Multiplex Fast Real-Time PCR for Quantitative Detection and Identification of cos- and pac-Type Streptococcus thermophilus Bacteriophages

IC.

The fermentation of milk by Streptococcus thermophilus is awidespread industrial process that is susceptible to bacteriophageattack. In this work, a preventive fast real-time PCR methodfor the detection, quantification, and identification of typesof S. thermophilus phages in 30 min is described.

INTRODUCTION

Top

INTRODUCTION

IC.

Streptococcus thermophilus is a gram-positive thermophilic lacticacid bacterium used, along with Lactobacillus spp., as a starterculture for the manufacture of important fermented dairy foods,including yogurt and Swiss- or Italian-type hard cooked cheeses(5). Unfortunately, these bacteria are susceptible to infectionby bacteriophages during the fermentation process, a phenomenonthat ultimately results in fermentation failure. The commonfeatures of S. thermophilus phages include double-stranded DNAgenomes that are 31 to 45 kb long, small isometric heads, longnoncontractile tails, and affiliation with the Siphoviridaefamily, corresponding to Bradley's group B (3). They are currentlydivided into two groups (cos and pac types) based on the genomeencapsidation machinery (11). In the case of yogurt isolates,these types are also related to host range and serotype (4).

Fast detection methods are an important tool to avoid phageattacks in dairy factories. Detection of bacteriophages in milkis normally carried out using standard microbiological methods(plaque assays, activity tests, etc.) (8), but these methodsare time-consuming. To speed up the analysis, PCR techniqueshave been used to detect phages in different kinds of dairysamples (1, 4, 6, 7, 10, 12). Increasing demand for quantitative,more sensitive, and quicker procedures is prompting the developmentof real-time quantitative PCR (qPCR) methods. The objectiveof the present study was to develop a fast multiplex qPCR methodthat allows quantitative detection and identification of cos-and pac-type S. thermophilus bacteriophages in milk samples.

Primer and probe design.

Top

Primer and probe design.

IC.

In the first step, databases were screened to select the mostconserved genes of cos- and pac-type S. thermophilus phages.orf1510 encoding the putative minor tail protein of the Sfi11bacteriophage (pac type) and orf18 encoding the antireceptorprotein of the Sfi21 bacteriophage (cos type) were selectedand aligned, using the CLUSTAL W algorithm (14), with the sequencesof the orthologous genes available in the GenBank database.Highly similar sequences were selected to design primers qPac1,qPac2, qCos1, and qCos2 and probes mgbPac2 and mgbCos (Table1) using Primer Express software (Applied Biosystems, Warrington,United Kingdom). The species specificity of the primers wasassessed by using BLAST 2.2.15 (Basic Local Alignment SearchTool) to ensure that they amplify only the corresponding S.thermophilus bacteriophage sequences. Both the mgbPac2 and mgbCosprobes were synthesized with a minor groove binder (MGB) nonfluorescentquencher attached to the 3' end and with a different reporterdye attached to the 5' end (VIC and 6-carboxyfluorescein [FAM],respectively) in order to combine them in the same sample.

View this table:
[in this window]
[in a new window]

TABLE 1. PCR primers and TaqMan MGB probes used in this study

IC.

Top

IC.

A general and important advantage of the qPCR based on fluorescentprobes is the possibility of including an internal positivecontrol (IC) in every reaction. pEM125, a plasmid containingan unrelated sequence (EMBL database accession no. X64695),was constructed as an IC. Primers IC-FW and IC-R and the TaqManMGB probe mgbIC were selected (Table 1). The probe was NED labeledat the 5' end and had an MGB nonfluorescent quencher attachedto the 3' end. A total of 10⁶ copies of plasmid pEM125 (3 logarithmicunits greater than the determined level of detection) was addedto all the reaction mixtures as an IC. The reaction was consideredto be inhibited if the cycle threshold (C_T) value increasedmore than 3 U. Correct amplification of the IC indicated thatthe whole biochemistry machinery worked properly and that therewere no PCR inhibitors in the samples. Therefore, negative phagedetection results were much more reliable than the results obtainedusing previous PCR methods.

Quantification range and sensitivity.

Top

IC.

Quantification range and...

pEM212, the plasmid used as a standard and a positive controlin the qPCR for pac-type bacteriophages, was constructed bycloning a 1,196-bp fragment of the orf1510 gene from the Sfi11bacteriophage into the pCR-2.1 TOPO vector (Invitrogen, Carlsbad,CA). pEM213, the plasmid used as a standard and a positive controlin the qPCR for cos-type bacteriophages, was constructed bycloning a 147-bp fragment of the orf18 gene from the Sfi21 bacteriophageinto the same vector. Triplicate experiments with serial 10-folddilutions ranging from 10¹¹ to 1 copy of plasmid per ml of milkwere performed to generate the standard curves. The qPCR conditionsare described in the supplemental material. A linear functionbetween the average C_T values and the logarithm of the genecopy number was established (Fig. 1A and 1B for plasmids pEM212and pEM213, respectively). The results showed that the detectionlimit was one plasmid molecule in 33.22 cycles with a standarddeviation (SD) of ±0.7 for plasmid pEM212 and one plasmidmolecule in 33.23 cycles with an SD of ±1.0 for plasmidpEM213. The assay variability increased when less than 100 copieswere present. However, the dynamic range of the qPCR assay waswide (from 1 to 10⁸ copies of the standard plasmids). Consequently,the quantification limit was determined to be 10 copies perreaction. Another important parameter, the reaction efficiency(9), was obtained from the standard curves. In both cases theamplification efficiency was high (0.96 and 0.94, respectively).

View larger version (22K):
[in this window]
[in a new window]

FIG. 1. Real-time qPCR analysis of 10-fold serial dilutions in skim milk of (A) plasmid pEM212 DNA digested with BglII and bacteriophage ${phi}$ P13.2 and (B) plasmid pEM213 DNA digested with BglII and bacteriophage ${phi}$ ipla124. C_T values were plotted against the logarithm of the calculated plasmid copy number or the number of PFU included in each reaction mixture.

To test the precision of the standard curves using phages astemplates, two new curves were generated using milk artificiallycontaminated with known titers of ${phi}$ P13.2 (pac type) and ${phi}$ ipla124(cos type) ranging from 1 to 10⁶ PFU per PCR mixture (Fig. 1Aand 1B). As expected, the results revealed that the slopes ofthe curves were similar to the slopes of curves previously generatedwith the pEM212 and pEM213 control plasmids. Thus, the S. thermophilusbacteriophage titer of a milk sample could be determined bymeans of the pEM212 and pEM213 regression functions.

Reproducibility and specificity of primers and probes.

Top

IC.

Reproducibility and specificity...

To determine the reproducibility of the proposed method, quadruplicatereactions with two independent ${phi}$ P13.2 and ${phi}$ ipla124 suspensionswere performed. Tenfold milk dilutions containing from 10³ to10⁹ PFU ml^–1 were used as templates. The SD of the C_Tvalues obtained were calculated and ranged from a minimum of±0.07 to a maximum of ±0.6 for ${phi}$ P13.2 (pac type)and from a minimum of ±0.04 to a maximum of ±0.92for ${phi}$ ipla124 (cos type). C_T values obtained for the same dilutionson three different days were used to determine the interassayvariability.

The specificity was assessed by testing 27 different S. thermophilusbacteriophages previously isolated in Europe from failed industrialfermentations and characterized in our laboratory (unpublishedresults), 15 different S. thermophilus phages isolated in America(1, 13), and the type phages Sfi11 and Sfi21. Four differentbacteriophages infecting Lactobacillus delbrueckii and ninebacteriophages infecting Lactococcus lactis were also tested.All these bacteriophages are listed in Table S2 in the supplementalmaterial. The qPCR method designed in this work was extremelyspecific since only the S. thermophilus phages were detected.Moreover, using the function of the fluorescent dye detected,VIC or FAM, it was possible to identify the type of phage (pacand cos, respectively).

Although milk is unlikely to be contaminated with differentphages in practice (2), the method was used successfully todetect both S. thermophilus phage types in the same sample.Milk samples simultaneously contaminated with titrated suspensionsof ${phi}$ P13.2 (pac type) and ${phi}$ ipla124 (cos type) were used as templatesources. No interference was observed in these multiple qPCRassays.

Phage detection is still mainly done by the plaque assay, whichdepends on knowledge of the identity of the starter strain.PCR methodology is an interesting alternative since it doesnot depend on the starter strain that is being used. On theother hand, PCR cannot distinguish viable phage particles fromDNA, but the presence of viral DNA is an indication of the potentialpresence of infective virions. Moreover, soluble DNA is rapidlydegraded in milk and dairy products. To our knowledge, thisis the first phage detection method based on fast real-timePCR technology. The most important advantages compared to previouslydescribed methods (plaque assays, activity tests, conventionalPCR, etc.) are probably the considerable time reduction andsimplicity of the analysis, since it is possible to detect phagein no more than 30 min without previous or subsequent sampletreatments. In addition, this method is able to classify S.thermophilus phages in one of the two groups that were establishedbased on the DNA packaging mechanism (11). Fast and simple classificationtechniques are useful for obtaining epidemiological data forthe industrial environment.

In conclusion, the proposed qPCR procedure is easy, sensitive,and specific and allows detection, quantification, and identificationof the type of S. thermophilus phages in a short period of timeand thus is suitable for routine use in factory-associated laboratories.Since milk storage time plays an important strategic role witheconomic implications, fast qPCR detection of phages would beprofitable for dairy industries. Correct and rapid identificationof bacteriophages potentially able to attack starter culturesallows speedy decisions concerning the destination of contaminatedmilk. Such milk might be earmarked for use in processes in whichphages are deactivated, processes that do not require starters,or processes that employ starter bacteria not sensitive to thedetected phage. In addition, qPCR would also be useful for detectionand characterization of phages at all stages of milk productmanufacture and in all the niches of the dairy industries.

ACKNOWLEDGMENTS

We thank Corporación Alimantaria Peñasanta S.A.(CAPSA) for its support. B.D.R. and M.C.M. were beneficiariesof I3P CSIC contracts financed by the European Social Fund.A.H.M. was a recipient of a fellowship from the Spanish Ministryof Education and Science. This research was supported by projectPC-04-14 from FICYT, Asturias, Spain (cofinanced by CAPSA) andproject BIO 2002-01458 from MEC, Spain (cofinanced by the FEDERPLAN of the European Union).

We thank Juan E. Suarez and Carmen Madera for providing thelactococcal phages, Jorge A. Reinheimer for providing the S.thermophilus and L. delbrueckii phages, and Nestec Ltd. (NestléResearch Center, Lausanne, Switzerland) for providing phagesSfi11 and Sfi21. We are also grateful to María Fernándezfor her critical revision of the manuscript.

FOOTNOTES

* Corresponding author. Mailing address: Instituto de Productos Lácteos de Asturias (CSIC), 33300 Villaviciosa, Asturias, Spain. Phone: 34 985 89 21 31. Fax: 34 985 89 22 33. E-mail: maag{at}ipla.csic.es

Published ahead of print on 6 June 2008.

Supplemental material for this article may be found at http://aem.asm.org/.

B.D.R. and M.C.M. contributed equally to this work.

REFERENCES

Top

IC.