thermal amplification (LAMP)

Subscriber access provided by - Access paid by the | UCSB Libraries

Simplified real-time multiplex detection of loopmediated isothermal amplification (LAMP) using novel mediator displacement probes with universal reporters Lisa Becherer, Mohammed Bakheit, Sieghard Frischmann, Silvina Stinco, Nadine Borst, Roland Zengerle, and Felix von Stetten Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b05371 • Publication Date (Web): 06 Mar 2018 Downloaded from http://pubs.acs.org on March 7, 2018

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 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 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.

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 9

Simplified real-time multiplex detection of loop-mediated isothermal amplification (LAMP) using novel mediator displacement probes with universal reporters Lisa Becherer,*,1 Mohammed Bakheit,3 Sieghard Frischmann,3 Silvina Stinco,4 Nadine Borst,2 Roland Zengerle1,2 and Felix von Stetten1,2 1

Laboratory for MEMS Applications, IMTEK - Department of Microsystems Engineering, University of Freiburg, GeorgesKoehler-Allee 103, 79110 Freiburg, Germany 2 Hahn-Schickard, Georges-Koehler-Allee 103, 79110 Freiburg, Germany 3 Mast Diagnostica GmbH, Feldstraße 20, 23858 Reinfeld, Germany 4 Max von Pettenkofer-Institute, Diagnostic virology department, Pettenkoferstr. 9A, 80336 Munich, Germany Supporting Information, *Phone: +49 761 20373227. E-Mail: [email protected]

Me d Me dc

ABSTRACT: A variety of real-time detection techniques for LAMP mediator based on the change in fluorescence intensity during DNA amplifica5' displaceuniversal reporter ment probe tion enable simultaneous detection of multiple targets. However these 5' LF primer 5' Med techniques depend on fluorogenic probes containing target-specific 3' Medc 3' DNA 5' sequences. That complicates the adaption to different targets leading to time-consuming assay optimization. Here, we present the first + rev. Primer 2 1 universal real-time detection technique for multiplex LAMP. The - DNA novel approach allows simple assay design and is easy to implement 5' Me d 5' Med for various targets. The innovation features a mediator displacement 5' rev. Primer Medc 3' Medc 5' probe and a universal reporter. During amplification of target DNA cDNA 3' the mediator is displaced from the mediator displacement probe. Then it hybridizes to the reporter generating a fluorescence signal. The novel Mediator Displacement (MD) detection was validated against state-of-the-art molecular beacon (MB) detection by means of a HIV-1 RT-LAMP: MD surpassed MB detection by accelerated probe design (MD: 10 min, MB: 3-4 h), shorter times to positive (MD 4.1±0.1 min shorter than MB, n = 36), improved signal to noise fluorescence ratio (MD: 5.9±0.4, MB: 2.7±0.4; n = 15) and showed equally good or better analytical performance parameters. The usability of one universal mediator-reporter set in different multiplex assays was successfully demonstrated for a biplex RT-LAMP of HIV-1 & HTLV-1 and a biplex LAMP of Haemophilus ducreyi & Treponema pallidum, both showing good correlation between target concentration and time to positive. Due to its simple implementation it is suggested to extend the use of the universal mediator-reporter sets to the detection of various other diagnostic panels. F2

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

Loop-mediated isothermal amplification (LAMP) first described in 2000 by Notomi et al1 has recently emerged as a powerful in vitro nucleic acid amplification technique and represents an alternative diagnostic tool to the gold standard polymerase chain reaction2,3. As major advantage, LAMP overcomes the need of elongated thermal cycling protocols and sophisticated thermal management. One of the most significant shortcomings of LAMP, however, is the lack of easy to implement sequence-dependent detection techniques that would allow the multiplex detection of different target molecules. Multiplex detection during LAMP is an indispensable requirement for many applications, ranging from clinical diagnosis of infectious diseases, including for example the identification4 and subtyping5 of viruses, over monitoring of cancer treatment6 to the detection of foodborne pathogens7. Common techniques for real-time detection during LAMP like fluorescence intercalating dyes8, turbidity measurement9 or color-change of fluorescence metal-sensitive indicators10 are well-established but are not capable of distinguishing between multiple target sequences and may even

detect unspecific by-products. By contrast, multiplex LAMP with subsequent gel electrophoreses4 permits detection of multiple targets but suffers from time-consuming postamplification processes and the risk of carryover contamination caused by opening of reaction vessels11. Two-stage amplification methods, like dubbed rapid amplification (RAMP) presented by Song et al.12, stand out due to a very high level of multiplexing, but require additional reagents for a second amplification reaction leading to higher costs per assay. Simultaneous detection of multiple targets without the opening of reaction vessels is achieved by using a microfluidic chip12. In addition, a few sequence-dependent multiplex detection techniques based on fluorescence resonance energy transfer (FRET) have been developed of late: LAMP detection methods utilizing molecular beacons containing primer sequences were introduced by Liu et al.13 and Nyan et al.14. This technique depends strongly on primer design which is why the molecular beacons are prone to exist in open conformation under LAMP conditions causing false positive signals at low GC-contents. Alternatively QUASR (Quench-

ACS Paragon Plus Environment

Analytical Chemistry 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

ing of Unincorporated Amplification Signal Reporters) LAMP developed by Ball et al.15,16 for end-point detection or DARQ (Detection of Amplification by Release of Quenching) LAMP established by Tanner et al.17,18 for real-time detection can be applied in multiplex assays. In QUASR and DARQ one primer is modified with a 5’ fluorophore and hybridized with a short oligonucleotide labeled with a quencher. During target amplification the quencher oligonucleotide is displaced leading to an increase in fluorescence. Further methods make use of the combination of fluorophore-labeled primer and intercalating dye19 or require additional enzymes like restriction endonucleases20–22 leading to higher costs per assay. What all these detection methods have in common is that the fluorogenic oligonucleotides comprise target-specific sequences, hence the fluorescence yield of the fluorophore depends on the guanine content of the adjoining base sequence. Apart from that the GC-content of involved sequences influences the equilibrium between closed and opened conformation of signal generating oligonucleotides. Consequently above listed methods do not represent universal detection techniques and have to be redesigned for different targets individually leading to timeconsuming assay design and optimization work. Despite the variety of methods capable of multiplex LAMP detection, there is still the unmet need for a universal detection technique which can be adapted easily to various target panels. In order to overcome the disadvantages of the available multiplex detection methods for LAMP, we have developed a novel approach for real-time detection during LAMP which is not only capable of monitoring different targets simultaneously, but can also be implemented easily for different target panels without the need of advanced assay design nor elaborate optimization work. In this work we present the first universal detection technique for LAMP, named Mediator Displacement (MD) detection. The MD detection is based on the strand displacement of a universal mediator during amplification and its subsequent interaction with a universal fluorogenic reporter oligonucleotide, in which the last step was first described by Faltin et al. 23–25 and further investigated and applied for pentaplex assays by Wadle et al.26,24. We start this research by demonstrating the simple implementation of MD detection compared to the state-of-the-art technique based on molecular beacons. Focal point of the discussion is the elaborate design of target-specific molecular beacons accompanied by the need for time-consuming optimization work. To verify the MD detection regarding assay performance we investigate signal to noise ratio, time to positive, LOD, linearity, inter- and intraassay variance of the reverse transcription (RT)-LAMP of HIV-1 and compare the herein introduced MD technique with the detection based on molecular beacons. In the next step the simultaneous detection of two targets is presented by applying the MD detection for biplex RT-LAMP of HIV-1 and HTLV-1. Finally we close with the demonstration of the universal applicability of the same mediator-reporter set used for HIV-1 and HTLV-1 by applying it for the detection of the biplex LAMP of Haemophilus ducreyi and Treponema pallidum, again underlining the easy implementation of the MD detection for various target sets.

EXPERIMENTAL SECTION Chemicals and Oligonucleotides. All oligonucleotides (Table S1) were synthesized and HPLC purified by Biomers (biomers.net, Ulm, Germany). Bst 2.0 WarmStart DNA

Polymerase, 10x Isothermal Amplification Buffer, 100 mM magnesium sulfate (MgSO4), 20 mg/ml bovine serum albumin (BSA) molecular biology grade and deoxynucleotide (dNTP) mix 10 mM each were obtained from New England Biolabs (Frankfurt, Germany). 5 M betaine was purchased from Sigma-Aldrich (Munich, Germany). Transcriptor reverse transcriptase was bought from Roche Diagnostics (Mannheim, Germany) and DNase/RNase-free distilled water was obtained from Life Technologies GmbH (Darmstadt, Germany). MagaZorb® DNA Mini-Prep Kit for nucleic acid extraction was purchased from Promega (Mannheim, Germany). QubitTM dsDNA HS assay Kit was bought from InvitrogenTM, Thermo Fisher Scientific (Waltham, MA, USA). Primer and MD Oligonucleotide Design. LAMP primers for HIV-1 targeting the protein p24 of the gag gene were taken from Curtis et al. 27. LAMP primers for Treponema Pallidum (T. pallidum) targeting the polA gene (TP_0105) were taken from (in preparation, Ref1). LAMP primers for HTLV-1 targeting the tax gene and primers for Haemophilus ducreyi (H. ducreyi) targeting the 16S ribosomal RNA gene were designed using Primer Explorer version 5 software (Fujitsu, Tokyo, Japan)28. All primer sequences are listed in Table S1. Fluorogenic universal reporter (UR) molecules were taken from (in preparation, Ref2). Non-fluorogenic MD oligonucleotides were designed in silico using nearestneighbor thermodynamic parameters29 for the calculation of DNA duplex binding strengths. In silico computation was carried out with the software VisualOMPTM (DNA Software®, USA, Version 7.8.42.0) 30,29. Parameters were adjusted for LAMP reaction conditions (common solution conditions, 63 °C assay temperature, 80 mM [Monovalent+], 8 mM [Mg2+]). Non-fluorogenic MD oligonucleotides comprise two oligonucleotides, a universal mediator (Med) and a modified Loop F (LF) primer as illustrated in Scheme 1. (LF_Medc, black box). The modified primer contains a target specific primer sequence (LF) and a mediator hybridizing side (Medc), which is complementary to the mediator. The dimer consisting of mediator and modified primer is called mediator displacement probe (MD probe). Loop primer sequences prove as appropriate components for the probe part of MD probes as the modification of the 5’ end does not influence LAMP mechanism unlike F3/B3 or FIP/BIP primers. LF and Loop B (LB) primers both show similar results regarding time to positive as well as signal to noise ratio when modified and used for MD detection (data shown in SI). In this work LF was chosen for MD probe design. Based on the target specific LF sequence, the novel MD probe is designed as follows: Design Guidelines for MD probes: A universal sequence tag (Medc, 29-30 nucleotides) is added to the 5’ end of the target specific LF sequence presented as LF_Medc in Scheme 1. The complete mediator sequence hybridizes to LF_Medc forming the novel MD probe. Medc and mediator sequences are illustrated by underlined oligonucleotides in Table 1. The mediator is also designed to hybridize with 14 nucleotides of a fluorogenic UR starting from the 3’-end of the mediator and leaving a dangling end of about 15-16 nucleotides at the 5’-end of the mediator. Complementary regions between mediator and UR are illustrated by bold nucleotides in Table 1. The consequence of the complete hybridization between mediator and the Medc sequence of LF_Medc is, that the bond strength between LF_Medc and mediator is higher than between mediator and UR under


Page 2 of 9

Page 3 of 9 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 LAMP conditions. This ensures that the mediator binds to the LF_Medc and not to the UR before amplification avoiding prematurely interaction between mediator and UR. Following this protocol recommended bond strengths (∆GLF_Medc-mediator=-24 kcal/mol and ∆GUR-mediator=-9 kcal/mol) optimized in previous experiments (data shown in Figure S1) are obtained. Table 1 shows the sequences of LF_Medc, mediator and UR for the targets HIV-1, HTLV-1, H. ducreyi and T. pallidum. One and the same mediator-reporter set is applied for the detection of HIV-1 and T. pallidum (each detected with UR2/Med2) as well as for HTLV-1 and H. ducreyi (each detected with UR1/Med1)) indicated by the indices in URi, Medi and Medci. The index i designates universal sequences assigning mediator, tag with mediator hybridizing side and UR to one defined UR-mediator set. The preconfigured URi-Medi sets can be applied for different targets by simply adding the underlined oligonucleotides at

the 5’-end of LF_Medi shown in Table 1 to a target-specific Loop primer. For multiplex detection URs with different fluorogenic labels are used to differentiate between distinct targets. Reference Assay. As a reference, a molecular beacon (MB) for HIV-1 detection was designed using VisualOMPTM (DNA Software®, USA, Version 7.8.42.0)30,30,29 based on nearest-neighbor thermodynamic parameters29 with parameters adjusted for LAMP reaction conditions. The MB for HIV-1 LAMP was directly modified from LF as described by Liu et al. 13. To ensure a closed conformation of the molecular beacon under LAMP conditions the original HIV-1 LF primer had to be reduced by 4 nucleotides at the 3’-end and the MB arm consisted of 7 instead of the 6 nucleotides contrary to the recommendation by Liu et al. 13. The sequence of the MB for RT-LAMP of HIV-1 used in this article is listed in Table 1. a Table 1 Oligonucleotides used for Mediator Displacement (MD) and molecular beacon (MB) detection during (RT-)LAMP. Description

Sequence (5’–3’)

LFHTLV_Medc1

GGTCGTAGAGCCCATTGCGCGATGAGTGGGAGGGGAGTCGAGGGATAAG

LFHIV_Medc2

CACTGACCGAACTGAGCTCCTGAGGCATGGTTTAACATTTGCATGGCTGCTTGAT

LFH. ducreyi_Medc1

GGTCGTAGAGCCCATTGCGCGATGAGTGGGTCACCCAAGGAGCAAGCC

LFT. pallidum_Medc2

CACTGACCGAACTGAGCTCCTGAGGCATGGCGATAAATACCATCAAGTGTGCCAAA

Med1

CCACTCATCGCGCAATGGGCTCTACGACC

Med2

CCATGCCTCAGGAGCTCAGTTCGGTCAGTG

UR1(Ref2: in preparation)

BMN-Q-535-ATTGCGGGAGATGAGACCCGCAA-dT-FAM-TGTTGGTCGTAGAGCCCAGAACGA-C3

UR2(Ref2: in preparation)

BMN-Q-535 –CACCGGCCAAGACGCGCCGG-dT-Atto-647N-GTGTTCACTGACCGAACTGGAGCA-C3

MB (HIV)

6-Fam-AGCAGCCTTTAACATTTGCATGGCTGCT-BMN-Q535

a

Complementary regions between LF_Medc and mediator are illustrated by underlined oligonucleotides. Bold nucleotides visualize complementary regions between mediator and UR. Complementary sequences between Med and Medc or UR are indicated by indices. Targetspecific sequences (LF primer sequences) are written in italics.

Viral and Bacterial Templates. Isolated genomic HIV-1 (group M subtype B) and HTLV-1 RNA were provided by Max von Pettenkofer-Institute, (Dr. Hans Nitschko, Munich, Germany). HIV-1 and HTLV-1 RNA were extracted and purified with MagaZorb® DNA Mini-Prep Kit from cell cultures (HIV-1: MVP89931, HTLV-1: infected MT-2 cells32). Reference DNA samples of H. ducreyi and T. pallidum DNA were provided by by University of Washington. The concentration of HIV-1 RNA was determined with the Abbott m2000rt system (Abbott Laboratories, North Chicago, IL, USA). HTLV-1 nucleic acid concentration and DNA concentrations of H. ducreyi and T. pallidum were determined fluorometrically (QubitTM dsDNA HS assay). RT-LAMP and LAMP Assays. A stock solution of the MD probe was prepared by mixing mediator and LF_Medc in the ratio of 1:2 at room temperature without further treatment. Stock solutions can be stored at -20 °C. All LAMP assays were designed according to Nagamine et al. 33 including six primers for accelerated reaction. LAMP reactions were performed in a 25 µl reaction mixture containing 1x Isothermal Amplification Buffer, 8 mM MgSO4, dNTP Mix 1.4 mM each, 8 U Bst 2.0 WarmStart DNA Polymerase and 1 g/l BSA. RT-LAMP detected by MD contained additionally 0.4 mM betaine and 10 U Transcriptor reverse transcriptase. RT-LAMP detected by MB contained additionally 0.8 mM betaine and 10 U Transcriptor reverse transcriptase. Singleplex (RT-)LAMP combined with MD detection included 1.6 µM of each FIP and BIP, 0.8 µM LB (except RTLAMP of HTLV-1, there LB was not needed), 0.6 µM LF

and 0.2 µM LF_Medc, 0.2 µM of each F3 and B3, 0.1 µM mediator and 0.05 µM UR. Singleplex RT-LAMP detected by MB included 1.6 µM of each FIP and BIP, 0.8 µM of each LB and LF, 0.2 µM of each F3 and B3 and 0.32 µM MB. Primer concentrations for biplex (RT-)LAMP reactions were optimized (data not shown) and are listed in Table S2. The concentrations of MD oligonucleotides in biplex assays were the same as in singleplex assays with 0.2 µM LF_Medc, 0.1 µM mediator and 0.05 µM UR for each assay. All reactions were performed at 63 °C. Immediately before amplification RNA was linearized by heating the isolated sample suspension to 70 °C for 10 minutes followed by immediate cooling on ice. RNA and DNA samples were diluted in 10 mM Tris (pH 8). Instrumentation, Data Collection and Processing. Realtime (RT-)LAMP assays were carried out in a Rotor-Gene 6000 (Corbett Research Pty., now Qiagen). The fluorescence signal was acquired every minute with signal acquisition using the Cy5- and FAM-readout gain, depending on the fluorescence modification at the UR or MB. Data processing was performed as previously described 34 by calculating time to positive (tp) as the time of the maximum increase of fluorescence. The time of the maximum fluorescence increase was taken from the first derivative of the fluorescence intensity, correlating directly with template concentration in the sample. The calculation is based on the raw fluorescence intensity data as described in Figure 2C. Data processing was carried out using OriginPro 9.0 (OriginLab Corporation,



Northampton, USA). Limit of detection (LOD) was calculated with Probit regression analysis allowing prediction of the DNA concentration having a 95% probability of detection. Probit regression analysis was conducted by using IBM SPSS Statistics 19 (IBM Company, Chicago, USA).

RESULTS AND DISCUSSION

by a universal sequence tag (Medc), to which a complementary oligonucleotide, the mediator (Med), is hybridized (Scheme 1: black box). MD probes participate in the amplification just almost as well as LF primers and consequently incorporate a Medc into amplification products (Scheme 1B and C). Scheme 1B and C illustrate the elongation of the MD probe and the resulting displacement of the 5’ terminus of the first stem-loop structure building up the before mentioned dumbbell-like DNA. In the next step the initial DNA is displaced and the 3’ end of the elongated MD probe can refold and form a stem-loop structure (Scheme 1D). Either hybridization and extension of a BIP primer, or extension of the refolded 3’ end induces displacement of the mediator (Scheme 1E). The mediator displaced in the previous step anneals to the Medc at the UR and displaces the 5’ terminus of the UR during mediator extension (Scheme 1F). The unfolding of the hairpin structure of the UR impedes FRET and leads to dequenching of the fluorophore. This mechanism described above is one of many possible reaction paths for mediator displacement during amplification as loop primers do not only hybridize with dumbbell-like structures described in Scheme 1, but also with other amplicons generated during LAMP. MD detection enables multiplex realtime detection during amplification by combining various URi (i = 1, …, n), modified with different fluorophores, and related mediators Medi with defined targets. Owing to the universal, not target-specific sequence of mediator and UR, the MD technique can be applied for different targets without the need for optimization.

Working Principle of the MD Detection of LAMP. MD detection provides real-time detection during target amplification in a homogeneous reaction mix. The working principle of MD detection during LAMP is illustrated in Scheme 1 and is based on the strand displacement activity of the polymerase used in LAMP. The signal generation of the MD technique relies on the displacement of a universal mediator oligonucleotide during amplification. The released mediator then anneals to a universal reporter (UR) modified with fluorophore and quencher. Subsequent mediator elongation at the UR leads to signal generation. In a first step LAMP amplification is initialized as described by Notomi et al.1: Dumbbell-like DNA structures, containing two stem-loop parts, are generated during synthesis of daughter strands using F3 and B3 as well as FIP and BIP primers (Scheme 1A). Once these dumbbell-like structures are generated Loop primers anneal to the loop of the stem-loop structure and accelerate the amplification reaction (Scheme 1B) as described in 33. MD detection relies on the usage of mediator displacement probes (MD probes). The novel MD probe consists of a Loop F (LF) primer extended Scheme 1. Mechanism of the Mediator Displacement (MD) detectiona of loop-mediated isothermal amplification (LAMP). (A) Primer annealing during LAMP. (B) Dumbbell-like DNA structure is generated. (C) Elongation of the MD probe at the stem-loop structure. (D) Annealing of backward primer. (E) Mediator is displaced from the incorporated MD probe. (F) Mediator triggers signal generation at the universal reporter.

a Oligonucleotides required for amplification and detection are listed in the box on top. The universal reporter (UR) is modified with fluorophore and quencher and a sequence taq at the 3' end (Medc), which is complementary to the mediator (Med). The novel MD probe consists of a LoopF (LF) primer, modified with a Medc taq at the 5' end (LF_Medc), and a mediator hybridized to Medc at LF_Medc.


Page 4 of 9

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 Mediator Displacement Compared to Molecular Beacon Detection: Systematic Verification. In the course of a critical and systematic verification of the proposed method, the analytical performance of MD detection was analyzed and compared to a renowned state-of-the-art detection technique based on molecular beacons (MB) as proposed by Liu et al.13 on the example of a RT-LAMP of HIV-1. First, assay set up for MD and MB detection is discussed with focus on oligonucleotide design and assay composition. In the next step analytical performance parameters as signal to noise ratio, limit of detection (LOD), inter- and intraassay variance were investigated for both detection methods. Further experiments addressing the specificity of MD detection are listed in the Supporting Information. Unless otherwise stated, MD and MB reagents were used as listed in Table 1. Key difference between these two methods is the elaborate and time-consuming design of MB compared to MD oligonucleotides arising from target-specific sequences in the MB. The Loop primer sequence serves as probe and arm region of the MB13, which in case of a low GC content of the Loop primer causes issues concerning the equilibrium between closed and opened configuration of the MB under LAMP conditions. MB design for RT-LAMP of HIV-1, presented in this article, requires special expertise and the usage of special software as both Loop primers have low GC contents. Estimated duration needed for MB design is settled between one and three hours per MB depending on the GC content of the Loop primers, whereas MD detection can be adapted to different targets in less than 10 minutes. Moreover using MBs for detection entails a potential risk of the need for further design iterations in case that the MB exists mainly in open configuration under LAMP conditions. In contrast MD technique ensures successful detection by applying the novel MD probe described in the Experimental Section, Primer and MD Oligonucleotide Design, due to its universal character. Similarities between the two considered methods can be found in assay procedure as the assay compositions of MD and MB reactions were identical except of differing oligonucleotides for detection and increased betaine concentration used for MB detection as recommended by Liu et al.13. Increased betaine concentration proves problematic as betaine is able to manipulate the relative stability of base pairs35. The need for additional betaine in MB detection may lead to different primer interactions and therefore changed assay performance pointing out another disadvantage of using MBs. In the next step analytical performance parameters were investigated for both detection techniques. Frist, the signal to noise ratio of MD detection was compared to the signal to noise ratio of MB detection. Figure 1 presents a dilution series for the RT-LAMP of HIV-1, spanning a range from 1x102 to 1x106 copies per reaction. Assays were conducted in triplicates. To enable a direct comparison of signal to noise ratios between MD and MB detection, the fluorophore in UR2 (Table 1) was exchanged with FAM. Consequently universal reporter and molecular beacon comprised the same fluorophore and quencher molecules. The results shown in this figure indicate, that compared to MB detection, MD detection enables a low background fluorescence signal (noise) resulting in a high signal to noise ratio. In contrast to MD detection, surpassing with a signal to noise ratio of about 5.9±0.4 (n = 15), MB detection generates a signal to noise ratio of about 2.7±0.4 (n = 15) that is more than two times lower than the ratio for MD detection. n stands for the number of data points considered for calculation. This result

indicates evidently the advantage of using a universal detection technique over target-specific fluorogenic probes. The benefit of using a target-independent UR as a universal fluorogenic probe is that the reaction for signal generation is molecular decoupled from the interaction between probe and target. Therefore the signal generation step can be optimized independently of the target sequence and enables universal applicability of the optimized UR for multiple assays. 10 9 Norm. Fluorescence

Page 5 of 9

MD

8 7

1×106 cp/rxn 1×105 cp/rxn 1×104 cp/rxn 1×103 cp/rxn 1×102 cp/rxn

NTC MD NTC MB

6 5 4 3

MB

2 1 0

10

20

30 40 Time (min)

50

60

Figure 1 Normalized fluorescence intensity, plotted against time, of 1x102 - 1x106 copies per reaction (cp/rxn) HIV-1 RNA detected by MD and MB detection. To enable a comparison of signal to noise rations between MD and MB detection the fluorophore in UR2 was exchanged with FAM.

Secondly, the LOD was determined as recommended in the MIQE Guidelines36 by amplifying various HIV-1 RNA concentrations (5x102, 1x102, 5x101, 1x101, 1x100 copies per reaction) and no template controls (NTC) in 8 replicates each. With the fraction of positive amplifications generating a detectable fluorescence signal and Probit regression analysis the 95 % detection limit was determined to be 131 copies per reaction (95 % confidence interval (CI): 87 – 284 copies per reaction) for MD detection and 141 copies per reaction (95 % CI: 95 – 300 copies per reaction) for MB detection generating a comparable result for both methods (Figure 2A). MD detection surpasses MB detection regarding time to positive, whereby MD detection is on average 4.1±0.1 min (n = 36) faster. n stands for the number of data points considered for calculation. Decreased reaction speed of MB detection is for one thing based on higher betaine concentration required during MB (RT-)LAMP. And secondly, the Loop primer competes with the MB containing identical sequences, whereas the MB is not able to start amplification and therefore slows down reaction speed. On the contrary LF_Medc in MD detection is capable of participating in the amplification reaction and therefore does not significantly slow down the reaction. Investigations of the reaction speed of an HIV-1 RT-LAMP showed, that 0,2 µM LF_Medc does not influence reaction speed (data shown in Supporting Information). In the next step the repeatability (intraassay variance) and reproducibility (interassay variance) of HIV-1 RT-LAMP using MD detection was examined and compared to MB detection. Figure 2B, C and D illustrate the data analysis of the recorded real-time fluorescence data during amplification. Time to positive (tp) was exemplarily determined for MD detection as illustrated in Figure 2B and C and plotted against initial RNA copy number (Figure 2D). Error bars reflect SD values for tp calculated by the scattering of measurement values between triplicates. It can be observed, that the lower the copy number per reaction, the higher the SD for tp as demonstrated in Figure 2D. To assess


Analytical Chemistry the interassay reproducibility three batches of RT-LAMP reaction mixes were individually prepared and processed in different runs. Different concentrations of HIV-1 RNA ranging from 1x103 to 1x106 copies per reaction were amplified in triplicates by the use of both detection techniques in parallel. The data for 1x102 copies per reaction is excluded from analysis because of high intraassay variance caused by inconsistent detection at low copy numbers below the LOD. Interassay variance of tp for 1x103 – 1x106 copies per reaction is presented in Figure 2E. The percentage of CV ranged from 34.0 % (1x103 copies per reaction) to 0.7 % (1x106 copies per reaction) for MD detection and from 25.8 % (1x103 copies per reaction) to 3.4 % (1x106 copies per reaction) for MB detection. Intraassay variance was determined by calculating the average value of SD for each

A

10

MD MD lower/upper bound MB MB lower/upper bound 60

70 80 Probability (%)

90

4 1×106 cp/rxn 1×105 cp/rxn 1×104 cp/rxn 1×103 cp/rxn 1×102 cp/rxn NTC

3 2 1 0

100

D

∆Fluorescence /∆Time (a.u./min)

Fluorescence (a.u.)

Copies/reaction

C

5 MD

2

101 50

of the three runs obtained from triplicates within one run. The percentage of CV ranged from 16.0 % (1x103 copies per reaction) to 0.7 % (1x106 copies per reaction) for MD detection and from 21.2 % (1x103 copies per reaction) to 0.9 % (1x106 copies per reaction) for MB detection. The average values of R2 were calculated from the three runs conducted for inter- and intraassay variance analysis. The results indicate the same linearity between 1x103 and 1x106 copies per reaction with R2 = 0.94 for MD detection and R2 = 0.94 for MB detection. These results confirm that our proposed method features equally good LOD, linearity, repeatability and reproducibility and even shorter time-to-positive as well as higher signal to noise ratio compared to molecular beacons.

B

103

0

10

20 30 Time (min)

40

50

1×106 cp/rxn 1×105 cp/rxn 1×104 cp/rxn 1×103 cp/rxn 1×102 cp/rxn

0,5 0,4 0,3 0,2 0,1 0,0 0

E

10

20

30 40 Time (min)

50

60

F 26

27

MD MB

28 26 24 22 20 18 16 102

103 104 105 copies/reaction

106

1×103 cp/rxn

26 MB RT-LAMP (tp (min))

30

107

MB RT-LAMP (tp (min))

32

101

0,6 MD

60

34

tp (min)

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 9

3

1×10 cp/rxn

25 24 1×104 cp/rxn

23 22

1×105 cp/rxn

21

1×106 cp/rxn

20 19 16

17 18 19 20 21 MD RT-LAMP (tp (min))

22

23

25 24

1×104 cp/rxn

23 1×105 cp/rxn

22 21 20 16

1×106 cp/rxn 17

18

19

20

21

22

MD RT-LAMP (tp (min))

Figure 2. Verification of the Mediator Displacement (MD) detection of the real-time RT-LAMP of HIV-1 with molecular beacon (MB) based detection as reference method. MD and MB reagents were used as listed in Table 1. (A) Limit of detection (LOD). The probability of the successful amplification of a given input copy number was calculated with Probit analysis and plotted in grey for MD detection and in black for MB detection. Dashed lines illustrate 95 % confidence interval (CI). (B) Fluorescence intensity, plotted against time, of 1x102 1x106 copies per reaction (cp/rxn) HIV-1 RNA for MD detection. (C) The derivative of the raw data of fluorescence intensity given in B, measured with MD detection, is plotted against time to illustrate the calculation of time to positive (tp). tp is determined by calculating the peak maximum of the derivative of fluorescence intensity. (D) tp values for MD and MB detection have been determined with the method described in (C) and plotted against the number of copies per reaction. The data shows linear dependence between tp and copy number per reaction between 1x103 and 1x106 cp/rxn for MD detection. 1x103 cp/rxn deviates from linearity during MB detection. 1x102 cp/rxn show a deviation from linear behavior for both techniques due to high variance between triplicates. (E) Interassay (reproducibility) and (F) intraassay (repeatability) variance of MD detection compared to MB detection. The data for 1x102 cp/rxn is excluded because of high intraassay variance at the LOD.

Universal Applicability of MD Detection for Multiplex (RT-)LAMP. A set of different universal mediator and reporter molecules can be used for multiplex detection of various targets without the need for target-specific modifications or adaption of molecular structures. To verify universal applicability of MD detection for multiplex (RT-)LAMP assays two clinical relevant target panels were chosen. For the biplex detection of the two retroviruses HIV-1 and HTLV-1 a universal MD set, comprising two different universal reporter and corresponding mediator molecules, was investigated, whereas each universal reporter and mediator were assigned with one defined target, respectively. The

same MD set was applied in a second biplex LAMP for the detection of Haemophilus ducreyi (H. ducreyi) and Treponema pallidum (T. pallidum) DNA, both are predominant aetiological agents for genital ulcer disease 37. The universal MD set investigated in this work consists of two universal reporter molecules UR1 and UR2 modified with FAM/BMNQ-535 and Atto-647N/BMN-Q-535 (in preparation, Ref2), respectively, as well as two universal mediator molecules Med1 and Med2. The mediator Med1 for HTLV-1 is capable to hybridize selectively with UR1 and the mediator Med2 for HIV-1 hybridizes selectively with UR2.


Page 7 of 9

HIV-1 RNA (1x103 - 1x106 copies per reaction) was reverse transcripted and coamplified with HTLV-1 RNA (each assay comprises 2000 copies per reaction). The results are shown in Figure 3A by plotting tp for both targets against HIV-1 RNA copy number per reaction. Coamplification of various amounts of HIV-1 RNA and HTLV-1 RNA with constant copy number as well as the correlation between template concentration and tp is successfully demonstrated. Error bars reflect SD values for tp calculated by the scattering of measurement values between triplicates. High error bars for 1x103 copies per reaction HIV-1 RNA result from the delayed signal increase of one out of the three replicates. The repetition of this experiment, whereas each assay comprised 20 copies per reaction HTLV-1 RNA, showed that a low HTLV-1 copy number can still be detected in a biplex RTLAMP with HIV-1 (data not shown). Increased tp values for HIV-1 amplification compared to singleplex assays (Figure 2D) occur as a result of reduced primer concentration as common in biplex LAMP. To show universal applicability of the same MD set, UR1 and UR2 as well as corresponding mediators Med1 and Med2 were used for the MD detection of H. ducreyi and T. pallidum biplex LAMP, respectively. Various H. ducreyi DNA concentrations (4, 40 and 400 cp/rxn) were coamplified with constant T. pallidum DNA concentration (100 copies per reaction). Figure 3B illustrates the results of tp for each target plotted against H. ducreyi DNA concentration. Again, successful application of the MD set as well as the correlation between template concentration and tp was demonstrated successfully. High template concentrations of H. ducreyi lead to increased time to positive for T. pallidum but do not result in false negative signals for low T. pallidum concentrations (Supporting Information). These results provide evidence for the universal applicability of one MD oligonucleotide set, comprising mediator and corresponding universal reporter molecules, for different multiplex LAMP assays without the need of advanced assay redesign or additional optimization work.

tp (min)

A 36 34 32 30 28 26 24 22 20 18 16 10

tp HIV-1 tp HTLV-1 (constant 2000 cp/rxn)

2

B

10

3

4

5

10 10 10 HIV-1 copies/reaction

6

10

7

Figure 3. Application of the same universal mediator-reporter set in two different biplex (RT-)LAMP assays. (A) MD detection of the biplex RT-LAMP of various HIV-1 RNA concentrations and HTLV-1 RNA. Time to positive (tp) for HIV-1 (black squares) and HTLV-1 (grey triangles) amplification is plotted against HIV-1 copy number per reaction for 2000 copies per reaction (cp/rxn) HTLV-1 RNA. NTCs are not illustrated as no signal increase occured during the detection time of 60 min. (B) MD detection of the biplex LAMP of various H. ducreyi DNA concentrations and T. pallidum DNA. tp for H. ducreyi (black squares) and T. pallidum (grey triangles) amplification is plotted against H. ducreyi copy number per reaction for 100 cp/rxn T. pallidum DNA. NTCs are not illustrated as no signal increase occured during the detection time of 60 min.

CONCLUSION In summary, this article presented the first universal detection method for LAMP assays, called Mediator Displacement (MD) detection, surpassing state-of-the-art detection techniques with respect to easy adaptability to various target panels. The proposed MD detection technique was applied for RT-LAMP of HIV-1 whereas effort for assay design as well as assay performance parameters were compared to the state-of-the-art detection technique based on molecular beacons (MB). MD detection exceled not only through simplified assay design, but also showed 4.1 min on average shorter times to positive than MB detection. Signal to noise fluorescence ratio was three times higher for MD compared to MB detection, underlining the advantage of using universal reporter molecules over target-specific fluorogenic probes. Both detection techniques showed similar results for limit of detection (MD: 131 copies/reaction, MB: 141 copies/reaction), repeatability, reproducibility and linearity (R2 = 0.94 for MD and MB). To demonstrate the capability of simultaneous and multiple target detection and furthermore the universal applicability of one mediator-reporter set for different target panels, the MD detection was successfully implemented for the biplex RT-LAMP of HIV-1 & HTLV-1 and the biplex LAMP of H. ducreyi & T. pallidum. As an outlook, the novel approach holds the potential for the simultaneous detection of multiple targets, merely restricted by the number of fluorescence channels of available detection devices. To increase the degree of multiplexing MD detection may also be combined with immobilized universal reporter molecules, whereas in addition to fluorescence, detection can be based on different signal generation methods. In future, the MD technique may also be expanded to further amplification methods that use strand displacement polymerases, serving as an universal detection technique.

ASSOCIATED CONTENT

24 22

Supporting Information

20

tp (min)

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


The supporting information is available. Primer sequences; Primer concentrations in biplex assays; Optimization of MD RT-LAMP reaction conditions; Influence of LF_Medc on reaction kinetics; Specificity of RT-LAMP of HIV-1; Biplex MD LAMP of H. ducreyi and T. pallidum.

18

tp H. ducreyi tp T. pallidum

16 14

(constant 100 cp/rxn))

12

AUTHOR INFORMATION

10 0

10

1

2

10 10 H. ducreyi copies/reaction

3

10

Corresponding Author



*Phone: +49 761 20373227. E-Mail: [email protected]

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Notes The authors declare the following competing financial interest(s): A patent covering the technique described in the paper has been applied for by the University of Freiburg and HahnSchickard.

ACKNOWLEDGEMENTS This work was financially supported by BMBF EuroTransBio project “Light-Up“, Grant agreement number: 031B0132B. We gratefully acknowledge Dr. Hans Nitschko from Max von Pettenkofer-Institute, Ludwig Maximilians University (LMU), Munich, Germany and Dr. Sheila A. Lukehart, University of Washington, Seattle, Washington, USA, for providing sample material. We want to thank Dr. Simon Wadle and Dr. Michael Lehnert from the University of Freiburg for their advice and support. References (1) Notomi, T. Nucleic Acids Research 2000, 28 (12), 63e-63. (2) Li, Y.; Fan, P.; Zhou, S.; Zhang, L. Microbial pathogenesis 2017, 107, 54–61. (3) Parida, M.; Sannarangaiah, S.; Dash, P. K.; Rao, P. V. L.; Morita, K. Reviews in medical virology 2008, 18 (6), 407–421. (4) Nyan, D.-C.; Swinson, K. L. Scientific reports 2015, 5, 17925. (5) Chi, Y.; Ge, Y.; Zhao, K.; Zou, B.; Liu, B.; Qi, X.; Bian, Q.; Shi, Z.; Zhu, F.; Zhou, M.; Cui, L.; Su, C. Scientific reports 2017, 7, 44924. (6) Liu, H.; Tian, T.; Zhang, Y.; Ding, L.; Yu, J.; Yan, M. Biosensors & bioelectronics 2017, 89 (Pt 2), 710–714. (7) Oh, S. J.; Park, B. H.; Jung, J. H.; Choi, G.; Lee, D. C.; Kim, D. H.; Seo, T. S. Biosensors & bioelectronics 2016, 75, 293–300. (8) Oscorbin, I. P.; Belousova, E. A.; Zakabunin, A. I.; Boyarskikh, U. A.; Filipenko, M. L. BioTechniques 2016, 61 (1), 20–25. (9) Mori, Y.; Kitao, M.; Tomita, N.; Notomi, T. Journal of biochemical and biophysical methods 2004, 59 (2), 145–157. (10) Tomita, N.; Mori, Y.; Kanda, H.; Notomi, T. Nature protocols 2008, 3 (5), 877–882. (11) Hsieh, K.; Mage, P. L.; Csordas, A. T.; Eisenstein, M.; Soh, H. T. Chemical communications (Cambridge, England) 2014, 50 (28), 3747–3749. (12) Song, J.; Liu, C.; Mauk, M. G.; Rankin, S. C.; Lok, J. B.; Greenberg, R. M.; Bau, H. H. Clinical chemistry 2017, 63 (3), 714– 722.

(13) Liu, W.; Huang, S.; Liu, N.; Dong, D.; Yang, Z.; Tang, Y.; Ma, W.; He, X.; Ao, D.; Xu, Y.; Zou, D.; Huang, L. Scientific reports 2017, 7, 40125. (14) Nyan, D.-C.; Swinson, K. L. Scientific reports 2015, 5, 17925. (15) Ball, C. S.; Light, Y. K.; Koh, C.-Y.; Wheeler, S. S.; Coffey, L. L.; Meagher, R. J. Analytical chemistry 2016, 88 (7), 3562–3568. (16) Priye, A.; Bird, S. W.; Light, Y. K.; Ball, C. S.; Negrete, O. A.; Meagher, R. J. Scientific reports 2017, 7, 44778. (17) Tanner, N. A.; Evans, T. C. Curr Protoc Mol Biol., 2014, 105. (18) Tanner, N. A.; Zhang, Y.; Evans, T. C. BioTechniques 2012, 53 (2), 81–89. (19) Kouguchi, Y.; Fujiwara, T.; Teramoto, M.; Kuramoto, M. Molecular and cellular probes 2010, 24 (4), 190–195. (20) Wang, Y.; Li, D.; Wang, Y.; Li, K.; Ye, C. Molecules (Basel, Switzerland) 2016, 21 (1), E111. (21) Wang, Y.; Wang, Y.; Lan, R.; Xu, H.; Ma, A.; Li, D.; Dai, H.; Yuan, X.; Xu, J.; Ye, C. The Journal of molecular diagnostics : JMD 2015, 17 (4), 392–401. (22) Wang, Y.; Wang, Y.; Luo, L.; Liu, D.; Luo, X.; Xu, Y.; Hu, S.; Niu, L.; Xu, J.; Ye, C. Frontiers in microbiology 2015, 6, 1400. (23) Faltin, B.; Wadle, S.; Roth, G.; Zengerle, R.; Stetten, F. von. Clinical chemistry 2012, 58 (11), 1546–1556. (24) Wadle, S.; Lehnert, M.; Schuler, F.; Köppel, R.; Serr, A.; Zengerle, R.; Stetten, F. von. BioTechniques 2016, 61 (3), 123–128. (25) Wadle, S.; Rubenwolf, S.; Lehnert, M.; Faltin, B.; Weidmann, M.; Hufert, F.; Zengerle, R.; Stetten, F. von. Methods in molecular biology (Clifton, N.J.) 2014, 1160, 55–73. (26) Wadle, S.; Lehnert, M.; Rubenwolf, S.; Zengerle, R.; Stetten, F. von. Biomolecular detection and quantification 2016, 7, 1–8. (27) Curtis, K. A.; Rudolph, D. L.; Owen, S. M. Journal of virological methods 2008, 151 (2), 264–270. (28) PrimerExplorer V5, http://primerexplorer.jp/lampv5e/index.html. (29) SantaLucia, J. Methods in molecular biology (Clifton, N.J.) 2007, 402, 3–34. (30) Visual OMP DNA Software, http://www.dnasoftware.com/ourproducts/visual-omp/. (31) Gürtler, L. G.; Hauser, P. H.; Eberle, J.; Brunn, A. von; Knapp, S.; Zekeng, L.; Tsague, J. M.; Kaptue, L. Journal of virology 1994. (32) Morozov, V. A.; Weiss, R. A. Virology 1999, 255 (2), 279– 284. (33) Nagamine, K.; Hase, T.; Notomi, T. Molecular and cellular probes 2002, 16 (3), 223–229. (34) Martineau, R. L.; Murray, S. A.; Ci, S.; Gao, W.; Chao, S.-H.; Meldrum, D. R. Analytical chemistry 2017, 89 (1), 625–632. (35) Rees, W. A.; Yager, T. D.; Korte, J.; Hippel, P. H. von. Biochemistry 1993, 32 (1), 137–144. (36) Bustin, S. A.; Benes, V.; Garson, J. A.; Hellemans, J.; Huggett, J.; Kubista, M.; Mueller, R.; Nolan, T.; Pfaffl, M. W.; Shipley, G. L.; Vandesompele, J.; Wittwer, C. T. Clinical chemistry 2009, 55 (4), 611–622. (37) Suntoke, T. R.; Hardick, A.; Tobian, A. A. R.; Mpoza, B.; Laeyendecker, O.; Serwadda, D.; Opendi, P.; Gaydos, C. A.; Gray, R. H.; Wawer, M. J.; Quinn, T. C.; Reynolds, S. J. Sexually transmitted infections 2009, 85 (2), 97–101.


Page 8 of 9

Page 9 of 9

For TOC only

5'

Me d Me dc

mediator displacement probe

universal reporter

5' LF primer

5' Med 3' Medc

F2

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


DNA 5'

3'

1

5' rev. Primer cDNA 3'

+ rev. Primer - DNA

2

5' Med Medc 5'

5' Med 3' Medc


thermal amplification (LAMP)

Recommend Documents