Rapid Identification of Biothreat and Other Clinically Relevant Bacterial Species by Use of Universal PCR Coupled with High-Resolution Melting Analysis |
| |
| Authors: | Samuel Yang Padmini Ramachandran Richard Rothman Yu-Hsiang Hsieh Andrew Hardick Helen Won Aleksandar Kecojevic Joany Jackman Charlotte Gaydos |
| |
| Affiliation: | Johns Hopkins University, Department of Emergency Medicine, Baltimore, Maryland,1. Johns Hopkins University, Division of Infectious Diseases, Baltimore, Maryland,2. Applied Physics Laboratory, Johns Hopkins University, Laurel, Maryland3. |
| |
| Abstract: | ![]()
A rapid assay for eubacterial species identification is described using high-resolution melt analysis to characterize PCR products. Unique melt profiles generated from multiple hypervariable regions of the 16S rRNA gene for 100 clinically relevant bacterial pathogens, including category A and B biothreat agents and their surrogates, allowed highly specific species identification.Rapid and accurate diagnostic tools are critical for infectious disease surveillance and early diagnosis of disease (8, 12). A simple platform which could deliver broad-based screening and specific pathogen identification would be invaluable for the timely recognition of emerging and biothreat (BT) outbreaks, as well as commonly encountered clinical infections (2, 7, 9, 11, 12).We previously reported a probe-based PCR assay, which utilizes conserved and variable 16S rRNA gene sequences for initial broad-based eubacterial detection and subsequent identification of specific bacterial agents (11). The assay demonstrated high analytical sensitivity but was limited by an inability to differentiate closely related pathogens due to decreased specificity of the TaqMan probe chemistry and high sequence homology within selected hypervariable regions of the 16S rRNA gene. Probe-based amplicon characterization accordingly limits testing to a finite number of anticipated pathogens. Alternative strategies for amplicon analysis, such as sequencing and mass spectrometry, allow broader-scale product characterization but are costly, time-consuming, and lacking in throughput (1, 6). High-resolution melt analysis (HRMA) offers a simple, low-cost, closed-tube approach to amplicon analysis with the capacity for single-nucleotide discrimination and easy integration with PCR analysis (10). We report a unique strategy for the rapid, highly specific identification of BT- related and non-BT-related bacterial pathogens which couples eubacterial PCR with HRMA.Three hypervariable regions (V1, V3, and V6), each flanked by highly conserved sequences within the 16S rRNA gene, were selected for primer design (3). Sequence data for clinically or BT-relevant bacteria were obtained from GenBank and aligned using ClustalW (www.ebi.ac.uk/clustalw/) to determine sequence variability. Primer pairs used to target hypervariable regions were as follows: V1-F (5′-GYGGCGNACGGGTGAGTAA-3′) and V1-R (5′-TTACCCCACCAACTAGC-3′), V3-F (5′-CCAGACTCCTACGGGAGGCAG-3′) and V3-R (5′-CGTATTACCGCGGCTGCTG-3′), and V6-F (5′-TGGAGCATGTGGTTTAATTCGA-3′) and V6-R (5′-AGCTGACGACANCCATGCA-3′).One hundred common, BT-related, and BT-surrogate organisms composed of 58 different bacterial species of American Type Culture Collection (ATCC) strains, clinical isolates, or inactivated or nonpathogenic strains were used for analysis (Table ) and cerebral spinal fluid samples collected from patients suspected of having septic arthritis or bacterial meningitis, respectively, were also used for blinded analyses.TABLE 1.Melting analysis of non-BT-related and BT-related organisms| Organism group | Organism or strain | Grouping code of analysis subseth
| Signature codei |
|---|
| V1 | V3 | V6 |
|---|
| Non-BT related | Acinetobacter sp. strain ATCC 5459 | a | b | a | aba | | Acinetobacter calcoaceticus | b | d | a | bda | | Aerococcus viridans | f | h | c | fhc | | Bacteroides fragilisa | a | a | e | aae | | Bordetella pertussisa | c | c | f | ccf | | Bordetella parapertussis | a | c | h | ach | | Campylobacter jejunia | c | a | e | cae | | Clostridium difficile | g | f | a | gfa | | Clostridium perfringens | b | d | d | bdd | | Corynebacterium sp.a | c | c | e | cce | | Chlamydia pneumoniaea | g | c | a | gca | | Chlamydia trachomatisa | f | a | b | fab | | Citrobacter freundiia | b | c | a | bca | | Enterobacter aerogenes | c | b | a | cba | | Enterococcus gallinarum | i | i | h | iih | | Enterococcus faecium | b | a | e | bae | | Enterobacter faecalis ATCC 29212 | i | i | a | iia | | Escherichia coli ATCC 25927 | e | d | c | edc | | Helicobacter pyloria | g | b | a | gba | | Haemophilus influenzae ATCC 49247 | b | g | d | bgd | | Klebsiella pneumoniaea | h | c | a | hca | | Legionella pneumophila ATCC 33495 | a | a | b | aab | | Listeria monocytogenes ATCC 7648 | b | e | a | bea | | Micrococcus sp. strain ATCC 14396 | b | b | b | bbb | | Moraxella catarrhalis | h | i | d | hid | | Mycobacterium kansasii | i | c | a | ica | | Mycobacterium gordonae | d | i | i | dii | | Mycobacterium fortuitum | a | i | b | aib | | Mycoplasma pneumoniaea | b | d | g | bdg | | Mycoplasma hominisa | a | b | e | abe | | Neisseria meningitis ATCC 6250 | d | f | c | dfc | | Neisseria gonorrhoeaea | a | c | a | aca | | Oligella urethralis | b | a | i | bai | | Pasteurella multocida | b | i | a | bia | | Pseudomonas aeruginosa ATCC 10145 | b | b | c | bbc | | Propionibacterium acnes | e | i | e | eie | | Proteus mirabilisa | b | a | f | baf | | Proteus vulgarisa | c | a | i | cai | | Salmonella sp. strain ATCC 31194 | c | e | a | cea | | Serratia marcescens ATCC 8101 | b | j | c | bjc | | Staphylococcus aureus ATCC 25923 | c | b | h | cbh | | Staphylococcus epidermidis ATCC 12228 | a | a | h | aah | | Staphylococcus lugdunensis | g | i | i | gii | | Staphylococcus saprophyticus | h | i | h | hih | | Streptococcus pneumoniae ATCC 49619 | g | d | g | gdg | | Streptococcus pyogenesa | b | e | b | beb | | Streptococcus agalactiae ATCC 13813 | b | e | d | bed | | Treponema palliduma | f | b | e | fbe | | Viridans group streptococci, ATCC 10556 | c | e | f | cef | | Category A BT agent, near-neighbor, and/or surrogate | Bacillus anthracisc | c | a | a | caa | | Strain 3001 | c | a | a | caa | | Bacillus cereusa | a | a | d | aad | | Strain BC 9634 | a | a | d | aad | | Strain BC 12480 | a | a | d | aad | | Strain BC 27877 | a | a | d | aad | | Strain BC 7064 | a | a | d | aad | | Strain BC B33 | a | a | d | aad | | Strain BC 1410-1 | a | a | d | aad | | Strain BC 1410-2 | a | a | d | aad | | Strain BC T | a | a | d | aad | | Strain BC 2599 | a | a | d | aad | | Strain BC 2464 | a | a | d | aad | | Strain BC 7687 | a | a | d | aad | | Strain BC 10329 | a | a | d | aad | | Strain BC 11143 | a | a | d | aad | | Strain BC 11145 | a | a | d | aad | | Strain BC 1414 | a | a | d | aad | | Strain BC 7089 | a | a | d | aad | | Strain BC 6464 | a | a | d | aad | | Strain BC 6474 | a | a | d | aad | | Strain BC 7004 | a | a | d | aad | | Strain BC 10987 | a | a | d | aad | | Strain BC 23674 | a | a | d | aad | | Strain BC 9189 | a | a | d | aad | | Strain BC 246 | a | a | d | aad | | Strain BC 13472 | a | a | d | aad | | Bacillus subtilis 110 NA | a | a | g | aag | | Strain SB168 | a | a | g | aag | | Strain W168 | a | a | g | aag | | Strain W23 | a | a | g | aag | | Strain her 148 | a | a | g | aag | | Strain T6 | a | a | g | aag | | Strain ATCC 27505 | a | a | g | aag | | Strain ATCC 15841 | a | a | g | aag | | Coxiella burnettib | d | b | g | dbg | | Strain “9 mile” | d | b | g | dbg | | Francisella philomiragia (GAO1-2810)d | a | g | g | agg | | Francisella tularensis (LVSB)e | b | h | g | bhg | | Strain Fran 0001 | b | h | g | bhg | | Yersinia pseudotuberculosis (PB1/+)f | a | g | c | agc | | Schutze''s group type B strain/ATCC 6903 | a | g | c | agc | | Schutze''s group II strain/ATCC 27802 | a | g | c | agc | | Strain CDC P62 strain/ATCC 29910 | a | g | c | agc | | Schutze''s group III strain/ATCC 13980 | a | g | c | agc | | Raffinose-positive strain, ATCC 4284 | a | g | c | agc | | Strain ATCC 13979 | a | g | c | agc | | Yersinia enterocolitica, O:9 serotype | a | g | d | agd | | Strain WA.C | a | g | d | agd | | Yersinia pestis (P14−)g | a | b | d | abd | | Strain 1122 | a | b | d | abd |
Open in a separate windowaClinical isolate.bCoxiella burnetti DNA was obtained from Steven Dumbler, Department of Pathology, School of Medicine, Johns Hopkins University, Baltimore, MD.cInactivated nonpathogenic strain.dNonpathogenic strain obtained from the Centers for Disease Control and Prevention, Fort Collins, CO, via the Walter Reed Army Medical Hospital, Washington, DC.eLVSB, live vaccine strain type.fWild-type strain.gDepigmented and virulence pCD1 negative.hDifference plots generated for each organism were grouped based on curve similarity within each analysis subset (V1, V3, or V6), and a unique letter code was assigned to each group as well as each individual organism with a distinct curve shape.iCombined grouping code letters assigned in each analysis subset.Extracted DNA from each organism or clinical sample was subjected to three PCR analyses, targeting V1, V3, and V6 hypervariable regions, respectively. Every PCR analysis was performed in a 10-μl total volume comprised of 8 μl of PCR master mix and 2 μl of target input. The PCR master mix contained 4 μl of 2× Universal PCR mix (Idaho Technology, Salt Lake City, UT) and LC Green dye (Idaho Technology) for high-resolution melting. A total of 1.0 μl of 1.5-μM forward primer and reverse primer was added to the master mix. Each PCR analysis contained one primer pair. The PCR was performed using a GeneAmp Thermocycler (ABI, Foster City, CA). Cycling conditions were as follows: denaturation at 95°C for 30 s, followed by 45 cycle repeats at 95°C for 30 s and annealing/extension at 60°C/72°C for 60 s, and 1 cycle at 95°C for 30 s and 28°C for 30 s.Each post-PCR sample amplicon was subjected to HRMA on the LightScanner instrument (Idaho Technology). Melting temperatures ranged from 60°C to 95°C. Data acquisition was performed for every 0.1°C increase in temperature. HRMA for each PCR sample was performed in triplicate and analyzed using the LightScanner software version 2.0 (Idaho Technology). The software function “negative filter” was first used to identify negative controls and any failed PCRs. Melt analysis of the positive samples was then subjected to fluorescence normalization and temperature shift to obtain the minimum inter- and intra-run variabilities (LightScanner version 2.0 operator''s manual; Idaho Technology, Salt Lake City, UT). Specifically, normalization minimized the variations in fluorescence magnitude between samples due to differences in starting template or optics, and a temperature shift will overcome the effect of absolute temperature variation from position to position across the plate. Derivative plots were generated to assess the number of melting peaks. Analysis subsets (V1, V3, and V6) were defined by the primer sets used for amplification. Using the melting curves of Staphylococcus aureus as the reference curve, the difference plot for each positive sample was generated for subsequent grouping analysis. “Auto grouping” was performed on the difference plots to group all positive samples with a similar curve shape within the same analysis subset. A unique letter code was manually assigned for each group identified, starting with the letter “a” and progressing alphabetically. A combination of each letter from each of the variable regions was then accumulated to provide a signature code for each organism.Each of the 100 bacterial organisms tested had a melting curve generated from HRMA for each of the analysis subsets (V1, V3, and V6) based on the primer set used. Each derivative plot revealed a single dominant peak, which was absent in the nontemplate control, indicating the presence of a single amplified sequence. The melting curves were demonstrated to be reproducible from run to run despite various target DNA concentrations over a 10,000-fold range (data not shown). Using the melting curve of Staphylococcus aureus as the reference, difference plots of the 100 tested organisms generated were compared within their analysis subset. The S. aureus melting curve was chosen as the reference curve, due mainly to the high sequence homology between various S. aureus strains (n = 8) compared within our target amplified regions. After grouping analysis, each difference plot was assigned a unique code letter and only plots with similar characteristics within the same analysis subset shared the same code letter (Fig. ; Table ). Identical signature codes were observed among various strains of the same species (Table Open in a separate windowThe difference plots of all the category A BT bacterial organisms and their surrogates. A grouping code letter (indicated on the top left corner of each graph) is assigned for each plot based on similarity in curve shape with other organisms under the same analysis subset (V1, V3, or V6).We also performed HRMA on eubacterial PCR products from 40 blinded archived clinical samples, which included synovial fluids and cerebral spinal fluids previously collected from patients suspected of having septic arthritis or bacterial meningitis, respectively. HRMA correctly identified all 20 culture-negative samples as being negative. The signature codes generated from each of the 20 remaining positive samples were compared to our reference database of 58 different bacterial species for identification (Table | Open in a separate windowaTwenty blinded culture-negative samples were tested and were identified as negative by HRMA. N. meningitidis, Neisseria meningitidis.In this study, we demonstrate as proof of concept a simple, powerful approach to amplicon analysis for rapid bacterial species identification and differentiation of BT agents from their related surrogates. This approach relies on eubacterial real-time PCR analysis followed by HRMA. Unlike probe-based approaches to amplicon analysis, melt curve analysis can characterize PCR products without a priori knowledge of anticipated organisms. Further work will be required to develop a comprehensive database of signature codes from all common pathogens. Once established, nonmatching code generated from a positive amplification reaction may signify the presence of an uncommon, mutant, or emerging pathogen. This approach offers a simple work flow with a total turnaround time of 2 h (from sample collection to species identification) and obviates the need for laborious post-PCR procedures. Due to the ease of integrating the melt analysis, this approach has the potential to be used as a point-of-care test and may be feasible in resource-deficient clinical settings.Despite the high discriminatory precision of HRMA, we found that amplicons of very different sequences may generate similar melt curves. These findings have been reported previously (4). To resolve “melting groups,” Cheng et al. performed heteroduplex melt analyses between amplicons of unknown and reference bacterial species (4). A potential drawback with this approach is that closely related species with identical sequences within the amplified region may not be readily differentiated. We chose to analyze the melt profiles based on three instead of one of the 16S, hypervariable regions (3, 5). This yielded a unique set of melt plots for every non-BT or BT-relevant bacterial organism tested, with even closely related species able to be discerned (13). As expected, different strains of the same species with identical target sequences shared similar melt profiles. Future studies will determine whether the triple-PCR analyses are more cost-effective when performed in parallel or in a series for routine diagnostic testing and/or surveillance.Potential limitations of using melt analysis for pathogen identification include nucleotide polymorphisms, which may exist between intragenomic copies of the 16S rRNA gene in some bacterial species, as well as polymicrobial infections. The number of peaks in the derivative plot may allow discrimination of single versus multiple infections. Future studies will focus on assay reproducibility and specificity using expanded panels of clinically relevant bacterial species, animal studies with BT agents, and human clinical validation studies of patients with suspected systemic bacterial infections. |
