Direct colorimetric detection of unamplified pathogen DNA by dextrin-capped gold nanoparticles

Biosensors and Bioelectronics 101 (2018) 29–36

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Direct colorimetric detection of unamplified pathogen DNA by dextrincapped gold nanoparticles

MARK

Amy M. Baetsen-Younga, Matthew Vasherb, Leann L. Mattab, Phil Colgana, ⁎ ⁎⁎ Evangelyn C. Alociljab, , Brad Daya,c, a b c

Department of Plant, Soil and Microbial Sciences, Michigan State University, East Lansing, MI, United States Department of Agricultural Biosystems Engineering, Michigan State University, East Lansing, MI, United States Plant Resilience Institute, Michigan State University, East Lansing, MI, United States

A R T I C L E I N F O

A B S T R A C T

Keywords: Dextrin gold nanoparticles DNA Salt Stabilization Genomic Pathogen

The interaction between gold nanoparticles (AuNPs) and nucleic acids has facilitated a variety of diagnostic applications, with further diversification of synthesis match bio-applications while reducing biotoxicity. However, DNA interactions with unique surface capping agents have not been fully defined. Using dextrincapped AuNPs (d-AuNPs), we have developed a novel unamplified genomic DNA (gDNA) nanosensor, exploiting dispersion and aggregation characteristics of d-AuNPs, in the presence of gDNA, for sequence-specific detection. We demonstrate that d-AuNPs are stable in a five-fold greater salt concentration than citrate-capped AuNPs and the d-AuNPs were stabilized by single stranded DNA probe (ssDNAp). However, in the elevated salt concentrations of the DNA detection assay, the target reactions were surprisingly further stabilized by the formation of a ssDNAp-target gDNA complex. The results presented herein lead us to propose a mechanism whereby genomic ssDNA secondary structure formation during ssDNAp-to-target gDNA binding enables d-AuNP stabilization in elevated ionic environments. Using the assay described herein, we were successful in detecting as little as 2.94 fM of pathogen DNA, and using crude extractions of a pathogen matrix, as few as 18 spores/µL.

1. Introduction In many parts of the world, emerging diseases account for huge losses in human life, crops, and livestock, and thus, rapid, accurate and reliable monitoring technologies are needed to prevent further impacts on human, plant, and animal health (Foley et al., 2011; Rodrigues et al., 2017). At present, molecular- and biochemical-based techniques, such as PCR and ELISA, are arguably the most reliable methods for the identification of plant and pathogen traits (Kong et al., 2016; Tsui et al., 2011; Wang et al., 2015). Additionally, recent advances in genomeenabled technologies have facilitated the development of probes to rapidly identify genetic markers for trait identification of some of the most devastating pathogens of humans and plants, including Phytophthora infestans (potato; Hussain et al., 2014), E. coli O157 (Desmarchelier et al., 1998), Magnaporthe oryzae (rice; Sun et al., 2015), and Mycobacterium tuberculosis (Drosten et al., 2003). However, while PCR-based assays offer sensitivity and specificity, they lack pointof-contact portability and functionality. In recent years, numerous nanoparticle-based assays have been

⁎

developed which facilitate the detection of both amplified and purified genomic DNA (Bakthavathsalam et al., 2012; Deng et al., 2013; Zanoli et al., 2012). By exploiting the unique properties of gold nanoparticles (AuNPs), which includes highly specific spectral absorption properties (Sepúlveda et al., 2009), their ability to adhere to DNA, and large surface to volume ratios, AuNPs have emerged as a robust assay for colorimetric biosensing and diagnostics applications, including at single nucleotide polymorphism levels (Daniel and Astruc, 2004; Li and Rothberg, 2004; Upadhyayula, 2012). For example, the use of surface plasmon resonance (SPR) to characterize the interaction between single stranded (ss) and double-stranded (ds) DNA by AuNPs in the presence of salt has illuminated an understanding of the complex association(s) between citrate ions, DNA, and AuNPs (Gearheart et al., 2001; Li and Rothberg, 2004). At a mechanistic level, ssDNA-AuNP interactions are mediated by the stabilization of the nucleotide-nanoparticle complexes in low salt concentrations. In these environments, dsDNA does not adsorb to AuNPs and they therefore aggregate from disruption of SPR. By utilizing these interactions, a variety of DNA nanobioassays have been described, each of which were designed upon salt-induced gold

Corresponding author. Corresponding author at: Department of Plant, Soil and Microbial Sciences, Michigan State University, East Lansing, MI, United States. E-mail addresses: [email protected] (E.C. Alocilja), [email protected] (B. Day).

⁎⁎

http://dx.doi.org/10.1016/j.bios.2017.10.011 Received 4 August 2017; Received in revised form 22 September 2017; Accepted 3 October 2017 Available online 07 October 2017 0956-5663/ © 2017 Elsevier B.V. All rights reserved.

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

prepared according to the method of Anderson et al. (2010). In brief, 5 mL of 20 mM HAuCl4 was added to 20 mL of 25 g/L of dextrin stock in a 250-mL flask. The pH of the solution was adjusted to 9.0 with 10% sodium carbonate (Na2CO3) and the final reaction volume was adjusted to 50 mL with sterile distilled water (pH 9.0). Particle formation occurred as the flask was incubated at 50 °C for 8 h in the dark. The synthesized nanoparticles were evaluated by TEM using a concentration of d-AuNPs of 7.6 × 10−9 M. This value was derived from Beer's Law based on a molar extinction coefficient of 2.7 × 108 M−1cm−1 for 13 nm AuNPs (Jin et al., 2003). C-AuNPs (10 nm, 9.93 × 10−9 M) were obtained from Cytodiagnostics (Ontario, Canada).

nanoparticle aggregation through the use of ssDNA probes (ssDNAp) and un-modified or functionalized AuNPs (e.g., citrate-capped AuNPs (c-AuNPs)) (Fang et al., 2013; Li and Rothberg, 2004a; Torres-Chavolla and Alocilja, 2011; Sattarahmady et al., 2015). Several technological limitations have prevented widespread adoption of AuNP-DNA nanobiosensors, including the economical and sustainable synthesis of nanoparticles for target detection (Rodrigues et al., 2017). In instances where assays have been developed, the overall detection limits are still relatively low (i.e., ca. 18 ng of genomic DNA) (D’Agata and Spoto, 2012). Additionally, with optimal reaction conditions in the low molar range of salt (i.e., ca. 0.05 M), the use of AuNPs are still limited for most point-of-care assays, as many biological salt concentrations are higher than 0.1 M (Liandris et al., 2009; Munns, 2002, Liu et al., 2015). In recent years, several of these limitations have been resolved, and the use of AuNPs in reaction conditions that parallel native biological conditions have been extended through the use of carbohydrate-coated AuNPs (glyco-AuNPs), which resulted in increased stability and uniformity of the modified nanoparticles, while decreasing the environmental biotoxicity (Anderson et al., 2010; Hastings and Eichelberger, 1937; Wang et al., 2010). In support of greener chemistries, methods have also been developed to synthesize glyco-AuNPs for use in diagnostic applications such as detection of the chemical analyte dihydralazine sulfate by dextran-AuNPs in high ionic biological mediums (Wang et al., 2010). Similarly, a recent study by Torres-Chavolla and Alocilja (2011) demonstrated DNA functionalized dextrin-AuNPs (d-AuNPs) can be used to electrochemically detect the IS16110 gene from Mycobacterium tuberculosis at concentrations as low as 0.01 ng/µL from isothermally amplified DNA (Torres-Chavolla and Alocilja, 2011). Thus, the use of glyco-AuNPs offers the potential for DNA detection in complex biological matrices from demonstrated enhanced stability (Anderson et al., 2010; Torres-Chavolla and Alocilja, 2011; Wang et al., 2010). In the current study, we used d-AuNPs to detect a unamplified DNA sequence from Pseudoperonospora cubensis, the causal agent of cucurbit downy mildew, currently the primary threat in the United States limiting cucumber production (Summers et al., 2015). As described above, previous reports have demonstrated the use of functionalized d-AuNPs to electrochemically signal DNA target capture; however, the study presented herein represents the first report of colorimetric, sequencespecific, unamplified gDNA detection using unmodified d-AuNPs. To dissect and benchmark this interaction, we compared salt induced aggregation of citrate- and dextrin-capped AuNPs with and without a ssDNAp. These basic explorations into DNA-d-AuNPs interactions enabled us to develop a sequence-specific gDNA-based detection assay utilizing the interactions of ssDNA, dsDNA and d-AuNPs in an elevated ionic environment. Using a combination of UV–vis absorption spectra, aggregation ratios, and transmission electron microscopy (TEM), we have uncovered a putative mechanism underpinning the functionality of this assay. Herein, we propose a model describing the interaction between ssDNA, dsDNA, and d-AuNPs.

2.3. DNA extraction Genomic DNA was extracted from Pseudoperonospora cubensis sporangia isolated from cucumber plants using the Machanery Nagel Nucleospin DNA Kit (Düren, Germany). Sporangia were flash frozen in liquid nitrogen and homogenized in a tissue grinder for 40 s at 4.0 M/S using a FastPrep-24 tissue homogenizer (MP-Biomedical, Santa Ana, CA). DNA was purified according to the manufacturer's protocol and quantified by Qubit (ThermoFisher, Waltham, MA). The extracted DNA was stored at −20 °C until use. For non-target DNA reactions, gDNA was extracted from five cucumber leaf discs collected using a #3 cork borer (1 cm2) and flash frozen in liquid nitrogen. DNA was isolated and purified as described above. 2.4. Assay procedure To evaluate the stability of the dextrin- and c-AuNPs in the presence and absence of a 66 nM ssDNAp, 20 µL of NaCl (0, 50, 100, 150, 200, 250, and 300 mM final reaction concentration) was added to 10 µL of each of the AuNPs. After a 10-min incubation at 21 °C, the visible absorption spectrum of the d-AuNP aggregation was quantified as described below. ssDNAp-to-target hybridization was initiated by the addition of 2 µL of 1 µM ssDNAp and 5 µL of P. cubensis or C. sativus extracted gDNA in 3 µL hybridization buffer [10 mM phosphate buffered saline (PBS) at 0.4 M NaCl (pH 7.0)]. Next, the reaction was denatured at 95 °C for 5 min, followed by annealing for 1 min at 57.5 °C. The reaction was cooled for 10 min at 21 ± 1 °C before adding 10 µL of d-AuNPs, followed by 10 µL of 0.8 M NaCl to initiate particle aggregation (NaCl is further denoted as salt). The final reaction solution contained 66 nM ssDNAp, and the concentration of 24 fM P. cubensis gDNA or 4 fM of C. sativus gDNA, 0.66 nM ssDNAp, 2.5 nM d-AuNPs and 0.3 M NaCl. The reaction was then incubated for 10 min at 21 °C, and the aggregation of AuNPs was quantified by measuring the absorption spectrum of the reaction from 400 to 700 nm. 2.5. Characterization of AuNP aggregation PCR tubes (200 µL) were used as a reaction vessel. A SpectraMax M2e plate reader (Molecular Devices, Sunnyvale, CA) was used to measure the 520 and 620 nm absorbance values for AuNPs salt and oligonucleotide interactions and genomic DNA sensitivity in a 96 well 200 µL plate. A NanoDrop 2000 (Thermo-Fisher, Waltham, MA) was used to assess the UV–vis absorption spectrum for ssDNA oligomer-totarget hybridization and crude matrix sensitivity. Means of aggregation were separated with a one-way ANOVA using aov in CRAN.R-project. Means were separated at P ≤ 0.05 using Tukey's honestly significance difference test (R Development Core Team, 2013). Particle dispersion was determined by TEM images and were collected with a JEOL 100CS TEM from 20 µL final reaction volumes containing d-AuNPs in water, in 66 nM ssDNAp, in 66 nM ssDNAp in the presence of 4 fM non-target gDNA, and in 66 nM ssDNAp in the presence of 29 fM target gDNA. All TEM reactions were conducted in 1.5 mM PBS containing 60 mM NaCl. Reactions were incubated at 95 °C for 5 min, followed by annealing for one min at 57.5 °C, and then cooled for 10 min at 21 °C. d-AuNPs

2. Material and methods 2.1. Biological reagents Genomic DNA and sporangia from the plant pathogenic oomycete Pseudoperonospora cubensis and its cucumber host, Cucumis sativus cv Eureka, were used in this study. The ssDNA oligonucleotide 5′AATCACAGCTTCTATGTTTTACAT-3′ was synthesized by Integrated DNA Technologies (Coralville, IA). The target sequence of the ssDNAp was 5′- ATG TAA AAC ATA GAA GCT GTG ATT −3′ contained within the genomic DNA of P. cubensis. 2.2. Gold nanoparticle synthesis Dextrin-capped gold nanoparticles (13 nm in diameter) were 30

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

Scheme 1. Illustration of the proposed mechanism of interaction of DNA with dextrin-capped AuNPs for sequence-specific genomic DNA detection.

AuNPs in a variety of diagnostic applications (Koo et al., 2015; Nelson and Rothberg, 2011; Sedighi et al., 2014; Zhang et al., 2012). To further these studies, we conducted a simple experiment to elucidate the impact of salt concentration on the aggregation of d-AuNPs, and from this, optimized a method using salt-induced AuNP aggregation for a dsDNApg-containing reaction. To do this, and as shown in Fig. 1, the stability of c- and d-capped AuNPs was analyzed over a salt gradient ranging from 0 to 300 mM to identify concentrations critical for AuNP aggregation. As shown, we observed a visible color change (i.e., from red to blue) at NaCl concentrations ranging between 50 mM for the cAuNPs (Figs. 1A) and 0.5 M for the d-AuNPs (Supplemental Fig. 1). This difference was highlighted by a concomitant increase in AuNP aggregation (the ratio of UV–Vis absorbance at 620 and 520 nm; A620/ A520), indicated by a shift from the gold SPR peak at 520–620 nm with increasing concentrations in NaCl (Fig. 1B). At a mechanistic level, AuNP salt-induced aggregation results from inter-particle plasmon coupling during the reduction of electrostatic forces between particles (Wang et al., 2010; Li and Rothberg, 2004). Our results demonstrate that the SPR of c-AuNPs was disrupted by the addition of 40–60 mM of salt, a result previously observed (Hussain et al., 2013; Liu et al., 2015) (Fig. 1). As shown, however, we did not observe AuNP aggregation until a final concentration of 250 mM NaCl was achieved, indicating that the dextrin capping agent stabilized AuNP SPR. Similar previous results were observed using dextran-capped AuNPs at NaCl concentrations as high as 100 mM NaCl (Wang et al., 2010). Furthermore, Katti et al. (2009) described this interaction and also observed SPR stabilization at increased salt concentrations, a mechanism hypothesized from molecular crowding of the polysaccharide chains surrounding the glyco-AuNPs. In total, these data demonstrate that glyco-coating of AuNPs facilitate a reduction in NaCl-induced aggregation of AuNPs. To determine if dextrin-capping agent altered ssDNAp adsorption to AuNPs, thereby affecting salt-induced aggregation we defined the interaction(s) between oligonucleotide-d-AuNPs. As a point of comparison, we used c-AuNPs in parallel reactions, as these AuNPs have previously been shown to adsorb ssDNAp, a function that stabilizes the colloidal state of the reaction (Li and Rothberg, 2004). As shown in Fig. 1C, we observed a reduction of salt induced aggregation of dAuNPs, as indicated by the maintenance of a red hue in the reaction, when comparing citrate-capped and d-AuNPs over a 300 mM salt concentration range. Similarly, the addition of ssDNAp resulted in further

(10 µL) were added to cooled reactions. 2.6. Detection of unpurified pathogen extracts and quantitative PCR Pseudoperonospora cubensis sporangia were serial diluted in sterile microcentrifuge tubes in amounts ranging from 185 to 1.85/µL in 200 µL of hybridization buffer. Samples were pulverized using 3 mm glass beads, and 5 µL of the resultant sporangial extract from each serial dilution was used in the in the AuNP assay. All samples were analyzed in triplicate. Quantitative PCR specific for P. cubensis DNA was completed on the serial dilution of sporangia used for the crude DNA sample. DNA was extracted in triplicate from each dilution in hybridization solution and completed qPCR as denoted in Summers et al. (2015). 3. Results and discussion 3.1. Assay scheme: A mechanism for target d-AuNP-DNA interactions As shown in Scheme 1, the underlying principle, and specificity, of the interaction between DNA and d-AuNPs relies on the induced stability of d-AuNPs in the presence of sequence-specific gDNA targets. Using this approach, we hybridized a denatured gDNA target with a complementary sequence-specific ssDNAp, thereby exploiting the electrostatic and hydrophobic properties (Deng et al., 2013; Li and Rothberg, 2004; Sedighi et al., 2014) of generated genomic ssDNA (ssDNAg) and dsDNA of the probe-target complex (dsDNApg). The underlying principle of this approach revealed that ssDNAg complex stabilized the d-AuNPs under high ionic conditions. Based on this, we hypothesized that during annealing, a ssDNAp would bind to the denatured target gDNA, displacing a ssDNAg, which in turn, through electrostatic interactions in high ionic conditions, would generate a ssDNAg stabilized d-AuNP complex. Thus, in the presence of non-target gDNA, in a high salt environment, the ssDNAg adsorption to AuNPs will lead to moderate aggregation and differentiation from greater destabilization of d-AuNPs bound with ssDNAp. 3.2. Dextrin AuNPs display reduced aggregation in elevated salt, both in the presence and absence of DNA oligonucleotides Previous studies have demonstrated the function and utility of c31

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

Fig. 1. Dextrin-capped AuNPs (d-AuNPs) resist aggregation at elevated NaCl concentrations and are stabilized in the presence of ssDNA. A. Photograph of citrate and d-AuNPs solutions with increasing salt concentration. B. Aggregation of citrate- and d-AuNPs in the presence of increasing NaCl concentrations. C. Visualization of citrate- and d-AuNPs solutions with increasing NaCl concentration, with and without a ssDNA probe. D. Aggregation of citrate- and d-AuNPs with and without ssDNA probe with increasing salt concentration. Circles represent means, ± standard error of the mean. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.).

(2004) using citrate-capped AuNP and dsDNA. The further development of a hand-held assay resulted in an increase of aggregation of the d-AuNPs (Fig. 2) due to increased quantification time when compared to salt optimization (Fig. 1). As noted above, we observed that ssDNAp alone, or when mixed with non-target gDNA, resulted in a distinguishable absorbance profile at 620 nm while maintaining a reduced SPR peak at 520 nm (Fig. 2A), a process that stems from ssDNAp adsorbing to the d-AuNPs, preventing total aggregation (Fig. 2B). Conversely, we observed a statistically enhanced SPR peak at 520 nm (P ≤ 0.05) when the target gDNA and ssDNAp interacted, compared to the aforementioned reaction, or in reactions containing non-target gDNA-ssDNAp combinations. Taken together, these data demonstrate the high specificity of the reaction, and moreover, support a mechanism whereby a reduction of d-AuNP aggregation in samples containing both ssDNAp and target gDNA leads to enhanced stabilization of d-AuNPs. In short, this yields a reaction that is visually distinguishable without any instrumentation (Fig. 2C). The observation of an increase in the overall stabilization of dAuNPs by the ssDNAp and complementary gDNA target was surprising, as previous studies using c-AuNPs reported that aggregation is not disrupted when ssDNAp binds to the DNA target (Deng et al., 2013; Liandris et al., 2009; Li and Rothberg, 2004; Liu et al., 2015). The simplest explanation for this apparent discrepancy stems from divergence of the citrate-capped gold nanoparticle salt-induced aggregation by the dextrin surface chemistry and increased salt concentration. However, our results shown in Fig. 1 clearly demonstrate that ssDNAp adsorbs to d-AuNPs, and previous results from Torres-Chavolla and Alocilja (2011) demonstrated that dextrin capping did not alter AuNP adsorption of ssDNAp, including thiol modified ssDNAp. Based on the sum of these data, we posit that the direct interaction of the dextrin surface chemistry with DNA is not responsible for stabilization of the d-

decreased aggregation of both citrate-capped and d-AuNPs in higher salt concentrations; this result was not observed when the AuNPs were incubated in the absence of the ssDNAp. As shown in Fig. 1D, we observed a shift in absorbance to 620 nm was delayed in the ssDNA oligomer treatments in both AuNPs resulting in a lower aggregation ratio. This observation suggests that ssDNAp adsorbs, and further stabilizes, the AuNP colloidal state independently of the surface chemistry. Similarly, unmodified d-AuNPs were previously found to become as efficiently functionalized with thiol modified ssDNAp as c-AuNPs, (Torres-Chavolla and Alocilja, 2011) illustrating that the surface chemistry of the AuNPs did not alter functionality. Based on these data, we posit that d-AuNPs can be utilized in an unmodified state to detect target DNA sequences much in the same manner as c-AuNPs, yet in higher ionic environments. 3.3. Specificity of oligonucleotide-target genomic DNA binding stabilizes dAuNPs As noted above, the optical properties of unmodified AuNPs in the presence of DNA enables sequence-specific DNA detection by differentiation with salt. Based upon this, we next investigated the development of a DNA sequence-specific assay to examine ssDNA and dsDNA interactions with d-AuNPs. First, we investigated using PCR synthesized target and non-target dsDNA (Supplemental Fig. 2). As shown, using in vitro amplified DNA, we were able to detect a specific interaction between AuNPs and target DNA. Next, as shown in Fig. 2A, in the presence of unamplified gDNA, as well as reactions containing no DNA, we observed a visual colorimetric shift of d-AuNPs towards 620 nm, indicating a disruption of d-AuNP SPR by the salt. Additionally, we did not observe a disruption of d-AuNPs aggregation by gDNA in the presence of salt (Fig. 2B), an observation similar to that of Li and Rothberg 32

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

Fig. 2. Complementary genomic DNA (gDNA) target stabilizes dextrin-capped AuNPs (d-AuNPs) in the presence of NaCl. A. UV–vis absorption spectrum of d-AuNPs in the presence of water, ssDNA P. cubensis (Pc) probe (ssDNAp). B. Corresponding aggregation d-AuNPs. C. Visualization of DNA-d-AuNP reactions. Bars represent means ± standard error of the mean. Means within each treatment followed by different letters are significantly different at P ≤ 0.05 according to Tukey's honestly significant difference test.

AuNPs following ssDNAp annealing to denatured gDNA before salt induced aggregation by using TEM. As shown in Fig. 3A, we observed a uniform dispersion of d-AuNPs in the absence of DNA interactions. However, in the presence of a ssDNAp, we observed a slight aggregation of AuNPs (Fig. 3B). A similar dispersion pattern to ssDNAp treatment was observed in the non-target gDNA-ssDNAp treatment (Fig. 3C). However, when AuNPs were incubated in the presence of target gDNAssDNAp, we observed enhanced aggregation of AuNPs, with few, unaggregated, d-AuNPs (Fig. 3D). These data suggest that d-AuNPs aggregate in the vicinity of target gDNA strands displaced by the ssDNAp, which was also colorimetrically revealed by isolated interactions of synthesized target or non-target dsDNA and ssDNAp with d-AuNPs (Supplemental Fig. 2). This is in agreement with Sedighi et al. (2014) which proposed that the binding of AuNPs in the presence of ssDNA is likely mediated by the formation of de-hybridization bubbles. Taken together, these data support the hypothesis that d-AuNP physical aggregation is enhanced (i.e., d-AuNPs stabilized) within these ssDNAg environments.

AuNPs in the presence of the ssDNAp and the target gDNA. Our results support the hypothesis that the dextrin-capping agent enhances the stability of AuNPs under high ionic conditions, even more so than c-AuNPs (Fig. 1). In fact, the d-AuNPs assay developed herein are in a five-fold greater ionic concentration than is typically conducted for c-AuNPs-DNA detection assays. With increasing salt concentrations, DNA-DNA and DNA-AuNP interactions change by quenching destabilizing negative charges on the phosphodiester DNA backbone and promoting DNA base stacking (Bosco et al., 2014; Nelson and Rothberg, 2011). This indicates that non-native secondary structures likely form when the ssDNAp and complementary target bind, creating a secondary structure with the non-binding strand of gDNA. Supported by our results herein, the addition of d-AuNPs into an elevated ionic solution would likely increase the rate of ssDNA adsorption (Nelson and Rothberg, 2011), thereby creating a more favorable environment for dAuNPs to bind to gDNA secondary structures. In total, this reaction would lead to enhanced stabilization of the d-AuNPs upon ssDNAg secondary structure target-complex formation, more than ssDNAp alone (Fig. 2).

3.5. Sensitivity of DNA detection by d-AuNPs 3.4. Physical association of d-AuNPs and oligonucleotide-target DNA precludes stabilization in salt

3.5.1. Extracted genomic DNA The ability to detect low levels of pathogens, including under asymptomatic conditions, requires the development of diagnostic assays that offer both specificity and sensitivity. Previously, we demonstrated the specificity of this assay through the detection of plant-extracted DNA. Next, to define the limits and utility of the current DNAAuNP-based assay for the detection of pathogens, we investigated the sensitivity of the assay using gDNA from the cucurbit downy mildew

AuNP-DNA interactions can be visualized through aggregation-dispersion characteristics of the individual AuNPs (Samanta and Medintz, 2016). In brief, AuNP size, aggregation clustering, or shape will change upon bias of DNA interaction (Jung et al., 2010; Zhang et al., 2012). Therefore, to test if the physical interaction of the ssDNAp and gDNA target stabilized the d-AuNPs, we next investigated the dispersion of d33

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

Fig. 3. Transmission electron microscopy of dextrincapped AuNPs (d-AuNPs) dispersion in the presence of genomic DNA (gDNA). A. d-AuNPs + water only control. Bar = 100 nm; B. d-AuNPs + ssDNA probe (ssDNAp). Bar = 200 nm; C. AuNPs + non-target gDNA and ssDNAp. Bar = 200 nm; D. d-AuNPs + target gDNA-ssDNAp complex. Bar = 200 nm.

pathogen Pseudoperonospora cubensis ranging from 29 fM to 2.9 aM As shown in Fig. 4A, we observed a discernable, visual, reduction in dAuNP aggregation from 29 fM to 2.9 fM, as compared to control reactions (i.e., no DNA). This decrease in d-AuNP aggregation relative to an increase in target DNA supports our hypothesis that the d-AuNP-DNA interaction is a viable assay not only for the quantitative detection of DNA, but also represents an advance over current approaches (Lin et al., 2013) (Fig. 4B).

3.5.2. Crude DNA from pathogen matrix Next, to determine the sensitivity of this assay in the detection of crude DNA (i.e., non-extracted) samples, we investigated the sensitivity of the reaction using serially-diluted sporangia – the wind-dispersed spores – of P. cubensis. As shown in Fig. 5A, sporangia were serially diluted from 185 to 1.8 spores/µL, and the ground, crude lysates were incubated in the presence of AuNPs and the ssDNA probe to reveal optically distinct target detection of approximately 18.5 sporangia/µL when compared to no DNA control reactions. The linear relationship (R2 = 0.999) between AuNP stabilization in the presence of DNA from pathogen spores demonstrates the quantitative abilities of this assay to detect pathogen DNA within crude matrices (Fig. 5B). Moreover, given the obligate nature of P. cubensis, the DNA samples analyzed herein may also contain contaminating non-target cucumber DNA, thus skewing the true detection limits of our assay. To explore the extent of contaminating host DNA and the true sensitivity of the d-AuNP-DNA assay, we used qPCR to quantify the amount of P. cubensis DNA in the crude samples. The extracted P. cubensis DNA represented one fifth portion of total extracted DNA from samples, confirming the detection between 0.026 fM and 0.016 fM of P. cubensis DNA (Fig. 5). Thus, we estimate

Fig. 4. Sensitivity of the colorimetric dextrin-capped AuNPs (d-AuNPs) nanobiosensor assay with Pseudoperonospora cubensis extracted genomic DNA. A. Photograph d-AuNP visual aggregation. B. Aggregation response of serially diluted DNA. Bars represent means ± standard error of the mean of three reactions. The inset depicts the linear range of the DNA detection assay.

34

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

cannot be deployed as point-of-contact methods, and they remain costprohibitive and timely. In this regard, the proposed method offers many advantages over current methods. As presented herein, the DNA-dAuNP assay provides a platform allowing the use of crude matrices, ~30-min visual assay, low cost (i.e., < $0.01 per reaction), and limited equipment (i.e., water bath), thus enabling a specific and quantitative point-of-care assay. Through future optimizations of salt and probe concentrations, ultra-sensitive DNA detection levels approaching qPCRbased methods can likely be developed. 4. Conclusions In total, the DNA-based nanoparticle assay described herein permits the detection of DNA at limits several magnitudes lower than existing point of care approaches, including with the added advantage of not requiring expensive temperature-sensitive reagents. This method captures the unique properties of gold nanoparticles coated with dextrin. As the first report describing the application of unmodified d-AuNPs to directly detect specific DNA sequences in reaction solution containing genomic DNA, we highlight 1) d-AuNPs have a wider range of stability to salt than c-AuNPs, 2) d-AuNPs adsorbed ssDNA enabling a DNA sequence-specific detection assay, and 3) the elevated ionic assay concentration alters DNA-DNA and d-AuNPs-DNA interactions allowing target stabilization of d-AuNPs and DNA detection within crude matrices. We posit that DNA detection by d-AuNPs overcomes many challenges currently limiting nanotechnology adoption for field-deployable-based detection of pathogens affecting human health and food security through the cost of this assay as well as the increased synthesis sustainability over c-AuNPs (Torres-Chavolla and Alocilja, 2011). In addition, the work described addresses several key gaps in our understanding the principle(s) underlying the interaction between AuNPs and DNA. Acknowledgments A.B.Y. was supported by a C.S. Mott Predoctoral Fellowship in Sustainable Agriculture from Michigan State University. Research in the laboratory of E.C.A. was supported by the Midland Research Institute for Value Chain Creation and the Michigan State University College of Natural Resources Undergraduate Research Program. Research in the laboratory of B.D. was supported by the Michigan State University Rackham Foundation and Michigan State University Project GREEEN (GR14-017). We would like to thank M. Haus for critical reading of the manuscript. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2017.10.011. References Fig. 5. Colorimetric crude detection of Pseudoperonospora cubensis DNA. A. Photograph dextrin-capped AuNP visual aggregation. B. Aggregation of dextrin-capped AuNPs serially diluted sporangia from P. cubensis. Bars represent means ± standard error of the mean. Inset graph demonstrates the linear range of the detection assay. C. Quantitative PCR detection of extracted P. cubensis DNA from serially diluted sporangia. Inset graph demonstrates the linearity of the detection assay.

Anderson, J.M., Torres-Chavolla, E., Castro, A.B., Alocilja, E.C., 2010. J. Nanopart. Res. 13, 2843–2851. Bakthavathsalam, P., Rajendran, V.K., Ali, B.M.J., 2012. J. Nanobiotechnol. 10, 3155–3165. Bosco, A., Camunas-Soler, J., Ritort, F., 2014. Nucleic Acids Res. 42, 2064–2074. D’Agata, R., Spoto, G., 2012. In: Spoto, G., Corradini, R. (Eds.), In Detection of NonAmplified Genomic DNA. Springer, New York, pp. 150–153. Daniel, M.C., Astruc, D., 2004. Chem. Rev. 104, 293–346. Deng, H., Zhang, X., Kumar, A., Zou, G., Zhang, X., Liang, X., 2013. J. Chem. Commun. 49, 51–53. Desmarchelier, P.M., Bilge, S.S., Fegan, N., Mills, L., Vary, J.C., Tarr, P.I., 1998. J. Clin. Microbiol. 36, 1801–1804. Drosten, C., Panning, M., Kramme, S., 2003. Clin. Chem. 49, 1659–1661. Fang, X., Jaing, W., Han, X., Zhang, Y., 2013. Microchim. Acta. 180, 1271–1277. Foley, J.A., Ramankutty, N., Brauman, K.A., Cassidy, E.A., Gerber, J.S., Johnston, M., Mueller, N.D., O’Connell, C., Ray, D.K., West, P.C., et al., 2011. Nature 487, 337–342. Gearheart, L.A., Ploehn, H.J., Murphy, C.J., 2001. J. Phys. Chem. 105, 12609–12615. Hastings, A.B., Eichelberger, L., 1937. J. Biol. Chem. 117, 73–93.

that the sensitivity of this assay is approximately 1 log less than the qPCR-based methods. The assay described herein represents a marked improvement over currently available diagnostic methods. Indeed, current diagnostic technologies (e.g., ELISA and PCR) are the most reliable detection methods available (Kong et al., 2016; Tsui et al., 2011; Wang et al., 2015). However, while these assays are quantitative and sensitive, they

35

Biosensors and Bioelectronics 101 (2018) 29–36

A.M. Baetsen-Young et al.

〈http://www.R-project.org/〉. Rodrigues, S.M., Demokritou, P., Dokoozlian, N., Hendern, C.O., Karn, B., Mauter, M.S., Sadik, O.A., Safarpour, M., Unrine, J.M., Viers, J., et al., 2017. Environ. Sci-Nano. 4, 767–781. Samanta, A., Medintz, I.L., 2016. Nanoscale 8, 9037–9095. Sattarahmady, N., Tondro, G.H., Gholchin, M., Heli, H., 2015. Biochem. Eng. J. 97, 1–7. Sedighi, A., Li, C.H.P., Pekcevik, I.C., Gates, B.D., 2014. ACS Nano 7, 6765–6777. Sepúlveda, B., Angelomé, P.C., Lechuga, L.M., Liz-Marzán, L.M., 2009. Nano Today 4, 244–251. Summers, C.F., Adair, N.L., Gent, D.H., McGrath, M.T., Smart, C.D., 2015. Can. J. Plant Pathol. 37, 315–330. Sun, G., Liu, J., Li, G., Zhang, X., Chen, T., Chen, J., Zhang, H., Wang, D., Sun, F., Pan, H., 2015. Plant Dis. 99, 219–224. Torres-Chavolla, E., Alocilja, E.C., 2011. Biosens. Bioelectron. 26, 4614–4618. Tsui, C.K., Woodhall, J., Chen, W., Levesque, C.A., Lau, A., Schoen, C.D., Najafzadeh, M.J., de Hoog, G.S., 2011. IMA Fungus 2, 177–189. Upadhyayula, V.K., 2012. Anal. Chim. Acta. 715, 1–18. Wang, J., Jacobs, J.L., Byrne, J.M., Chilvers, M.I., 2015. Phytopathol 105, 378–387. Wang, Y., Zhan, L., Huang, C.Z., 2010. Anal. Methods 2, 1982–1988. Zanoli, L.A., D’Agata, R., Spoto, G., 2012. Anal. Bioanal. Chem. 402, 1759–1771. Zhang, X., Servos, M.R., Liu, J., 2012. Langmuir 28, 3896–3902.

Hussain, M.M., Samir, T.M., Azzazy, H.M.E., 2013. Clin. Biochem. 46, 633–637. Hussain, T., Singh, B.P., Anwar, F., 2014. Saudi. J. Biol. Sci. 21, 380–386. Jin, R., Wu, G., Li, Z., Mirkin, C.A., Schatz, G.C., 2003. J. Am. Chem. Soc. 125, 1643–1654. Jung, Y.L., Jung, C., Parab, H., Li, T., Park, H.G., 2010. Biosens. Bioelectron. 25, 1941–1946. Katti, K.K., Kattumuri, V., Bhaskaran, S., Katti, K.V., Kannan, R., 2009. Int J Green Nanotechnol Biomed 1, B53–B59. Koo, K.M., Sina, A.A.I., Carrascosa, L.G., Shiddiky, M.J.A., 2015. Anal. Methods 7, 7042–7054. Kong, X., Qin, W., Huang, X., Kong, F., Schoen, C.D., Wang, Z., Zhang, H., 2016. Sci. Rep. 6, 28935. Li, H., Rothberg, L., 2004. Proc. Natl. Acad. Sci. USA 101, 14036–14039. Li, H., Rothberg, L.J., 2004a. J. Am. Chem. Soc. 126, 10958–10961. Liandris, E., Gazouli, M., Andreadou, M., Comor, M., Abazovic, N., Sechi, L.A., Ikonomopoulos, L., 2009. J. Microbiol. Methods. 78, 260–264. Lin, C., Zhang, Y., Zhou, X., Yao, B., Fang, Q., 2013. Biosens. Bioelectron. 47, 515–519. Liu, Z., Zhu, J., Yang, C., Li, X., 2015. Anal. Methods 7, 8159–8164. Munns, R., 2002. Plant Cell Environ. 25, 239–250. Nelson, E.M., Rothberg, L.J., 2011. Langmuir 27, 1770–1777. R Core Team. R Foundation for Statistical Computing, Vienna, Austria. 2013, URL

36