Improved Performance of Loop-Mediated ... - ACS Publications

Subscriber access provided by RYERSON UNIVERSITY

Article

Improved performance of loop-mediated isothermal amplification assays via Swarm priming Rhett L Martineau, Sarah Anne Murray, Shufang Ci, Weimin Gao, Shih-Hui Chao, and Deirdre R. Meldrum Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b02578 • Publication Date (Web): 04 Nov 2016 Downloaded from http://pubs.acs.org on November 4, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Improved performance of loop-mediated isothermal amplification assays via Swarm priming Rhett L. Martineau*, Sarah A. Murray, Shufang Ci, Weimin Gao, Shih-hui Chao, and Deirdre R. Meldrum Center for Biosignatures Discovery Automation, The Biodesign Institute, Arizona State University, Tempe, AZ, 85287, USA ABSTRACT: This work describes an enhancement to the loop-mediated isothermal amplification (LAMP) reaction which results in improved performance. Enhancement is achieved by adding a new set of primers to conventional LAMP reactions. These primers are termed ‘Swarm primers’ based on their relatively high concentration and their ability to create new amplicons despite the theoretical lack of single-stranded annealing sites. The primers target a region upstream of the FIP/BIP primer recognition sequences on opposite strands, substantially overlapping F1/B1 sites. Thus, despite the addition of a new primer set to an already complex assay, no significant increase in assay complexity is incurred. Swarm priming is presented for three DNA templates: Lambda phage, Synechocystis sp PCC 6803 rbcL gene, and human HFE. The results of adding Swarm primers to conventional LAMP reactions include increased amplification speed, increased indicator contrast, and increased reaction products. For at least one template, minor improvements in assay repeatability are also shown. In addition, Swarm priming is shown to be effective at increasing the reaction speed for RNA amplification via RT-LAMP. Collectively, these results suggest that the addition of Swarm primers will likely benefit most if not all existing LAMP assays based on state-of-the-art, 6-primer reactions.

INTRODUCTION The original loop-mediated isothermal amplification (LAMP) technique introduced in 2000 by Notomi et al. was based on four primers which recognized six distinct regions in the target DNA1. Since then, two publications detail the improvements associated with adding new primers targeting novel regions on the target DNA sequence2,3. The first new primer set, termed “Loop primers”, targets single-stranded loops existing in hairpin structures of early stage amplicons2. The second new primer set, termed “Stem primers”, takes advantage of a transiently single-stranded region on an early stage amplicon—in this case the region located between the inner-most binding sites of the conventional reaction (the “Stem” region). See Figure 1.

in response time, limit of detection, and repeatability. Even when added to 6-primer reactions (LAMP with Loop primers, the state-of-the-art), performance improvements remain significant and include increased reaction speed, increased repeatability for some primer sets, and, with certain indicators, increased production of readout signal. Importantly, adding Swarm primers to conventional reactions does not appear to increase rates of false positives. The advantages of adding Swarm primers to a LAMP assay facilitate biomolecular sensing devices with decreased size, wait times, power requirements, and cost when compared with standard 6-primer reactions.

For both Loop and Stem primers, performance enhancements were mechanistically ascribed to events occurring at the stage of reaction cycling. In both cases, single-stranded DNA was noted as key to performance enhancements. Binding to these single-stranded regions on the early stage amplicons reportedly enabled new amplification reaction pathways, resulting in increased amplicon production speed and faster assays. Corroborating this mechanistic understanding, supplementing LAMP reactions with either of the new primer sets does in fact show new amplicon species on gels3. This paper presents a third new primer set that can be added to the LAMP reaction. This new primer set targets a region on growing amplicons that is not expected to exist for any appreciable amount of time as a single strand. See Figure 1. Despite this, addition of the new primers results in improved reaction performance. When added to a four-primer LAMP mixture, performance improvements include a significant improvement

Figure 1. Loop, Stem, and Swarm primer positions. Primers from each set are located upstream of the FIP/BIP binding sites on opposite strands. Loop primers target the area between F1-F2 and B1-B2. Stem primers act anywhere in the area between F1 and

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 10

B1. F1S/B1S Swarm primers completely or substantially overlap F1 and B1.

MATERIALS AND METHODS Swarm primers LAMP reactions are typically performed on DNA templates that have undergone denaturation, either as a result of the nucleic acids extraction process, or as a result of an explicit denaturation step. Denaturation of the template is not strictly required4, although LAMP performance is often better when denaturation has taken place5. Given that the LAMP reaction requires the annealing of primers to template DNA in order to initiate amplification, we hypothesized that two phenomena might be enabling amplification on non-denatured, presumably double-stranded DNA (dsDNA) templates: 1) Sufficient amounts of DNA, either single-stranded or sufficiently destabilized, might be present in sample solutions despite the lack of explicit attempts at denaturation, or 2) LAMP primers are themselves directly capable of associating with target regions of dsDNA long enough to enable extension polymerization and dsDNA destabilization, facilitating LAMP reactions. If either of these mechanisms is true, then we reasoned that it might be possible to increase reaction initiation rates on dsDNA templates by increasing the concentration of primers. However, simply increasing the concentrations, especially of FIP/BIP primers, might also increase the rates of false positives. On the other hand, if a new primer set—one that does not participate in the cycling stage of the LAMP reaction—can be included in the reaction for the explicit purpose of facilitating reaction initiation, then reaction performance on dsDNA templates might be improved. It would be expected that these new primers, if not involved in the cycling stage of the LAMP reaction, could be added to LAMP reactions at high concentration without any adverse effect on reaction specificity. Figure 2 illustrates the priming concept which motivated this work. A high concentration of primers that specifically lack the loop-back feature of FIP/BIP primers (so as to not promote self-priming dimerization) is introduced into the traditional reaction cocktail (see panel A). The primers (designated ‘F1S’ or ‘B1S’, denoting Swarm primer at the F1 or B1 regions) are designed to be complementary to regions upstream of and on the strand opposite to the FIP/BIP binding sites, so that Swarm primer annealing and extension (panel B) might disrupt the dsDNA and facilitate the binding of BIP primers (panel C). According to this description, two features define ‘Swarm primers’: 1) binding positions located upstream of FIP/BIP but on opposite strands, and 2) ‘high’ concentration.

Figure 2. Hypothetical priming mechanism. (A) The concept that motivated this investigation starts with the addition of a high concentration of new primers to conventional LAMP reactions. (B) The primers anneal to the template upstream of the FIP/BIP and F3/B3 binding sites on opposite strands. (C) Extension exposes single-stranded primer binding sites, improving the accessibility of those sites to standard LAMP primers.

Upon close examination of primers from these considerations, it soon became apparent that both Loop and Stem primers share the first defining feature of the proposed Swarm primers: strand position and synthesis orientation. This is shown in Figure 1. Primers in any of these positions should be expected to be capable of facilitating reaction initiation if any of the proposed mechanisms as described above are correct. New questions thus emerged from this observation. First, what is the actual mechanism of the increased performance observed using both Loop and Stem primers? Is it an increase in amplification efficiency due to the creation of parallel amplification pathways, occasioned by the binding of single-stranded regions on early stage amplicons (as hypothesized by earlier authors)? Or, is it an increase in amplification efficiency due to facilitated initiation? Is it both? We supposed that a comparison of amplification results obtained by using all three primer systems at both high and conventional concentrations would yield mechanistic insight. In particular, primers targeted to the F1S/B1S regions provide unique insight into the mechanisms of enhanced performance since stable, ssDNA to which F1S/B1S Swarm primers can bind does not appear in any predicted amplicon species. (See Supplementary Figure S-1). Since both Loop and Stem primers act on ssDNA regions, the mechanisms ascribed to Loop and Stem primer efficacy increases are not possible with Swarm primers. Swarm primer design Three primer sets targeted sequences in Lambda phage DNA (λ DNA), one targeted the rbcL gene of Synechocystis sp. PCC 6803, and a fifth set targeted the HFE gene in humans. One primer set targeting λ DNA was taken from the literature6. The primer sets consisted of the four base LAMP primers (FIP,


Page 3 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


BIP, F3, and B3), supplemented with any or all of Loop, Stem, and/or F1S/B1S Swarm primers. All primers were synthesized by Integrated DNA Technologies, Inc (Coralville, IA). Manual accessory primer design consisted of: first, constructing FIP/BIP, F3/B3 primers using Primer Explorer V4 software (Eiken Chemical Co. Ltd., Tokyo, Japan; https://primerexplorer.jp/e/index.html), 2) identifying the canonical LAMP primer recognition sequences (F1, F2, F3, etc.), and 3) selecting sequences in the defined Loop, Stem, or F1S/B1S primer regions as shown in Figure 1. Primers were generally 16-25 bases long, contained a 3’ GC clamp where possible, and had calculated melting temperatures between 66 and 70 °C (designed for 65 °C LAMP reactions). The primers used in this study are detailed in Supplementary Table 1. Note that primers targeting the F1S/B1S regions are substantially similar to FIP/BIP 5’ ends, keeping assay complexity low. LAMP conditions LAMP reactions were conducted using Bst 2.0 warm start polymerase (New England BioLabs, Ipswich, MA) in the manufacturer’s recommended buffer, supplemented to 8 mM Mg++ using MgSO4. All reactions were performed at 65 °C. Standard LAMP reactions included 1.6 µM FIP/BIP, 0.8 µM LF/LB, and 0.2 µM F3/B3 primers. Experimental concentrations of the Loop, Stem, and F1S/B1S Swarm primers commonly ranged from 0.1-10 µM but are detailed with each ex-

periment’s description. All reactions presented in this work were conducted without heat pre-denaturation. Experiments using 96-well plates were loaded with random positioning of treatments. Purified λ DNA was obtained commercially (New England BioLabs, Ipswich, MA). Synechocystis and human DNA were obtained from cell cultures, extracted with commercial kits (ZR fungal/bacterial DNA miniprep, Zymo Research, Irvine, CA.) Indicator reagents For all reactions in this study, reaction procession was monitored in real-time using either the intercalating dye EvaGreen (Biotium, Inc., Hayward CA), Hydroxynaphthol blue (HNB) (Sigma-Aldrich, St. Louis, MO, CAS number 63451-35-4)6, or FAM-tagged primers according to the method described by Tanner et al7. For assays based on intercalating dyes or FAM fluorescence, signal intensity was normalized using ROX reference dye (InvitrogenTM, Carlsbad, CA). HNB experiments were done using 120 µM HNB, measured by taking the difference in absorbance at 670 nm from the absorbance at 540 nm. Experiments based on fluorescence were performed using 10 µl reaction volumes, whereas 20 µl reactions were used for absorbance-based experiments.

Figure 3. Primer effect comparisons. (A) Increasing amounts of Swarm primer added to standard LAMP assays (which already contain 0.8 µM Loop primer) result in increasing reaction speed. (B) The derivative of the fluorescence trend associated with the reaction of (A). (C) The effect of Swarm dose on reaction performance indicates an optimum concentration that far exceeds the concentrations used in conventional LAMP reactions. (D) Comparison of primer increases using L2 primers on λ DNA. Swarm primers can be added at high concentration with increasing effect. In contrast, the displacement primers (F3/B3) show little effect with increased concentration, and the Loop primers show inhibitory effects with modest additions over standard concentrations. (E) Combinatorial examination of LAMP with Loop and F1S/B1S primers indicates the order of increasing effectiveness: Loop + F1S/B1S > Loop > F1S/B1S > 4-primer LAMP. The data indicates that the time to detection, the rate of fluorescence production, and whether or not a reaction actually occurs can all depend on the primer treatment. (F) Comparison of primer designs that support Stem priming with primer designs that do not. Adjusting the F2-B2 spacing enables Stem primers (L4 10 µM Stem) to be compared with no Stem (L4 LAMP). Although Stem primers in this configuration have a dramatic effect on assay performance, the performance is poorer than standard 6-primer LAMP with optimal F2-B2 spacing (L2 LAMP).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Data collection and processing

Page 4 of 10

Combinatorial Loop and Swarm primers TM

Amplification curves were collected using a StepOnePlus real-time PCR System (Life Technologies, Inc., Carlsbad, CA) or a SynergyTM H4 multi-mode microplate reader (Biotek Instruments, Inc., Winooski, VT). Fluorescent LAMP assays were conducted on the StepOnePlusTM, whereas colorimetry assays were conducted on the SynergyTM H4. For StepOnePlusTM measurements, the raw data were typically used to create time courses of green intensity, red intensity, the green to red intensity ratio, or the rate of change of the green to red intensity ratio. Collection of the rate of change data enabled the calculation of the time of the peak rate of change in fluorescence, T∆Fmax, which is used in this study, like the Ct value of qPCR, to quantify the template concentration in the sample. The processing of T∆Fmax values from raw fluorescence intensity data is detailed in Figure 3 and is similar to data analysis using bioluminescence approaches8,9. Colorimetric or turbidity data collected on the SynergyTM H4 were processed using Ct values based on arbitrary threshold crossings. T∆Fmax values were not computed for microplate data because the noisier signals on that platform under those indicator schemes were not amenable to rate of change analyses. RESULTS Effect of Swarm primers The effect of adding Swarm primers targeting the F1S/B1S region was tested at a range of concentrations. Figure 3A-C shows the results of adding 0, 0.8 µM, 1.6 µM, 3.2 µM, 6.4 µM, and 10 µM F1S/B1S Swarm primers to a conventional 6primer LAMP reaction (including Loop primers) at standard concentrations. Panel A shows the normalized fluorescence response over time for the various reactions. Panel B plots the data as the change in fluorescence, and Panel C shows the results when the time of peak fluorescence change, T∆Fmax, is plotted as a function of Swarm concentration. Primers targeting this region could be used effectively at concentrations significantly exceeding the concentrations typically used for LAMP, in a dose-dependent fashion.

Figure 3D contrasts the effect of increasing Swarm primer concentrations with increasing concentrations of conventional primers. In this particular case using the L2 primer set on λ DNA, the time to reaction detection could be cut to 50% using Swarm primers. Over the concentration range tested, no inhibition to the reaction was detected. This contrasts with the results of increasing either Displacement primers (F3/B3) or Loop primers (LF/LB); for Displacement primers, increasing amounts of primer have no impact on reaction performance, whereas Loop primers in this case actually inhibit the reaction. As earlier, these reactions used EvaGreen and targeted λ DNA at a concentration of 20,000 copies per 10 µl reaction with no added background DNA. Performance enhancement at the F1S/B1S regions using high concentrations of Swarm primer was observed for 5 of 5 primer sets tested. On the other hand, increasing the concentration of any one primer pair (F3/B3, or LF/LB, or FIP/BIP) over conventional concentrations for 2 of 2 primer sets tested did not result in improved assay performance.

Having observed increased performance with Loop and Swarm primers combined, an experiment was designed to indicate the relative contributions of each. λ DNA at 2,000 copies per 10 µl reaction was targeted using L2 primers. The template concentration was chosen at a level that would typically not react with a 4-primer system but that would react to that 4-primer system supplemented with Swarm primers. As earlier, no background DNA was added to the reaction and EvaGreen was used as the indicator. Treatments consisted of 4-primer LAMP reactions (FIP/BIP, F3/B3 primers), LAMP plus Loop primers, LAMP plus Swarm primers (targeting F1S/B1S), and LAMP plus Loop plus Swarm. Primer concentrations were as follows: FIP/BIP (1.6 µM), F3/B3 (0.2 µM), Loop (0.8 µM), Swarm (3.2 µM). The results are shown in Figure 3E, performed in triplicate but plotted singly for clarity. The results reinforce previous observations that Swarm primers can be used effectively in combination with Loop primers. New findings include: 1) both Swarm and Loop primers are more effective at initiating amplification compared with the 4-primer LAMP reaction. 2) The improvement in reaction speed associated with Loop primers alone is greater than that due to Swarm primers alone. 3) The rate of reaction post-initiation is greater with Loop primers alone than Swarm primers alone. These results support a hypothesis that Loop primers act at both stages of initiation and cycling, potentially disrupting or destabilizing dsDNA, and by also creating new, parallel amplification pathways. Importantly, Figure 3E demonstrates that Swarm primers at F1S/B1S are capable of enhancing initiation; however, they do so less effectively than Loop primers. Finally, these data indicate that Swarm primer effects post-initiation are weak by comparison with Loop primers, as noted by the lower rate of fluorescence production post-initiation. Swarm primers versus Stem primers Standard LAMP primers are designed so that the distance between 5’ ends of the F2 and B2 regions (on opposite strands) is optimally between 120-160 base pairs (see https://primerexplorer.jp/e/v4_manual/index.html, “Guide to LAMP primer designing”, Eiken Chemical Co. Ltd., Tokyo, Japan.). In general, this is too short to accommodate Stem primers. In order to test the effect of Stem primers, it was not possible to add the Stem primers to any primer sets designed using Primer Explorer under recommended conditions. Rather, a new target needed to be selected with an increased distance between the F2 and B2 binding regions (see Figure 1). A primer set (L4) was designed with F2-B2 spacing of 233 base pairs, sufficient to allow Stem priming, using Lambda phage DNA as a target. Approximately 35 bases separated the closest ends of the Stem primers. Figure 3F presents the results of an experiment comparing reactions based on optimal F2-B2 spacing (L2 primers), suboptimal spacing with Stem primers (L4 Stem), and suboptimal spacing without Stem primers (L4). For these primer sets, the comparisons were striking: the primer set with suboptimal spacing (L4) showed far inferior performance compared with the primer set with optimal spacing (L2). The ef-


Page 5 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fect of Stem priming was substantial, however it was not sufficient to bridge the performance gap. Note that all reactions contained Loop primers at standard concentrations, but varied only in their target region spacing and presence/absence of Stem primers. The data presented used Stem primers at a concentration of 10 µM. Analytical EvaGreen

performance

of

Swarm-LAMP

assay

in

A new assay termed Swarm-LAMP was examined for its analytical performance. The assay consists of supplementing the conventional 6-primer LAMP assay with Swarm primers targeting the F1S/B1S regions. Preliminary optimizations indicated that adding 3.2 µM Swarm primers for most primer sets produced peak reaction rate increases without incurring any loss in performance (in some cases, increasing Swarm primers past 3.2 µM resulted in decreasing performance). Serial dilutions of λ DNA, ranging from 20,000,000 to 2 copies per 10 µl reaction, were tested. Assays were conducted in the presence of 100 ng/µl salmon sperm DNA to simulate a mixed envi-

ronmental sample. A 96-well plate was incubated at 65 °C with a standard 6-primer LAMP primer set or the same set supplemented with 3.2 µM Swarm primers at the F1S/B1S locations. In response to observations of highly variable reaction speeds for lower template concentration reactions, a greater number of replicates were performed for the lower concentrations. See Supplementary Material for more detail on the variable replication design. Figure 4A plots the time of peak rate of fluorescence increase vs. the template copy number in log units using the LDNA primers targeting λ DNA. The data clearly indicate an increase in the reaction speed when Swarm primers are added. The data also suggest that the addition of Swarm primers may improve repeatability or even lower the reaction’s limit of detection (only 1 of 6 reactions for the No Swarm treatment were positive at the single copy level, compared with 4 of 6 for the 3.2 µM Swarm treatment. This is addressed in greater detail later). Results are plotted as the mean ± 1 standard deviation. Notably, the advantages of Swarm priming are not offset with an increase of false positives using this primer set, or any others tested throughout this study.

Figure 4. Performance results of Swarm primer assays. (A) Performance comparison of Swarm-LAMP with LAMP using EvaGreen on λ DNA. Swarm-LAMP is faster with the suggestion of improved repeatability. (B) Swarm priming increases performance against a target in Synechocystis sp PCC 6803, rbcL (C) Human HFE is likewise detected more rapidly using Swarm primers. (D) Comparison of Swarm-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 10

LAMP with LAMP (n=48) indicates that repeatability can be improved with the addition of Swarm primers. (E) HNB-based SwarmLAMP reactions produce a greater color change than conventional LAMP, as measured by taking the difference in absorbance at 670 and 540 nm on a microplate reader. (F) Improved color change of HNB-based Swarm-LAMP, seen visually and as computed from RGB components from a digital image. The plus signs indicate the addition of Swarm primers, the minus signs indicate their absence.

Swarm primers were also tested on two additional targets. Figure 4B shows the results of Swarm priming on the Synechocystis rbcL gene. As with λ DNA, Synechocystis DNA was not pre-treated with denaturing heat. In agreement with the λ DNA data, an increase in response speed is observed using Swarm primers, and no increase in false positives confound the assay. Figure 4C illustrates the effect of Swarm priming on human DNA. In this case, the human gene for iron transport, HFE, was targeted with and without Swarm primers. Again, an increase in reaction speed was observed with no increase in the number of false positives. Interestingly, this primer set produced a relatively rare false positive that was characterized by an extremely rapid reaction. Although the data set illustrated in Figure 4C shows such a reaction only for the Swarm priming case, the fast-responding false positives occur with this primer set with or without Swarm primers. To further examine the possibility for improved repeatability suggested by the experiment of Figure 4A, a follow-on experiment compared a 3.2 µM Swarm treatment with a 0 µM Swarm treatment (n=48) for a sample with 200 copies of target per reaction. An improvement in repeatability is shown in the histogram of Figure 4D, where the time of peak rate of fluorescence change is more tightly clustered in the presence of Swarm primers. A similar repeatability test was also conducted for the primer set targeting human HFE, although in that case no improvement in repeatability was observed (n=32). Further experiments comparing single-template detection in λ DNA with and without Swarm primers did not indicate an improvement in limit of detection. Thus, adding Swarm primers to a reaction that uses EvaGreen, and likely other intercalating dyes, has two effects: 1) A possible reduction in the number of replicate reactions that are needed to produce highconfidence quantification estimates, at least in some primer systems, and 2) a reduction in the amount of time required to run the reactions to obtain those estimates. Importantly, no change in the assay’s limit of detection or dynamic range was consistently observed using any of the primer sets developed in this work, when tested using 10-fold dilutions. However, when supplementing conventional 6-primer LAMP reactions with Swarm primers, a few experiments did show statistically significant increases in the limit of detection, but the results were not consistent enough to support a conclusion that Swarm priming increases the assay limit of detection when used in conjunction with Loop primers. Examination using a lower concentration gradient (perhaps 2-fold dilutions) might yield a different conclusion.

ured on a microplate reader in a differential absorbance mode. Figure 4F shows a photograph indicating the visual results of Swarm primer usage, wherein the final color after a reaction is a deeper sky blue compared with the reactions with no Swarm primer additions. The change in color depth is also apparent when using image analysis approaches to differentiate the reactions, shown in Figure 4F. By extracting RGB components from the images and then plotting the difference in Red and Green, a greater signal change is achieved using the Swarm primer systems. Details of the image analysis are included in the Supplementary material. Swarm primers and RT-LAMP LAMP can be used to amplify RNA targets by using a polymerase with native RNA amplification capability or by supplementing standard reactions with reverse transcriptase. The utility of Swarm priming during RNA amplification was investigated next. To facilitate comparisons with amplification on DNA templates (and to enable the use of the same primers), a fragment of DNA containing portions of the human HFE gene was cloned into a plasmid that enabled in vitro transcription (IVT) of DNA into RNA (pGEM-T Easy Vector System I, Promega, Madison, WI). IVT was performed using a HiScribe T7 High Yield RNA Synthesis kit (New England Biolabs, Ipswich, MA). Reactions with RNA templates were setup identically to those explained elsewhere herein for DNA except that reverse transcriptase was added at manufacturerrecommended concentrations (Warm-start Reverse Transcriptase, New England Biolabs, Ipswich, MA). Swarm primers were added at concentrations of 0-10 µM. The results shown in Figure 5 indicate that Swarm primers are capable of improving reaction times during RT-LAMP. This data, with RNA at 5 ng/reaction, shows potential speed increases of about 15%. Similar speed increases with dose-responses were also observed when the template was diluted to approximately 1000 copies per reaction. Lower concentrations weren’t tested.

Analytical performance of Swarm-LAMP using HNB The effect of Swarm primers on assays using the colorimetric indicator HNB was also examined. Experiments were performed similar to EvaGreen experiments. The results of Swarm priming using HNB are qualitatively similar to those using EvaGreen except that Swarm priming in HNB produces an additional benefit- an increase in color production. Figure 4E shows that the color change is more dramatic when meas-

Figure 5. Swarm RT-LAMP. The effect of adding Swarm primers to an RT-LAMP assay targeting human HFE is shown above


Page 7 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(n=6). An improvement in speed of approximately 15% is shown for Swarm concentrations greater than 0.8 µM.

DISCUSSION This research set out to test whether high concentrations of primer could be added in locations that facilitate LAMP reaction initiation, especially for non-denatured templates. Close examination of earlier publications revealed that existing primer sets should also be capable of functioning according to the mechanism postulated in this work, should it be possible at all. Examination of the locations of both Loop and Stem primers on starting dsDNA reveals that neither primer set is excluded from the first criterion of Swarm primers: they bind to a region upstream of, and on the opposite strand to, the FIP/BIP recognition sites (see Figure 1). If Swarm priming affects the reaction at the stage of initiation, then each of these primers should also be able to ‘Swarm’. Loop and Stem primers have two important differences when compared with Swarm primers. The first regards their performance during the post-initiation, cycling stage of LAMP: Loop primers were designed to target a single-stranded portion of an early stage hairpin structure, binding between the F1 and F2 regions, and Stem primers were designed to target a transiently single-stranded region located in the ‘Stem’ region. Both primer sets produce new amplicon structures which are capable of undergoing new cycling reactions, and thus offer performance increases at the post-initiation stage. In contrast, Swarm primers for the F1S/B1S sites are not expected to bind to any stable or transiently single-stranded regions of any of the conventional, intermediate reaction products, as none appear to be available (see Supplementary Material, Figures S-1, S-2, and S-3). Nevertheless, Swarm primers do produce new amplicons, as observed on gel runs. Notably, a large number of small amplicons less than 200 bps are seen on gel runs and are in relatively good agreement with mechanistic predictions (see Supplementary Material). The second difference between Loop/Stem and Swarm primers concerns the reaction at the stage of initiation. Although all of these primer sets are positioned both upstream of and on strands opposite to the FIP/BIP binding sites, they differ in their distance from the FIP/BIP binding sites and other key sites of the LAMP reaction. Not all three primer sets enable priming at very high concentrations. Notably, Loop primers added at concentrations above a commonly published optimum (0.8 µM) have not been observed to significantly increase performance. Indeed, other studies use a lower concentration, for example 0.4 µM10, including the original Loop primers publication11. Inhibition at higher concentrations has been observed for 2 of 2 primer sets tested in this study. Inhibition may be due to steric hindrances caused by proximity to key reaction sites in the case of Loop primers. On the other hand, both Stem primers and Swarm primers located at the F1S/B1S regions are capable of improved performance with very high (>1.6 µmolar, tested up to 10 µmolar) concentrations. Consequently, primer sets at either the Stem or F1S/B1S locations can be thought of as ‘Swarm’ primers, as defined in this work, as they occupy similar functional positions and are useful at unconventionally-high concentrations.

the data presented in Figure 3E. In this figure, it is seen that the addition of Swarm primers to a 4-primer conventional LAMP assay (an assay without Loop primers) enables a reaction to occur; otherwise the reaction does not occur. This experiment was performed with triplicate reactions, and all replicates were consistent with this observation. This shows that, in at least some cases, Swarm primers can initiate a reaction better than a LAMP reaction without any ‘Swarm’ primers. Even with reactions that did contain Loop primers, on occasion an experiment supplemented with Swarm primers resulted in statistically lower limits of detection. These experiments were not consistent, however, and thus were not shown. However, these observations might also support the notion that under some circumstances, initiation of reaction is improved using Swarm primers. A second piece of evidence partially supporting the mechanism of Swarm disruption of dsDNA is the improved repeatability observed with Swarm priming Lambda phage DNA. If Swarm priming acts only after reaction initiation, then it is hard to see how repeatability is increased. In contrast, if Swarm priming increases the effective target concentration by exposing ssDNA targets, then it is conceivable that the reaction kinetics can be shifted from a stochastic regime into a deterministic regime12. This might manifest as an improved reaction repeatability. The third line of evidence regards the comparison of Swarm efficacy between DNA and RNA targets. Swarm efficacy during RNA amplification of HFE indicated approximate speed increases of about 15%, compared with 25% increases observed with corresponding DNA sequences. The targets differ in length and presumed structural complexity such that primer disruption of reverse-transcribed double-stranded RNA templates (

Improved Performance of Loop-Mediated ... - ACS Publications

Recommend Documents