Integrated Graphene Oxide Purification-Lateral Flow Test Strips (iGOP

Oct 26, 2017 - An integrated graphene oxide purification-lateral flow test strip (iGOP-LFTS) was developed for on-strip purifying and visually detecti...
46 downloads 15 Views 2MB Size
Subscriber access provided by READING UNIV

Article

Integrated graphene oxide purification-lateral flow test strips (iGOP-LFTS) for direct detection of PCR products with enhanced sensitivity and specificity Shanglin Li, Yin Gu, Yi Lyu, Yan Jiang, and Peng Liu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b02769 • Publication Date (Web): 26 Oct 2017 Downloaded from http://pubs.acs.org on October 27, 2017

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 16

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

Integrated graphene oxide purification-lateral flow test strips (iGOPLFTS) for direct detection of PCR products with enhanced sensitivity and specificity Shanglin Li,1 Yin Gu,1 Yi Lyu,2 Yan Jiang,2 Peng Liu1,* 1

Department of Biomedical Engineering, School of Medicine, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Tsinghua University, Beijing, 100084, China 2 National HIV/HCV Reference Laboratory, National Center for AIDS/STD Control and Prevention, Chinese Center for Disease Control and Prevention, Beijing, 102206, China ABSTRACT: An integrated graphene oxide purification-lateral flow test strip (iGOP-LFTS) was developed for on-strip purifying and visually detecting polymerase chain reaction (PCR) products with an improved sensitivity as well as a more stringent specificity. PCR products amplified with a pair of biotin- and digoxin-labeled primers were directly pipetted onto GO pads, on which graphene oxide selectively adsorbed residual primers and primer-dimers with the aid of a running buffer containing MgCl2 and Tween 20. By stacking up three GO pads to increase the surface area for adsorption, 83.4% of double-stranded DNA with a length of 30 bp and 98.6% of 20-nt primers could be removed from a 10-µL DNA mixture. Since no primers interfered with detection, the increase of the sample loading volume from 5 to 20 µL could improve the signal-to-noise ratio of the test line 1.6 fold using the iGOP-LFTS while no changes were observed using the conventional LFTS. The limit of detection of the iGOP-LFTS was determined to be 30 copies of bacteriophage λ-DNA with naked eyes and this limit could be further decreased to 3 copies by loading 20 µL of the sample, which corresponded to a 1000-fold improvement compared to that of the LFTS detected by naked eyes. When the ImageJ analysis was employed, a 100-fold decrease of the detection limit can be obtained. In addition, due to the removal of the primer-dimers, the dim test line observed in the negative control of the LFTS was eliminated using the iGOP-LFTS. A mock clinical specimen spiked with defective HIV-1 (human immunodeficiency virus) viruses was successfully analyzed using a two-step reverse transcription-PCR with 30 amplification cycles followed by the iGOPLFTS detection. These significant improvements were achieved without introducing any additional hands-on operations and instrumentations.

A sensitive yet easy-to-use detection method is indispensable for realizing nucleic acid testing (NAT) in point-of-care diagnostics.1,2 While many detection strategies, such as real-time polymerase chain reaction (PCR),3 electrophoresis,4 microarray,5 mass spectrometry,6 and DNA sequencing,7 have been routinely employed for nucleic acid analysis, the dependence on bulky instruments and the complicated operations together with high costs in reagents and consumables make these methods not suitable for using outside clinical laboratories. Lateral flow test strip (LFTS) has been recognized as a promising detection approach due to its advantages of colorimetric analysis with naked eyes, simple operations that can be performed by untrained personnel, and the compact sizes of the paper strips.8-10 One of the most straightforward methods for detecting nucleic acids with LFTS was achieved by employing biotin and fluorescein isothiocyanate (FITC) labeled primers for PCR amplification and then by detecting the double-labeled amplicons with anti-FITC antibodies on a test strip.11 Since then, this double labeling scheme as well as the variants have been widely

employed as they were easy to realize and required no additional hands-on preparations for detection.12,13 However, this method has two critical problems: first, the residual primers can compete with the amplicons for the binding sites of the test line on the test strip, resulting in a low sensitivity. Second, the inevitable primerdimers containing both labels of the primer pair can produce false positive results.11 Several signal amplification approaches have been developed to improve the sensitivity of the LFTS, including fluorescent detection,14,15 electrochemical sensing,16 enzymatic amplification,17 etc. Meanwhile, antibody-free lateral flow devices have been successfully invented for the detection of nucleic acids. These hybridization-based test strips can to some extent eliminate the interferences of residual primers and primer-dimers due to the additional sequence recognition by capture oligos.18,19 However, the abovementioned methods usually either introduce additional manual operations, such as DNA denaturation and oligo incubation, or depend on extra instrumentations, undermining the “dip-to-use” concept of the LFTS.20

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

Recently, a variety of nucleic acid amplification methods have been developed to couple with lateral flow test strips for rapid genetic analysis with low costs.21-23 For examples, strand displacement amplification mediated by aptamers was analyzed using test strips, realizing a sensitivity of 10 CFU (colony forming unit) of S. enteritidis.24 Loop mediated amplification (LAMP) was integrated with test strips in a cartridge for the detection of Mycobacterium tuberculosis genomic DNA.25 Unfortunately, in contrast to PCR, these amplification methods were still not widely adopted in clinical diagnosis owing to the lack of the versatility in amplifying various genetic targets, the proneness to contamination, and the lengthy process of optimization. Therefore, a lateral flow test strip that can take the advantages of PCR while eliminating the interference of primers and primerdimers is highly desired. Here we describe an integrated graphene oxide purification-lateral flow test strip (iGOP-LFTS), on which graphene oxide (GO) is immobilized on three stacked GO pads to purify PCR products directly, leading to enhanced sensitivity and specificity. It has been well known that graphene oxide has a much stronger adsorption to single-stranded DNA (ssDNA) than that to doublestranded DNA (dsDNA).26,27 A variety of biosensors based on the high affinity of GO to ssDNA have been reported with a high sensitivity at a low cost.28 More recently, Huang et al. demonstrated the separation of ssDNA from dsDNA in a liquid phase using graphene oxide by centrifugation.29 Similarly, we believe GO immobilized on the LFTS should be able to remove residual single-stranded primers from PCR products. Moreover, it has been reported that dsDNA could also bind to GO forming dsDNA/GO complex in the presence of certain cations, such as Mg2+, which reduce the electrostatic repulsion between dsDNA and GO and bind to both dsDNA and carboxylic groups on GO.3032 Stacked chemically converted graphene sheets can also capture dsDNA under the influence of salts.33 By contrast, some surfactants, such as Tween 20 and Triton X-100, strongly interfere with the formation of dsDNA/GO complexes.30,34 We deduced that the short primer-dimers (usually less than 40 bp) might be selectively adsorbed by GO from PCR products under a carefully optimized condition. As a result, this new test strip coupled with the on-strip GO purification should be able to overcome the troubles associated with the conventional antibody-based method without causing any inconvenience to the overall detection process. We believe the iGOP-LFTS will play a significant role in point-of-care diagnosis in the future.

EXPERIMENTAL SECTION Reagents. Streptavidin (S4762), Tween 20, sodium azide, and tetrachloroauric acid were purchased from Sigma-Aldrich (St. Louis, MO). Graphene oxide, agarose, and bovine serum albumin (BSA) were obtained from Aladdin Bio-Chem (Shanghai, China). Sodium chloride (NaCl), magnesium chloride (MgCl2), and TE buffer (10 mM Tris, 1 mM EDTA, pH=8.0) were from Sangon (Shanghai, China). All chemicals were of analytical grade. All of the primers and the HIV-1 (human immunodeficiency virus) plasmids were synthesized by Sangon. Anti-digoxin antibody (ab20814) was purchased from Abcam (Cambridge, MA). Bacteriophage λ-DNA was obtained from Promega (Madison, WI). AmpliTaq Gold® 360 Master Mix was from Thermo Fisher (Waltham, MA). A 20-bp DNA ladder (3420A) and a GeneGreen nucleic acid dye kit were purchased from TaKaRa (Shiga, Japan) and Tiangen Biotech (Beijing, China), respectively. Nitrocellulose membrane (NC membrane, HF13502S25) was obtained from Merck Millipore (Massachusetts, MA). Glass fiber membrane, absorbent paper, and plastic backing card were all from Shanghai Kinbio Tech (Shanghai, China).

Page 2 of 16

Synthesis and labeling of gold nanoparticles. The 15-nmdiameter gold nanoparticles (AuNPs) were synthesized using the Frens method which can be found elsewhere.35 Briefly, 100 mL of 0.01% (w/v) tetrachloroauric acid solution was prepared and heated to boil. Then, 4.5 mL of 1% (w/v) sodium citrate solution was added with a vigorous mechanical stirring. When the color of the solution turned into wine red, the heating was stopped and the solution was continuously stirred for another 10 min. After that, the solution was cooled to room temperature and stored in a dark bottle at 4 oC. For labeling the AuNPs with anti-digoxin antibodies, the pH of the AuNP solution was first adjusted to 8.5 using 200-mM K2CO3. Then, the labeling reaction was carried out by adding 10 µL of the anti-digoxin antibody solution at a concentration of 5.78 mg/mL to 5 mL of the AuNPs solution, followed by an incubation at 37 oC for an hour. After that, a 100-µL sealing solution (10% BSA and 20 mM Na4B2O7) was added into the reaction and incubated at room temperature for 0.5 hour to stabilize the gold nanoparticles. Finally, the excess reagents were removed by centrifugation for 15 min at 18,000 g, and the AuNPs was dispersed again with a suspension solution (0.1 M Tris, 10% sucrose, 5% BSA, 0.25% Tween 20, 0.05% NaN3 pH=8.0). Preparation of graphene oxide pads and lateral flow test strips. Sample loading pad: a glass fiber membrane with dimensions of 50×15×0.3 mm was soaked into a PBS solution containing 100 mM NaCl and 0.25% Tween 20 for 3 hours and then dried at 65 oC for 4 hours. After that, the treated glass fiber membrane was stored in a desiccator until use. Graphene oxide (GO) pad: first, 1% (w/v) GO stock solution was diluted by absolute ethyl alcohol into 0.25%. Then, 400 µL of the 0.25% GO solution was pipetted over a glass fiber membrane with dimensions of 50×15×0.3 mm (corresponding to 0.06 mg on a 3×15×0.3 mm pad) and dried at 50 oC for 3 hours. Finally, a 150µL salt solution containing 20 mM MgCl2, 100 mM NaCl, and 40 mM NaAc (pH=5.0) was added onto the pad and dried at 50 oC for 0.5 hour. Conjugation pad: the AuNPs labeled with anti-digoxin antibodies were first mixed with an equal volume of a conjugation solution (20 mM Na4B2O7, 2% BSA, 3% sucrose, 600 mM NaCl, 0.2% Tween 20, and 0.05% NaN3). Then, 160 µL of the mixed solution was deposited onto a glass fiber conjugation pad with dimensions of 50×10×0.3 mm, which was dried overnight at 37 oC. Finally, the prepared conjugation pad was stored in a desiccator at 4 oC. Nitrocellulose (NC) membrane: a test and a control line were formed by manually drawing 0.24 µL/strip of 2 mg/mL streptavidin and 0.24 µL/strip of 2 mg/mL anti-mouse IgG antibodies onto the NC membrane, respectively. Then, the NC membrane was dried at 37 oC overnight and stored in a desiccator until use. Assembly of lateral flow test strips: to assemble a conventional lateral flow test strip, a NC membrane was first placed on a plastic adhesive backing. Then, a conjugation pad was placed next to the NC membrane with a 2-mm overlap. A sample loading pad was placed on the backing with a 2-mm overlap with the conjugation pad. Finally, an absorbent pad was pasted on the other side of the NC membrane with a 2-mm overlap. After assembly, test strips were cut to 3 mm wide and stored in a desiccator at 4°C until use. The assembly of the iGOP-LFTS is similar to that of the conventional test strip. The only difference is that three GO pads were stacked up to replace the sample loading pad on the strip.

ACS Paragon Plus Environment

Page 3 of 16

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

Preparation of PCR samples. For the gel electrophoretic analysis, a pair of primers (FAM-labeled FFP and NRP listed in Table S1) was employed to amplify a 106-bp fragment from λDNA. For the LFTS detection, another pair of primers (BFP and DRP in Table S1), labeled with biotin and digoxin, respectively, was used for PCR from λ-DNA. A 20-µL PCR mixture was composed of 0.4 µL of each primer (10 µM), 10 µL of 2× AmpliTaq Gold® 360 Master Mix, 1 µL of λ-DNA with different concentrations, and 8.2 µL of DI water. The thermal cycling protocol included an initial activation of Taq polymerases at 95 oC for 5 min, followed by 35 cycles of 95 oC for 30 s, 60 oC for 30 s, and 72 oC for 30 s, and a final extension step at 72 oC for 5 min. Preparation of standard DNA mixture. The 30-bp doublestranded DNA labeled with FAM was prepared as follows: first, two 30-nt oligos (FAM-labeled OA and unlabeled OB in Table S1) were synthesized by Sangon and dissolved in 1× TE buffer to a concentration of 10 µM. Then, the equal volumes of both oligos were mixed and heated to 95 oC for 5 min followed by cooling to room temperature at a rate of 1 oC/min. The solution was stored at 4 oC. To prepare a standard DNA mixture for testing, the 106-bp amplicons obtained from 1000 copies of λ-DNA with 35 PCR cycles, the 30-bp dsDNA (5 µM) shown above, and the FAMlabeled primer (10 µM) were mixed together in a volume ratio of 47: 1: 2.

incubated at 37 oC for 15 min, followed by enzyme inactivation at 98 oC for 5 min. In the PCR step, 1 µL of the RT product was used for amplification using the same protocol as that in the HIV1 plasmid analysis with either 30 or 35 PCR cycles. Quantitative analysis of gel electrophoreses and lateral flow test strips. The gel electrophoresis images were taken by the Azure c150 Biosystem (Dublin, CA) and the intensities of the DNA bands were measured using the ImageJ software. The images of the lateral flow test strips were taken by a digital camera, and then the signal intensities along the central crosssection line of the test strip were measured using the ImageJ. The peak areas of the test and the control line were calculated and the peak area ratio between these two lines was subsequently obtained. The signal-to-noise ratio (S/N) of the test line was also calculated by dividing the peak height of the test line with the standard deviation of the baseline signals from the central crosssection line. All the experiments were independently repeated three times and the results were expressed as means ± standard deviation.

Optimization of running buffer and immobilized GO. To optimize the running buffer and the immobilized GO, 10 µL of the standard DNA mixture was pipetted onto a GO pad, immediately followed by the loading of 10 µL of the running buffer. After that, the purified sample was aspirated out from the side of the GO pad without any incubation. The samples were then analyzed using agarose gel electrophoresis along with the 20bp DNA ladder. Preparation of HIV-1 samples. The plasmids containing HIV-1 gag proviral DNA sequences were provided by Sangon and a pair of primers (SK38 and SK39) was synthesized for amplification.36 A 20-µL PCR mixture was composed of 1 µL of each primer (10 µM), 10 µL of 2× AmpliTaq Gold® 360 Master Mix, 1 µL of the HIV-1 plasmids with different concentrations, and 7 µL of DI water. The thermal cycling protocol was the same as that of the λDNA amplification except that the cycle number was increased to 40. The 8E5 cell line carrying a single defective proviral genome of HIV-1 was provided by the National HIV/HCV Reference Laboratory of the Chinese Center for Disease Control and Prevention. 8E5 cells were cultured in complete RPMI 1640 medium (Gibco, Grand Island, NY) supplemented with 10% (v/v) fetal bovine serum (PAA Laboratories, Pasching, Austria) and 1% (v/v) penicillin/streptomycin (Invitrogen, Carlsbad, CA) at 37 °C with saturated humidity and 5% CO2. When the cells were grown to about 106 cells/mL, the culture medium containing about 108 copies of defective HIV-1 viruses per µL was collected and centrifuged at 1000 rpm for 5 min to remove cell debris. The defective virus has a single-base mutation in the pol gene that precludes the expression of reverse transcriptase and integrase. A mock clinical specimen was prepared by spiking healthy human plasma (obtained with informed consent) with the viruses to a concentration of ~100 viruses/µL. A two-step reverse transcription-PCR (RT-PCR) was performed to amplify from the gag gene with the SK38 and SK39 primer pair. In the first step, a 10-µL reverse transcription mixture containing 1 µL of spiked human plasma and 9 µL of ReverTra Ace® qPCR RT mix (0.5 µL of Enzyme Mix, 0.5 µL of Primer Mix, 2 µL of 5× RT buffer, and 6 µL DI water) (FSQ-101, TOYOBO, Osaka, Japan) was

Figure 1. Schematic of the integrated graphene oxide purification- lateral flow test strip (iGOP-LFTS). (A) Structure of the iGOP-LFTS. (B) Direct loading of PCR products onto the stacked GO pad. (C) Purification and detection mechanism of the iGOP-LFTS.

RESULTS AND DISCUSSION Purification and detection mechanisms of iGOP-LFTS. Although diverse amplification methods of nucleic acids have been developed with many superb properties, such as isothermal reaction, ultra-high sensitivity, simplified sample preparations, etc., polymerase chain reaction remains one of the best choices for clinical diagnosis.37,38 A molecular diagnosis kit based on PCR can be easily designed and optimized to provide a robust amplification performance with a high specificity. As a result, a detection method that requires minimum modifications to the PCR system is highly desired. In our study, PCR was carried out using a pair of primers, which were labeled with biotin and

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

digoxin, respectively. This simple labeling strategy preserved the advantages of PCR and enabled the direct use of PCR products for detection. As illustrated in Figure 1, the iGOP-LFTS was comprised of four overlaid pads: three stacked GO pads, a conjugation pad, a nitrocellulose membrane, and an absorbent pad. The PCR products were directly loaded onto the GO pads, on which the primers and the primer-dimers were selectively adsorbed by graphene oxide. Right after the sample loading, a 100-µL running buffer (0.1% (w/v) Tween 20, 20 mM MgCl2, 200 mM NaCl, and 80 mM NaAc (pH=5.0)) was pipetted onto the GO pad to drive the purified amplicons to the conjugation pad, where the gold nanoparticles with anti-digoxin antibodies bound to the amplicons via the antibody-antigen interaction. The formed AuNP-DNA complexes as well as the excess AuNPs were further transferred to the NC membrane, which contained a test (T line) and a control line (C line). On the test line, the immobilized streptavidin captured the biotin-end of the amplicons, leading to the aggregation of the AuNPs. On the control line, anti-mouse IgG antibodies captured the excess AuNPs, validating the proper function of the lateral flow test strip. Since no incubation was involved, the entire process took only 10-15 min.

Page 4 of 16

Optimization of graphene oxide purification. To optimize the on-strip purification of PCR products, we envisioned that the running buffer should play an important role as the large amount of the buffer provided a liquid phase for the selective DNA adsorption by graphene oxide. A DNA mixture containing FAMlabeled double- and single-stranded DNA (106-bp amplicons, 30bp synthesized dsDNA, and 20-nt primers shown in Table S1) was employed as a standard sample for testing. A series of the running buffers containing different concentrations of MgCl2 and Tween 20 was evaluated for their purification efficiencies. As illustrated in Figure S1 and S2, the gel electrophoreses of the recovered samples from the GO pads followed by the quantitative analyses determined that the combination of 20 mM MgCl2 and 0.1% (w/v) Tween 20 yielded the best efficiency in terms of the amplicon-to-primer ratio of recovery. Next, we optimized the amount of GO pre-dried on the pad and found that, generally speaking, more graphene oxide in the pads provided higher purification efficiencies. As shown in Figure 2, 0.12 mg of GO on a 15×3×0.3 mm pad can remove 97.5% of the 20-nt singlestranded primers from a 10-µL DNA mixture. However, since GO was simply pre-dried within the pads (Figure S3), such a large amount of GO could be washed off by the running buffer to interfere with the downstream detection on the strip. As a result, the immobilization amount of 0.06 mg was chosen due to its acceptable efficiency (the amplicon-to-primer ratio = 5.1) and the negligible interference with the detection. To further enhance the purification capability of the strip, we stacked up multiple graphene oxide pads to increase the surface area for adsorption while maintaining the concentration of GO in the pads. Figure 3 demonstrated that 83.4% of the 30-bp dsDNA and 98.6% of the primers could be removed from the 10-µL DNA mixture using three layers of the GO pads. Although 39.5% of the 106-bp amplicons were lost, the amplicon-to-dsDNA and the amplicon-to-primer ratios went up to 3.6 and 43.0, respectively. Therefore, this three-layer design was adopted for purifying PCR products in the rest of the experiments.

Figure 2. Purification efficiency of different amounts of graphene oxide immobilized in the pads. (A) Gel electrophoresis image of recovered DNA mixtures from the GO pads. (B) Relative band intensities of the 106-bp amplicons, the 30-bp dsDNA, and the 20-nt primers to the positive control (PC). When 0.12 mg of GO was immobilized on the pad, the best purification can be achieved with an amplicon-to-primer ratio of 29.0. Unfortunately, such a large amount of GO on the pad could be washed out to interfere with the downstream detection. Therefore, 0.06 mg of GO was immobilized on a pad for the rest of the study.

ACS Paragon Plus Environment

Page 5 of 16

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

Figure 3. Purification efficiency of one, two, or three stacked graphene oxide pads. (A) Gel electrophoresis image of recovered DNA mixtures from stacked GO pads. (B) Relative band intensities of the 106-bp amplicons, the 30-bp dsDNA, and the 20-nt primers to the positive control. Up to 83.4% of the 30-bp dsDNA and 98.6% of the 20-nt primers were adsorbed by three stacked GO pads. Sample volume effect to sensitivity. Since the residual primers were eliminated by GO, the increase of the sample loading volume should lead to the aggregation of more gold nanoparticles on the test line, resulting in an improve sensitivity of the iGOPLFTS. First, as a comparison, the conventional LFTS without the GO pad was employed to detect 5, 10, and 20 µL of PCR products amplified from 1000 copies of λ-DNA with the biotin- and the digoxin-labeled primer set (Table S1). As shown in Figure 4A, the dim T lines on all the test strips demonstrated no changes or even slightly decreases of the intensities with the increase of the sample volumes, which were confirmed by the quantitative analyses (Figure 4B). This is because larger volumes of the samples without purification have more primers, leading to the saturation of AuNPs and streptavidin. By contrast, the iGOP-LFTS containing three stacked GO pads can eliminate all residual primers, resulting in significantly increased signals on the T lines (Figure 4A). The four-fold volume increase (5 to 20 µL) led to 1.6-fold increase of the signal-to-noise ratios of the test lines and 3-fold increase of the peak area ratios of the test to the control line, proving the effectiveness of this on-strip purification method (Figure 4B).

Figure 4. Signal improvement by increasing sample loading volumes. (A) Photos of the LFTS and the iGOP-LFTS with different sample loading volumes (5, 10, and 20 µL). (B) Peak area ratio of test to control line and signal-to-noise ratio of test line as functions of the sample loading volume. When the loading volumes were increased from 5 to 20 µL, the peak area ratio and the S/N were improved 3 and 1.6 fold, respectively, using the iGOP-LFTS. Limit of detection. To test the sensitivity of the iGOP-LFTS, we prepared a series of PCR products amplified from 3000 down to 3 copies of λ-DNA. As illustrated in Figure 5A, the conventional LFTS without the GO purification can only reach a sensitivity of 3000 copies of template determined by naked eyes. When the iGOP-LFTS device with three layers of the GO pads was used to analyze 10-µL PCR products, the limit of detection was decreased to 30 copies, leading to a 100-fold improvement in the sensitivity. Interestingly, when 20 µL of the PCR products was loaded onto the strip, the sensitivity of iGOP-LFTS detected by naked eyes can reach 3 copies of template. By contrast, the sensitivity of the conventional LFTS cannot be improved with more samples due to the competence of residual primers. As a result, a remarkable 1000-fold improvement in sensitivity can be obtained with the visual detection. In addition, with the aid of an image analysis software, the sensitivities of both the LFTS and the iGOP-LFTS could be determined to be 300 and 3 copies, respectively, as shown in Figure 5B (S/N >3), corresponding to a 100-fold increase of the limit of detection.

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

Page 6 of 16

resolve the problem of the false positives (S/N=1.9) and provide results without any ambiguity.

Figure 5. Limit of detection of the conventional LFTS and the iGOP-LFTS. (A) Photos of the strips tested using PCR products amplified from 3-3000 copies of λ-DNA. The sample loading volumes (10 or 20 µL) were listed above the photos. * indicates the limits of detection of the strips determined by naked eyes. (B) Signal-to-noise ratios of the test lines analyzed with ImageJ. When 20 µL of PCR products was loaded onto the iGOP-LFTS, the limit of detection was 3 copies of template, corresponding to a 1000-fold increase compared to the conventional LFTS. * indicates the limits of detection of the strips determined by image analysis software (S/N >3). Enhanced specificity. The false positives caused by primer-dimer artifacts is one of the most critical drawbacks of the lateral flow test strip employing the double-labeling scheme for PCR amplification and detection. To evaluate the capability of the iGOP-LFTS of resolving the primer-dimer problem, we tested the visual detection of PCR products amplified from plasmids containing HIV-1 gag proviral DNA sequences with a slight primer-dimer issue (primers listed in Table S1).36 Although the nucleic acid testing of HIV has a shorter window period than that of the HIV immunoassay,39 the wide adoption of NAT in clinical diagnosis was still limited due partially to the complicated operations and the high costs which could be alleviated by the use of lateral flow test strips. Here we prepared a series of PCR products amplified from the HIV-1 plasmids from 300 down to 3 copies with 40 cycles. As shown in Figure 6, although the sensitivity of the LFTS without the GO pads can reach the detection limit of 3 copies, a dim T line on the strip (S/N=5.5) appeared in the negative control. This false positive result was produced by the primer-dimers generated during the amplification. When the iGOP-LFTS was employed, most of the primer-dimers were removed by the three stacked GO pads on the strip and no false positive was observed, validating the effectiveness of the on-strip purification. The quantitative analyses of these strips confirmed that the iGOP-LFTS could

Figure 6. Visual detection of HIV-1 plasmids using the conventional LFTS and the iGOP-LFTS. (A) Photos of the strips tested using PCR products amplified from 3-300 copies of HIV-1 plasmids. * indicates the limits of detection determined by naked eyes. A dim test line was observed in the negative control of the LFTS. (B) Signal-to-noise ratios of the test lines. The S/N in the negative control of the LFTS was higher than 3, while that in the iGOP-LFTS is far less than 3. * indicates the limits of detection of the strips determined by image analyses (S/N >3). Analysis of mock clinical specimen. To more critically evaluate the capability of our iGOP-LFTS for HIV testing, we prepared a mock clinical specimen by spiking healthy human plasma with defective HIV-1 viruses to a concentration of ~100 viruses/µL. One-µL specimen was amplified using a two-step RT-PCR protocol with either 30 or 35 PCR cycles. The products were then analyzed using the conventional LFTS and the iGOP-LFTS in parallel. As shown in Figure 7A, the iGOP-LFTS revealed clear test lines on the strips for both the PCR products amplified with 30 and 35 cycles, while the conventional LFTS can only detect the sample prepared with 35 PCR cycles. The ImageJ analyses shown in Figure 7B confirmed that the S/N ratios of all the positive test lines were higher than 3. This test with a mock clinical specimen clearly proved the iGOP-LFTS could be used to detect HIV-1 viruses existing in human plasma.

ACS Paragon Plus Environment

Page 7 of 16

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 the iGOP-LFTS can significantly improve the robustness of the LFTS-based nucleic acid testing.

CONCLUSION In summary, we have successfully developed an integrated graphene oxide purification-lateral flow test strip for the detection of PCR products directly. Since the residual primers and the primer-dimer artifacts can be effectively removed by three stacked GO pads on the strip, the sensitivity of the test strip was improved up to 1000 fold in visual detection, and the chance of obtaining false positive results was dramatically reduced. More importantly, these improvements were achieved without introducing any additional hands-on operations and instruments. The “dip-to-use” concept of the lateral flow test strip was well preserved by this integrated PCR purification and lateral flow detection method. Our iGOP-LFTS provides a sensitive, easy-touse, and inexpensive detection approach for checking PCR products in a qualitative or semi-quantitative way. The combination of the iGOP-LFTS with direct PCR (no nucleic acid extraction required) will allow the patient’s diagnoses to be performed in a physician’s office, an ambulance, the home, and the field by users with minimum trainings. In the future, it is possible to further enhance the sensitivity and the specificity of the iGOP-LFTS by permanently immobilizing GO within the pad and by carefully selecting appropriate materials of the pads for GO modification. In addition, the integration of the test strip with nucleic acid extraction and PCR amplification on a single microfluidic device will resolve the contamination concern of the current “open-tube” detection by the strip, realizing a “sample-in-answer-out” analysis.

ASSOCIATED CONTENT Figure 7. Analysis of spiked clinical specimens containing defective HIV-1 viruses using the conventional LFTS and the iGOP-LFTS. (A) Photos of the strips tested using RT-PCR products amplified from ~100 copes of HIV-1 viruses. * indicates the limits of detection observed by naked eyes. (B) Signal-tonoise ratios of the test lines. The conventional LFTS can only detect the PCR products amplified with 35 cycles, while the iGOP-LFTS allowed the cycle number was reduced to 30. * indicates the limits of detection analyzed by the ImageJ software.

Supporting Information

Since the lateral flow test strip still depends on the conventional PCR for sample preparations, PCR parameters, such as primer design, annealing temperature, and cycle number, can significantly affect the performance of the strip. For example, the tests shown in Figure 6 demonstrated that both the conventional LFTS and the iGOP-LFTS can detect 3 copies of HIV-1 plasmids amplified in a 20-µL reaction with 40 PCR cycles. However, the improvement in sensitivity achieved by the cycle number sacrificed the assay specificity. Primer-dimers or non-specific amplifications did occur, causing false positives in the LFTS test. Therefore, a PCR condition that keeps the balance between the sensitivity and the specificity must be carefully optimized for the conventional LFTS. In contrast, the iGOP-LFTS has the capability of eliminating residual primers and primer-dimers. As a result, a PCR condition that is more towards the sensitivity could be chosen. For instance, the iGOP-LFTS can detect 3 copies of HIV-1 plasmids with a 40-cycle amplification while the interference of the primer-dimers was eliminated. In addition, Figure 7 illustrated that the iGOP-LFTS can successfully analyze the mock clinical specimen amplified with only 30 PCR cycles. Not only did the reduced cycle number lower the chance of generating non-specific products but also shortened the analytical time. Overall, the enhanced sensitivity and specificity provided by

Notes

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author * Email: [email protected]. Phone: +86-10-62798732. Fax: +86-10-62798732.

The authors declare no competing financial interest.

ACKNOWLEDGEMENTS Financial supports were provided by the National Key Research and Development Program of China (No. 2016YFC0800703) from the Ministry of Science and Technology of China.

REFERENCES (1) St John, A.; Price, C. P. Clin. Biochem. Rev. 2014, 35, 155167. (2) Pai, N. P.; Vadnais, C.; Denkinger, C.; Engel, N.; Pai, M. PLoS Med. 2012, 9, e1001306. (3) Bustin, S. A. J. Mol. Endocrinol. 2002, 29, 23-39. (4) Zhuang, B.; Gan, W.; Wang, S.; Han, J.; Xiang, G.; Li, C. X.; Sun, J.; Liu, P. Anal. Chem. 2015, 87, 1202-1209. (5) Wang, S.; Sun, Y.; Gan, W.; Liu, Y.; Xiang, G.; Wang, D.; Wang, L.; Cheng, J.; Liu, P. Biomicrofluidics 2015, 9, 024102. (6) Ross, P.; Hall, L.; Smirnov, I.; Haff, L. Nat. Biotechnol. 1998, 16, 1347-1351. (7) Goodwin, S.; McPherson, J. D.; McCombie, W. R. Nat. Rev. Genet. 2016, 17, 333-351.

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

(8) Posthuma-Trumpie, G. A.; Korf, J.; van Amerongen, A. Anal. Bioanal. Chem. 2009, 393, 569-582. (9) Zhou, W.; Gao, X.; Liu, D.; Chen, X. Chem. Rev. 2015, 115, 10575-10636. (10) Cordeiro, M.; Ferreira Carlos, F.; Pedrosa, P.; Lopez, A.; Baptista, P. V. Diagnostics 2016, 6, 43. (11) Kozwich, D.; Johansen, K. A.; Landau, K.; Roehl, C. A.; Woronoff, S.; Roehl, P. A. Appl. Environ. Microbiol. 2000, 66, 2711-2717. (12) Kim, Y. T.; Jung, J. H.; Choi, Y. K.; Seo, T. S. Biosens. Bioelectron. 2014, 61, 485-490. (13) Seidel, C.; Peters, S.; Eschbach, E.; Fessler, A. T.; Oberheitmann, B.; Schwarz, S. Vet. Microbiol. 2017, 200, 101106. (14) Corstjens, P. L. A. M.; Zuiderwijk, M.; Nilsson, M.; Feindt, H.; Niedbala, R. S.; Tanke, H. J. Anal. Biochem. 2003, 312, 191200. (15) Wang, Y. H.; Nugen, S. R. Biomed. Microdevices 2013, 15, 751-758. (16) Liu, G.; Lin, Y. Y.; Wang, J.; Wu, H.; Wai, C. M.; Lin, Y. Anal. Chem. 2007, 79, 7644-7653. (17) Rastogi, S. K.; Gibson, C. M.; Branen, J. R.; Aston, D. E.; Branen, A. L.; Hrdlicka, P. J. Chem. Commun. 2012, 48, 77147716. (18) Aveyard, J.; Mehrabi, M.; Cossins, A.; Braven, H.; Wilson, R. Chem. Commun. 2007, 4251-4253. (19) Mao, X.; Ma, Y.; Zhang, A.; Zhang, L.; Zeng, L.; Liu, G. Anal. Chem. 2009, 81, 1660-1668. (20) Yin, R.; Sun, Y.; Yu, S.; Wang, Y.; Zhang, M.; Xu, Y.; Xue, J.; Xu, N. Food Control 2016, 60, 146-150. (21) Rohrman, B. A.; Richards-Kortum, R. R. Lab Chip 2012, 12, 3082-3088. (22) Wu, W.; Zhao, S.; Mao, Y.; Fang, Z.; Lu, X.; Zeng, L. Anal. Chim. Acta 2015, 861, 62-68. (23) Jung, J. H.; Park, B. H.; Oh, S. J.; Choi, G.; Seo, T. S. Lab Chip 2015, 15, 718-725. (24) Fang, Z.; Wu, W.; Lu, X.; Zeng, L. Biosens. Bioelectron. 2014, 56, 192-197. (25) Roskos, K.; Hickerson, A. I.; Lu, H. W.; Ferguson, T. M.; Shinde, D. N.; Klaue, Y.; Niemz, A. Plos One 2013, 8, e69355. (26) Lu, C. H.; Yang, H. H.; Zhu, C. L.; Chen, X.; Chen, G. N. Angew. Chem. Int. Ed. 2009, 48, 4785-4787. (27) Park, J. S.; Goo, N. I.; Kim, D. E. Langmuir 2014, 30, 12587-12595. (28) Gao, L.; Lian, C. Q.; Zhou, Y.; Yan, L. R.; Li, Q.; Zhang, C. X.; Chen, L.; Chen, K. P. Biosens. Bioelectron. 2014, 60, 22-29. (29) Huang, P. J. J.; Liu, J. Nanomaterials 2013, 3, 221-228. (30) Lei, H.; Mi, L.; Zhou, X.; Chen, J.; Hu, J.; Guo, S.; Zhang, Y. Nanoscale 2011, 3, 3888-3892. (31) Xing, X. J.; Xiao, W. L.; Liu, X. G.; Zhou, Y.; Pang, D. W.; Tang, H. W. Biosens. Bioelectron. 2016, 78, 431-437. (32) Tang, L.; Chang, H.; Liu, Y.; Li, J. Adv. Funct. Mater. 2012, 22, 3083-3088. (33) Liu, M.; Zhao, H. M.; Chen, S.; Yu, H. T.; Quan, X. Chem. Commun. 2012, 48, 564-566. (34) Kim, S.; Park, C.; Gang, J. J. Nanosci. Nanotechnol. 2015, 15, 7913-7917. (35) Xia, H.; Xiahou, Y.; Zhang, P.; Ding, W.; Wang, D. Langmuir 2016, 32, 5870-5880. (36) Ou, C. Y.; Kwok, S.; Mitchell, S. W.; Mack, D. H.; Sninsky, J. J.; Krebs, J. W.; Feorino, P.; Warfield, D.; Schochetman, G. Science 1988, 239, 295-297. (37) Li, J.; Macdonald, J. Biosens. Bioelectron. 2015, 64, 196211. (38) Deng, H.; Gao, Z. Anal. Chim. Acta 2015, 853, 30-45. (39) Pilcher, C. D.; Christopoulos, K. A.; Golden, M. J. Infect. Dis. 2010, 201, S7-S15.

ACS Paragon Plus Environment

Page 8 of 16

Page 9 of 16

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

Figure 1 83x98mm (300 x 300 DPI)

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

Figure 2 76x117mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 10 of 16

Page 11 of 16

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

Figure 3 76x112mm (300 x 300 DPI)

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

Figure 4 77x112mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 12 of 16

Page 13 of 16

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

Figure 5 78x108mm (300 x 300 DPI)

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

Figure 6 78x112mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 14 of 16

Page 15 of 16

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

Figure 7 77x120mm (300 x 300 DPI)

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

Table of Contents Graphic 79x38mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 16 of 16