands at the concentration of 10,000, intervalvular liquid and digestive glands at the concentration of 1000 of M. galloprovincialis samples positive to Toxoplasma at 24 h post-contamination are reported in Supplementary Figs. 1 and 2, respectively.
For R. philippinarum, the prevalence/sensitivity and the positive predictive values of RT-PCR and Q3 at 24 h and 72 h for Toxoplasma, Cryptosporidium and Giardia are shown in Table 3.
Table 3.
PCR | Toxoplasma | Cryptosporidium | Giardia | |||
---|---|---|---|---|---|---|
P/Sa | PV− | P/S | PV− | P/S | PV− | |
RT 24 h |
79.2% (62.9–95.4) |
82.8% (69.0–96.5) |
33.3% (14.5–52.2) |
60.0% (44.8–75.2) |
50.0% (30.0–70.0) |
66.7% (51.3–82.1) |
Q3 24 h |
25.0% (7.7–42.3) |
57.1% (42.2–72.1) |
8.3% (2.7–19.4) |
52.2% (37.7–66.6) |
41.7% (21.9–61.4) |
63.2% (47.8–78.5) |
RT 72 h |
33.3% (14.5–52.2) |
60.0% (44.8–75.2) |
12.5% (− 0.7 to 25.7) |
53.3% (38.8–67.9) |
12.5% (− 0.7 to 25.7) |
53.3% (38.8–67.9) |
Q3 72 h |
20.8% (4.6–37.1) |
55.8% (41.0–70.7) |
0% |
50.0% (35.9–64.1) |
0% |
50.0% (35.9–64.1) |
RT 24 h + 72 h |
56.3% (42.2–70.3) |
69.6% (58.7–80.4) |
22.9% (11.0–34.8) |
56.5% (45.9–67.0) |
31.3% (18.1–44.4) |
59.3% (48.6–70.0) |
Q3 24 h + 72 h |
22.9% (11.0–34.8) |
56.5% (45.9–67.0) |
4.2% (− 1.5 to 9.8) |
51.1% (41.0–61.2%) |
20.8% (9.3–32.3) |
55.8% (45.3–66.3) |
aPrevalence coincides with the sensitivity because no false positives were detected
RT-PCR detected more positives to Toxoplasma than Q3 at 24 h (79.2% vs 25%; p < 0.0001). No significant differences in positivity among the tissues were detected for RT-PCR and Q3. There were also no significant differences in relation to concentrations, except that RT-PCR performed better than Q3 at 72 h at the concentration of 100,000 Cryptosporidium oocysts.
The RT-PCR and Q3 amplification curves of hemolymph at the concentration of 100,000, gills and digestive glands at the concentration of 10,000 and intervalvular liquid at the concentration of 1000 of R. philippinarum samples positive to Toxoplasma at 24 h post-contamination are reported in Supplementary Figs. 3 and 4, respectively.
For O. edulis, the prevalence, sensitivity and the positive predictive values of RT-PCR and Q3 at 24 h and 72 h are shown only for Toxoplasma and Cryptosporidium (Table 4).
Table 4.
PCR | Toxoplasma | Cryptosporidium | ||
---|---|---|---|---|
P/Sa | PV− | P/Sa | PV− | |
RT 24 h | 100% | 100% |
33.3% (14.5–52.2) |
60.0% (44.8–75.2) |
Q3 24 h |
58.3% (38.6–78.1) |
70.6% (55.3–85.9) |
0% |
50.0% (35.9–64.1) |
RT 72 h |
16.7% (1.8–31.6) |
54.5% (39.8–69.3) |
37.5% (18.1–56.9) |
61.5% (46.3–76.8) |
Q3 72 h |
16.7% (1.8–31.6) |
54.5% (39.8–69.3) |
8.3% (− 2.7 to 19.4) |
52.2% (37.7–66.6) |
RT 24 h + 72 h |
58.3% (38.6–78.1) |
70.6% (55.3–85.9) |
35.4% (21.9–48.9) |
60.8% (50.0–71.5) |
Q3 24 h + 72 h |
37.5% (18.1–56.9) |
61.5% (46.3–76.8) |
4.2% (− 1.5 to 9.8) |
51.1% (41.0–61.2%) |
aPrevalence coincides with sensitivity because no false positives were detected
Giardia in O. edulis could not be investigated due to technical problems. RT-PCR performed better than Q3 in detecting Toxoplasma (100% vs 58.3%, respectively, p < 0.001) and Cryptosporidium (33.3% vs 0%; p < 0.005) at 24 h. Concerning concentration, both techniques detected Toxoplasma oocysts only at the highest concentration (100,000 oocysts) (Supplementary Table 1).
The RT-PCR and Q3 amplification curves of digestive gland and intervalvular liquid at the concentration of 100,000, intervalvular liquid and digestive glands at the concentration of 10,000, digestive glands, intervalvular liquid and gills at the concentration of 1000 of O. edulis samples positive to Toxoplasma at 24 h post-contamination are reported in Supplementary Figs. 5 and 6, respectively.
On the whole, for M. galloprovincialis, a very good concordance was found between RT-PCR and Q3 for Toxoplasma at 24 h and 72 h. For R. philippinarum, a very good concordance was found between RT-PCR and Q3 for Giardia at 24 h, and a substantial concordance was found for Toxoplasma at 72 h. For O. edulis, a discrete concordance was found between RT-PCR and Q3 for Toxoplasma at 24 h and a very goodconcordance was found for Toxoplasma at 72 h (Table 5).
Table 5.
Mollusk species | PCR | Toxoplasma | Cryptosporidium | Giardia | |||
---|---|---|---|---|---|---|---|
K | p | K | p | K | p | ||
Mytilus galloprovincialis | RT 24 h versus Q3 24 h | 0.824 | < 0.01 | nc | – | 0.308 | < 0.05 |
RT 72 h versus Q3 72 h | 1.000 | < 0.01 | nc | – | – | – | |
Ruditapes philippinarum | RT 24 h versus Q3 24 h | 0.161 | ns | 0.308 | < 0.05 | 0.833 | < 0.01 |
RT 72 h versus Q3 72 h | 0.690 | < 0.01 | nc | – | – | – | |
Ostrea edulis | RT 24 h versus Q3 24 h | 0.651 | < 0.01 | nc | – | np | |
RT 72 h versus Q3 72 h | 1.000 | < 0.01 | 0.263 | ns | np | – |
Concordance (K) value and its interpretation: > 0.81 almost perfect concordance; 0.61–0.80 substantial concordance; 0.41–0.60 discrete concordance; 0.20–0.40 moderate concordance; 0.1–0.20 weak concordance
np not performed, ns not significant, nc not computable because one of the variables is a constant (all negatives)
Discussion
The efficiency of the Q3 “lab-on-chip” molecular platform developed by STMicroelectronics for detection of the investigated zoonotic protozoans (Toxoplasma, Cryptosporidium and Giardia) in three mollusk species (M. galloprovincialis, R. philippinarum and O. edulis) was good in comparison with the results obtained by RT-PCR. This is the first study on the use of lab-on-chip prototype technology in a food matrix, i.e., edible marine mollusks.
We expected to find a lower number of samples contaminated by spiked oo/cysts than the number of tested pools (Table 1), since laboratory conditions simulate the situation in the natural environment, i.e., not all mollusks assume protozoan parasites. Although many pools tested negative because of the very low concentration, Q3 demonstrated a good ability to detect all the tested pathogens (Ct ranged from 24.5 to 38), which was comparable to standard RT-PCR procedures (Ct ranged from 30.4 to 38); Q3 was particularly efficient for Toxoplasma in mussels (Ct ranged from 31 to 36), and in oysters (Ct ranged from 24.5 to 35.5) compared to RT-PCR (Ct ranged from 34 to 38 in mussels and 30.4 to 37.8 in oysters), especially at the highest protozoan concentration and after a longer exposure time.
While RT-PCR and Q3 were found to be highly efficient in detecting Toxoplasma, their lower efficiency for Giardia and Cryptosporidium might be due to the gene portions used. Unlike concentration and exposure time, which obviously increase the chances of detecting the pathogens, the reasons for a higher correlation between pathogens and mollusk species are unknown.
The lack of differences between the anatomic sites—independently of the species—is a very interesting point both in terms of laboratory practice and of platform applicability, since it means that whole mollusks can be used for pathogen detection.
Although few documented cases of infection are actually linked to the consumption of mollusks (mostly oysters contaminated by Giardia and/or Cryptosporidium) (Baumgartner et al. 2000; Potasman et al. 2002; Robertson 2007), the lack of epidemiological information for shellfish is likely related to inadequate diagnosis, detection or reporting, due to the long time lapse between contamination and the onset of clinical signs (usually 1–2 weeks for both Giardia and Cryptosporidium infections). Consequently, medical practitioners are unlikely to associate these infections with mollusks eaten several days previously, particularly because shellfish are not commonly recognized as potential sources of infection involving these parasites (Ryan et al. 2016). Even if the cause is actually suspected, the food item has already been consumed in most cases and is therefore unavailable for testing (Robertson 2007). The recent conclusion of Sutthikornchai et al. (2016) that oysters can be an effective transmission vehicle for Cryptosporidium oocysts, especially within 24–72 h of contamination, with viable oocysts present at up to 7 days post-contamination, further demonstrates the reason for concern about public health. The association between shellfish consumption and Toxoplasma infection is even less likely to be recognized, because immunocompetent individuals are unlikely even to recognize that they have been infected. When infection is suspected, and congenital transmission is excluded, it is more likely that water sources, undercooked meat products or direct transmission from cat feces are investigated as possible sources.
Current European legislation does not include the investigated protozoans (Giardia, Cryptosporidium and Toxoplasma) in the list of routinely monitored pathogens, including in mollusk species. However, according to the European Technology Platform “Food for Life”—which emphasizes the need not only to detect parasites, but also to develop specific innovative methods to include in HACCP practice—the Q3 platform might be a useful and innovative tool for screening the safety of these food products.
Conclusion
The Q3 platform can be considered a valuable molecular tool for detecting all the investigated pathogens, with a particularly high level of efficiency in detecting Toxoplasma. Q3 detection efficiency for Cryptosporidium could be improved by using different target genes and/or by the use of different PCR protocols.
This work represents a first technological approach and a challenge for implementing pathogen detection in different mollusk species. Pending improvements to its performance (even at a lower range of pathogen concentration) on all mollusk species, the “lab-on-chip” molecular integrated Q3 platform provides a feasible tool that is small, rapid and easy-to-use, and thus suitable for in-the-field testing by companies producing and/or marketing mollusks.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
The study was funded by “New Strategies for Improvement of Food Safety: Prevention, Control, Correction” (S.I.Mi.S.A.)—PON02_00186_3417512—PON Ricerca e Competitività 2007–2013” (PO Puglia FESR 2007e2013 Asse I, Linea 1.2dPO Puglia FSE 2007e2013 Asse IV). The authors wish to thank Anna Lass of the Department of Tropical Parasitology, Medical University of Gdansk (Poland), and Anja Joachim of the Institute of Parasitology, University of Wien (Austria) for providing Toxoplasma oocyst strains, Tiziana Caradonna for her helpful work in the lab, Tommaso Marazia for his outstanding assistance in aquarium system management, and Alessandra Barlaam and Sarah Christopher for the English revision of the MS.
Authors’ contributions
AG, GN and DO conceived the study design; LP performed in vitro culture; FF and MC developed the Q3 system and trained for the Q3 analysis; MM, MSL and GA collected samples and performed the Real Time and Q3 analysis; GC performed the statistical analysis; AG, GN, MM, LP, MC and DO interpreted the data and wrote the paper. All authors contributed to editing the manuscript. All authors read and approved the final version of the manuscript.
Footnotes
Aquarium system (Acqua and Co. S.r.l., Cadelbosco Sopra, Ravenna, Italy). Tanks with independent hydraulics, cool chiller module units with 0.5cv power each; 2 high-frequency 55 W UVC lamps with two spare lamps; hydraulics complete with PVC pipes, valves, fittings and discharges—double single-phase electric panel with magnetic circuit breakers according to EEC standards; 1500 L salt water tank in bins c680 with lid and module filling system.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Aksoy U, Marangi M, Papini R, Ozkoc S, Bayram Delibas S, Giangaspero A. Detection of Toxoplasma gondii and Cyclospora cayetanensis in Mytilus galloprovincialis from Izmir Province coast (Turkey) by Real time PCR/High-Resolution Melting analysis (HRM) Food Microbiol. 2014;44:128–135. doi: 10.1016/j.fm.2014.05.012. [DOI] [PubMed] [Google Scholar]
- Baumgartner A, Marder HP, Munzinger J, Siegrist HH. Frequency of Cryptosporidium spp. as cause of human gastrointestinal disease in Switzerland and possible sources of infection. Schweiz Med Wochenschr. 2000;130:1252–1258. [PubMed] [Google Scholar]
- Biava M, Colavita F, Marzorati A, Russo D, Pirola D, Cocci A, Petrocelli A, Delli Guanti M, Cataldi G, Kamara TA, Kamara AS, Konneh K, Cannas A, Coen S, Quartu S, Meschi S, Valli MB, Mazzarelli A, Venditti C, Grassi G, Rozera G, Castilletti C, Mirazimi A, Capobianchi MR, Ippolito G, Miccio R, Di Caro A. Evaluation of a rapid and sensitive RT-qPCR assay for the detection of Ebola Virus. J Virol Methods. 2018;252:70–74. doi: 10.1016/j.jviromet.2017.11.009. [DOI] [PubMed] [Google Scholar]
- Coupe A, Howe L, Burrows E, Sine A, Pita A, Velathanthiri N, Vallée E, Hayman D, Shapiro K, Roe WD. First report of Toxoplasma gondii sporulated oocysts and Giardia duodenalis in commercial green-lipped mussels Perna canaliculus in New Zealand. Parasitol Res. 2018;117:1453–1463. doi: 10.1007/s00436-018-5832-8. [DOI] [PubMed] [Google Scholar]
- Dubey JP, Jones JL. Toxoplasma gondii infection in humans and animals in the United States. Int J Parasitol. 2008;38:1257–1278. doi: 10.1016/j.ijpara.2008.03.007. [DOI] [PubMed] [Google Scholar]
- El Sheikha AF, Montet D. How to determine the geographical origin of seafood? Crit Rev Food Sci Nutr. 2016;56:306–317. doi: 10.1080/10408398.2012.745478. [DOI] [PubMed] [Google Scholar]
- Eurostat (2015) http://ec.europa.eu/eurostat/statistics-explained/index.php/Aquaculture_statistics/. Accessed 13 Dec 2018
- FAO (2015) Fisheries and aquaculture statistics http://www.fao.org/3/a-i7989t.pdf/. Accessed 13 Dec 2018
- FAO-WHO (2014a) http://apps.who.int/iris/bitstream/10665/112672/1/9789241564700_eng.pdf/. Accessed 13 Dec 2018
- FAO-WHO (2014b) http://www.fao.org/tempref/codex/Meetings/CCFFP/ccffp34/Shellfish%20Sanitation%20Initiation%20meeting%20Report%20%20final%20for%20CCFFP.pdf. Accessed 13 Dec 2018
- Giangaspero A, Cirillo R, Lacasella V, Lonigro A, Marangi M, Cavallo P, Berrilli F, Di Cave D, Brandonisio O. Giardia and Cryptosporidium in inflowing water and harvested shellfish in a lagoon in Southern Italy. Parasitol Int. 2009;58:12–17. doi: 10.1016/j.parint.2008.07.003. [DOI] [PubMed] [Google Scholar]
- Giangaspero A, Papini R, Marangi M, Koehler AV, Gasser RB. Cryptosporidium parvum genotype IIa and Giardia duodenalis assemblage A in Mytilus galloprovincialis on sale at local food markets. Int J Food Microbiol. 2014;171:62–67. doi: 10.1016/j.ijfoodmicro.2013.11.022. [DOI] [PubMed] [Google Scholar]
- Graczyk TK, Fayer R, Lewis EJ, Trout JM, Farley CA. Cryptosporidium oocysts in Bent mussels (Ischadium recurvum) in the Chesapeake Bay. Parasitol Res. 1999;85:518–521. doi: 10.1007/s004360050590. [DOI] [PubMed] [Google Scholar]
- Ghozzi K, Marangi M, Papini R, Lahmar I, Challouf R, Houas N, Ben Dhiab R, Normanno G, Babba H, Giangaspero A. First report of Tunisian coastal water contamination by protozoan parasites using mollusk bivalves as biological indicators. Maar Pollut Bull. 2017;117:197–202. doi: 10.1016/j.marpolbul.2017.01.057. [DOI] [PubMed] [Google Scholar]
- Hohweyer J, Dumètre A, Aubert D, Azas N, Villena I. Tools and methods for detecting and characterizing Giardia, Cryptosporidium and Toxoplasma parasites in marine mollusks. J Food Prot. 2013;76:1649–1657. doi: 10.4315/0362-028X.JFP-13-002. [DOI] [PubMed] [Google Scholar]
- Kane AV, Ward HD, Keusch GT, Pereira ME. In vitro encystation of Giardia lamblia: large-scale production of in vitro cysts and strain and clone differences in encystation efficiency. J Parasitol. 1991;77:974–981. doi: 10.2307/3282752. [DOI] [PubMed] [Google Scholar]
- Keister DB. Axenic culture of Giardia lamblia in TYI-S-33 medium supplemented with bile. Trans R Soc Trop Med Hyg. 1983;77:487–488. doi: 10.1016/0035-9203(83)90120-7. [DOI] [PubMed] [Google Scholar]
- Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–174. doi: 10.2307/2529310. [DOI] [PubMed] [Google Scholar]
- Marangi M, Giangaspero A, Lacasella V, Lonigro A, Gasser RB. Multiplex PCR for the detection and quantification of zoonotic taxa of Giardia, Cryptosporidium and Toxoplasma in wastewater and mussels. Mol Cell Probes. 2015;29:122–125. doi: 10.1016/j.mcp.2015.01.001. [DOI] [PubMed] [Google Scholar]
- Marziliano N, Notarangelo MF, Cereda M, Caporale V, Coppini L, Demola MA, Guidorossi A, Crocamo A, Pigazzani F, Boffetti F, Del Giudice F, Orsini F, Pirola D, Cocci A, Manzalini C, Casu G, Bianchessi M, Ardissino D, Merlini PA. Rapid and portable, lab-on-chip, point-of-care genotyping for evaluating clopidogrel metabolism. Clin Chim Acta. 2015;451:240–246. doi: 10.1016/j.cca.2015.10.003. [DOI] [PubMed] [Google Scholar]
- Miller WA, Gardner IA, Atwill ER, Leutenegger CM, Miller MA, Hedrick RP, Melli AC, Barnes NM, Conrad PA. Evaluation of methods for improved detection of Cryptosporidium spp. in mussels (Mytilus californianus) J Microbiol Methods. 2006;65:367–379. doi: 10.1016/j.mimet.2005.08.011. [DOI] [PubMed] [Google Scholar]
- Mmbaga BT, Houpt ER. Cryptosporidium and Giardia infections in children: a review. Pediatr Clin North Am. 2017;64:837–850. doi: 10.1016/j.pcl.2017.03.014. [DOI] [PubMed] [Google Scholar]
- Nazeer JT, Khalifa KES, von Thien H, El-Sibaei MM, Abdel-Hamid MY, Tawfik RA, Tannich E. Use of multiplex real-time PCR for detection of common diarrhea causing protozoan parasites in Egypt. Parasitol Res. 2013;112:595–601. doi: 10.1007/s00436-012-3171-8. [DOI] [PubMed] [Google Scholar]
- Potasman I, Paz A, Odeh M. Infectious outbreaks associated with bivalve shellfish consumption: a worldwide perspective. Clin Infect Dis. 2002;35:921–928. doi: 10.1086/342330. [DOI] [PubMed] [Google Scholar]
- Putignani L, Mancinelli L, Del Chierico F, Menichella D, Adlerstein D, Angelici MC, Marangi M, Berrilli F, Caffara M, di Regalbono DA, Giangaspero A. Investigation of Toxoplasma gondii presence in farmed shellfish by nested-PCR and real-time PCR fluorescent amplicon generation assay (FLAG) Exp Parasitol. 2012;127:409–417. doi: 10.1016/j.exppara.2010.09.007. [DOI] [PubMed] [Google Scholar]
- Robert-Gangneux F, Dardé ML. Epidemiology of and diagnostic strategies for toxoplasmosis. J Clin Microbiol. 2012;25:264–296. doi: 10.1128/CMR.05013-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Robertson LJ. The potential for marine bivalve shellfish to act as transmission vehicles for outbreaks of protozoan infections in humans: a review. Int J Food Microbiol. 2007;120:201–216. doi: 10.1016/j.ijfoodmicro.2007.07.058. [DOI] [PubMed] [Google Scholar]
- Ryan U, Zahedi A, Paparini A. Cryptosporidium in humans and animals a one health approach to prophylaxis. Parasite Immunol. 2016;38:535–547. doi: 10.1111/pim.12350. [DOI] [PubMed] [Google Scholar]
- Sutthikornchai C, Popruk S, Chumpolbanchorn K, Sukhumavasi W, Sukthana Y. Oyster is an effective transmission vehicle for Cryptosporidium infection in human. Asian Pac J Trop Dis. 2016;9:562–566. doi: 10.1016/j.apjtm.2016.04.018. [DOI] [PubMed] [Google Scholar]
- Thompson RC, Olson ME, Zhu G, Enomoto S, Abrahamsen MS, Hijjawi NS. Cryptosporidium and cryptosporidiosis. Adv Parasitol. 2005;59:77–158. doi: 10.1016/S0065-308X(05)59002-X. [DOI] [PubMed] [Google Scholar]
- Yoon JY, Kim B. Lab-on-a-chip pathogen sensors for food safety. Sensors (Basel) 2012;12:10713–10741. doi: 10.3390/s120810713. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.