sensors Article Electrochemical Genosensing of E. coli Based on Padlock Probes and Rolling Circle Amplification Alejandra Ben Aissa 1 , Narayanan Madaboosi 2,3 , Mats Nilsson 3 and Maria Isabel Pividori 1,4, * 1 Grup de Sensors i Biosensors, Departament de Química, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain; alejandra.benaissa@eurecat.org 2 IIT Madras Bioincubator, Indian Institute of Technology, Chennai 600113, India; narayanan.srinivasan@scilifelab.se or narayanan@bioincubator.iitm.ac.in 3 Science for Life Laboratory, Department of Biochemistry and Biophysics, Stockholm University, 11419 Stockholm, Sweden; mats.nilsson@scilifelab.se 4 Institute of Biotechnology and Biomedicine, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain * Correspondence: Isabel.Pividori@uab.cat; Tel.: +34-93-581-1976 Abstract: Isothermal amplification techniques are emerging nowadays for the rapid and accurate detection of pathogenic bacteria in low resource settings, where many infectious diseases are endemic, and the lack of reliable power supply, trained personnel and specialized facilities pose critical barriers for timely diagnosis. This work addresses the detection of E. coli based on DNA isothermal amplification performed on magnetic particles (MPs) followed by electrochemical genosensing on disposable electrodes by square-wave voltammetry. In this approach, the bacterial DNA is preconcentrated using a target-specific magnetic probe and then amplified on the MPs by rolling circle amplification (RCA). Two different electrochemical readout methods for the RCA amplicons   are tested. The first one relied on the labelling of the magnetic RCA product with a digoxigenin probe followed by the incubation with antiDIG-HRP antibody as electrochemical reporter. In the Citation: Ben Aissa, A.; Madaboosi, second case, the direct detection with an HRP-probe was performed. This latter strategy showed an N.; Nilsson, M.; Pividori, M.I. improved analytical performance, while simultaneously avoiding the use of thermocyclers or bulky Electrochemical Genosensing of E. coli Based on Padlock Probes and bench top equipment. Rolling Circle Amplification. Sensors 2021, 21, 1749. https://doi.org/ Keywords: isothermal amplification; rolling circle amplification; padlock probes; electrochemical 10.3390/s21051749 genosensing; magnetic particles Academic Editor: Giovanni Neri Received: 30 December 2020 1. Introduction Accepted: 25 February 2021 Escherichia coli is the most common enteric Gram-negative bacteria. However, some Published: 3 March 2021 strains, as extraintestinal pathogenic E. coli (ExPEC) [1] can cause illnesses in humans, including neonatal meningitis, sepsis, pneumonia, surgical site infections [2,3], and the Publisher’s Note: MDPI stays neutral urinary tract infections (UTIs) being the most common one [4]. These UTIs cause morbidity with regard to jurisdictional claims in in all age groups and can be categorized according to the clinical syndrome, such as cystitis, published maps and institutional affil- pyelonephritis, catheter-related infection, or asymptomatic bacteriuria [5]. Importantly, iations. UTIs caused by ExPEC represents a large burden in terms of health care costs [6], resulting in an estimated cost of billions of dollars annually [7]. Furthermore, an additional issue is related with the increasing antimicrobial resistance developed in the last years among the E. coli strains [8]. Copyright: © 2021 by the authors. The incidence and the cost related to these infections would be reduced if reliable tests Licensee MDPI, Basel, Switzerland. for the accurate and rapid diagnosis of the causative bacteria become widely available [9,10]. This article is an open access article Conventional microbiological techniques are currently the gold standard for identification distributed under the terms and of pathogenic bacteria, although they are time-consuming [11]. Common approaches for conditions of the Creative Commons the rapid detection of bacteria usually include immunoassays and nucleic-acid amplifi- Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ cation tests (NAATs). The first generation of NAATs based on PCR is not amenable with 4.0/). rapid diagnostic tests (RDTs), since it requires thermocycling platforms, trainee personnel Sensors 2021, 21, 1749. https://doi.org/10.3390/s21051749 https://www.mdpi.com/journal/sensors Sensors 2021, 21, 1749 2 of 13 and infrastructure including reliable power supply, which constitute a critical barrier for PCR in resource-limited settings (RLS) [12]. Recent work on implementing RDTs in RLS has focused on portable platforms operated by batteries [13], with smartphone-based diagnostics widening their scope [14,15]. Many companies have launched portable PCR platforms that can be operated with batteries, including Palm PCR™ (Ahram Biosystems Inc.), Freedom4 (Ubiquitome), miniPCR (Amplyus), among others, that are commercially available. Another approach involves isothermal amplification techniques that do not require temperature cycling [16]. These methods use a variety of reaction principles to specifically amplify nucleic acid targets through isothermal conditions, and the ampli- cons can be detected even without the need for any instrument. Important advances in the integration of Rolling Circle Amplification (RCA) in an ultrasensitive electrochemical genosensor were recently achieved and described by our research group [17] and have also been demonstrated using clinical samples [18]. This technique is based on the endlessness of a circle [19]. The circle DNA complementary to the target is formed using a padlock probe (PLP), by circularization with a highly specific enzymatic ligation, which is followed by an amplification using DNA polymerase. Upon amplification, long DNA repeats that get formed collapse into a condensed blob-like structure, which can then be readily and sensitively detected by hybridization with a tagging probe. Importantly, RCA has no requirement of special instrumentation to cycle the temperature, contrary to PCR-based DNA diagnostics. This work addresses the detection of E. coli based on DNA isothermal amplifica- tion [20] using PLPs and electrochemical genosensing on commercial screen-printed elec- trodes by square-wave voltammetry, with outstanding sensitivity. Moreover, this approach can be integrated with magnetic separation for a further increase of the sensitivity, by using an oligonucleotide coupled to magnetic particles (MPs) [21]. The integration of magnetic particles to the assay allows the use of high concentration of PLPs, which reduces the assay and ligation time from hours to several minutes [22,23]. Once the ligation is finished, the unreacted PLPs are easily removed by magnetic actuation. Progress in this area will, without any doubt, have direct impact in the detection of communicable diseases, since even with the best current technologies, amplification of the genetic material for certain applications cannot be avoided. 2. Materials and Methods 2.1. Instrumentation The electrochemical measurements were performed on carbon screen-printed elec- trodes (ref. DRP-110) using a portable bipotentiostat DRP-STAT200 operated by DropView 2.2 for instrument control and data acquisition (Metrohm DropSens, Oviedo, Spain). Fluores- cently labelled rolling circle products were detected using Aquila 400 (Q-linea AB, Sweden). 2.2. Chemicals and Biochemicals Streptavidin magnetic particles (MPs) (Dynabeads® MyOne Streptavidin T1 Prod. No. 65601) were purchased from Life Technologies. Synthetic DNA, padlock probe and Cy3 labelled detection probes were produced by Integrated DNA technologies (Belgium), and their sequences described in Table 1. Peroxidase-modified probes were provided by Biomers, Germany. All other solutions were prepared with Milli-Q water and other reagents were in analytical reagent grade (supplied from Sigma and Merck), as described in Supplementary Materials. The composition of these solutions was: (i) phosphate buffer for electrochemical measurement (ePBS): 0.1 mol L−1 Na2 HPO4 , 0.1 mol L−1 NaCl, (ii) Hybridization buffer: 6.6 mmol L−1 Tris-acetate pH 7.9, 2 mmol L−1 magnesium acetate, 13 mmol L −1 potassium acetate, 0.02% Tween-20, 0.2 mmol L −1 DTT and (iii) washing buffer: 10 mmol L−1 Tris-HCl pH 7.5, 5 mmol L−1 EDTA, 0.1% Tween-20 and 0.1 mol L−1 NaCl. Sensors 2021, 21, 1749 3 of 13 Table 1. Oligonucleotide sequences used in this study. Description Sequence Label [Phos]GTTACCCGCAGAAGAAGAG Padlock probe TGTACCGACCTCAGTATCTTGC- none GACGTCAGTGGATAGTGTCTTACAC- GATTTATACCTTTGCTCATTGAC TAACGCTTGCACCCTCCGTAT TACCGCGGCTGCTGGCACGGAGT- 16s ribosomal E.coli synthetic target TAGCCGGTGCTTCTTCT- none GCGGGTAACGTCAATGAG- CAAAGGTATTAACTTTACTCCCTTCC CTCTCTCTCTCGTGCCAGCAGCCGCGG 50 Biotin Capture probe TAATACGGAGGGTGCAAGCGTTA (magnetic separation) CTTGCGACGTCAGTGG Readout probe 1 50 HRP (direct electrochemical readout) ATAGTGTCTTACACGATTT CTTGCGACGTCAGTGG Readout probe 2 50 DIG (indirect readout) ATAGTGTCTTACACGATTT CTTGCGACGTCAGTGG Readout probe 3 50 Cy3 (direct fluorescence readout) ATAGTGTCTTACACGATTT Readout probe 4 AGAGTGTACCGACCTC 50 Cy3 (direct fluorescence readout) 2.3. Design of the Padlock, Capture and Readout Probes A PLP is a linear, 80–90 bases long, oligonucleotide that consists of at least three aligned segments: two complementary sequences located at the 50 and 30 ends which hybridize the DNA target, and a central non-complementary region, which hybridizes the readout probes. Specifically, as shown in Figure 1A, the two complementary sequences located at the 50 and 30 ends are indicated as a1 (17 bp) and a4 (18 bp). Further details can be found in Supplementary Materials, including Figure S1 and Table S1, in which each sequence is highlighted. The selected DNA target sequence is specific for Escherichia coli 16S ribosomal gene (Genbank accession Nº CP016182.2). The 16S ribosomal gene from E. coli was selected as this gene is directly involved in various events of the central dogma (including discrimination of mRNA initiation sites, tRNA binding and association of the two ribosomal units) [24]. This gene is comprised of highly conserved regions flanking hypervariable regions [25]. By blasting the selected sequence, high specificity towards E. coli is found in different isolates, including 100% homology from different sources, ranged from environmental samples isolates (GenBank: LR738970.1), clinical isolates (GenBank: CP062970.1), food safety (GenBank: AP023286.1), among many others. Besides, the high copy number repetitions of this sequence reported (GenBank CP058342.1, AP023286.1, CP047710.1) envisages an increased the sensitivity for the detection of the whole bacteria. After hybridization of both ends of the PLP to the DNA target sequence for E. coli (Figure 1B) and further ligation (Figure 1C), the linear probe is then converted to a DNA circle (Figure 1D). The PLP also comprises two sequences located in the center indicated as a2 (14 b) and a3 (35 bp), in Figure 1A, non-complementary to the target but which hybridize with the readout probes, modified with different tags for indirect (digoxigenin) and direct (Cy3 and HRP) labelling and to achieve fluorescence and electrochemical readouts. More details can be also observed in in Supplementary Materials, including Figure S1 and Table S1. The PLP also comprises two sequences located in the center indicated as a2 (14 b) and a3 (35 bp), in Figure 1A, non-complementary to the target but which hybridize with the readout probes, modified with different tags for indirect (digoxigenin) and direct (Cy3 and HRP) labelling and to achieve fluorescence and electrochemical readouts. More de- Sensors Sensors 21,21, 2021, 2021, x 1749 tails can be also observed in in Supplementary Materials, including Figure S1 and FOR PEER REVIEW 4 of Table 13 4 of 13 S1. Table 1. Oligonucleotide sequences used in this study. Description Sequence Label [Phos]GTTACCCGCAGAAGAA- GAGTGTACCGACCTCAG- Padlock probe TATCTTGCGACGTCAGTGGA- none TAGTGTCTTACACGATTTATAC- CTTTGCTCATTGAC TAACGCTTGCACCCTCCG- TATTACCGCGGCTGCTGGCAC- GGAG- 16s ribosomal E.coli synthetic target TTAGCCGGTGCTTCTTCTGCGGG none TAACGTCAATGAG- Figure 1. (A) Design of the specific padlock probe for the detection of E. coli comprising two Figure 1. (A) Design of the specific CAAAGGTATTAACTTTACTCCCT padlock complementary sequences to the target (a1 probe and a4) for thetwo and detection of E. sequences (a2coli andcomprising a3) to achievetwothe com- plementary sequences to the target (a1 and a4) and TCC two sequences (a2 and a3) to achieve the readout. (B) Hybridization of the padlock probe to the 16S ribosomal E. coli gene. (C) Ligation by readout. CTCTCTCTCTCGTGCCAG- the T4(B) DNAHybridization ligase enzyme. Capture of the padlock probe (D) Circularized probe DNA to templates the 16S ribosomal CAGCCGCGGTAATAC- E. coli obtained after gene. of ligation (C) Ligation 5′the padlockby Biotin the probes T4 DNA ligase enzyme. (D) Circularized DNA and used for the rolling circle amplification. templates obtained after ligation of the padlock (magnetic separation) GGAGGGTGCAAGCGTTA probes and used for the rolling circle amplification. CTTGCGACGTCAGTGGA- 5′ HRP (direct electrochemical Readout In order probe 1 the target isolation and preconcentration based on magnetic actu- to achieve TAGTGTCTTACACGATTT readout) In order ation, to achieve a biotinylated the target capture probe,isolation was and preconcentration designed CTTGCGACGTCAGTGGA- to hybridize a specific based on magnetic sequence located ac- upstream Readout of the probe E. coli2recognition tuation, a biotinylated capture probe, sequence (detailed in Figure S1 and 5′ was designed to hybridize a specific sequence DIG Table(indirect S1). readout) Thus, lo- TAGTGTCTTACACGATTT streptavidin-magnetic particles cated upstream of the E. coli recognition can bind to the biotinylated probe and preconcentrate sequence (detailed in Figure S1 and Table S1). CTTGCGACGTCAGTGGA- Readout probe 3 16s rRNA5′was Cy3designed. (direct fluorescence Table 1 readout) the streptavidin-magnetic Thus, sample. Finally, synthetic canEscherichia targetTAGTGTCTTACACGATTT particles for bind to the colibiotinylated probe and preconcen- summarizes Readoutthe probe oligonucleotide sequences 4 synthetic target used in this study. 5′ AGAGTGTACCGACCTC Further Cy3was details (direct can be readout) fluorescence trate the sample. Finally, for Escherichia coli 16s rRNA designed. Table found in Table S1, in which the different sequences involved in hybridization and readout 1 summarizes the oligonucleotide sequences used in this study. Further details can be are highlighted. 2.4. found Selection in of the Table S1, Readout in which theProbe Sequence different and Optimization sequences involved inofhybridization the Rolling Circle and readout areAmplification highlighted. 2.4. Selection of the Readout Probe Sequence and Optimization of the Rolling Circle Amplification Differentparameters Different parameters were were firstly firstly evaluated evaluated forRCA for the the based RCA based on the on the fluorescent fluorescent readout (schematically shown in Figure 2). readout (schematically shown in Figure 2). Figure Figure 2.2.Schematic Schematicrepresentation representation of the of the rolling rolling circlecircle amplification amplification readoutreadout based onbased on fluores- fluorescence. cence. (A) Padlock probe is hybridized on the bacterial DNA target, followed by ligation; (B) Rolling(B) Roll- (A) Padlock probe is hybridized on the bacterial DNA target, followed by ligation; ing circle circle amplification amplification (C) Hybridization (C) Hybridization with thewith Cy3the Cy3 readout readout probeasacting probe acting as fluorescent fluorescent reporter; re- porter; (D) Detection (D) Detection with with Aquila 400Aquila 400equipment. detection detection equipment. To achieve that, a dilution series of 16S ribosomal E. coli synthetic target was per- formed from 1 fM to 1 pM. PLP ligation mix was added to each sample, comprising 10 nmol·L−1 padlock probe, 0.2 mg mL−1 BSA, 0.68 mmol L−1 ATP, 1x T4 ligase reaction buffer Sensors 2021, 21, 1749 5 of 13 To achieve that, a dilution series of 16S ribosomal E. coli synthetic target was performed from 1 fM to 1 pM. PLP ligation mix was added to each sample, comprising 10 nmol·L−1 padlock probe, 0.2 mg mL−1 BSA, 0.68 mmol L−1 ATP, 1x T4 ligase reaction buffer and 5 U T4 ligase. The reaction was incubated at 37 ◦ C for 15 min. After that, RCA mix containing 1x phi29 DNA polymerase buffer, 125 µM dNTPs, 0.2 mg ml −1 BSA and 6 U phi29 DNA polymerase was added to the circularized DNA and RCA was performed at 37 ◦ C for 60 min. RCPs cannot be resolved by gel electrophoresis due to the broad smear of the high molecular weight DNAs. In all instances, in order to achieve the fluorescence readout, RCPs were fluorescently labelled by using readout probe 3 or 4 (Table 1), labelled with the fluorescence tag Cy5. Aquila 400 amplified single molecule counter (Q-linea, Uppsala) was used to record the fluorescence readout and to count the RCA products. Finally, 20 µL of the total sample was analyzed using the Aquila 400 amplified single molecule counter. This method was used to achieve the optimal conditions, including the composition of hybridization buffer, labelling temperature, readout probe sequence and concentration, as described below: Hybridization buffer. Four hybridization buffers were studied for the labelling process, being their compositions. hybridization buffer A. 6.6 mmol L−1 Tris-acetate pH 7.9, 2 mmol L−1 magnesium acetate, 13 mmol L −1 potassium acetate, 0.02% Tween-20, 0.2 mmol L −1 DTT. hybridization buffer B. 20 mmol L−1 EDTA, 40 mmol L−1 Tris-HCl, 2.8 mol L−1 NaCl, and 0.2% Tween-20. hybridization buffer C. 20 mmol L−1 EDTA, 40 mmol L−1 Tris-HCl, 2.8 mol L−1 NaCl, 0.2% Tween-20, 30 mM trisodium citrate and 40% formamide. hybridization buffer D. 40 mM Tris-HCl pH 8.3, 50 mM KCl, 20 mM MgCl2 , 1 mM NAD, 0.02% Triton® X-100. Labelling temperature optimization. A binding step at high temperatures is usually included during the hybridization process due to its demonstrated improvement in terms of time. In order to evaluate the effect on the yield of the labelling, 20 µL of RCP were incubated in hybridization buffer with 25 nmol L−1 of readout probe 4. In procedure 1, the mix was incubated at 75 ◦ C for 2 min, after that another incubation at 37 ◦ C for 45 min was performed. The procedure 2 consisted in one only incubation at 37 ◦ C for 45 min. Readout probe sequence. A comparison of two different readout probes (as shown in Table 1) labelled in 50 with Cy3 was performed in order to achieve the direct fluores- cence readout. Readout probe 3 CTTGCGACGTCAGTGGATAGTGTCTTACACGATTT Readout probe 4 AGAGTGTACCGACCTC Readout probe concentration. Concentrations ranging from 10 to 25 nmol L−1 of the readout probe were studied in order to evaluate its influence during the labelling process. 2.5. Rolling Circle Amplification on Streptavidin Magnetic Particles and Electrochemical Genosensing The procedure is schematically described in Figure 3 and detailed described in Supple- mentary Materials, involving the following steps, as previously described by our research group [17]: (A) Hybridization and Ligation. Each sample of DNA target was incubated with the biotinylated capture probe, in order to achieve the preconcentration of the RCPs on streptavidin magnetic particles in (B) Coupling on streptavidin magnetic particles, followed by (C) Rolling Circle amplification and (D) Labelling with the readout probe labelled with HRP. After each incubation or washing step, a magnetic separator was posi- tioned under the tubes until pellet formation occurred on the tube’s side wall, followed by supernatant separation. Finally, (E) electrochemical readout of RCPs attached on the streptavidin-MPs was performed on carbon screen-printed electrodes with a portable bipotentiostat connected by a universal USB port to a laptop computer operated by batteries (as detailed described in Supplementary Materials and Figures S3 Figure S4 therein), by square wave voltammetry (SWV) measurements. The maximal signal obtained in the cathodic peaks was used for the electrochemical signal plotted in the results. ensors 2021, 21, Sensors x FOR 2021, PEER REVIEW 21, 1749 6 of 13 Figure 3.3.Schematic Figure representation Schematic of the electrochemical representation genosensing forgenosensing of the electrochemical the rolling circlefor ampli- the rolling fication products detection on streptavidin-MP. (A) The biotinylated capture probe as well as the fication products detection on streptavidin-MP. (A) The biotinylated capture probe padlock probe are both hybridized to the DNA target (B) The DNA target is then coupled on strep- padlock probeparticles; tavidin magnetic are both (C)hybridized to the DNA is RCA on the streptavidin-MP target (B) The DNA then performed; target is then co (D) Hybridization streptavidin with the readoutmagnetic probe (HRP particles; labelled) to(C) RCAtheon achieve the streptavidin-MP electrochemical is then reporter is then performed; performed; zation with the readout (E) Electrochemical readout by probe (HRP square wave labelled)(SWV) voltammetry to achieve the electrochemical upon addition of the mediator HQreporte formed; (E) Electrochemical and the substrate H2 O2 for the HRP readout by square on screen-printed wave voltammetry (SWV) upon add electrodes. mediator HQIndirect 2.6. Direct and and the substrate Readout ApproachH2O2 for the HRP on screen-printed electrodes. In order to optimize the electrochemical detection of RCPs, two different electrochemi- 2.6. Directmethods cal readout for the RCA and Indirect amplicons Readout on MPs were tested. The first one relied on the Approach indirect labelling of the magnetic RCA product with a digoxigenin probe (readout probe 2, Table In order by 1) followed tothe optimize incubationthe with electrochemical antiDIG-HRP antibody detection of RCPs, as electrochemical two diff reporter. chemical In the second readout case, themethods for the direct detection withRCA amplicons an HRP-probe on probe (readout MPs 1,were Tabletested. 1) was The f performed. In both instances, 16S ribosomal E. coli synthetic target (1 nM) was incubated on the indirect labelling of the magnetic RCA product with a digoxigenin p for 20 min at 55 ◦ C with 50 nmol L−1 of biotinylated capture oligo in PLP ligation mix. probe 2, Table In the direct 1) followed labelling, by the this procedure incubation was performed with by the antiDIG-HRP incubation antibody as with the readout − 1 ◦ ical probereporter. 1 labelled In withtheHRPsecond case, (20 nmol L ) the direct detection in hybridization buffer A with at 37 an HRP-probe C for 45 min. (r 1, Table 1) was performed. In both instances, 16S ribosomal E. coli synthetic was incubated for 20 min at 55 °C with 50 nmol L−1 of biotinylated capture ligation mix. In the direct labelling, this procedure was performed by the inc Sensors 2021, 21, 1749 7 of 13 The indirect labelling was performed by the incubation with the readout probe 2 labelled with digoxigenin (20 nmol L−1 ) in the same hybridization buffer A at 37 ◦ C for 45 min. After that, to come up with the indirect label, a further incubation step was performed with antiDIG-HRP antibody in PBS buffer with 1% BSA at 37 ◦ C for 30 min. In both instances, labelled RCA products were washed (x3) to remove the unbound reagents. Negative controls, performed with all the components with the exception of the template bacterial DNA, were in all instances prepared. After each incubation or washing step, a magnetic separator was positioned under the tubes until pellet formation on the tube’s side wall, followed by supernatant separation. 2.7. Electrochemical Genosensing of E. coli. The electrochemical genosensor was evaluated for the detection of Escherichia coli. For that, rolling circle amplified DNA coming from an overnight culture of E. coli DH5α were evaluated. The bacteria E. coli was routinely grown in sterile liquid Luria Bertani (LB) broth for 18 h at 37 ◦ C under aerobic conditions. After that, serial dilutions from the culture were performed and 100 µL of each dilution was spiked on LB agar plates. After incubating the plate at 37 ◦ C for 24 h, the culture colonies on the plates were counted to estimate the number of viable bacteria in CFU mL−1 . On the other hand, cells from 1 mL of each sample were lysed and the DNA was extracted. For the DNA extraction, 1 mL of each sample were centrifuged, and the cell pellets were resuspended in 1 mL of sterile water. The resuspended cells were re-centrifuged at 12,500 g for 15 min, after that, the cells were resuspended in 50 µL of Tris–EDTA buffer and then kept in a boiling water bath for 10 min. Immediately after heat denaturation, the samples are cooled on ice for 5 min. The hybridization of the capture probe and the ligation was immediately performed by incubating the E coli template (10 µL of the extracted E coli DNA) at the same time with the padlock and capture probe, which was after coupled to 4 µL of streptavidin-magnetic particles for 5 min at RT. Then, the sample- bead complex was washed in order to eliminate the padlock excess and to remove the bacterial matrix and resuspended in RCA mix for the isothermal amplification. RCA was performed for 60 min at 37 ◦ C. A negative control without E coli DNA, which had the same treatment as the positive sample, was processed. The electrochemical genosensing of the RCPs was performed by the hybridization with the readout probe 1 (Table 1) labelled with HRP (20 nmol L−1 ) (as schematized in Figure 3) in hybridization buffer A for 45 min at 37 ◦ C, followed by washing steps under magnetic actuation, to eliminate the excess of readout probe. 3. Results and Discussion 3.1. Selection of the Readout Probe Sequence and Optimization of the Rolling Circle Amplification Different parameters were firstly evaluated for the RCA based on fluorescence readout and using Aquila 400 (Q-linea AB, Sweden). The results, shown in Figure 4, include the optimized conditions, such as the composition of hybridization buffer (Figure 4A), labelling temperature (Figure 4B), readout probe sequence (Figure 4C) and concentration (Figure 4D). To summarize, Buffer A exhibits improved results for labelling at 37 ◦ C for 45 min with 20 nmol L−1 of readout probe with sequence as 3 (the same as sequence 1 and 2, with different tags) (Table 1). Based on the obtained results, all further studies for the RCA on streptavidin magnetic particles and electrochemical genosensing were performed by using these optimized conditions. 3.2. Rolling Circle Amplification on Streptavidin Magnetic Particles and Electrochemical Genosensing. Direct and Indirect Readout Approach Figure 5 shows schematically the different approaches (direct and indirect) for the labelling of the rolling circle amplification products (RCPs) for the subsequent electrochem- ical readout. Sensors 2021, 21, x FOR PEER REVIEW Sensors 2021, 21, 1749 8 of 13 Figure 4. Results for the optimization of the rolling circle amplification. The results are shown bars including the Unit counter signal. (A). Hybridization buffer optimization. (B). Labelling perature optimization (Procedure 1: incubation at 75 °C for 2 min + 37 °C for 45 min; Procedu incubation at 37 °C for 45 min); (C). Readout probe sequence. (D): Readout probe optimizatio Further experimental details can be found in the main text. In all instances, the negative contr without the DNA template, were performed (blue bar). Figure 4. Results for the optimization of the rolling circle amplification. The results are shown Figure 4. Results for the optimization of the rollingMagnetic circle amplification. The results are show as3.2. barsRolling includingCircle Amplification the Unit on (A). counter signal. Streptavidin Particles and Hybridization buffer optimization. Electrochemical (B). Labelling bars including Genosensing. the Unit Direct and counter signal. Indirect (A). Hybridization Readout Approach buffer optimization. (B). Labelling temperature optimization (Procedure 1: incubation at 75 ◦ C for 2 min + 37 ◦ C for 45 min; Procedure perature optimization 2: incubation at 37 ◦ C for(Procedure 1: incubation 45schematically min); (C). Readout probe at 75 °C (D): for 2Readout sequence.approaches min + probe 37 °Coptimization. for 45 min; Proced incubation Figure at 375°Cshows for 45 min); (C). the Readout different probe sequence. (D): (direct and Readout probeindirect) optimizatfo Further experimental details can be found in the main text. In all instances, the negative control labelling of the rolling Further circlebeamplification products (RCPs) for thethe subsequent ele without experimental the DNA template, details were can performedfound (bluein the main bar). text. In all instances, negative con chemical readout. without the DNA template, were performed (blue bar). 3.2. Rolling Circle Amplification on Streptavidin Magnetic Particles and Electrochemical Genosensing. Direct and Indirect Readout Approach Figure 5 shows schematically the different approaches (direct and indirect) labelling of the rolling circle amplification products (RCPs) for the subsequent e chemical readout. Figure 5. Schematic representation of the electrochemical genosensing of E. coli by rolling circle Figure 5. Schematic amplification. (A) RCA onrepresentation of the(B1) the streptavidin-MPs; electrochemical Direct labelling.genosensing Hybridization of E.the with coli by rolling cir readout amplification. probe (A) RCA 1 (HRP labelled) actingon the streptavidin-MPs; as electrochemical (B1)Indirect reporter. (B2) Directlabelling. labelling. Hybridization Hybridization with with the readout probe 1 (HRP labelled) acting as electrochemical reporter. (B2) Indirect the readout probe 2 (digoxigenin modified) and incubation with the electrochemical reporter antiDIG-labelling. Hy zation with the readout probe 2 (digoxigenin modified) and incubation with HRP. In both cases, the electrochemical determination is performed by square wave voltammetry the electrochem reporter upon antiDIG-HRP. addition In HQ of the mediator bothandcases, the electrochemical the substrate determination is performed by squa H2 O2 for the HRP. wave voltammetry upon addition of the mediator HQ and the substrate H2O2 for the HRP. Figure 5. Schematic representation of the electrochemical genosensing of E. coli by rolling ci amplification. (A) RCA on the streptavidin-MPs; (B1) Direct labelling. Hybridization with th Sensors 2021, 21, x FOR PEER REVIEW 9 of 13 Sensors 2021, 21, 1749 9 of 13 The direct labelling consists of the hybridization of the magnetic RCP (obtained by The direct synthesizing labelling the RCP on consists a MP)ofwith the hybridization of thedirectly a readout probe magneticmodified RCP (obtained by bulky with the synthesizing the RCP on a MP) with a readout probe directly modified with the HRP (readout probe 1). The second approach relied on the indirect labelling with a di- bulky HRP (readout probe 1). The second approach relied on the indirect labelling with a digox- goxigenin probe (readout probe 2, Table 1) followed by the incubation with antiDIG-HRP igenin probe (readout probe 2, Table 1) followed by the incubation with antiDIG-HRP antibody antibody as electrochemicalreporter. as electrochemical reporter. In both In both instances, instances, the RCPs the RCPs on theon thewere MPs MPsthen were then electrochemically detected.The electrochemically detected. The results results areare shown shown in Figure in Figure 6. 6. Figure 6. Comparative study on direct and indirect readout systems for the rolling circle amplification Figure 6. Comparative study on direct and indirect readout systems for the rolling circle amplifi- and electrochemical genosensing. For both systems, 1 nM of circularized bacterial DNA were used, cation and electrochemical genosensing. For both systems, 1 nM of circularized bacterial DNA except for the negative controls. (A) Raw square wave voltammetry data. (B) Graph bars of the were used, except for the negative controls. (A) Raw square wave voltammetry data. (B) Graph SWV maximal signal. The bars represent the mean value of the maximal signal and the error bars, bars of the SWV maximal signal. The bars represent the mean value of the maximal signal and the the standard deviation for n = 3. Medium: phosphate buffer. Mediator: hydroquinone 1 mmol L−1 . error bars, the standard deviation for n = 3. Medium: phosphate buffer. Mediator: hydroquinone 1 Substrate: H2 O2 0.25 mmol L−1 . The potential range was 0.1 to −0.7 V with potential step and mmol L−1. Substrate: H2O2 0.25 mmol L−1. The potential range was 0.1 to −0.7 V with potential step amplitude of 10 mV and frequency of 1 Hz. For both systems, the negative control without the DNA and amplitude of 10 mV and frequency of 1 Hz. For both systems, the negative control without the template, were performed (n = 3). DNA template, were performed (n = 3). Although a higher hybridization rate was expected for the readout probe 2 based on Although digoxigenin tagaandhigher hybridization the indirect approachrate duewas expected to the for the smaller size, readout it seems to beprobe 2 based on hindered digoxigenin tag and the indirect approach due to the smaller size, it by the subsequent recognition of the antibody, thereby providing poorer performance than seems to be hindered thethe by direct labelling with subsequent the readoutofprobe recognition 1. For the concentration the antibody, of 1 nM (10 thereby providing fmol) of poorer 16S performance ribosomal than E. colilabelling the direct synthetic with target,the a mean value readout of 9.52 probe For the0.4 1.µA/SD µA (direct labelling) concentration of 1 nM and(10 fmol) 1.76 µA/SD 0.5 µA (indirect labelling) were obtained for the two methods. The of 16S ribosomal E. coli synthetic target, a mean value of 9.52 μA/SD 0.4 μA (direct label- mean values for the background signals were 0.32 µA/SD 0.02 µA (for direct labelling), 0.42 µA/SD ling) and 1.76 μA/SD 0.5 μA (indirect labelling) were obtained for the two methods. The 0.09 µA (for indirect labelling). The signal-to-background ratios were 29.75 and 4.19 for mean values for the background signals were 0.32 μA/SD 0.02 μA (for direct labelling), direct and indirect labelling, respectively, confirming an increased performance of the direct 0.42 μA/SD labelling for 0.09 μA (for of the detection indirect the RCPs.labelling). The signal-to-background It is important ratioswere to highlight that the RCPs were 29.75 and previously described as a blob of DNA [26] in which some of the repeated hybridization perfor- 4.19 for direct and indirect labelling, respectively, confirming an increased mance of the sequences direct might labelling be in for and the interior, the detection thus being oflessthe RCPs. Itfor accessible isfurther important to highlight reactions. This that the RCPs direct weresystem readout previously was useddescribed as experiments. in further a blob of DNA [26] in which some of the repeated hybridization sequences might be in the interior, and thus being less accessible for further 3.3. Rolling Circle Amplification on Streptavidin Magnetic Particles and reactions. ThisGenosensing Electrochemical direct readout system was used in further experiments. To determine the performance and limit of detection (LOD) of the RCA on streptavidin- 3.3. MPs,Rolling Circle Amplification the electrochemical on Streptavidin genosensing Magnetic was performed Particles on a serial and of dilution Electrochemical 16s ribosomal Genosensing E. coli synthetic target, ranging from 1 pM to 10 nM (10 amol to 0.1 pmol), and detected by square Towave voltammetry. determine The results are the performance andshown in Figure limit of 7. (LOD) of the RCA on streptav- detection The maximal idin-MPs, signals obtained the electrochemical in the voltagramms genosensing are presented was performed in Figure on a serial dilution7A. of The 16s ribo- LOD was estimated somal E. coli syntheticby fitting the raw data using a nonlinear regression (Two site target, ranging from 1 pM to 10 nM (10 amol to 0.1 pmol), and binding –hyperbola) (R2 = 0.9965), as shown in Figure 7C, by processing the negative control samples detected by square wave voltammetry. The results are shown in Figure 7. (n = 6) obtaining a mean value of 0.32 µA with a standard deviation (SD) of 0.02. The cut-off valueThe wasmaximal signals obtained then determined in the voltagramms with a one-tailed t test at a 95% are presented confidence levelin(t Figure = 2.015),7A. The LOD was estimated by fitting the raw data using a nonlinear regression (Two site binding –hyperbola) (R2 = 0.9965), as shown in Figure 7C, by processing the negative control sam- ples (n = 6) obtaining a mean value of 0.32 μA with a standard deviation (SD) of 0.02. The cut-off value was then determined with a one-tailed t test at a 95% confidence level (t = 2.015), giving a value of 0.36 μA. The LOD was found to be 6.7 amol (4,03 × 106 DNA 11 −1 Sensors 2021, 21, x FOR PEER REVIEW 10 of 13 It is important to highlight that the limit of detection reached in this work with the Sensors 2021, 21, 1749 electrochemical detection is similar or even improved than other reported bacterial 10 ofdetec- 13 tion methods where RCA had been used for the signal amplification [27] (Table S2, Sup- plementary Materials, and references therein). Further improvement of the LOD of E. coli atgiving clinically relevant a value of 0.36levels, µA. Thecan be was LOD achieved foundby to re-amplification of 10 be 6.7 amol (4,03 × the RCA copies) 6 DNA products in by C2CA 10 µL of [17]. 11 sample (0.67 pM, corresponding to 4.03 × 10 DNA copies L ). − 1 Figure 7. (A) Voltagramms for the RCA on MPs and electrochemical genosensing based on direct Figure 7. (A) Voltagramms for the RCA on MPs and electrochemical genosensing based on direct labelling, performed labelling, performed bybyisothermal isothermal amplification amplificationof 16S ribosomal of 16S E. coliE.synthetic ribosomal target target coli synthetic ranging from ranging 1pM to 10 nM, and including the negative controls. All other conditions as Figure 6. (B) from 1pM to 10 nM, and including the negative controls. All other conditions as Figure 6. (B) Cali- Calibration plot of plot bration the SWV of themaximal signal. signal. SWV maximal The barsTherepresent the mean bars represent thevalue, mean and theand value, errorthebars, errorthe SD the bars, for n = 3. (C) Fitted SWV signals by a nonlinear regression, nonlinear regression SD for n = 3. (C) Fitted SWV signals by a nonlinear regression, nonlinear regression (Two site(Two site binding –hyperbola). binding The negative –hyperbola). controls are The negative also shown controls (n =shown are also 6). (n = 6). It is importantGenosensing 3.4. Electrochemical to highlight of that the limit of detection reached in this work with E. coli the electrochemical detection is similar or even improved than other reported bacterial Finally, detection the electrochemical methods where RCA had genosensor been usedwas for evaluated for the detection the signal amplification [27] of E. coli. (Table S2,The samples were treated according to the experimental section. Figure 8 shows Supplementary Materials, and references therein). Further improvement of the LOD of the maximum signal E. coli obtained inrelevant at clinically the voltagramms forachieved levels, can be the E coli. byThe LOD was found re-amplification of the to be products RCA 3 × 103 CFU inby10C2CA μL of[17]. sample, corresponding to 10 CFU mL . 4.3 −1 It was estimated as above, by fitting the raw data using a nonlinear regression (Two 3.4.binding site Electrochemical Genosensing –hyperbola) of E. coli Regarding the stability, it is determined by the sta- (R2 = 0.9513). Finally, the electrochemical genosensor bility of the reagents (oligonucleotides, was evaluated strepAv-MP, RCAformix the detection of E. coli. and enzymes), thatThe were samples kept at 4 were °C astreated accordingby recommended to the the experimental manufacturers. section. As theFigure 8 shows theelectrodes screen-printed maximum are signal not obtained modified, biologically in the voltagramms they werefor the E coli. storage at RT,The asLOD was found by recommended to be × 103 CFU the3 manufacturer, in 10use of sample, µLthem corresponding todate. 10 CFU mL . 4.3 −1 and within the expiration According to these results, it was demonstrated that the system presented in this work, combining the electrochemical genosensing strategy with RCA, offers a sensitive method for the bacterial detection and promising features to be used in Low resource set- tings, when compared with other methods (Table S2 and references therein, Supplemen- tary Materials). Figure 8. Rolling circle amplification and electrochemical genosensing of E. coli samples. The plots show the SWV maximal signal. The bars represent the mean value of the maximal signal and the error bars, the standard deviation for n = 3. All other conditions as Figure 6, including the negative controls are also shown (n = 6). Sensors 2021, 21, 1749 11 of 13 It was estimated as above, by fitting the raw data using a nonlinear regression (Two site binding –hyperbola) (R2 = 0.9513). Regarding the stability, it is determined by the stability of the reagents (oligonucleotides, strepAv-MP, RCA mix and enzymes), that were kept at 4 ◦ C as recommended by the manufacturers. As the screen-printed electrodes are not biologically modified, they were storage at RT, as recommended by the manufacturer, and use them within the expiration date. According to these results, it was demonstrated that the system presented in this work, combining the electrochemical genosensing strategy with RCA, offers a sensitive method for the bacterial detection and promising features to be used in Low resource settings, when compared with other methods (Table S2 and references therein, Supplementary Materials). 4. Conclusions The design of low cost, user friendly and point of care devices able to provide rapid results with high sensitivity and specificity is nowadays a major challenge. With recent advances in molecular methods, the combination of NAAT using isothermal assays with electrochemical sensor is a highly suitable approach to reach this objective. This work reports an electrochemical genosensor for the detection of bacterial DNA by using PLPs and the subsequent amplification with RCA. This strategy is a powerful combination for highly specific and sensitive E. coli DNA detection that can be applied in RLS. The incorporation of MPs in the assay brings several advantages besides the preconcentration under magnetic actuation, including their large surface to volume ratio, facilitating high capture efficiencies and reaction kinetics. Other important advantages include the ease of washing procedure, simplified assay steps and reduced assay time. The specificity provided by the ligation of PLPs and the isothermal nature of the RCA make it an attractive method for the application at the practitioner site. Moreover, the use of electrochemical readout offers a robust test that can be performed with hand-held equipment operated with batteries. This approach requires minimal training for final user and provide rapid results for taking action immediately, for example prescribing without any dilation the treatment with antibiotics at first visit. Supplementary Materials: The following are available online at https://www.mdpi.com/1424-822 0/21/5/1749/s1. Table S1. Oligonucleotide sequences used in this study. In the padlock probe, the sequences are tagged in blue (dark a1 and light a4), corresponding to the regions complementary to the DNA target, while orange (a2) and yellow (a3) regions correspond to the readout probes. Pink shows the capture oligonucleotide region complementary to the target. The different complemen- tary regions have been underlined in the same way. Table S2. Analytical performance and main characteristics of other related biosensors or platforms based on isothermal amplification for the detection of E. coli or other bacteria. Figure S1. Design of the specific padlock probe for the detection of E. coli comprising two complementary sequences to the target (a1 and a4) and two sequences (a2 and a3) to achieve the readout. The schematic location of the capture probe is also shown. Figure S2. Schematic representation of the electrochemical genosensing for the rolling circle amplification products detection on streptavidin-MP. (A) The biotinylated capture probe as well as the padlock probe are both hybridized to the DNA target (B) The DNA target is then coupled on streptavidin magnetic particles; (C) RCA on the streptavidin-MP is then performed; (D) Hybridization with the readout probe (HRP labelled) to achieve the electrochemical reporter is then performed; (E) Electrochemical readout by square wave voltammetry upon addition of the mediator HQ and the substrate H2O2 for the HRP on screen-printed electrodes. Figure S3. Configuration of the commercial screen-printed electrodes used in this work and image of the Boxed Connector for Screen-Printed Electrodes. It can be appreciated the drop on the Surface of the working carbon electrode (images obtained from Dropsens). Figure S4. Enzymatic mechanism of the HRP enzyme conjugated to the readout probes on the surface of the carbon screen-printed electrode, upon addition of H2O2 as a substrate and hydroquinone (HQ) as a mediator. Sensors 2021, 21, 1749 12 of 13 Author Contributions: Conceptualization, N.M., M.N. and M.I.P.; Data curation, A.B.A.; Formal analysis, A.B.A.; Funding acquisition, M.N. and M.I.P.; Investigation, A.B.A. and N.M.; Methodology, A.B.A.; Project administration, M.N. and M.I.P.; Resources, M.N.; Supervision, N.M. and M.I.P.; Writing—original draft, M.I.P.; Writing—review & editing, N.M. and M.I.P. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded the Ministry of Economy and Competitiveness (MINECO), Madrid (Project BIO2016-75751-R) and Ministry of Science and Innovation (Project PID2019-106625RB-I00). The coauthors from Sweden thank the Swedish Research Council, VR and the SSF project, FLU-ID. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conflicts of Interest: The authors declare no conflict of interest. References 1. Russo, T.A.; Johnson, J.R. A proposal for an inclusive designation for extraintestinal pathogenic Escherichia coli: ExPEC. J. Infect. Dis. 2000, 181, 1753–1754. [CrossRef] 2. Bonacorsi, S.; Bingen, E. Molecular epidemiology of Escherichia coli causing neonatal meningitis. Int. J. Med. Microbiol. 2005, 295, 373–381. [CrossRef] 3. Breland, E.J.; Eberly, A.R.; Hadjifrangiskou, M. An Overview of Two-Component Signal Transduction Systems Implicated in Extra-Intestinal Pathogenic E. coli Infections. Front. Cell. Infect. Microbiol. 2017, 7, 162. [CrossRef] 4. Russo, T.A.; Johnson, J.R. Medical and economic impact of extraintestinal infections due to Escherichia coli: Focus on an increasingly important endemic problem. Microbes Infect. 2003, 5, 449–456. [CrossRef] 5. Bergsten, G.; Wullt, B.; Svanborg, C. Escherichia coli, fimbriae, bacterial persistence and host response induction in the human urinary tract. Int. J. Med. Microbiol. 2005, 295, 487–502. [CrossRef] 6. Biran, D.; Ron, E.Z. Extraintestinal pathogenic Escherichia coli. Curr. Top. Microbiol. Immunol. 2018, 416, 149–161. [PubMed] 7. Simmering, J.E.; Tang, F.; Cavanaugh, J.E.; Polgreen, L.A.; Polgreen, P.M. The Increase in Hospitalizations for Urinary Tract Infections and the Associated Costs in the United States, 1998–2011. Open Forum Infect. Dis. 2017, 4, ofw281. [CrossRef] [PubMed] 8. von Baum, H.; Marre, R. Antimicrobial resistance of Escherichia coli and therapeutic implications. Int. J. Med. Microbiol. 2005, 295, 503–511. [CrossRef] 9. Land, K.J.; Boeras, D.I.; Chen, X.S.; Ramsay, A.R.; Peeling, R.W. REASSURED diagnostics to inform disease control strategies, strengthen health systems and improve patient outcomes. Nat Microbiol. 2019, 4, 46–54. [CrossRef] [PubMed] 10. Smith, S.; Korvink, J.G.; Mager, D.; Land, K. The potential of paper-based diagnostics to meet the ASSURED criteria. RSC Adv. 2018, 8, 34012–34034. [CrossRef] 11. Cho, I.-H.; Ku, S. Current Technical Approaches for the Early Detection of Foodborne Pathogens: Challenges and Opportunities. Int. J. Mol. Sci. 2017, 18, 2078. [CrossRef] 12. Urdea, M.; Penny, L.A.; Olmsted, S.S.; Giovanni, M.Y.; Kaspar, P.; Shepherd, A.; Wilson, P.; Dahl, C.A.; Buchsbaum, S.; Moeller, G.; et al. Requirements for high impact diagnostics in the developing world. Nature 2006, 444 (Suppl. 1), 73–79. [CrossRef] [PubMed] 13. Marx, V. PCR heads into the field. Nat. Methods 2015, 12, 393–397. [CrossRef] 14. Kühnemund, M.; Wei, Q.; Darai, E.; Wang, Y.; Hernández-Neuta, I.; Yang, Z.; Tseng, D.; Ahlford, A.; Mathot, L.; Sjöblom, T.; et al. Targeted DNA sequencing and in situ mutation analysis using mobile phone microscopy. Nat. Commun. 2017, 8, 13913. [CrossRef] [PubMed] 15. Hernández-Neuta, I.; Neumann, F.; Brightmeyer, J. Smartphone-based clinical diagnostics: Towards democratization of evidence- based health care. J. Int. Med. 2019, 285, 19–39. [CrossRef] [PubMed] 16. Niemz, A.; Ferguson, T.M.; Boyle, D.S. Point-of-care nucleic acid testing for infectious diseases. Trends Biotechnol. 2011, 5, 240–250. [CrossRef] [PubMed] 17. Carinelli, S.; Kühnemund, M.; Nilsson, M.; Pividori, M.I. Yoctomole electrochemical genosensing of Ebola virus cDNA by rolling circle and circle to circle amplification. Biosens Bioelectron. 2017, 93, 65–71. [CrossRef] [PubMed] 18. Ciftci, S.; Cánovas, R.; Neumann, F.; Paulraj, T.; Nilsson, M.; Crespo, G.A.; Madaboosi, N. The sweet detection of rolling circle amplification: Glucose-based electrochemical genosensor for the detection of viral nucleic acid. Biosens Bioelectron. 2020, 151, 112002. [CrossRef] 19. Nilsson, M.; Malmgren, H.; Samiotaki, M.; Kwiatkowski, M.; Chowdhary, B.P.; Landegren, U. Padlock probes: Circularizing oligonucleotides for localized DNA detection. Science 1994, 265, 2085–2088. [CrossRef] 20. Zhao, Y.; Chen, F.; Li, Q.; Wang, L.; Fan, C. Isothermal Amplification of Nucleic Acids. Chem. Rev. 2015, 115, 12491–12545. [CrossRef] Sensors 2021, 21, 1749 13 of 13 21. Carinelli, S.; Xufré, C.; Martí, M.; Pividori, M.I. Interferon gamma transcript detection on T cells based on magnetic actuation and multiplex double-tagging electrochemical genosensing. Biosens Bioelectron. 2018, 117, 183–190. [CrossRef] 22. Brandão, D.; Liébana, S.; Pividori, M.I. Multiplexed detection of foodborne pathogens based on magnetic particles. New Biotechnol. 2015, 32, 511–520. [CrossRef] 23. Carinelli, S.; Marti, M.; Alegret, S.; Pividori, M.I. Biomarker detection of global infectious diseases based on magnetic particles. New Biotechnol. 2015, 32, 521–532. [CrossRef] 24. Janda, J.M.; Abbott, S.L. 16S rRNA Gene Sequencing for Bacterial Identification in the Diagnostic Laboratory: Pluses, Perils, and Pitfalls. J. Clin. Microbiol. 2007, 45, 2761–2764. [CrossRef] [PubMed] 25. Johnson, J.S.; Spakowicz, D.J.; Hong, B.Y. Evaluation of 16S rRNA gene sequencing for species and strain-level microbiome analysis. Nat. Commun. 2019, 10, 5029. [CrossRef] [PubMed] 26. Melin, J.; Jarvius, J.; Göransson, J.; Nilsson, M. Homogeneous amplified single-molecule detection: Characterization of key parameters. Anal. Biochem. 2007, 368, 230–238. [CrossRef] [PubMed] 27. Russell, C.; Welch, K.; Jarvius, J.; Cai, Y.; Brucas, R.; Nikolajeff, F. Gold Nanowire Based Electrical DNA Detection Using Rolling Circle Amplification. ACS Nano 2014, 8, 1147–1153. [CrossRef] [PubMed]