RAPID DNA EXTRACTION AND PCR VALIDATION FOR DIRECT DETECTION OF Listeria monocytogenes IN RAW MILK

.

The routine method for detecting L. monocytogenes in food samples involves the use of selective enrichments and subsequent culturing on selective media; it is a laborious and time-consuming procedure.Therefore, a rapid, sensible and reliable method for L. monocytogenes detection is desirable (10,11).The aim of this study was to validate a method for removing PCR inhibitory factors, like milk fat, and other PCR techniques for differentiating L. monocytogenes from other Listeria spp.In order to find out the usefulness of combining an extraction procedure with the PCR method.The aim of this study was to validate a method for detecting L. monocytogenes in raw milk.

Strains and master mell banks (MCB)
Sixteen bacterial genera distributed in 39 strains were used: 6 strains of Listeria and 33 of other contaminant and/or normal biota of raw milk (Table 1).All the strains were preserved in culture media, supplemented with glycerol at 30% (v/v) and stored at -70ºC.The whole MCB was followed up with viability and microbial purity tests every two months of conservation (12,13).

INTRODUCTION
Listeria monocytogenes has been recognized as an important transmission route leading to human listeriosis, becoming a public health problem.It is a food-borne pathogen, responsible for severe and fatal infections.This microorganism can enter into the food chain through animals that shed cells in their milk and faeces.

PCR reaction
Two primer sets were used: L1 (CTCCATAAAGGTGACCCT) and U1 (CAGCMGCCGCGGTAATWC), that recognize the Listeria genus amplifying a fragment of about 938bp of the 16S rDNA ; and LF (CAAACGTTAACAACGCAGTA) and LR (TCCAGAGTGATCGATGTTAA) that recognize the hlyA gene of L. monocytogenes, amplifying a 750bp fragment.The final volume of the PCR reaction was 50µl, and the mixture was composed of 1X PCR buffer, 1.5mM of MgCl 2 , 0.2mM (each) dNTP, 20pmol of primer, and 2U TaqDNA polymerase (Invitrogene TM ).Five µl of DNA (~200ng) of the extracted sample were used.Temperature cycling was controlled in a BioRad Gene Cycler TM programmed as follows: 95 0 C for 1min, followed by 30 cycles of 94 0 C for 30s, 51 0 C for 20s, 72 0 C for 30s and a final extension step at 72 0 C for 8min (15,17,18).

Steps for PCR validation Physical and chemical analysis of the raw milk samples
Twenty samples of raw milk were submitted to physical and chemical analyses according to the techniques recommended by the National Health Institute of Bogotá Colombia: alcohol test, pH, percentage of lactic acid, reductase assay, and tests for starch, sucrose, chlorides, formaldehyde and fatty detection, etc.

Microbiological analysis of the raw milk samples
The raw milk samples were analyzed by using the traditional Gold Standard method for L .monocytogenes detection (19).Briefly, 25ml of raw milk were taking to a final volume of 250ml with enrichment broth for Listeria (Palcam broth); the suspensions were incubated at 30 o C for 48 hours.Later, the cultures were plated on Palcam agar and incubated at 30 o C from 24 to 48h.After the incubation time, the positive colonies for esculine hydrolysis were selected and then re-isolated on TSAYE agar (tryptone soy agar, yeast extract) and incubated at 30 o C for 24h.Colonies were Gram and catalyze tested.A second purification was made in TSAYE agar, and incubated at 35 o C for 24h; later on, Henry's illumination test was made; positive colonies for this test were assayed for motility, CAMP, RMVP and fermentation of mannitol, ramnose and xylose (10,20).

Reproducibility determination of the PCR method
The reproducibility test started with a pilot analysis with 15 replications of the PCR detection for L. monocytogenes ATCC 1915, L. innocua (Swiss), Staphylococcus aureus ATCC 29213 and Escherichia coli O127, followed by 3 replications of the PCR for each one of the 39 strains of the study.
Preliminary determination of the PCR densitivity and comparison with the gold standard and the immunological methods PCR, Gold Standard Method and an immunological method (available in Colombia) were compared through double-blind assay.For this purpose, 15 screw cap tubes with 5ml of tryptone soy broth were artificially contaminated with 10 8 CFU/ml or equivalent to tube 5 of the MacFarland scale; the strains employed for this assay were L. monocytogenes ATCC 1915, L. innocua (Swiss), E. coli O127 and S. aureus ATCC 29213.The samples were processed following the methodology described for each one of the methods mentioned above; later on, a statistical analysis was made.

DNA extraction from raw milk samples
200µl of raw milk, homogenate by vortex sample, were mixed with 400µl of lysis buffer (0.5% (w/v) N-laurylsarcosine, 50mM Tris-HCl, 25mM EDTA, pH 8.0 ± 0.2).After vortexing for 1min, the mixture was centrifuged at 15,000 rpm for 5min.Pellet was dissolved in 200µl of lysis buffer containing 0.03µg/µl of glycogen and 4µl of proteinase K (2 mg/ml).After incubation for 1h at 37 o C, 300µl of NaI solution (6M NaI in 50mM Tris-HCl, 25mM EDTA, and pH 8.0 ± 0.2) and 500µl of isopropyl alcohol were aggregated to the suspension and then centrifuged at 10000g for 5min.The pellet was washed with 35% (v/v) of isopropyl alcohol, dried for 5min at 37 o C, and then suspended in 20µl of sterilized water for PCR (21).The samples were then PCR tested.
Sensitivity of the extraction method from raw milk samples 0.5 ml of raw milk samples were artificially contaminated with 0.5ml of diluted samples of L. monocytogenes; the concentration of dilutions ranged between 10 1 and 10 9 CFU/ml; later on, the extraction procedure was executed as previously mentioned and 5µl (~200ng of DNA) of the extracted sample were taken for PCR assay following the conditions described before.

Reproducibility of PCR from raw milk samples
The PCR was made three times with each raw milk samples artificially contaminated with the smallest L. monocytogenes concentration (detection limit) that was detected in the sensitivity assay.

Statistical analysis
The results were analyzed in Epi-info 6.0d to observe the relevance of combining the Makino DNA extraction method and the Bansal modified PCR method (15,17,18,21).A previous test, composed of 10 repetitions, was used to guarantee measurement reliability; continuous variable agreement in the measures was estimated by using two measure systems.The statistic parameter used to estimate the agreement between two measures in the binary variable was the Kappa coefficient (K), which is defined as the agreement beyond chance divided by the possible agreement (22).A negative PCR (no amplification) or a non-specific amplification of sample DNA was considered as negative for L. monocytogenes; considerations for positive results were the amplification fragments of 938 and 750bp.The qualitative terms assigned to Kappa are the following: (0 -0.2 = weak; 0.2 -0.4 = good; 0.4 -0.6 = moderate; 0.6 -0.8 = substantial; 0.8 -1 = almost perfect (22,23).

Physical and chemical analysis of raw milk simples
For a microbial hazard, temperature is the major critical measure in several steps in dairy production; for this reason milk samples were aseptically taken and transported to the laboratory at 4°C, to avoid the multiplication of pathogens (24).Raw milk contains approximately 4% lactose, 3% protein and 3% fat, which could be used by several microorganisms as substrates for growth.In order to show the feasibility of the DNA extraction procedure and PCR method combination reported in this paper, it was essential to determine that raw milk samples were free of L .monocytogenes as well as testing that the physical and chemical characteristics were normal.Normal values of physical and chemical characteristics of the raw milk samples are summarized in Table 2, showing that normal physical and chemical characteristics do not interfere with the DNA extraction procedure or PCR (Figures 1 and 2).

Specificity of the PCR technique
The genomic DNA of 39 strains under study was tested in a PCR reaction.Results evidenced the high specificity of the technique since all the strains of L. monocytogenes amplified the fragments 938bp and 750bp, specifically for genus and species, respectively (15,18,25).L. innocua strains only amplified the 938bp f r a g m e n t , a n d t o o k p l a c e d u e t o t h e hybridization of the primers LU1 and U1 with the specific sequence of the coding region of the 16S rDNA of the Listeria genus.Negative amplification fragments were observed for other bacterial genera (Figure 1).In 1990, Border amplified L. monocytogenes DNA by PCR using 5 primers, two of them based on the sequences of the hlyA gene and other 3 primers based on the region 16S of the rDNA; however, the results showed fragment lengths of 702pb and 938pb (26).Allman (1995) used four primers, LO1, LO2, LO3 and LO4, to recognize the gene hlyA that produced fragment lengths of 234pb, 207pb and 204pb; this led to confirm PCR detection using the Nested PCRs technique (27).Moyra et al. (1996) used primers prfA and prfB with complementary sequences to the prfA gene, which is involved in the regulation of the listeriolysine synthesis (28).Both Allman (1995) and Moyra (1996) could only identify in their works the species, a fact that did not eliminate the interference of crossed reactions with other species of non pathogenic Listeria spp.(29).

Reproducibility of the PCR technique
The pilot test revealed that the technique is highly reproducible and statistically reliable.Since the analysis made by taking into account a kappa of a null hypothesis of 0.75 and alternative hypothesis of kappa 0.25 and a correct classification probability of 85%, generates a final Kappa of 0.85; according to some authors, this value is classified as "nearly perfect" in terms of agreement among the assayed variables (22,23).

Sensitivity of the PCR technique
Data showed an association of 95% in the reliability interval, a sensitivity of 100%, a specificity of 100%, a predictive positive value of 100%, and a predictive negative value of 100%.These results demonstrate that the detection of Listeria spp. is possible by using any of the three methods since they share the same sensitivity and specificity.One hundred percent of the predictive value for PCR (alternative method) provides high reliability, and   (30).The DNA extraction method from milk using NaI allowed bacterial DNA to be recovering.The detection limit of the PCR was 10 1 CFU/ml; subsequent treatments were not necessary for extraction to eliminate the interference caused by food components (Figures 2).The traditional method (Gold Standard) reports a detection limit above 10 2 CFU/ml.In the case of the immunology method, microorganism populations higher than 10 5 CFU/ml were required.
Finally, the PCR technique used in this study was highly reproducible and statistically reliable, showing a kappa of 0.85 in the level of agreement among assayed variables.The major finding of this paper was the possibility of reproducing Bansal PCR (17) and Makino DNA extraction (21) procedures, adapted to our conditions, and of validating the PCR technique for L. monocytogenes detection in raw milk samples.In recent years L. monocytogenes has been detected in several dairy products (31,32).Many Latin American countries have reported L. monocytogenes as food-borne emerging microorganism, causing abortions, sepsis and meningitis among other clinical signals leading to death (33,36).Taking into account that several foods appeared as vehicle of infections, we proposed an extraction procedure combined with a PCR method that reduced the time of identification of L. monocytogenes in raw milk, from 15 days (gold standard method) to ~5h.This PCR technique could be adapted and validated to be used for other types of food, such as cheeses, poultry and meat products.Moreover, the proposed technique could also be used in body fluids such as Cerebral Spinal Fluid (CSF) with the aim of rapid and reliable detection of the pathogen, which would be an important diagnostic tool to make prompt decisions for convenient treatment of patients.
Listeria species, while the PCR generated highly reliable, sensitive and reproducible results in 4 hours able 2. Results of microbiological, physical and chemical analysis of raw milk samples before the artificial contamination with Listeria monocytogenes L. monocytogenes detection were negative; reductase, sucrose, chloride and starch tests were negative.