Method for Optimizing Pulsed-Field Gel Electrophoresis Banding Pattern Data

The genomic DNA of 47 strains of TSST-1 toxin-producing Staphylococcus aureus were cleaved with SmaI restriction endonuclease and resolved in an agarose gel by pulsed-field gel electrophoresis (PFGE). An algorithm was designed to standardize the band weights or brightness (trace quantity) produced to a bounded region between 0 and 1 regardless of DNA fragment size while simultaneously reducing gel-to-gel variability. The algorithm allows for classification of isolates by band intensity as well as DNA mobility without a numerical hierarchy of band intensity that is caused by ranging DNA fragment lengths. On analysis many isolates were classified as separate entities on the basis of DNA co-migration only. Isolates differing by only DNA co-migration were subjected to a second digestion with restriction enzyme SacII. These isolates were characterized similarly to the standardized trace quantity analysis of SmaI PFGE patterns. The standardization method proposed in this article permits characterization of isolates on the basis of band differences, regardless of DNA co-migration, thus increasing the discriminatory power (0.79 to 0.89) of PFGE by increasing band-associated information. An established unbiased approach to the partitioning of data were also explored. The genomic DNA of 47 strains of TSST-1 toxin-producing Staphylococcus aureus were cleaved with SmaI restriction endonuclease and resolved in an agarose gel by pulsed-field gel electrophoresis (PFGE). An algorithm was designed to standardize the band weights or brightness (trace quantity) produced to a bounded region between 0 and 1 regardless of DNA fragment size while simultaneously reducing gel-to-gel variability. The algorithm allows for classification of isolates by band intensity as well as DNA mobility without a numerical hierarchy of band intensity that is caused by ranging DNA fragment lengths. On analysis many isolates were classified as separate entities on the basis of DNA co-migration only. Isolates differing by only DNA co-migration were subjected to a second digestion with restriction enzyme SacII. These isolates were characterized similarly to the standardized trace quantity analysis of SmaI PFGE patterns. The standardization method proposed in this article permits characterization of isolates on the basis of band differences, regardless of DNA co-migration, thus increasing the discriminatory power (0.79 to 0.89) of PFGE by increasing band-associated information. An established unbiased approach to the partitioning of data were also explored. Pulsed-field gel electrophoresis (PFGE) has been one of the most useful developments in molecular epidemiology for the past few decades and is now regarded as the gold standard for molecular typing of microorganisms.1Goering RV Molecular epidemiology of nosocomial infection: analysis of chromosomal restriction fragment patterns by pulsed-field gel electrophoresis.Infect Control Hosp Epidemiol. 1993; 14: 595-600Crossref PubMed Scopus (141) Google Scholar, 2Goering RV The molecular epidemiology of nosocomial infection: in overview of principles, application, and interpretation.in: Specter S Bendinelli M Friedman H Rapid Detection of Infectious Agents. Plenum Press, New York1998: 131-157Google Scholar, 3Linhardt F Ziebuhr W Meyer P Witte W Hacker J Pulsed-field gel electrophoresis of genomic restriction fragments as a tool for the epidemiological analysis of Staphylococcus aureus and coagulase-negative staphylococci.FEMS Microbiol Lett. 1992; 74: 181-185Crossref PubMed Google Scholar, 4Saulnier P Bourneix C Prevost G Andremont A Random amplified polymorphic DNA assay is less discriminant than pulsed-field gel electrophoresis for typing strains of methicillin-resistant Staphylococcus aureus.J Clin Microbiol. 1993; 31: 982-985PubMed Google Scholar PFGE is capable of resolving large fragments of DNA with a practical range of 10 kb to ∼7 Mb.1Goering RV Molecular epidemiology of nosocomial infection: analysis of chromosomal restriction fragment patterns by pulsed-field gel electrophoresis.Infect Control Hosp Epidemiol. 1993; 14: 595-600Crossref PubMed Scopus (141) Google Scholar, 5Chu G Pulsed field electrophoresis in contour-clamped homogeneous electric fields for the resolution of DNA by size or topology.Electrophoresis. 1989; 10: 290-295Crossref PubMed Scopus (33) Google Scholar Highly discerning endonuclease-treated Staphylococcus aureus genomic DNA resolved by PFGE generates a high degree of discrimination.3Linhardt F Ziebuhr W Meyer P Witte W Hacker J Pulsed-field gel electrophoresis of genomic restriction fragments as a tool for the epidemiological analysis of Staphylococcus aureus and coagulase-negative staphylococci.FEMS Microbiol Lett. 1992; 74: 181-185Crossref PubMed Google Scholar, 4Saulnier P Bourneix C Prevost G Andremont A Random amplified polymorphic DNA assay is less discriminant than pulsed-field gel electrophoresis for typing strains of methicillin-resistant Staphylococcus aureus.J Clin Microbiol. 1993; 31: 982-985PubMed Google Scholar, 6Hunter PR Reproducibility and indices of discriminatory power of microbial typing methods.J Clin Microbiol. 1990; 28: 1903-1905PubMed Google Scholar, 7Schlichting C Branger C Fournier JM Witte W Boutonnier A Wolz C Goullet P Doring G Typing of Staphylococcus aureus by pulsed-field gel electrophoresis, zymotyping, capsular typing, and phage typing: resolution of clonal relationships.J Clin Microbiol. 1993; 31: 227-232PubMed Google Scholar, 8Struelens MJ Deplano A Godard C Maes N Serruys E Epidemiologic typing and delineation of genetic relatedness of methicillin-resistant Staphylococcus aureus by macrorestriction analysis of genomic DNA by using pulsed-field gel electrophoresis.J Clin Microbiol. 1992; 30: 2599-2605PubMed Google Scholar, 9Tenover FC Arbeit RD Goering RV Mickelsen PA Murray BE Persing DH Swaminathan B Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing.J Clin Microbiol. 1995; 33: 2233-2239PubMed Google Scholar Traditionally, band velocity or relative front (Rf), the distance a band travels in a lane divided by the total lane length, is the determining variable used to discriminate band type. The band is then designated as present or absent, indicated by a 1 or 0, respectively (a form of data standardization). Pattern recognition methods are then used to identify groups of isolates with similar banding patterns. This method of numerically reducing PFGE pattern data, although quite successful, overlooks much of the diversity present in PFGE banding patterns. PFGE, like many other methods of sieving molecules, separates only by size, isoelectric potential, or topology.5Chu G Pulsed field electrophoresis in contour-clamped homogeneous electric fields for the resolution of DNA by size or topology.Electrophoresis. 1989; 10: 290-295Crossref PubMed Scopus (33) Google Scholar Because DNA fragments of similar lengths are likely to resolve to the same Rf during electrophoresis, a band is produced that is more intense or broader than other bands of similar migration distance; a phenomenon called “DNA co-migration.” Band typing by Rf alone does not recognize the difference between a band produced in this way and a band produced by a single fragment. One method used in an attempt to prevent DNA co-migration is to run longer gels. This effectively increases the resolving power of electrophoresis, allowing similar, but not identical, lengths of DNA to form discrete bands. Unfortunately, the use of longer gels also results in a greater dispersion of DNA fragments within the gel during the molecular sieving process; bands produced by smaller DNA fragments disappear as the gel length increases. Clearly, the limitations of the resolving power of PFGE illustrate a need for alternative methods of increasing discrimination between closely related restriction patterns. This study presents a classification algorithm that increases the sensitivity of current PFGE techniques in detecting pattern differences despite the potential for DNA co-migration. The algorithm is not intended to illustrate the biological relevance of subtle PFGE pattern differences per se, but rather to provide an unbiased approach for detecting their existence using current PFGE methods. An established method of partition recognition (stopping rule) was also evaluated. Forty-seven TSST-1 toxin-producing strains of S. aureus were acquired from nasal, anal, or vaginal swabs taken from women living in various geographic locations including, Ohio, Florida, Arizona, and New Jersey in the United States, and Manitoba in Canada. Strains were isolated as part of a large epidemiological study to determine rates of S. aureus carriage and occurrence of TSST-1 toxin-producing strains. TSST-1 toxin-producing isolates obtained from women living in one geographical location were selected to illustrate the standardization algorithm described in this article. Standard microbial techniques were used for isolating S. aureus. Briefly, swabs from each subject were streaked onto mannitol salt agar plates (PML Microbiologicals, Tualatin, OR) and incubated for 48 hours at 37°C. Isolated colonies were then streaked for purity onto tryptic soy agar plates with 5% sheep blood (PML Microbiologicals) and incubated for 24 hours at 37°C. A gram stain, catalase test, and a rapid S. aureus-specific latex agglutination test (Staphaurex; Remel, Lenexa, KS) were performed to confirm the identity of isolates as S. aureus. Isolates so identified were grown in brain heart infusion broth overnight at 37°C and the supernatant was evaluated for the presence of the TSST-1 toxin via a competitive enzyme-linked immunosorbent assay.10Parsonnet J Mills JT Gillis ZA Pier GB Competitive, enzyme-linked immunosorbent assay for toxic shock syndrome toxin 1.J Clin Microbiol. 1985; 22: 26-31PubMed Google Scholar TSST-1 toxin-producing S. aureus isolates were then stored at −80°C for later analysis. S. aureus isolates were incubated overnight on an orbital shaker at 37°C in brain heart infusion broth. After incubation, 200 μl of the cell culture were harvested and washed with cell suspension buffer [10 mmol/L Tris, pH 7.2, 20 mmol/L NaCl, 50 mmol/L ethylenediaminetetraacetate (EDTA)] and resuspended in 100 μl of fresh cell suspension buffer. Two μl of RNase A stock (10 mmol/L Tris Base, 0.1 mmol/L EDTA, ribonuclease A 1.25 mg/ml; Sigma, St. Louis, MO) was added to the suspension, which was warmed in a 50°C water bath. The preparation was mixed with 100 μl of 2% CleanCut agarose (Bio-Rad, Richmond CA), vortexed lightly, and cast in disposable plug molds (Bio-Rad). The plugs were then placed into a lysis solution containing 237 μl of lysozyme buffer (10 mmol/L Tris, pH 7.2, 50 mmol/L NaCl, 0.2% sodium deoxycholate, 0.5% sodium lauryl sarcosine; Bio-Rad), 10 μl of lysozyme stock (25 mg/ml, Bio-Rad), and 2.5 μl of lysostaphin stock (100 mmol/L Tris base, 40 mmol/L magnesium sulfate, 0.8 mol/L sucrose, pH 7.6, lysostaphin 10 mg/ml; Ambi Inc.) and incubated at 37°C for 4 hours. The plugs were then washed briefly with 1× wash buffer (20 mmol/L Tris, pH 8.0, 50 mmol/L EDTA; Bio-Rad) and incubated overnight at 50°C in 250 μl of proteinase K reaction buffer (100 mmol/L EDTA, pH 8.0, 0.2% sodium deoxycholate, 1% sodium lauryl sarcosine; Bio-Rad) with 10 μl of proteinase K stock (>600 U/ml, Bio-Rad). These plugs were then washed four times with 1× wash buffer; the second and third wash were treated with 10 μl of phenylmethyl sulfonyl fluoride stock (100 mmol/L phenylmethyl sulfonyl fluoride in 100% isopropanol) and stored at 4°C for later enzymatic treatment. Plugs were cut to size (5 × 1.5 × 2.5 mm) and digested in 100 μl of restriction buffer (10 mmol/L Tris-HCl, 50 mmol/L KCl, 7 mmol/L MgCl2, 1 mmol/L dithiothreitol, pH 7.75) with 40 U SmaI (Promega Corp., Madison, WI) overnight at 24°C. The plugs were then washed in 1× wash buffer, equilibrated in 0.5× TBE buffer (45 mmol/L Tris, 45 mmol/L borate, 1.0 mmol/L EDTA, pH 8.3), and loaded into a 1.2% pulsed-field certified agarose (Bio-Rad) gel with the samples flanked by bacteriophage λ DNA concatemers CI857Sam7 (Roche Molecular Biochemicals, Indianapolis, IN). All gels were electrophoresed in 0.5× TBE buffer at 5.1 V/cm for 36 hours at 14°C with a pulse duration of 1 to 85 seconds ramped linearly in a CHEF-DR II system (Bio-Rad). Gels were stained with ethidium bromide, destained in distilled water, and photographed with a Gel Doc 2000 (Bio-Rad). Data were obtained from the digital image using the Diversity Database software (Bio-Rad). This included band typing by Rf and trace quantities, with trace quantity defined as the integration of the signal intensity over the width and height of a particular band in the image. Thus, a trace quantity directly reflects the band intensity or brightness. Band assignment was determined by Rf values plus or minus 5% error. Data standardization and classification were performed using Minitab for Windows (Minitab Inc., State College, PA). Standardized data can be obtained by dividing values of all trace quantities from all bands produced by a given isolate by the sum of those intensities from that isolate to produce adjusted trace quantities (e1a) and then dividing the resulting values by the maximum adjusted trace quantity of all like band types produced by all samples (e1b). This can be written in mathematical terms as a two-part algorithm (e1a then e1b): a)Ai,j=Mi,j∑(Mj)b)Si,j=Ai,jMax(Ai)(e1a,b) where M is a the data and and the DNA and The values in S can be as the standardized relative of the like band of the same band of all Thus, the band of each band is by and all other like band intensities a standardized of the band for a given band type. The for this algorithm a given DNA fragment is the of restriction DNA digestion by restriction enzyme is DNA band in a PFGE gel stained with ethidium bromide, is to the quantity of DNA the size for evaluated isolates are and an of the total genomic DNA is These are well to large of the same and for nosocomial an analysis of banding patterns that DNA co-migration. analysis was performed by the method on of Rf or data at an by partitioning method with the partition produced by the W analysis in 1992; Google Scholar The of determined using was closely or by the difference This the total by the method where the of and is the of The maximum of was then used as the where is the of and is the of The of groups is defined as the of that W analysis in 1992; Google A for determining the of groups in a data using Scopus Google An of for determining the of in a data 1985; Scopus Google Scholar a maximum of was to the of of the was and W analysis in 1992; Google Scholar The groups determined by the were evaluated for band differences the between or more all of isolate patterns in all within a were evaluated using where is a for the isolate and the total between the and The and maximum between a of isolates was also by to all of isolate for that The is an of genetic between isolates within a and is called the is a useful variable is a of genetic RV The molecular epidemiology of nosocomial infection: in overview of principles, application, and interpretation.in: Specter S Bendinelli M Friedman H Rapid Detection of Infectious Agents. Plenum Press, New York1998: 131-157Google Scholar, 9Tenover FC Arbeit RD Goering RV Mickelsen PA Murray BE Persing DH Swaminathan B Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing.J Clin Microbiol. 1995; 33: 2233-2239PubMed Google Scholar this study a maximum of which to genetic at most and RV The molecular epidemiology of nosocomial infection: in overview of principles, application, and interpretation.in: Specter S Bendinelli M Friedman H Rapid Detection of Infectious Agents. Plenum Press, New York1998: 131-157Google Scholar, 9Tenover FC Arbeit RD Goering RV Mickelsen PA Murray BE Persing DH Swaminathan B Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing.J Clin Microbiol. 1995; 33: 2233-2239PubMed Google was an for a values to or less than this a genetic to can be to a maximum genetic by the of be that the are from band quantity or data not be used for this of at numerical a band produced in PFGE is from or DNA fragments is and not on the need for to all trace quantity or values a was was that one band was the of one DNA fragment this is with current methods of PFGE pattern and more to current of a of genomic and DNA for a of a can only the genetic diversity for that the 47 isolates used to were 20 banding patterns on band only. a algorithm was at an determined by and band was the only variable taken into were band intensity was were determined an partitioning of from and The isolate determined by the methods were for the use of the partitioning methods described the that was to produce as determined by was and for Rf only and data, This partitioning not been by a of or with or The partition for data are not as defined as the for 1 many and maximum values between and analysis methods ranging from to that a between the methods or not be by maximum values at the same of but isolate be This at a of between and typing is determined by band migration only. typing is determined by the analysis where band migration and band intensities are that the which to the partitioning of their data are in a typing is determined by band migration only. typing is determined by the analysis where band migration and band intensities are that the which to the partitioning of their data are The use of band intensity and produced more than the use of band alone 1 and 4 were each cut into by the where the difference was a single was cut into on Rf or band intensity The differences on band intensity alone were most and the isolates were subjected to a endonuclease enzyme and PFGE each to or as determined by by were digested using endonuclease enzyme and resolved by PFGE using the same as described the one can subtle differences between and where SmaI digestion only differences, the digestion isolates are of a related gels and illustrate the subtle differences between which the analysis method is to The characterization by analysis was in high with analysis by (45 of 47 where was of 47 This was established by the PFGE patterns of the 47 isolates and in a to a with only the used to Band brightness was taken into and isolates were were on by with of PFGE pattern The was then with partitioning determined by and analysis. The only differences between and analysis methods were the of PFGE patterns that been into one of by or on which bands were not The of groups determined by the produced a maximum of ranging from to 7 This to one or genetic the most within defined These results that the of defined by is the of analysis one fragment of DNA resolved by PFGE be regardless of the fragment fragments are the and so then to the of DNA This directly to the and of endonuclease enzyme recognition the bacterial ranging from DNA and enzyme and gel-to-gel discrete of DNA fragment of form of standardization that for the of DNA as well as other data, band as at values data be or more fragments of DNA to be to the pattern all methods a data W analysis in 1992; Google Scholar patterns a hierarchy of band that the bands produced by fragments of DNA are more intense than the bands produced by smaller DNA a the values the band intensities of fragments of DNA an in analysis methods values are used directly PFGE band intensity data using methods is not PFGE data many that the of and the PFGE data is that of a of analysis methods. The method of standardization band patterns to a data by dividing trace quantities by the sum of those band were to as adjusted trace quantities This mathematical standardize trace quantities between all isolates regardless of DNA or gel-to-gel that of DNA and of size are the of band hierarchy all like values by the or numerical in a than 1 and smaller than 0 for given this for all values results in a standardized data that is practical for analysis patterns by the of PFGE are one isolate an high of DNA a After standardization and the isolate a band be characterized as from the other as this be a large used as a the with this band those with similar bands of other thus reducing the in characterization a mathematical this does not to be to the of the standardized data to be of the PFGE patterns. Two limitations of the analysis are all isolates be in the analysis of the standardization and that large data be as the standardization for a large be Data partitioning on a The of the is an data and this to be adjusted in with the of isolates evaluated. This is related to the of the data an the in a of isolates is well defined samples can in in the between This does not the groups or their to each other but does adjusted indices to an analysis for that are more and are but are not by methods is that groups be defined are the numerical classification is the methods. one of the few of that does is that are use of method is a in not where are application, this A for determining the of groups in a data using Scopus Google Scholar The one analysis is performed on A for determining the of groups in a data using Scopus Google Scholar a at the same while using of data the same one can be that an pattern in the data the as described by and A for determining the of groups in a data using Scopus Google Scholar is not the only on the data does to be to at determining the of most are that approach to determining the of is unbiased and with the used in the W analysis in 1992; Google Scholar the is an in the data as a into the maximum difference between and 1 of the total for each of is for the of groups present in an data The analysis method is for the of nosocomial and of same using Standardized band intensity as a for isolate and DNA co-migration, is the of this data patterns as of polymorphic DNA or cell patterns as that produced by sodium gel analysis is designed to data for each evaluated banding allowing a partitioning of the samples where use as to discriminatory the analysis a as the one described an unbiased approach to determining data in and at groups are This algorithm be into software as an method of data be to methods. and methods of also used by this algorithm is that be in a analysis using most software The of the analysis algorithm over current in software as and Diversity Database (Bio-Rad) is with as a by a capable of analysis as or and a can band pattern analysis at a over the analysis method used by Diversity Database a of band brightness. and bands are given more than smaller bands in the of patterns. This results in an partitioning of the data DNA co-migration is an of PFGE patterns. one isolate has fragments of DNA and does the isolates in all are DNA co-migration as an can the discriminatory power of PFGE by increasing the of for analysis. and determined by analysis of data in this study illustrate that The in band types and between the are for the method to separate the isolates into entities at a of This and other partitioning by analysis the of discriminatory power in this study from to with analysis. of the PFGE standardization methods described permits of DNA co-migration, in PFGE during analysis. the an unbiased tool for determining the of for an data The and for the S. aureus and TSST-1 for for of the standardization and for their in the

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.