WO2009036956A1

WO2009036956A1 - Detection of staphylococcus aureus in bovine mastitic milk

Info

Publication number: WO2009036956A1
Application number: PCT/EP2008/007735
Authority: WO
Inventors: Hans Ulrich Graber
Original assignee: University Of Bern
Priority date: 2007-09-19
Filing date: 2008-09-16
Publication date: 2009-03-26

Abstract

The invention relates to a method of detection and quantification of S. aureus in milk comprising the steps of trypsinizing a milk sample, extracting bacteria and isolating nucleic acids, and quantifying DNA specific for S. aureus using primers for the nuc gene in a polymerase chain reaction. The method is superior to the bacterial assays of the state of the art both in analytical specificity and sensitivity. The method is used to diagnose mastitis in cows. The invention further relates to a method of improved extraction and isolation of bacterial nucleic acids from milk and other biological fluids in the presence of L. casei.

Description

Detection of Staphylococcus aureus in bovine mastitic milk

Field of the invention

The invention relates to a method for detecting Staphylococcus aureus in milk and the diagnosis of bovine mastitis. It further relates to an improved method of extracting bacteria from any kind of biological fluid, e.g. body fluids such as milk.

Background of the invention

Staphylococcus aureus (S. aureus) is causing mastitis in cows and represents one of the most important death causes apart from reproduction disorders. This causes substantial financial damage to milk farming, and is estimated to amount to 200 Mio. $ per year in Switzerland alone.

Diagnostics of S. aureus based on bacteriological testing of milk is unsatisfactory. For a single test under optimal condition diagnostic sensitivity reaches only 75% on average and may be as low as 41% (Sears, P.M. et al., J. Dairy Sci., 73:2785-2789, 1990). A reasonable sensitivity (98%) can only be reached by analysis of three consecutive samples. Reasons for the unsatisfactory sensitivity are the high limit of detection in bacterial testing and the intermittent excretion of S. aureus into milk. The low diagnostic sensitivity of existing S. aureus tests make systematic survey of herds difficult, time- consuming and expensive. There is a clear need for improved, reliable S. aureus testing in bovine milk.

Summary of the invention

The invention relates to a method of detection and quantification of S. aureus in milk comprising the steps of trypsiniziπg a milk sample, extracting bacteria and isolating nucleic acids, and quantifying DNA specific for S. aureus using primers for a gene or gene fragment specific for S. aureus in a polymerase chain reaction. The method is superior to the bacterial assays of the state of the art both in specificity and sensitivity. The method is used to diagnose mastitis caused by S. aureus in cows. Preferably, trypsination of milk samples is performed in the presence of Lactobacillus casei (L casei) or related bacteria. The invention further relates to a method of improved extraction and isolation of bacterial nucleic acids from biological fluids, e.g. body fluids such as milk, or from dairy products, meat and meat products, in the presence of L casei or related bacteria. This method is suitable for any kind of bacteria to be detected and optionally quantified, e.g. to pathogenic bacteria. Body fluids considered are e.g. milk, urine, serum, plasma or liquor, swabs or lavages of mucosa, homogenates of body tissue, or homogenates of feces.

Brief description of the Figure

Correlation between the plate count (state of the art bacteriology) and quantitative realtime PCR (QPCR) methodology according to the invention. Data is expressed as log- transformed (logarithm base 10) values. Pearson's coefficient of correlation r = 0.925 (P < 0.001). CFU: colony forming units. SCE: Stapyhlococcal cell equivalents.

Detailed description of the invention

The invention relates to a method of detection and quantification of S. aureus in milk. At present diagnosis of mastitis in cows is mainly based on an indirect test, e.g. the so- called California Mastitis Test, which indicates inflammatory problems based on the number of leukocytes found in milk. Although this test is easy to handle, its results are unreliable and allow diagnosis of mastitis only at a point in time when inflammation has progressed considerably. A more extensive analysis uses bacteriological identification of S. aureus and other mastitis-related organisms in milk by plating milk specimens on nutritients supporting growth of S. aureus and other pathogens, and detection based on morphology, biochemical properties and hemolysis.

Since S. aureus is contagious, difficult to detect, and S. aureus infections difficult to treat, there is a need for a highly sensitive method which allows prospective screening of cow herds for S. aureus infection. This would allow to take measures before the majority of cows in a herd are infected with S. aureus. In the past, only bacteriological standard methods were available, with low sensitivity. As a consequence thereof, only part of infected cows were identified, and the extent of an infection in a herd was underestimated up to a very late stage of infection.

The present invention describes and claims a much improved diagnostic method based on quantitative polymerase chain reaction (QPCR). In particular, the invention relates to a method of detection and quantification of S. aureus in milk, preferably bovine milk, comprising the steps of trypsinizing a milk sample, extracting bacteria and isolating nucleic acids, and quantifying DNA of S. aureus using primers for a gene or gene fragment specific for S. aureus, for example a primer for the nuc gene, in a polymerase chain reaction.

Other suitable primers for a gene or gene fragment specific for S. aureus are, for example, primers for the coa (Coagulase) gene, spa (Protein A) gene, elf A (Clumping factor A) gene, or for S. aureus specific fragments of 23S rRNA.

The steps of trypsination, extraction of bacteria, nucleic acid isolation and quantitative polymerase chain reaction (QPCR) follow standard procedures well known in the art. However, the sensitivity of the method of the invention requires that the purity of the materials and reagents is exceptionally high and is checked regularly to avoid contaminations. If precautions to avoid contaminations are taken, the steps may be performed in a standard laboratory setting. For pipetting in the QPCR it is suggested to use a bench with laminar air flow and UV light.

The method is superior to the bacterial assays of the state of the art both in the analytical specificity and sensitivity. As is shown in the experimental part, the method of the invention shows a low intra-assay and inter-assay variability and is more sensitive than the standard bacterial assay by a factor of more than 25. Correlation of the inventive method with the state of the art bacterial assay is very high. Compared to state of the art phenotype bacteriology the method of the invention allows a safe decision with respect to presence and source of S. aureus in milk, and a reliable quantification without having to make use of dilution series.

Diagnosis of mammary infection in cows is based on the presence or absence of mastitis-related bacteria such as S. aureus in milk as determined with the method of the invention. To confirm mastitis it might be desirable to further test the cows for inflammation typical of mastitis, for example by measuring the content of leukocytes in milk. Such measurements may be based on machine or manual cell counting or the standard California Mastitis Test as described hereinabove.

The invention further relates to a method of diagnosis of mastitis caused by S. aureus in cows comprising the steps of drawing milk samples, quantifying S. aureus by trypsinizing the milk sample, extracting bacteria and isolating nucleic acids, and quantifying DNA of S. aureus using primers for a gene or gene fragment specific for S. aureus, for example a primer for the nuc gene, in a polymerase chain reaction, wherein mastitis is diagnosed when the count of S. aureus is higher than approx. 1400 staphylococcal cell equivalents (SCE) per ml of milk.

The invention further relates to a method of improved extraction and isolation of bacterial nucleic acids from biological fluids wherein the fluid is loaded with 10⁶ to 2 x 10¹⁰ colony forming units (CFU) of L casei or related bacteria per ml of fluid, treated with trypsin, incubated between 37°C and 60⁰C for 10 to 60 minutes, and distributed between an apolar solvent (optionally including chaotropic agents such as urea, guanidine or guanidine isothiocyanate) and an aqueous phase, followed by standard methods of bacterial extraction, nucleic acid isolation and optionally nucleic acid detection and/or quantification.

A preferred concentration of L. casei is between 10⁹ and 10¹⁰, e.g. around 5 x 10⁹ CFU per ml of biological fluid. Samples for extraction and isolation of bacterial nucleic acids are usually between 0.1 ml and 1000 ml, preferably around 0.4 ml. This means that around 1.5 x 10⁹ CFU for a sample size of 0.4 ml fluid, e.g. a body fluid such as milk, are used. Body fluids considered are e.g. milk, urine, serum, plasma, liquor cerebrospinalis, abdominal and thoracic fluid, lavages or swabs of mucosa, e.g. from mouth, throat, oesophagus, stomach, nose, larynx, trachea, bronchia, uterus, urinary bladder and the like, and homogenates of tissues or homogenates of feces obtained from human beings, cattle and other animals. Other biological fluids are dairy products, homogenates of meat and meat products, and related homogenates and suspensions. The method of the invention for an improved extraction of bacterial nucleic acids using L casei or related bacteria, for example Lactobacillus spp., Lactococcus spp., Streptococcus spp., Streptomyces spp., Propionibactehum spp., Bifidobacterium spp., or non pathogenic laboratory bacteria such as strains of Escherichia coli, which are used in molecular biology. The method is not restricted to detection of nucleic acids from S. aureus, but may be used for any problem to be solved involving detection and quantification of small amounts of any particular (potentially pathogenic) bacteria in combination with suitable probes for such bacteria and QPCR. Addition of L. casei or related bacteria (any bacterium is suitable as long as the detection by QPCR of the specific nucleic acid from the particular bacteria to be isolated is not hindered) makes sure that the particular bacteria to be detected and quantified are not lost during the procedural steps of extraction and isolation through non-specific sticking to tube and pipette walls or other equipment. A sufficient amount of L. casei or related bacteria further improves the repeatability and reliability of quantification results of QPCR. Furthermore, DNA from L casei or related bacteria supports the detection and analysis of the integrity of isolated DNA by gel electrophoresis. Use of the method of the invention for an improved extraction of bacterial nucleic acids using L casei or related bacteria is not restricted to cases of extraction and nucleic acid isolation by centrifugation but is equally well suited for filtration methods or filtration steps within extraction and isolation methods.

Trypsination is performed according to standard procedures in the art. Trypsin of different sources may be used, although Trypsin 250™ of Becton Dickinson is preferred. Alternatively, proteinase K, pronase (Roche; a mixture of proteases) or other trypsin/protease products may be used. Detergents such as Triton-X100 or other non- ionic detergents (e.g. Tween derivatives) or ionic detergents (e.g. sodium dodecyl sulfate or bile acid salts), chelating agents such as EDTA or EGTA (to eliminate calcium in milk), and various buffers (e.g. phosphate buffer) are added.

Trypsination is preferably accomplished for 10 to 20 minutes, e.g. 15 minutes, at temperatures around 55⁰C.

On extraction of bacterial nucleic acids a preferred apolar solvent is pentane. If pentane and temperatures around 25°C are used, then the apolar phase will remain fluid on centrif ligation. After treatment of the samples with an aqueous base such as sodium hydroxide solution, they are centήfuged at top speed (between 19'00Og and 30'00Og) at temperature near 0^cC, e.g. around 4⁰C, in order to ease separation of supernatant from damaged bacteria.

Isolation of DNA is accomplished according to standard procedures, for example as described by Chavagnat, F. et al., FEMS Microbiol. Lett., 217:177-183, 2002, using lysozyme, mutanolysin and the "High Pure PCR Template Preparation Kit" of Roche Diagnostics, but is not restricted to those particular reaction conditions. It is preferable to use standardized analysis conditions in order to obtain repeatable recovery and repeatable analytical sensitivity.

Examples

Milk samples

77 milk probes from cows in the Canton of Berne, Switzerland, were collected. Cows are forestripped, and the first 15-20 ml milk of each quarter analyzed for macroscopic detectable changes and tested with a California Mastitis Test (CMT) for abnormal cell content. The teat ends are thoroughly purified with gauze pledgets moistened with 70% ethanol, and individual quarter milk samples are collected aseptically in sterile vials. The obtained probes are cooled to 4°C, transported into the laboratory at 4°C and stored at - 2O⁰C.

Bacteriology The milk probes are thawed for 5 minutes at 37⁰C and tipped over several times in order to obtain a homogenous suspension. Samples of 100 μl und 10 μl milk are each plated out with an Eppendorf pipette on a Columbia agar plate containing 5% sheep blood (BA; BioMerieux Suisse S.A., Geneva, Switzerland) and a Baird Parker agar plate supplemented with rabbit plasma fibrinogen (BP; BioMerieux). At the same time 400 μl milk are taken for a S. aureus detection with QPCR (see below).

The agar plates are incubated aerobically at 37°C and are analyzed after 24 h and 48 h of incubation. Bacteria are identified according to the guidelines of the National Mastitis Council (NMC 1999) which include morphology, biochemical properties, and detection of hemolysis. On BA plates, colonies are considered typical for S. aureus if they are large, greyish-white to yellow showing β or double hemolysis. On BP plates, large black to grey colonies surrounded by an unambiguous, opaque precipitation halo are accepted as S. aureus. After 24h and 48h, colonies typical for S. aureus are counted on BP and BA plates. A total of 38 epidemiological^ unrelated S. aureus and 36 CNS isolates were obtained from the diagnostic laboratory Dr. Graeub AG in Bern, Switzerland. All the isolates are gram-positive, catalase-positive cocci which were collected from bovine mastitis milk. They are further identified by PCR (see below) for the presence of the spa (protein A), coa (coagulase) and elf A (clumping factor A) genes. In the case of S. aureus, the isolates had to be positive for all the 3 genes whereas the CNS had to be negative. In none of the two groups, strains had to be excluded due to unclear PCR results.

Extraction of bacterial nucleic acids (NA) from liquid cultures

Single colonies grown on BA plates are incubated aerobically in 4 ml Trypticase™ Soy Broth (Becton, Dickinson and Company, Basel, Switzerland) at 37°C over night. Total nucleic acids (NA) containing DNA and RNA are extracted from 1 ml of culture according to Chavagnat, F. et al., FEMS Microbiol. Lett., 217:177-183, 2002, using lysozyme (Merck, Berne, Switzerland), mutanolysin (Sigma, Buchs, Switzerland) and the "High Pure PCR Template Preparation Kit" (Roche Diagnostics, Rotkreuz, Switzerland). The eluates are checked for DNA integrity and quantity by agarose gel electrophoresis in

0.5x TBE (45 mM Tris-Borate, 1 rtiM EDTA, pH=8.3) together with GelRed stain (Biotium Inc., Hayward, CA, USA). The extracts are then stored at -20⁰C until further use.

Standard PCR PCR for 1 sample is performed in 25 μl containing 1x HotStarTaq® Master Mix (QIAGEN AG, Hombrechtikon, Switzerland), 2.5 μl of NA, 300 nM (final concentration) of each of the spa, coa or clfA gene primers (Akineden, O. et al., Clin. Diagn. Lab. Immunol., 8:959- 964, 2001) and 1 μM for both nuc gene primers, respectively (see table 1). All the primers are obtained from Microsynth, Balgach, Switzerland. For the detection of each of the 4 genes, a pre-PCR step is run at 95⁰C for 15 minutes followed by 35 cycles under following conditions: denaturation at 94°C for 1 minute, annealing at 6O⁰C for 1 minute and extension at 72⁰C for 1 minute. The reaction is completed by a 10 minutes step at 72⁰C followed by cooling down to 4⁰C. The PCR products are analyzed by agarose gel electrophoresis in 0.5x TBE including GelRed stain. Table 1. Primers and probes used for quantitative real-time PCR (QPCR), standard PCR and nucleotide sequencing, respectively. The genes nuc, coa, spa and clfA are specific for Staphylococcus aureus (S. aureus).

¹)This oligonucleotide served as a QPCR probe. It was labeled with 6-carboxyfluorescein (FAM) at the 5' end and with 6-carboxytetramethylrhodamin (TAMRA) at the 3' end.

²)These oligonucleotides served as QPCR primers. dsDNA is directly detected by SYBR® GreenER™ from Invitrogen.

³)Akineden et al., loc. cit. DNA sequencing nuc gene amplicons are purified using the "QIAquick PCR Purification Kit" (QIAGEN) according to the protocol of the manufacturer. The DNA concentration of the eluates is measured by spectroscopy using the Nanodrop ND-1000 Spectrophotometer (Nanodrop Technologies, Rockland, DE, USA). 100 ng are then added to a total volume of 10 μl containing 20 pmol of one of the sequencing primers (see table 1). The sequencing procedures are done at Microsynth. Each amplicon is sequenced once on both strands which is enough for reliable evaluation due to the very high quality of the obtained sequences.

Extraction of bacterial nucleic acids (NA) from bacteήa in milk

The thawed milk samples used to determine the S. aureus plate counts (see above) are simultaneously processed to extract bacterial NA. For that reason, 400 μl milk of each sample is pipetted to a solution containing 250 μl Triton-X100 2% (w/v; Merck), 150 μl L case/ (1.5x10⁹ CFU) and 125 μl 1% trypsin solution (for details see below). After incubating at 55⁰C for 15 minutes, 925 μl pentane (Merck) is added to the suspension followed by vortexing the samples three times for 10 seconds. Afterwards, the specimens are centrifuged at 18'00Og for 10 minutes (25°C). The supernatant is discarded and the cell wall of the pel letted bacteria is lysed according to Chavagnat et al. (loc. cit.) with a minor modification: after the NaOH treatment centrifugation is performed at 20'80Og for 10 minutes (4⁰C). NA is then extracted with the "High Pure PCR Template Preparation Kit" (Roche Diagnostics). The standard protocol as listed below is preferably followed. The eluates (200 μl) are checked for DNA integrity and quantity by agarose gel electrophoresis. The extracts are stored at -20⁰C until further use.

L casei, strain 18121 , was obtained from D. Isolini, Swiss Federal Research Station for Animal Production and Dairy Products, Liebefeld-Bern, Switzerland. A single colony of L casei grown on a De Man Rogosa Sharp (MRS) agar plate (Biolife S.r.l., Milano, Italy) is inoculated in 3 ml MRS medium (Biolife) and incubated aerobically at 3O⁰C for 48 h. 100 μl of the suspension is then added to 100 ml MRS medium and cultured as described. Quantification of L. casei is done spectroscopically using a wave length of 650 nm and an own standard curve obtained from 3 independent dilution experiments including plating on MRS agar plates, incubation at 3O⁰C for 48 h and counting the grown colonies. The bacteria are stored in MRS medium (10⁹ CFU/100 μl) at 4⁰C for up to 3 weeks.

The 1% trypsin solution is always freshly prepared by dissolving 100 mg Trypsin 250™ (Becton Dickinson AG, Basel, Switzerland) in 10 ml of 100 mM Tris/HCI, pH=7.8, followed by incubation at room temperature for 30 minutes. Before use, the solution is sterilized using a filter with a pore size of 0.22 μm.

Standard protocol for DNA extraction and isolation

A particular standardized procedure according Chavagnat et al. using the "High Pure PCR Template Preparation Kit" of Roche is as follows:

1. Prewarm the milk samples for 5 min. at 37°C

2. For each of the milk samples take an aliquot of 400 μl

3. To each aliquot add 250 μl Triton-X100 2%, 125 μl Trypsin 1 %, and 150 μl L. casei (1.5 x 10⁹ CFU) 4. Incubate for 15 min. at 55⁰C

5. Add 125 μl pentane

6. Vortex 3 times for 10 sek.

7. Centrifuge the samples at 18'00Og, 10 min., 25°C

8. Resuspend the pellet in 950 μl H₂O, add 50 μl NaOH 1 M 9. Incubate for 15 min., room temperature (RT)

10. Centrifuge at 20'80Og, 10 min., 4⁰C

11. Resuspend the pellet in 1 ml TES buffer (100 mM Tris/HCI, 10 mM Na₂EDTA, 25% sucrose (w/v), pH=8.0)

12. Centrifuge at 18'00Og, 10 min., 20⁰C 13. Resuspend the pellet in 0.9 ml TES buffer

14. Add 50 μl lysozyme 50 mg/ml and 50 μl mutanolysin 2000 U/ml H₂O

15. Incubate for 60 min., 37°C

16. Centrifuge at 18'00Og, 10 min., 20°C

17. Resuspend the pellet in 200 μl H₂O 18. Add 200 μl binding buffer (Roche Kit) and 40 μl Proteinase K (Roche Kit)

19. Incubate for 10 min., 72°C

20. Add 100 μl 2-propanol

21. vortex

22. Introduce High Pure Filter Tube (Roche Kit) into the collecting device 23. Pipet the sample into the filter tube

24. Centrifuge at 5'00Og, 1 min., 2O⁰C

25. Discard fluid and introduce filter tube into a new collecting device

26. Pipet 500 μl washing buffer (Roche Kit) into the filter tube 27. Centrifuge at 5'00Og, 1 min., 20⁰C

28. Discard fluid and introduce filter tube into a new collecting device

29. Pipet 500 μl washing buffer (Roche Kit) into the filter tube

30. Centrifuge at 5'00Og, 1 min., 2O⁰C

31. Discard fluid and introduce filter tube into 1.5 ml Eppendorf tube with cover 32. Pipet 200 μl elution buffer (Roche Kit) heated to 72⁰C into filter tube

33. Incubate sample for 5 min., 72°C

34. Centrifuge 5¹OOOg, 1 min., 20°C

35. Store eluate (200 μl) at -2O⁰C

Generation of an internal control for QPCR

To produce an internal control, a pCI plasmid vector was obtained from Prof. Dr. A. Zurbriggen, Department of Clinical Veterinary Medicine, Vetsuisse-Faculty of Berne, Berne, Switzerland containing the full length N gene of the canine distemper virus (CDV- N gene). The insert is verified by sequencing the sense strand using the T7 primer (Microsynth). Afterwards, it serves as a template for a standard PCR to produce enough spiking DNA. For that reason, 10 μg of plasmid DNA is added to a final volume of 25 μl containing 1x HotStarTaq® Master Mix (QIAGEN), and 300 nM (final concentration) of the CDV-N-S and CDV-N-AS primer (see table 1). Amplification is performed by a pre- PCR step at 95°C for 15 minutes followed by 33 cycles under following conditions: 94⁰C for 1 minute, 63⁰C for 1 minute and 72°C for 1 minute. The reaction is finished by a 10 minutes step at 72⁰C followed by cooling down to 4°C. The PCR product ((788 bp) is analyzed by agarose gel electrophoresis in 0.5x TBE including Gel Red stain and is purified with the "QIAquick PCR Purification Kit" (QIAGEN) according to the protocol of the manufacturer. Quantification of the eluate DNA is done with the Nanodrop spectrometer.

Quantitative PCR

DNA specific for S. aureus is quantified by QPCR. For that reason, 2 primers and a probe for the nuc target gene are designed (see table 1) amplifying a 166 bp fragment. The probe is labeled at the 5¹ end with 6-carboxyfluorescein (FAM) and at the 3' end with 6-carboxytetramethylrhodamin (TAMRA; Microsynth). Each 25 μl of QPCR mixture contains 1x TaqMan Universal PCR master mix (Applied Biosystems, Rotkreuz, Switzerland), 900 nM sense and antisense primer, 300 nM probe and 3.5 μl NA extracted from bacteria in milk. The QPCR conditions are 5O⁰C for 2 minutes and 95°C for 10 minutes followed by 45 cycles of 95⁰C for 15 seconds and 60° for 1 minute. Cycling is performed in the GeneAmp 5700 Sequence Detection System (Applied Biosystems). To ensure that negative results are not due to nonspecific inhibition of the QPCR assay, the same NA is used in a second QPCR in the presence of an internal control DNA

(CDV-N amplicon) amplified with a CDV-N specific QPCR (amplified fragment: 130 bp). For that purpose each reaction (25 μl) contains 1x TaqMan Universal PCR master mix (Applied Biosystems), 300 nM sense and antisense primer, 300 nM FAM/TAMRA labeled probe (see table 1 ; Microsynth), 3.5 μl NA and 10⁵ copies of the CDV-N amplicon.

QPCRs for both genes are run in duplicate with results deemed positive if both reactions are positive. If only one reaction shows a positive result or the difference of the 2 CT values is more than 1.5 cycles, the QPCRs are repeated. A reaction is considered negative if nuc gene amplification resulted in a value < 10 molecules/reaction. For each assay, a standard is included (in duplicates). For the nuc gene, it ranges between 10¹ to 3.8 x 10⁴ molecules/reaction, for the CDV-N gene between 10² to 10⁶ molecules/ reaction. In the case of the nuc gene, 1 molecule is considered as 1 staphylococcal cell equivalent (SCE).

Assay controls

Various controls are included to monitor the whole procedure. To evaluate the extraction of S. aureus from milk, 1 sample of S. aureus negative raw milk (RM) is coprocessed together with 1 sample spiked with 5000 and 10⁶ CFU S. aureus /ml raw milk, respectively. If the nuc gene QPCR is positive for the RM sample or one of the spiked samples is not within the adequate range (+/- 2.5 times the expected result), all the milk samples of the affected series are re-extracted.

For each QPCR run, no template controls are included. If they result positive, the run is repeated. All negative nuc gene QPCR results require detection of CDV-N DNA. CDV-N QPCR is also used to check for PCR inhibitors contained in the NAs. If the slope (gradient?) of the amplification curve (log-transformed fluorescence intensities) is obviously flatter or the CT value (CT: cycle to reach the threshold) is more than 2 cycles larger than the mean, the NAs are diluted 1 :10 with H₂O and re-analyzed by both the nuc and CDV-N gene QPCRs. In case of doubt, the NAs are diluted.

Statistics

Data is expressed either as non-transformed or as log-transformed (logarithm base 10) values. Counts of S. aureus on BA and BP plates are compared by the Wilcoxon signed- rank test for paired samples. Comparison of the QPCR and bacteriological results is done using log-transformed values together with the paired t test. In addition, linear least square regression analysis is performed with the QPCR results (log SCE/ml) forming the dependent, the bacteriological results (log CFU/ml) the independent variable. Normality of the difference is demonstrated by quantile plot followed by the Lilliefors test. To analyze whether the QPCR method reveals more frequently a positive result than the bacteriological method, the McNemar test is applied. For this purpose, the exact P value is calculated manually. For all the other computations, the Systat 10 software package (Systat Software Inc., CA, USA) is used. P values < 0.05 are considered to be significant.

nuc gene of bovine isolates of S. aureus

The nuc gene is used as a target for the novel assay. This gene coding for the thermonuclease is highly specific for S. aureus and is not found in other staphylococcal species or in further mastitis-relevant bacteria. Published nuc sequences are from human S. aureus isolates and might differ from the bovine cases. The nuc gene of 20 epidemiological^ unrelated S. aureus isolates from bovine mastitis milk is amplified by conventional PCR. The resulting full length PCR products are then sequenced on both strands.

Although the similarities between the bovine and (the known) human strains are consistently very high (between 4 (98.8%) and 8 substitutions per 649 sequenced nucleotides (99.4%)), 19 of the 20 bovine isolates showed 5 nucleotide substitutions at nucleotide 51 , 113, 129, 538 and 633 that exclusively occurred in the bovine strains (position relative to human isolate S. aureus MSSA476). At position 51 , the substitution results in an exchange of amino acid isoleucine versus methionine, at position 113 phenylalanine versus tyrosine and at position 538 glutamate versus lysine. At the two remaining sites, the nucleotide changes are silent. The mutations are present in 2 alternative patterns of combination and each of the 19 isolates can be attributed to one of the 2 patterns. The 20th isolate, however, is very closely related to the human N315 strain with only 2 mutations observed in the 649 nucleotides analyzed. Based on these results, two primers and a fluorescent probe for QPCR are constructed which are expected to detect the nuc gene in all the bovine and human S. aureus isolates.

Assay performance Using dilution series of a purified nuc gene amplicon, the assay is shown to be linear between 10¹ and 10⁷ SCE/assay. It is even possible to measure down to 6 SCE/assay, but it is difficult to reproduce the results consistently. As a QPCR standard, therefore, concentrations between 10¹ and 3.8 x 10⁴ SCE per reaction are chosen. For this range, a coefficient of correlation of 0.998 is obtained. Moreover, the lowest value (10 SCE/assay) could always be reproduced very well. The amplification efficiency is 0.96 (determined from the gradient).

Analytical specificity

For determination of the specificity DNA free from RNA from 38 S. aureus and 36 CNS isolates (see above) was used at 5000 and 100 SCE/assay (SCE: Staphylococcal Cell Equivalents). L casei NA from 1.5 x 10⁹ bacteria is added to each assay in order to mimic the composition of specimens resulting from S. aureus extraction from milk. Both at 100 and at 5000 SCE/assay all S. aureus isolates gave a positive and all CNS isolates a negative result. Hence the analytical specificity of the assay is 100%.

Analytical sensitivity

The sensitivity of the assay procedure is evaluated by serially diluting counted liquid cultures of S. aureus in raw milk. These spiking experiments show an analytical sensitivity as low as 460 CFU/assay which translates to 1.8 CFU/10 μl of milk.

Repeatability

For the determination of the intra-assay variability samples of 10⁶ CFU/ml milk and 5 x 10³ CFU/ml milk are used and 8 replicates measured for each sample. For the high SCE content the coefficient of variation (CV) was 1.2%, for the low SCE content 3.9% (log- transformed values). Considering the interassay variability, the CV is 3.7% for 10⁶ CFU/ml milk and 6.5% for 5 x 10³ CFU/ml (7 replicates each; log-transformed values). Recovery

To determine the efficiency of S. aureus extraction from milk, 7 samples of raw milk (400 μl each) containing 1 x 10⁶ and 5 x 10³ CFU S. aureus/m\, respectively, are used. Both experiments included 2 raw milk samples free from S. aureus. Bacteria are isolated from milk according to standard procedures, corresponding DNA purified and QPCR performed for the nuc gene. Milk samples without S. aureus always gave a negative result, and the spiked samples were determined to contain 2.13 x 10⁶ SCE/ml and 7300 SCE/ml, respectively. Assuming that 1 CFU is formed by approx. 2 SCE a recovery of 107% (1 x 10⁶ CFU S. aureus/m\) and 73% (5 x 10³ CFU S. aureus/m\), respectively, was found.

To determine the loss of DNA through binding to the extraction column, 24 milk specimens are used and DNA isolated according to the protocol. Extracts are spiked with CDV-N plasmid before applying to the column. As a reference 3 samples are used wherein the corresponding amount of plasmid is directly added to the QPCR assay. The CDV-N gene is then quantified by QPCR. For the spiked milk probes an average of 2.16 x 10^s molecules/assay (SE = ± 0.09 x 10⁵) was found, for the reference 3.30 x 10⁵ molecules/assay (SE = ± 0.7 x 10⁵). Since the difference is small a correction due to loss on the column does not seem to be required.

Assay results: Analysis of 77 specimens by bacteriology

For the bacterial analysis of milk 10 μl and 100 μl of each specimen are plated on BA and BP plates. The results are read after 24 h and after 48 h and the grown colonies are counted. If the bacteria content is too high the corresponding samples are diluted in PBS and 10 μl each of dilutions applied to both types of plates. For comparison with QPCR it is preferable to use counts from BP plates only since many specimens contain more than one bacteria type and an unambiguous determination of grown bacteria often fails on BA plates. Out of 77 specimens tested 27 were sterile, and 35 contained S. aureus. If only results from 10 μl sample plates were considered, S. aureus was proven to be present in 33 specimens. S. aureus was the only detected bacteria in 24 specimens, in 11 specimens combinations of S. aureus + CNS/ Streptococcus spp./C. bovis were detected. 24 milk specimens contained other bacteria than S. aureus, 18 contained CNS, partially combined with Streptococcus spp./C. bovis. Environmental contaminants were found in 5.6% of the samples (n=4).

In order to test if S. aureus are equally well determined on either BP or BA plates, counts were compared in those cases where BA plates allowed unambiguous determination of S. aureus (β or double hemolysis). In total 30 value pairs were compared in the Wilcoxon signed-rank test. It was clear that no difference existed between the two types of plates in relation to S. aureus counts (p > 0.90).

Comparison between QPCR and bacteriology The same 77 milk specimens were tested both with bacteriology and QPCR. The results are collected in table 2. Using bacteriology 35 specimens tested positive for S. aureus, using QPCR, 46 specimens were positive. Statistical analysis with a McNemar test demonstrates that positive results with QPCR are significantly more often than with bacteriology (p =0.00049). For QPCR a minimum of 0, a maximum von 8.07 x 10⁸ and a median of 1.41 x 10⁴ SCE/ml was reached. For bacteriology a minimum of 0, a maximum of 1.55 x 10⁸ and a median of 0 CFU/ml was found. For positive values after Iog10 transformation a mean of 5.48 log10_SCE/ml (SD = 1.29) was calculated for QPCR, whereas the corresponding value for bacteriology is 4.91 log10_CFU/ml (SD = 1.50). The t test for paired Iog10 values gave a p value of < 0.001. QPCR methodology gave values for pairs of data which were up to 166 times higher than values from bacteriology. Only in 4 cases the value from bacteriological determination was higher than determined by QPCR (lowest ratio = 0.282). Furthermore QPCR results were always positive in cases of positive results from bacteriology. In those 31 specimens being positively tested by bacteriology, the plate counts were correlated with QPCR data (see Figure 1 ). A linear dependence of Iog10 values was found over the whole range (7.0 x 10¹ to 1.55 x 10⁸ CFU/ml). The correlation coefficient is 0.925 (p < 0.001), the gradient 0.802 (p < 0.001), the constant 1.705 (p < 0.001). As can be deduced from Figure 1 the QPCR method gives consistently higher values than the bacteriological method (p < 0.001). The SCE value is determined to be 50.8 times the average CFU value. With respect to analytical sensitivity in clinical milk specimens this means that the QPCR method is 8.9 timers more sensitive than the bacteriological method (corresponds to 0.11 CFU/10 μl milk), when the analytical sensitivity of the bacteriological method is set to 1 CFU/10 μl milk. Since 1 CFU may, however, be the result of contamination, only more than 3 CFU of the same pathogen per plate as proof of pathogen originating from the udder. Under these preconditions the method of the invention represent an analytical sensitivity which is 27.1 times (corresponds to 0.037 CFU/10 μl) higher than the state of the art bacteriological method. For these calculations one has to consider that the change in analytical sensitivity of the QPCR method is not proportional since the correlation between SCE and CFU values is log linear.

Table 2. Comparison between the classical plate count and the quantitative real-time PCR (QPCR) methodology: number of mastitic milk samples which were positive and negative for Staphylococcus aureus (S. aureus), respectively. For these investigations, a total of 77 milk samples were analyzed by both methods.

McNemar test: P = 0.00049

Claims

1. A method of detection and quantification of S. aureus in milk comprising the steps of trypsinizing a milk sample, extracting bacteria and isolating nucleic acids, and quantifying DNA specific for S. aureus using primers for a gene or gene fragment specific for S. aureus in a polymerase chain reaction.

2. The method of claim 1 wherein the primer for a gene or gene fragment specific for S. aureus is a primer for the nuc gene, the coa gene, the spa gene, the elf A gene, or for S. aureus specific fragments of 23S rRNA.

3. The method of claim 1 wherein the primer for a gene or gene fragment specific for S. aureus is a primer for the nuc gene.

4. The method of claim 1 wherein Lactobacillus casei or related bacteria are added to the milk sample before trypsination.

5. A method of diagnosis of mastitis caused by S. aureus in cows comprising the steps of drawing milk samples, quantifying S. aureus by trypsinizing the milk sample, extracting bacteria and isolating nucleic acids, and quantifying DNA of S. aureus using primers for a gene or gene fragment specific for S. aureus in a polymerase chain reaction, wherein mastitis is diagnosed when the count of S. aureus is higher than approx. 1400 staphylococcal cell equivalents per ml of milk.

6. A method of improved extraction and isolation of bacterial nucleic acids from biological fluids wherein the fluid is loaded with between 10⁶ and 2 x 10¹⁰ colony forming units of L. casei or related bacteria per ml of fluid, treated with trypsin, incubated between 37⁰C and 6O⁰C for 10 to 60 minutes, and distributed between an apolar solvent or chaotropic agent and an aqueous phase, followed by standard methods of bacterial extraction, nucleic acid isolation and optionally nucleic acid detection and/or quantification.

7. The method of claim 6 wherein, in the distribution between an apolar solvent and an aqueous phase, pentane is used as the apolar phase.

8. The method of claim 6 or 7 wherein the concentration of L casei is between 10⁷ and 10¹⁰ colony forming units per ml of biological fluid.

9. The method of anyone of claims 6 to 8 wherein the biological fluid is milk, urine, serum, plasma, liquor cerebrospinaiis, abdominal or thoracic fluid, lavages or swabs of mucosa, homogenates of tissues or homogenates of feces obtained from human beings, cattle and other animals, dairy products, or homogenates of meat or meat products.

10. The method of claim 9 wherein the biological fluid is milk.