Development of Duplex Droplet Digital PCR for Rapid Detection of Blood Stream Infections Caused by S. aureus, E. coli, S. pneumoniae and P. aeruginosa

doi:10.21203/rs.3.rs-1760127/v1

Purpose: The aim of this study was to develop highly specific primer-probes pairs for the detection of species-specific genes, and establish a multiplex ddPCR assay for the detection of low copies of genomic DNA, of common bacteria causing bloodstream infections in our institution, i.e. Staphylococcus aureus, Streptococcus pneumonia, Escherichia coli, and Pseudomonas aerugnosa..

Methods: Pubmed and GenBank databases were used to select well-characterized candidate genes for each microorganism. The newly designed primer-probe pairs were designed using AllelID 7.85 software. The designed primer-probe pairs were tested and validated using Droplet Digital PCR QX200 by using quantitative ATCC genomic DNA from S. aureus (ATCC® 29213DQ™), S. pneumoniae (ATCC® 700669DQ™), E. coli (ATCC® 25922DQ™), and P. aeruginosa (ATCC® 27853DQ™).

Results: All primer-probe pairs demonstrated an optimal annealing temperature at 56°C. The ddPCR detected 162 copies for S. aureus and 152 copies S. pneumonia in singleplex reactions as compared to 152 copies and 132 copies in duplex reactions, respectively, and 324 copies for P. aeruginosa, 30 copies for E. coli in singleplex reactions as compared to 356 copies and 34 copies, respectively, in duplex reactions. The results of the limit of detection of the ddPCR assay was as low as 1 copy/reaction for S. aureus, E. coli, and P. aeruginosa and 2 copies/reaction for S. pneumonia.

Conclusion: In this study, we have designed highly specific PP pairs for use in a ddPCR assay for the accurate and reliable detection and quantification of species-specific genes for four common blood culture pathogens in our institution.

Clinical microbiology

ddPCR

molecular diagnostics

infectious diseases

nosocomial infections

Sepsis remains one of the leading causes of death worldwide, affecting more than 49 million people with an estimated 11 million deaths each year. The high mortality rate for patients with symptoms of sepsis (more than 26% for inpatient and 42% for ICU patients) is associated with a delay in appropriate antimicrobial treatment. (1–3) The blood culture method remains the gold standard and cornerstone for microbiological diagnosis of blood stream infections. Unfortunately, it is not a rapid diagnostic method with a detection and identification time of 24-72h, and has a high rate of negative results (failing to detect a pathogen in approx. 50%)certain limitations. The main limitations of blood cultures are the infections with non-cultivable or fastidious microorganisms, and antimicrobial treatment prior to blood collection. (4, 5) A rapid and highly sensitive diagnostic method and early treatment can prevent up to 80% of sepsis-related deaths. (6)

In recent years, the Droplet Digital Polymerase Chain Reaction (ddPCR), which is a technique where a reaction is partitioned into approximately ≤ 20,000 nano-sized oil droplets prior to amplification, has gained attention due to its high analytical sensitivity. (7) A study by Wouters et al. 2019 has shown a high concordance between blood culture and ddPCR by retrospectively analysing frozen whole blood samples using 16S rRNA and Gram staining for differentiation between Gram negative and Gram positive bacterial and fungal 28S rRNA. In this study with a limited number of known positive blood culture samples, the ddPCR assay had a limit of detection (LoD) of 1–2 bacteria or fungi per ddPCR reaction, and an overall sensitivity and specificity of 80% (95% CI 52–96) and 87% (95% CI 69–96), respectively. Using ddPCR, they were able to detect pathogens in whole blood within 4h. (8) A recent study by Zhang et al. 2021 has also demonstrated high detection accuracy in blood samples. They were able to simultaneously identify Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) nucleic acid with almost no interference between the targets and with a detection rate as high as 80–90% in simulated bacteraemia blood samples. (9) The development of this method is an essential prerequisite for the development of a multiplex ddPCR assay for routine detection of the most common sepsis causing pathogens in whole blood and/or blood culture with minimal incubation requirement. (10)

The aim of this study was to develop highly specific primer-probes pairs for the detection of species-specific genes, and establish a multiplex/duplex ddPCR assay for the detection of low copies of genomic DNA, of common bacteria causing bloodstream infections in our institution, i.e. S. aureus, cpsA Streptococcus pneumoniae (S. pneumoniae, E. coli, and Pseudomonas aeruginosa (P. aeruginosa). We also sought to determine the limit of detection (LoD) of the assay using a series of five-fold serial dilutions of well-characterized genomic bacterial DNA, and to demonstrate the potential for simultaneous multiplex detection.

Design of Primer-Probe Sequences

Pubmed and GenBank database was used to select well-characterized candidate genes for each microorganism; coa (staphylocoagulase gene) in S. aureus, cpsA (capsular polysaccharide gene) in S. pneumoniae, uidA (beta-D-glucuronidase gene) in E. coli, and oprL (peptidoglycan-associated lipoprotein gene) in P. aeruginosa. The primer-probe (PP) pairs for the selected target genes were designed following the guidelines for assay design for ddPCR (applications guide by Bio-Rad; USA) using AllelID 7.85 software (Premier Biosoft; USA). In addition, AllelID 7.85 was used to analyse candidate sequences to minimise the potential formation of self-dimers and cross primer-dimers. All PP pairs were synthesized by Bio-Rad (USA). For Accession number, PP-sequences, amplicon lengths, and American Type Culture Collection (ATCC)-strains used in this work, see Table 1.

To confirm 100% species-specificity of the designed PP-sequences with the bacteria and ATCC-strains utilised in the study, a systematic bioinformatics search was conducted using BLAST (Basic Local Alignment Search Tool; NCBI) and the ATCC genome-searching tool (https://genomes.atcc.org/sequence-search). The ATCC genome-searching tool confirmed a 100% match with at least five different ATCC strains for each PP design (data not shown).

DNA Extraction and Purification

The DNA from S. aureus (ATCC® 29213DQ™), S. pneumoniae (ATCC® 700669DQ™), E. coli (ATCC® 25922DQ™), and P. aeruginosa (ATCC® 27853DQ™) was extracted and purified using QIAamp DNA Mini Kit (Qiagen; Germany) following the manufacturer's instructions. Lysostaphin (200 µg/ml) (Thermo Fisher; USA) was used for the DNA extraction in Gram-positive bacteria. However, the step of incubation for 15 min. at 95°C (step D5 in the chapter ‘Isolation of genomic DNA from Gram-positive bacteria’; QIAamp DNA Mini and Blood Mini Handbook) was excluded. The DNA was eluted in 100µL and 10-fold serially diluted (up to 1:10⁵), and used for preliminary testing and validation of the PP pairs (data not shown). All stocks and dilutions were stored at -20°C. The DNA concentration and purity (ratio of absorbance at 260 nm and 280 nm) was measured using Nanodrop (Thermo Fisher; USA) – all samples used in this study had a ratio of 1.8-2.0 (data not shown).

Droplet Digital PCR Reaction

The PCR reaction mix contained 12 µL of 2x ddPCR Supermix for probes (no dUTP) (Bio-Rad; USA), 1.2 µL of PP-mix (900 nM of primers and 250 nM of probe), 2 µL of template, and 8.8 µL of nuclease-free water (VWR; USA) for a total volume of 24 µL per reaction. The droplets were generated using the Automated Droplet Generator (Bio-Rad; USA), DG32 Automated Droplet Generator Cartridges, Filter Tips, and 20 µL of Automated Droplet Generation Oil for Probes (Bio-Rad; USA), generating a total volume of 40 µL in each well. After the amplification of targets, the QX200 Droplet Reader (Bio-Rad; USA) was used to count the PCR-positive and negative droplets. For simultaneous detection of two targets, S. aureus and S. pneumoniae were detected in one well, and E. coli and P. aeruginosa in another well separately.

Thermo-Cycling Conditions

The thermo-cycling conditions for the amplification of the PCR reaction, using a Veriti 96-well Thermal Cycler (Applied Biosystems; USA), were as follows: 10 min at 95°C, 40 cycles of denaturation: 30 sec at 94°C and extension: 60 sec at 56°C, respectively. 10 min at 98°C, and infinite (hold) at 4°C. The ramp rate was set to 65% corresponding to 2°C/sec. The optimal annealing/extension temperature was determined based on a temperature gradient experiment (see Results section below),

Limit of Detection (LoD)

Quantitative ATCC genomic DNA (LGC Standards Ltd; USA) from well-characterized stocks (≥ 10⁵ copies/µL) were used. The stocks were diluted to ≥ 10⁴ copies/µL followed by a five-fold serial dilution from ≥ 10⁴ to ≥ 0.64 copies/µL and subsequently used for the determination of LoD for each target in singleplex reactions (for ATCC-strains, see Table 1). QuantaSoft™ Analysis Pro Software (Bio-Rad) was used for the analysis of ddPCR results.

Thermal Gradient Optimization and Cycling Conditions

To determine the optimal annealing/extension temperature, a temperature gradient (52–62°C), with an interval of 2°C going from A/B01 and 02 to A/B11 and 12 (Fig. 1), was performed. The results showed a clear separation between the positive and negative droplet ‘clouds’ for three of the targets i.e. S. aureus (coa), E. coli (uidA), and P. aeruginosa (oprL) on the whole temperature range while only from 52–56°C (from B02 to A/B05 and 06) for S. pneumoniae (cpsA).

The amplitude for FAM fluorescence positive droplets for the S. aureus (coa) (Fig. 1A) was 3–4 fold higher as compared to the negative droplets in the temperature range of 52–60°C (A01-A09), and 2-fold higher for HEX fluorescence positive droplets for S. pneumonia (cpsA) (Fig. 1B) in the temperature range of 52–56°C (A02-A06). For FAM fluorescence positive droplets for the P. aeruginosa (oprL) (Fig. 1C), the amplitude was 7–10 fold higher as compared to negative droplets in the temperature range of 52–60°C (B02-B09), and 3-fold higher for HEX fluorescence positive droplets for E. coli (uidA) (Fig. 1D) in the temperature range of 52–60°C. Based on these results, the optimal annealing temperature was determined to be 56°C (A05, A06, B05 and B06) and chosen for further analyses for all targets.

ddPCR Duplex Assay

Next, we validated the specificity of the assay comparing the detection of targets in singleplex and duplex reactions (simultaneous detection) performed in duplicate. The results showed similar detection in singleplex reactions for S. aureus (1D drop plot in Fig. 2A) with 162 copies and S. pneumonia (1D drop plot in Fig. 2B) with 152 copies as compared to the duplex reaction/simultaneous detection of both targets (2D drop plot in Fig. 2C) with 152 copies (blue droplets) and 132 copies (green droplets), respectively. Furthermore, the ddPCR detected 324 copies (Fig. 3A) for P. aeruginosa (oprL) and 30 copies (Fig. 3B) for E. coli (uidA) in singleplex reactions as compared to 356 copies and 34 copies, respectively, in duplex reaction (Fig. 3C). In both groups/duplex reactions, the results showed no interference between the targets when detected simultaneously. In addition, all PP pairs were tested and validated for primer-dimer and cross-dimer formation, and for non-specific amplification by running an experiment with only non-template controls (NTCs) containing all PP pairs in the same wells and by running the PP pairs against all other non-target bacterial DNA, respectively (data not shown).

Limit of Detection (LoD)

Finally, we examined the LoD of the assay using five-fold serial dilutions of well-characterised quantitative bacterial genomic DNA stocks/reference materials of all targets in different concentrations ranging from ≥ 10⁴ to ≥ 0.64 copies/µL. In this experiment, the ddPCR was able to detect genomic DNA ranging from: 5.732 x 10⁴ to 1 copies/reaction for S. aureus (Fig. 4A), 3.676 x 10⁴ to 2 copies/reaction for S. pneumonia (Fig. 4B), 3.572 x 10⁴ to 1 copies/reaction for P. aeruginosa (Fig. 4C), and 5.1020 x 10⁴ to 1 copies/reaction for E. coli (Fig. 4D). The ddPCR assay for each target detected 2–3 fold the amount of copy number as compared to the manufacturer-specified minimal amount that should be present in the reference materials at the high end of the LoD. For comparison between the copy number values of the reference and the measured copy number values for all targets, see Table 2. The NTC in all experiments demonstrated no contamination.

In this study, we successfully developed a multiplex ddPCR assay for the detection of four clinical important bacteria causing bloodstream infections in our institution. The species-specific genes for the chosen bacteria i.e. coa in S. aureus, cpsA in S. pneumonia, uidA in E. coli, and oprL P. aeruginosa were all selected using the GenBank database, and to our knowledge none of the designed PP-pairs have been previously published. Due to the variations that exist within the selected genes of each bacterial species, we ensured that our design was a 100% match for at least five different ATCC strains for each species (data not shown) utilised in this study (Table 1).

All the designed PP-pairs demonstrated a clear and sufficient separation between the positive and negative droplet clouds within the same temperature of 56°C (as shown in Fig. 1A-D) and no interaction between PP pairs or interference between the targets when detected simultaneously. This demonstrated a high specificity for each target making them suitable for a multiplex reaction. We further demonstrated that the separation of droplets and the amount of copies detected for each target were unaffected in a simultaneous detection of targets in duplex reactions, as compared to singleplex reactions (as shown in Figs. 2 and 3A-C). Zhang et al. 2021 has also demonstrated the suitability of ddPCR for use in multiplex reactions. They simultaneously identified E. coli and S. aureus in simulated bacteraemia blood samples with a detection rate as high as 80–90%, and with the actual copy numbers detected by ddPCR close to the theoretical maximum copy number. (9)

Using five-fold serial dilutions of well-characterized genomic bacterial DNA (ATCC reference materials), we were able to detect at least 2–3 fold the amount of copy number as compared to the manufacturer-specified minimal amount that should be present in the reference materials at the high end of the LoD. Furthermore, as low as 1–2 copies/reaction for each target at the low end (as shown in Fig. 4A-D and Table 2), demonstrating high sensitivity. These results are similar to the results of a study by Wouters et al. 2019, which also demonstrated a LoD of 1–2 copies/reaction when detecting 16S rRNA. (8)

The strengths of this study are that this is, The strengths of this study are that we have managed to develop PP pairs not previously described for the chosen bacteria for use in a ddPCR assay, and that the PP pairs can be successfully multiplexed. We have also demonstrated that the sensitivity of this assay is considerably higher than that of standard qPCR (LoD of approximately 64 CFU/mL). (11, 12) It is a limitation of this study that the Droplet Reader only has two channels, which makes it difficult to do a probe-based multiplexing with more than two targets. However, a recent study by Nyaruaba et al. 2021 has demonstrated that an amplitude multiplexing is possible for different SARS-CoV-2 applications by varying the primer-probe concentrations for each target. (13) It may be possible to apply this approach with our four assays and thereby be able to detect all pathogens simultaneously in one well in a quadruplex reaction. Another limitation of this study is that we have so far only developed primers for the four specified bacterial species.

In conclusion, in this study, we have designed highly specific PP pairs for use in a ddPCR assay for the accurate and reliable detection and quantification of species-specific genes for four common blood culture pathogens in our institution, i.e. coa in S. aureus, cpsA in S. pneumonia, uidA in E. coli, and oprL in P. aeruginosa. We have demonstrated that these single assays can be utilised in a multiplex ddPCR assay with a LoD as low as 1–2 copies/reaction using well-characterized genomic bacterial DNA. In the light of studies by Wouters et al. 2019 and Zhang et al. 2021 (8, 9) using the ddPCR in blood samples, the present study demonstrates the potential for the use of our assay in the rapid molecular diagnosis of sepsis and detection of a low number of copies of bacterial DNA for common pathogens causing bloodstream infections. In order to investigate the further use of this assay for rapid detection and identification of the chosen bacteria in patients with symptoms of sepsis, studies of the analytical and clinical sensitivity and specificity in both whole blood and blood cultures are warranted.

Ethics approval and consent to participate

No ethical considerations since no clinical samples were used.

Consent for publication

All authors give their consent to publish this article.

Competing interests

None to declare.

Funding

This work is supported and funded by Region of Southern Denmark (grant number 20/14813), Hospital South West Jutland (grant BF/LH), A.P. Moller Fund (grant 20-L-0274), and Knud & Edith Mindefond (grant 62786-2020).

Author contributions

Samir Badran performed the study with support from Rikke Fredslund Andersen for analysing and interpreting the data. All authors contributed equally to the conception and design of this study. The study was revised critically for important intellectual content by all authors.

Data availability

Data available by request to the corresponding author.

Acknowledgements

The study was conducted at the Department of Biochemistry and Immunology, Lillebaelt Hospital, Vejle, Denmark, providing laboratory assistance and ddPCR equipment. A special thanks to Laboratory Technician Tina Brandt Christensen for her support and technical assistance in the study.

Global regional. and national age–sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. The Lancet. 2015 Jan;10(9963):117–71. 385(.
Fleischmann-Struzek C, Mellhammar L, Rose N, Cassini A, Rudd KE, Schlattmann P, et al. Incidence and mortality of hospital- and ICU-treated sepsis: results from an updated and expanded systematic review and meta-analysis. Intensive Care Med. 2020 Aug;46(8):1552–62.
World Health Organization. Global report on the epidemiology and burden of sepsis: current evidence, identifying gaps and future directions [Internet]. Geneva: World Health Organization; 2020 [cited 2022 Mar 2]. 55 p. Available from: https://apps.who.int/iris/handle/10665/334216.
Opota O, Jaton K, Greub G. Microbial diagnosis of bloodstream infection: towards molecular diagnosis directly from blood. Clin Microbiol Infect. 2015 Apr 1;21(4):323–31.
Opota O, Croxatto A, Prod’hom G, Greub G. Blood culture-based diagnosis of bacteraemia: state of the art. Clin Microbiol Infect. 2015 Apr 1;21(4):313–22.
Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock*. Crit Care Med. 2006 Jun;34(6):1589–96.
Hindson BJ, Ness KD, Masquelier DA, Belgrader P, Heredia NJ, Makarewicz AJ, et al. High-Throughput Droplet Digital PCR System for Absolute Quantitation of DNA Copy Number. Anal Chem. 2011 Nov 15;83(22):8604–10.
Wouters Y, Dalloyaux D, Christenhusz A, Roelofs HMJ, Wertheim HF, Bleeker-Rovers CP, et al. Droplet digital polymerase chain reaction for rapid broad-spectrum detection of bloodstream infections. Microb Biotechnol. 2020;13(3):657–68.
Zhang T, Niu Z, Wu F, Chen Z, Xu J, Jiang K, et al. Qualitative and quantitative detection of surgical pathogenic microorganisms Escherichia coli and Staphylococcus aureus based on ddPCR system. Sci Rep. 2021 Apr;22(1):8771. 11(.
Badran S, Chen M, Coia JE. Multiplex Droplet Digital Polymerase Chain Reaction Assay for Rapid Molecular Detection of Pathogens in Patients With Sepsis: Protocol for an Assay Development Study. JMIR Res Protoc. 2021 Dec 13;10(12):e33746.
Hu L, Han B, Tong Q, xiao H, Cao D. Detection of Eight Respiratory Bacterial Pathogens Based on Multiplex Real-Time PCR with Fluorescence Melting Curve Analysis. Can J Infect Dis Med Microbiol J Can Mal Infect Microbiol Médicale. 2020 Feb 26;2020:2697230.
Gadsby NJ, McHugh MP, Russell CD, Mark H, Conway Morris A, Laurenson IF, et al. Development of two real-time multiplex PCR assays for the detection and quantification of eight key bacterial pathogens in lower respiratory tract infections. Clin Microbiol Infect. 2015 Aug;21(8):788.e1-788.e13.
Raphael Nyaruaba, Li C, Mwaliko C, Mwau M, Odiwuor N, Muturi E, et al. Developing multiplex ddPCR assays for SARS-CoV-2 detection based on probe mix and amplitude based multiplexing. Expert Rev Mol Diagn. 2021 Jan 2;21(1):119–29.

Table 1: Showing the sequences of the designed primers-probes for each target; including information on target genes, accession numbers, length of amplicons, and ATCC strains utilized in this study.

Target gene	Accession number	Primers-probe sequences (900 nm:250nm)	Amplicon length	ATCC-strains
coa	D00184	5’-GAAGGTCTTGAAGGTAGC-3’ (forward) 5’-GTCTCGGACCATATTGAG-3’ (reverse) 5’6-FAM-AACCACAAGGTACTGAATCAACG-3’IBFQ	148 bp	ATCC® 29213DQ™
uidA	NC_000913	5’-CGGGAAAAGTGTACGTATC-3’ (forward) 5’-GTGGTGTAGAGCATTACG-3’ (reverse) 5’HEX-TGAACAACGAACTGAACTGGCA-3’IBFQ	165 bp	ATCC® 25922DQ™
cpsA	HM448897	5’- CAGAGTATCCAGACTATGC-3’ (forward) 5’-GTGGTGTAGAGCATTACG-3’ (reverse) 5’HEX- CGACACCGAACTAATAGGACCATAG-3’IBFQ	186 bp	ATCC® 700669DQ™
oprL	AF177774	5’- AGCTGAAGAAGTAAGAAGTC-3’ (forward) 5’- CGTCATACACAGGAACTTC-3’ (reverse) 5’6-FAM- AAGCACCTGCGTGTCCTG-3’IBFQ	105 bp	ATCC® 27853DQ™

Table 2: Showing the LoD measured copy number values of a serial dilution of each target compared to the reference copy number values determined by the manufacturer.

Reference strains (ATCC)	Measured copy number values compared to reference
manufacturer-specified minimal amount	≥16,666.7	≥3333.3	≥666.7	≥133.3	≥26.7	≥5.3	≥1.1
S. aureus (coa)	57.320	9.440	1.588	230	48	6.8	1.2
S. pneumonia (cpsA)	36.760	6.400	1.354	202	32	6.8	2.2
P. aeruginosa (oprL)	35.720	7.080	1.460	238	29	7.6	1.2
E. coli (uidA)	51.020	11.140	2.014	378	74	19	1

Development of Duplex Droplet Digital PCR for Rapid Detection of Blood Stream Infections Caused by S. aureus, E. coli, S. pneumoniae and P. aeruginosa

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Design of Primer-Probe Sequences

DNA Extraction and Purification

Droplet Digital PCR Reaction

Thermo-Cycling Conditions

Limit of Detection (LoD)

Results

Thermal Gradient Optimization and Cycling Conditions

ddPCR Duplex Assay

Limit of Detection (LoD)

Discussion

Declarations

References

Tables

Status:

Version 1