“Mix-to-Go” Silver Colloidal Strategy for Prostate Cancer Molecular

Subscriber access provided by UNIV OF LOUISIANA

Article

“Mix-to-Go” Silver Colloidal Strategy for Prostate Cancer Molecular Profiling and Risk Prediction Jing Wang, Kevin Maisheng Koo, Yuling Wang, and Matt Trau Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b02959 • Publication Date (Web): 27 Sep 2018 Downloaded from http://pubs.acs.org on September 28, 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 21 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

“Mix-to-Go” Silver Colloidal Strategy for Prostate Cancer Molecular Profiling and Risk Prediction Jing Wang1,‡, Kevin M. Koo1,‡, Yuling Wang2,* & Matt Trau1,3,* 1

Centre for Personalized Nanomedicine, Australian Institute for Bioengineering and Nanotechnology,

The University of Queensland, Brisbane, QLD 4072, Australia 2

Department of Molecular Sciences and ARC Centre of Excellence for Nanoscale BioPhotonics,

Faculty of Science and Engineering, Macquarie University, Sydney, NSW 2109, Australia 3

School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, QLD

4072, Australia

1 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

ABSTRACT Molecular profiling via analysis of multiple disease biomarkers, is a powerful tool for disease diagnosis and risk prediction. Due to simplicity and minimal instrumentation requirement, colloidalbased colorimetric DNA/RNA assays are attractive for driving molecular profiling towards widespread clinical usage. Still, the reliability and speed of current colorimetric assays need to be further improved upon for eventual clinical use. Herein, we propose a “mix-to-go” colloidal strategy which utilizes the electrostatic attraction between negatively charged target sequences and positively charged silver nanoparticles (AgNPs) to induce aggregation of AgNPs, to profile a panel of clinically-validated urinary prostate cancer (PCa) RNA biomarkers (TMPRSS2:ERG, T2:ERG; prostate cancer antigen 3, PCA3; and kallikrein-related peptidase 2, KLK2). Our strategy is unique in inducing a rapid (10 sec), unambiguous solution color change in the presence of target sequences, without the additional NP aggregation agents that are used in existing electrostatic-mediated aggregation assays. Our strategy is analytically specific and sensitive for the detection of 105 and 104 target copies by the naked eye and UV-vis spectrophotometry, respectively. Analytical accuracies of our strategy in detecting T2:ERG, PCA3, and KLK2 RNA biomarkers were 95.9%, 97.3%, and 100% respectively, as validated by quantitative reverse transcription-polymerase chain reaction. To further evaluate clinical molecular profiling performance beyond conventional proof-of-concept demonstration, we tested our colloidal strategy for non-invasive PCa risk prediction of 73 patients, using the urinary RNA biomarker panel comprising of T2:ERG, PCA3, and KLK2. We found that elevated T2:ERG and PCA3 levels were positively associated with high-risk PCa and obtained corresponding area-under-the-curve values of 0.790 and 0.833 for predicting PCa and high-risk PCa on biopsy, respectively. We believe our “mix-to-go” strategy may serve as a reliable and accessible Ag colloidal-based molecular profiling approach for clinical applications.


Page 2 of 21

Page 3 of 21 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


Molecular profiling involves the analysis of a biomarker panel in patient samples to potentially improve disease diagnosis, prediction, and treatment monitoring. For instance, prostate cancer (PCa) RNA biomarkers such as TMPRSS2:ERG (T2:ERG), prostate cancer antigen 3 (PCA3), and kallikrein-related peptidase 2 (KLK2) have been shown to be promising for better disease-specific detection. T2:ERG is the most frequent fusion gene mutation in PCa, occurring in approximately 50% of prostate-specific antigen (PSA)-screened PCa patients.1 PCA3 is abnormally overexpressed in greater than 95% of PCa but absent in healthy prostate cells.2 KLK2 is highly expressed in the human prostate gland, and its abnormal overexpression is associated to PCa carcinogenesis and metastatic potential.3 Recent findings have shown that the molecular profiling of T2:ERG and PCA3 reduces the use of prostate biopsy by 51% among men referred for prostate biopsy.4 Currently, molecular profiling assays of these RNA biomarkers mainly require conventional laboratory techniques which necessitate technical experimental procedures and specialized instrumentation. Therefore, we were driven to develop an alternative strategy focusing on these promising PCa biomarkers to enable rapid, cost-effective, and highly accessible PCa molecular profiling. Colorimetric assays have indicated potential as a simple tool for visual DNA/RNA detection due to minimal instrumentation requirement and lack of costly detection labels.5-8 Gold (Au) and silver (Ag) nanoparticles (NPs) have emerged as the most widely-used substrates in colorimetric assays.9-10 Due to Au- and Ag-NPs possessing excellent localized surface plasmon resonance (LSPR), their colloidal solutions display high extinction coefficients and show visible color differences between dispersed and aggregated NP states.11 Recently-described AuNP-based colorimetric assays for DNA/RNA detection consist of two major classes: DNA probe-modified5, 12-17 and electrostaticmediated aggregation.6, 18-20 DNA probe-modified AuNPs assemble into aggregated networks from a dispersed state via hybridization events in the presence of target sequences. This method requires well-controlled NP surface functionalization and generally suffers from slow target hybridization. In contrast, electrostatic-mediated assays mainly rely on quick target sequence adsorption onto unmodified NP surface to resist against salt-induced AuNP aggregation.6, 18, 21 Despite being rapid and NP functionalization-free, the reliability of electrostatic-mediated aggregation assays is prone to be affected by multiple variables such as salt concentration, temperature, as well as sequence composition and length. Herein, we proposed a unique colorimetric molecular profiling strategy which AgNP aggregation is mainly caused by electrostatic attraction between positively charged AgNPs and negatively charged target sequences, without extra NP aggregation agents. Specifically, our strategy combines isothermal reverse transcription-recombinase polymerase amplification (RT-RPA) of RNA targets and electrostatic-mediated AgNP aggregation for amplified sequence detection. The



electrostatic-mediated AgNP aggregation yielded a yellow to grey solution color change which could also be directly quantified by UV-vis measurements. Compared to existing colorimetric DNA/RNA detection assays, our strategy offers distinct advantages: (i) positively charged AgNPs with high extinction coefficient11 are easily prepared for high detection sensitivity; (ii) the isothermal amplification process can rapidly and specifically enrich rare RNA target sequences from patient samples for detection; (iii) the “mix-to-go” AgNP aggregation readout is rapid (10 sec) and solely target-triggered without additional aggregation agents (e.g., salts). We validated analytical accuracies of our technique for T2:ERG, PCA3, and KLK2 detection in human urine samples through comparison with a high-quality reference test— quantitative reverse-transcription polymerase chain reaction (qRT-PCR). Additionally, to progress towards potential clinical translation, we took a further step beyond conventional proof-of-concept analytical detection demonstration and applied our strategy for non-invasive PCa molecular profiling and risk prediction in 73 urine samples.

EXPERIMENTAL SECTION Reagents. All reagents were purchased from Sigma Aldrich, unless otherwise stated. Synthetic oligonucleotide and primer sequences used in our experiments were obtained from Integrated DNA Technologies (Singapore), and sequences are shown in Table 1. AgNP Synthesis. Positively charged AgNPs were prepared as previously reported.22 Briefly, 20 µL of 0.5 M AgNO3 was added to 10 mL of Milli-Q water, followed by addition of 7 µL of 0.1 M spermine. The mixture was vigorously stirred for 1 min. A varying amount of 0.01M NaBH4 was then added into the mixture to synthesize different sizes of AgNPs. All reagents were prepared with ice-cold Milli-Q water.

AgNP Characterization. The particle size distribution (PSD) of AgNPs was measured by differential centrifugal sedimentation (DCS) using a DC24000 UHR disc centrifuge (CPS Instrument Inc, USA). The disc was loaded with 14.4 mL of sucrose gradient fluid comprising 8-24 wt.% sucrose in water. The average density, refractive index, and viscosity of the sucrose gradient fluid were 1.069 g/mL, 1.36, and 1.505 cP, respectively. A disc rotational frequency of 24 000 rpm was used for the AgNP measurements. Transmission electron microscopy (TEM) images were taken with a HT7700 microscope (Hitachi, Japan) operated at 120 kV. Zeta-potentials of AgNPs (diluted in water) were determined by Zetasizer Nano ZS (Malvern, UK) using standard settings (Viscosity = 0.8872 cP, temperature = 25 °C). 4 ACS Paragon Plus Environment

Page 4 of 21

Page 5 of 21 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


Extinction spectra of AgNPs were measured with a NanoDrop 1000 spectrophotometer (Thermo Scientific, USA).

RNA Extraction from Cell Lines. DuCap and LnCap cell lines were cultured in RPMI-1640 growth media (Life Technologies, Australia) supplemented with 10% fetal bovine serum (Life Technologies, Australia), 2 mM Glutamax (Gibco, Australia) and 1% PenStrep (Invitrogen, Australia) in a humidified incubator containing 5% CO2 at 37 °C. RNA was extracted using Trizol® reagent (Life Technologies, Australia). RNA purity was checked using a NanoDrop 1000 Spectrophotometer (Thermo Scientific, USA).

RNA Extraction from Patient Urine Samples. Ethics approval was obtained from The University of Queensland Institutional Human Research Ethics Committee (Approval No. 201400012), and informed consent was obtained from each patient before sample collection. Clinical assays were conducted according to approved guidelines. De-identified voided urinary samples were prospectively collected from 54 male patients prior to PCa needle biopsies, and 19 healthy men with no PCa family history. Total RNA in urine samples was extracted using the commercially-available ZR urine RNA isolation KitTM (Zymo Research, USA). Briefly, 30 mL of urine was passed through the supplied ZRC GFTM Filter, followed with 700 µL of supplied urine RNA buffer. The urinary cells collected in the flow-through were then lysed, washed, and eluted in 10 µL of RNase-free water.

RT-RPA and Amplicon Purification. To specifically amplify each RNA biomarker, the TwistAmp Basic RT kit (Twist-DX, UK) was used with slight modifications to manufacturer’s instructions. Briefly, 1 µL of extracted total RNA and 250 nM of target-specific primers (Table 1) were added to supplied reagents to make a 12.5 µL of reaction volume. This solution was then incubated at 41 °C for 15 min to generate target amplicons. Target amplicons were further purified using the Agencourt AMPure XP SPRI kit (Beckman Coulter, USA) and then eluted in RNase-free water.

Electrostatic-Mediated AgNP Aggregation. 1 µL of RT-RPA amplicons was mixed with 20 µL of AgNPs for 10 sec to induce AgNP aggregation. The AgNP aggregation resulted in a yellow to grey color change in solution which was qualitatively detected by the naked eye or quantitatively measured by UV-vis spectrophotometry.

qRT-PCR. The KAPA SYBR FAST One-Step qRT-PCR kit (KAPA BIOSYSTEMS, USA) was used to set up a single reaction volume of 10 µL for each sample. Each reaction volume consists 5 ACS Paragon Plus Environment


of 1× KAPA SYBR FAST qPCR Master Mix, 200 nM of each primer (Table 1), 1× KAPA RT Mix and 3 µL of input RNA target. Tubes were incubated at 42 °C for 10 min to synthesize cDNA, followed by 95 °C for 5 min to deactivate reverse transcriptase before cycling 35 times (95 °C for 30 s, 69 °C for 30 s, and 72 °C for 1 min) and finished with 72 °C for 10 min. The required number of cycles for detectable fluorescence was used to evaluate positive/negative target biomarker status in samples.

Data Analysis. All data were normalized to minimize variations caused by factors such as different batches of AgNPs. As the extinction band of 20 nm AgNPs locates at 394 nm, ∆IExt at 394 nm was used to represent changes of LSPR properties before and after the addition of resultant amplicons, according to the following equation: ∆ 394 = 394 − 394 Binary logistic regression models were applied to predict the presence probability of PCa and high-risk PCa on biopsy based on expression profiles of T2:ERG, PCA3, and KLK2 RNA biomarkers. The regression equation is in the form of: ln

= + ∙ !2: $%& + ' ∙ ()*3 + ∙ +,+2

where P is the predicted probability belonging to a specified class; a, b, and c are regression coefficients for T2:ERG, PCA3, and KLK2, respectively; intercept value and regression coefficients were estimated from the data set. Binary logistic regression was performed with SPSS 24.0 software package (SPSS Inc., USA). Diagnostic ability was evaluated by receiver operating characteristic (ROC) analysis. The area under the curve (AUC) derived from ROC analysis was used as a measure of diagnostic ability.23-24 ROC analysis was computed using GraphPad Prism 7.0 (GraphPad Software Inc., USA). In evaluating mean levels of RNA biomarkers among high-risk, low-risk, and healthy patients, one-way ANOVA was used, and a P value of < 0.05 was considered statistically significant. Oneway ANOVA was analyzed by GraphPad Prism 7.0 (GraphPad Software Inc., USA).

RESULTS AND DISCUSSION Principle of Electrostatic-Mediated Colloidal Strategy for PCa Molecular Profiling. Our strategy for PCa molecular profiling is illustrated in Scheme 1. Briefly, total RNA is extracted from cultured cells/patient urine samples, followed by rapid isothermal RT-RPA to exponentially increase the amount of T2:ERG, PCA3, and KLK2 RNA biomarkers in parallel reactions. Additionally, the specificity of our strategy is mainly conferred by RT-RPA in selective target amplification. Resultant 6 ACS Paragon Plus Environment

Page 6 of 21

Page 7 of 21 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


homogeneous amplicons in the stabilized DNA form are then magnetically-purified to minimize excess non-target molecule interference on detection performance.25-26 Amplicons possess a doublehelix geometry that exposes negatively charged phosphate backbones. Upon mixing with positively charged AgNPs, the electrostatic attraction between negatively charged amplicons and positively charged AgNPs leads to AgNP aggregation without additional aggregation agents (e.g., salts or conjugated polymers),6,

18-19

which is the most significant advantages over existing electrostatic-

mediated assays. The colloidal mixture exhibits a distinguishable solution color change from yellow to grey due to changes in the extinction spectra of AgNPs (LSPR properties). The color change mirrors a specific and digital yes or no answer for naked eye biosensing purposes, and quantitative molecular profiling is achieved by measuring changes in the extinction intensity at 394 nm using UV-vis spectrophotometry. The entire “mix-to-go” colloidal strategy (from RNA extraction to AgNP aggregation readout) is achieved within 40 min and requires minimal instrumentation, and may be feasible for facilitating widespread clinical molecular profiling applications.

AgNP Optimization. The size and surface charge of NPs may affect electrostatic-mediated AgNP aggregation, which will in turn influence LSPR modes of AgNPs. As such, to optimize the detection sensitivity of target RNA biomarkers, we synthesized AgNPs of different surface charges and sizes, and compared their LSPR properties before and after the target amplicon addition. Through adjustment of reducing agent (NaBH4) amount, we generated different AgNP sizes. TEM images indicated that synthesized AgNPs were of 25 ± 5.5 nm (Ag-1), 43 ± 11 nm (Ag-2), and 77 ± 22 nm (Ag-3) (Figure 1A). The PSDs of these three types of AgNPs were further characterized by a weight-based particle size characterization technique—DCS. In agreement with TEM data, the PSDs peak maxima of DCS measurements showed AgNPs of 23.42 nm (Ag-1), 39.9 nm (Ag-2), and 60.9 nm (Ag-3) (Figure 1B). The corresponding zeta-potentials were 13.2 ± 2.1 mV, 8.4 ± 0.5 mV, and 21.6 ± 0.3 mV, respectively (Figure 1C), thereby confirming that these three sizes of AgNPs were all positively charged. Extinction spectra of Ag-1, Ag-2, and Ag-3 NPs displayed corresponding LSPR bands at 394 nm, 400 nm, and 409 nm; and generated yellow, dark yellow, and off-white colloidal suspensions, respectively (Figure 1D-F). In the presence of target sequences, electrostatic-mediated AgNP aggregation occurred and led to concomitant color changes due to a decrease in the molar extinction of AgNPs. Color changes for Ag-3 NPs (Figure 1F) were not as easily distinguishable as compared to Ag-1 and Ag-2 NPs (Figure 1D, E), making Ag-3 NPs less desirable for colorimetric readouts. As a narrower extinction band is more reactive for reflecting LSPR shifts and molar extinction changes,27 Ag-1 NPs were selected for our study. 7 ACS Paragon Plus Environment


Detection Specificity and Sensitivity. To evaluate the specificity of our strategy, we investigated the detection of RNA biomarkers (T2:ERG, PCA3, and KLK2) in well-characterized PCa cell lines.1-2, 28 We observed significant aggregation responses for T2:ERG, PCA3, and KLK2 amplicons derived from T2:ERG-positive DuCap cells, PCA3-positive LnCap cells, and androgentreated KLK2-positive LnCap cells; and no aggregation responses from no template (No T) and no transcriptase (No RT) negative controls (Figure 2A-C), confirming specific detection of RNA targets. Correspondingly, extinction spectra (Figure 2D-F) and ∆IExt at 394 nm (Figure 2G) showed significant changes in the presence of amplicons but negligible for No T and No RT controls. Gel electrophoresis analysis indicated band sizes of T2:ERG, PCA3, and KLK2 RNA amplicons, further confirming the specific amplification of target RNA biomarkers (Figure 2H). These results demonstrated that our strategy has excellent detection specificity for RNA biomarker profiling in cell line cells. To evaluate the assay sensitivity, we tested different starting copies (0, 102, 103, 104, and 105 copies) of synthetic RNA targets. Upon the addition of target amplicons to induce AgNP aggregation (i.e., solution color changes), naked eye observation detected as low as 104 copies for T2:ERG, and 105 copies for PCA3 and KLK2 (Figure 3A-C). We also performed a quantitative and sensitive target detection by using UV-vis spectrophotometry to measure changes in extinction spectra (Figure 3D-F) and ∆IExt at 394 nm (Figure 3G-I). We found that the detection sensitivity was further improved to 103 copies for T2:ERG, and 104 copies for PCA3 and KLK2 (Figure 3G-I). Given that a single cell approximately contains more than 105 copies of each RNA biomarker,29 we believe our strategy has sufficient detection sensitivity for RNA biomarker profiling in patient samples.

Validation of Detection Accuracy in Patient Samples. To validate the accuracy of our strategy for identifying T2:ERG, PCA3, and KLK2 RNA biomarkers in patient samples, we tested our colloidal strategy in 73 urine samples and validated its outcomes against a high-quality qRT-PCR reference test (Figure 4). To ensure unbiased validation, qRT-PCR experiments were performed as a blind test. For the detection of T2:ERG, we found that 73 samples clustered into two separate groups which displayed high (∆IExt at 394 nm ≥ 0.80) and low (∆IExt at 394 nm ≤ 0.58) T2:ERG levels. These results strongly agreed (95.9%) with the positive and negative T2:ERG detection outcomes of qRT-PCR (Figure 4A). For PCA3 detection, we also observed a strong agreement (97.3%) between high/low PCA3 levels of our strategy (∆IExt at 394 nm ≥ 0.88 or ≤ 0.61) and positive/negative PCA3 detection outcomes of qRT-PCR (Figure 4B). As KLK2 is predominantly expressed in prostate 8 ACS Paragon Plus Environment

Page 8 of 21

Page 9 of 21 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


cells,30 it is unsurprising that all our PCa samples indicated high ∆IExt at 394 nm (≥ 0.70) and exhibited 100% agreement with qRT-PCR (Figure 4C). This suite of experiments demonstrated the excellent agreement after head-to-head qRT-PCR reference test comparisons and thus validated the reliability of our colloidal strategy for accurately detecting RNA biomarkers from patient urine samples.

Molecular Profiling of PCa Patient Samples. Beyond analytical proof-of-concept, we seek to identify a potential clinical molecular profile for predicting PCa and high-risk PCa on biopsy. A heat map diagram (Figure 5A) was plotted based on ∆IExt at 394 nm to delineate expression profiles of aforementioned RNA biomarkers for each of the 73 urine samples. We observed that PCa patients showed considerable heterogeneity in expression levels of individual RNA biomarkers. The Gleason grading of prostate biopsies is the strongest prognostic parameter in PCa and a standard method to stratify PCa risk.31 According to the Gleason grading system, patients with Gleason score (GS) < 7 were defined as low-risk PCa and ≥ 7 as high-risk PCa.31 Based on the GS of our patient cohort, we found that the fraction of T2:ERG and PCA3 double-positive patients was higher in high-risk PCa patients (68%) than in low-risk patients (20%) and healthy patients (5%). Consistent with previous studies that used different detection methodologies,32-33 our strategy indicated that significantly higher expression levels of T2:ERG (P < 0.0001) (Figure 5B) and PCA3 (P < 0.05) (Figure 5C) were associated with high-risk PCa. On the other hand, we found that KLK2 expression levels showed no significant difference among high-risk PCa, low-risk PCa, and healthy patients (Figure 5D). Our KLK2 findings differ from previous reports which indicated that an increased serum KLK2 protein level is associated with advanced PCa tumorigenesis.34-36 This could be due to a difference in KLK2 expression levels in urine vs. serum/tissue, thus affecting biomarker prognostic ability. Nonetheless, KLK2 is a valuable endogenous prostate-specific biomarker that is useful for PCa molecular profiling to inform cellular biomarker levels of prostate origin. Finally, to perform PCa diagnosis and risk assessment, we developed binary logistic regression models based on the combined expression profile of T2:ERG, PCA3, and KLK2. An AUC value was applied to quantify the overall performance of binary logistic regression models for PCa risk stratification (the closer AUC is to 1, the better the overall diagnostic performance of the test24). Our strategy achieved AUCs of 0.790 (95% CI, 0.686 to 0.893, P = 0.0002) and 0.833 (95% CI, 0.705 to 0.962, P < 0.0001) for predicting PCa and high-risk PCa, respectively (Figure 6A,B). Our AUC data are comparably better to an earlier study37 that used T2:ERG and PCA3 for PCa molecular profiling test (AUC of 0.751 for PCa and AUC of 0.772 for high-risk PCa). Although a difference in patient cohort limited definitive comparisons,23 we anticipate that our simple “mix-to-go” colloidal 9 ACS Paragon Plus Environment


strategy for T2:ERG, PCA3, and KLK2 may provide a potential option for molecular profiling and improving PCa risk assessment.

CONCLUSIONS In conclusion, we developed a unique and simple “mix-to-go” Ag colloidal strategy without additional NP aggregation agents and applied it for PCa molecular profiling. We demonstrated the detection capabilities of our colloidal strategy to be rapid (40 min workflow), specific, sensitive (105 copies by the naked eye and 104 copies by UV-vis spectrophotometry), and accurate (> 95% agreement with qRT-PCR). In a further step to evaluate the clinical performance of our colloidal strategy, we performed PCa molecular profiling in 73 urine samples and achieved AUCs of 0.790 and 0.833 for predicting PCa and high-risk PCa on biopsy. With the development of portable optical instruments, we believed our colloidal strategy has the potential to be applied widely for rapid molecular profiling of DNA/RNA biomarkers in a variety of diseases.

AUTHOR INFORMATION Corresponding Authors *

[email protected]

*

[email protected]

Author Contributions ‡ Jing Wang and Kevin M. Koo contributed equally to this work. Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work was supported by ARC Discovery Project awarded to Y.W. and M.T. (DP160102836), and the National Breast Cancer Foundation of Australia awarded to M.T. (CG-12-07). These grants have significantly contributed to the environment to stimulate the research described here. J.W. and K.M.K are supported by Australian Government Research Training Program Scholarships. We thank Junrong Li for TEM imaging.


Page 10 of 21

Page 11 of 21 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


REFERENCES (1) Tomlins, S. A.; Rhodes, D. R.; Perner, S.; Dhanasekaran, S. M.; Mehra, R.; Sun, X. W.; Varambally, S.; Cao, X. H.; Tchinda, J.; Kuefer, R.; Lee, C.; Montie, J. E.; Shah, R. B.; Pienta, K. J.; Rubin, M. A.; Chinnaiyan, A. M., Science 2005, 310, 644-648. (2) Bussemakers, M. J.; van Bokhoven, A.; Verhaegh, G. W.; Smit, F. P.; Karthaus, H. F.; Schalken, J. A.; Debruyne, F. M.; Ru, N.; Isaacs, W. B., Cancer Res. 1999, 59, 5975-5979. (3) Nam, R. K.; Zhang, W. W.; Klotz, L. H.; Trachtenberg, J.; Jewett, M. A.; Sweet, J.; Toi, A.; Teahan, S.; Venkateswaran, V.; Sugar, L.; Loblaw, A.; Siminovitch, K.; Narod, S. A., Clin. Cancer Res. 2006, 12, 6452-6458. (4) Leyten, G. H.; Hessels, D.; Jannink, S. A.; Smit, F. P.; de Jong, H.; Cornel, E. B.; de Reijke, T. M.; Vergunst, H.; Kil, P.; Knipscheer, B. C.; van Oort, I. M.; Mulders, P. F.; Hulsbergen-van de Kaa, C. A.; Schalken, J. A., Eur. Urol. 2014, 65, 534-542. (5) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J., Nature 1996, 382, 607-609. (6) Li; Rothberg, L. J., J. Am. Chem. Soc. 2004, 126, 10958-10961. (7) Koo, K. M.; Wee, E. J.; Mainwaring, P. N.; Trau, M., Sci. Rep. 2016, 6, 30722. (8) Koo, K. M.; Wee, E. J.; Trau, M., Theranostics 2016, 6, 1415-1424. (9) Wei, H.; Li, B.; Li, J.; Wang, E.; Dong, S., Chem. Commun. 2007, 3735-3737. (10) Wei, H.; Chen, C.; Han, B.; Wang, E., Anal. Chem. 2008, 80, 7051-7055. (11) Vilela, D.; Gonzalez, M. C.; Escarpa, A., Anal. Chim. Acta 2012, 751, 24-43. (12) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A., Science 1997, 277, 1078-1081. (13) Reynolds, R. A.; Mirkin, C. A.; Letsinger, R. L., J. Am. Chem. Soc. 2000, 122, 3795-3796. (14) Cui, L.; Ke, G.; Zhang, W. Y.; Yang, C. J., Biosens. Bioelectron. 2011, 26, 2796-800. (15) Zhao, J.; Liu, T.; Fan, Q.; Li, G., Chem. Commun. 2011, 47, 5262-5264. (16) Xu, W.; Xue, X.; Li, T.; Zeng, H.; Liu, X., Angew. Chem., Int. Ed. Engl. 2009, 48, 6849-6852. (17) Xu, W.; Xie, X.; Li, D.; Yang, Z.; Li, T.; Liu, X., Small 2012, 8, 1846-1850. (18) Li, H.; Rothberg, L., Proc. Natl. Acad. Sci. 2004, 101, 14036-14039. (19) Xia, F.; Zuo, X.; Yang, R.; Xiao, Y.; Kang, D.; Vallée-Bélisle, A.; Gong, X.; Yuen, J. D.; Hsu, B. B.; Heeger, A. J., Proc. Natl. Acad. Sci. 2010, 107, 10837-10841. (20) Liu, P.; Yang, X.; Sun, S.; Wang, Q.; Wang, K.; Huang, J.; Liu, J.; He, L., Anal. Chem. 2013, 85, 7689-7695. (21) Koo, K. M.; Sina, A. A. I.; Carrascosa, L. G.; Shiddiky, M. J. A.; Trau, M., Anal. Methods 2015, 7, 7042-7054. (22) van Lierop, D.; Krpetić, Ž.; Guerrini, L.; Larmour, I. A.; Dougan, J. A.; Faulds, K.; Graham, D., Chem. Commun. 2012, 48, 8192-8194. (23) Hajian-Tilaki, K., Caspian J. Intern. Med. 2013, 4, 627-635. (24) Park, S. H.; Goo, J. M.; Jo, C.-H., Korean J. Radiol. 2004, 5, 11-18. (25) Wang, J.; Koo, K. M.; Wee, E. J.; Wang, Y.; Trau, M., Nanoscale 2017, 9, 3496-3503. (26) Koo, K. M.; Wang, J.; Richards, R. S.; Farrell, A.; Yaxley, J. W.; Samaratunga, H.; Teloken, P. E.; Roberts, M. J.; Coughlin, G. D.; Lavin, M. F.; Mainwaring, P. N.; Wang, Y.; Gardiner, R. A.; Trau, M., ACS Nano 2018, 12, 8362-8371. (27) Jain, P. K.; Huang, X.; El-Sayed, I. H.; El-Sayed, M. A., Acc. Chem. Res. 2008, 41, 1578-1586. (28) Heemers, H. V.; Regan, K. M.; Schmidt, L. J.; Anderson, S. K.; Ballman, K. V.; Tindall, D. J., Mol. Endocrinol. 2009, 23, 572-583. (29) Velculescu, V. E.; Madden, S. L.; Zhang, L.; Lash, A. E.; Yu, J.; Rago, C.; Lal, A.; Wang, C. J.; Beaudry, G. A.; Ciriello, K. M.; Cook, B. P.; Dufault, M. R.; Ferguson, A. T.; Gao, Y.; He, T. C.; Hermeking, H.; Hiraldo, S. K.; Hwang, P. M.; Lopez, M. A.; Luderer, H. F.; Mathews, B.; Petroziello, J. M.; Polyak, K.; Zawel, L.; Kinzler, K. W.; et al., Nat. Genet. 1999, 23, 387-388. (30) Mengual, L.; Lozano, J. J.; Ingelmo-Torres, M.; Izquierdo, L.; Musquera, M.; Ribal, M. J.; Alcaraz, A., BMC Cancer 2016, 16, 76. 11 ACS Paragon Plus Environment


(31) Epstein, J. I.; Egevad, L.; Amin, M. B.; Delahunt, B.; Srigley, J. R.; Humphrey, P. A.; Committee, t. G., Am. J. Surg. Pathol. 2016, 40, 244-252. (32) Leyten, G. H.; Hessels, D.; Jannink, S. A.; Smit, F. P.; de Jong, H.; Cornel, E. B.; de Reijke, T. M.; Vergunst, H.; Kil, P.; Knipscheer, B. C., Eur. Urol. 2014, 65, 534-542. (33) Tomlins, S. A.; Aubin, S. M. J.; Siddiqui, J.; Lonigro, R. J.; Sefton-Miller, L.; Miick, S.; Williamsen, S.; Hodge, P.; Meinke, J.; Blase, A.; Penabella, Y.; Day, J. R.; Varambally, R.; Han, B.; Wood, D.; Wang, L.; Sanda, M. G.; Rubin, M. A.; Rhodes, D. R.; Hollenbeck, B.; Sakamoto, K.; Silberstein, J. L.; Fradet, Y.; Amberson, J. B.; Meyers, S.; Palanisamy, N.; Rittenhouse, H.; Wei, J. T.; Groskopf, J.; Chinnaiyan, A. M., Sci. Transl. Med. 2011, 3, 94ra72. (34) Haese, A.; Becker, C.; Noldus, J.; Graefen, M.; Huland, E.; Huland, H.; Lilja, H., J. Urol. 2000, 163, 1491-1497. (35) Recker, F.; Kwiatkowski, M. K.; Piironen, T.; Pettersson, K.; Huber, A.; Lümmen, G.; Tscholl, R., Urology 2000, 55, 481-485. (36) Haese, A.; Graefen, M.; Steuber, T.; Becker, C.; Noldus, J.; Erbersdobler, A.; Huland, E.; Huland, H.; Lilja, H., J. Urol. 2003, 170, 2269-2273. (37) Tomlins, S. A.; Day, J. R.; Lonigro, R. J.; Hovelson, D. H.; Siddiqui, J.; Kunju, L. P.; Dunn, R. L.; Meyer, S.; Hodge, P.; Groskopf, J., Eur. Urol. 2016, 70, 45-53.


Page 12 of 21

Page 13 of 21 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 1. Oligonucleotide sequences Oligos

5'-Sequence-3'

T2:ERG Fwd Primer

ATTTAGGTACAACTCTTTCCCTCGTC

T2:ERG Rev Primer

GTATAGGAATCCCACTGAATTTTTC

PCA3 Fwd Primer

CCTGATGATACAGAGGTGAG

PCA3 Rev Primer

GCACAGGGCGAGGCTCATCG

KLK2 Fwd Primer

GGGGGTCCACTTGTCTGTAA

KLK2 Rev Primer

GGTGAGTTCCAAGCTTCAGG



Scheme 1. Schematic illustration of “Mix-to-Go” Ag colloidal strategy for PCa molecular profiling via RNA target panel detection. Total RNA is extracted from urine samples and isothermally amplified in parallel reactions. Negatively charged amplicons are then mixed with synthesized positively charged AgNPs to generate colloidal aggregation. The aggregation mechanism resulted in a solution color change from yellow to grey for naked-eye detection or UV-vis quantitative detection. Neg, negative; Pos, positive.


Page 14 of 21

Page 15 of 21 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 1. Aggregation behaviors of AgNPs with different sizes and surface charges after the addition of T2:ERG amplicons. (A) TEM images of Ag-1 (25 ± 5.5 nm), Ag-2 (43 ± 11 nm), and Ag-3 NPs (77 ± 22 nm). (B) Normalized weight-based PSDs and (C) zeta-potentials of Ag-1, Ag-2, and Ag-3 NPs. (D-F) Extinction spectra of Ag-1, Ag-2, and Ag-3 NPs before and after the addition of T2:ERG amplicons. The upper insets depict color photographs of the corresponding colloidal mixtures. Scale bars, 50 nm. Error bars represent the standard deviation of three technical replicates.



Figure 2. Detection specificity. (A-C) Naked eye detection, (D-F) Extinction spectra, and (G) ∆IExt at 394 nm of AgNP aggregation upon addition of amplicons, and (H) gel electrophoresis of T2:ERG, PCA3, and KLK2 amplicons from T2:ERG-positive DuCap cells, PCA3-positive LnCap cells, and androgen-treated KLK2-positive LnCap cells, respectively, and corresponding no template (No T) and no reverse transcriptase (No RT) controls. Error bars represent the standard deviation of three technical replicates.


Page 16 of 21

Page 17 of 21 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 3. Detection sensitivity. (A-C) Naked eye detection, (D-F) extinction spectra, and (G-I) ∆IExt at 394 nm upon addition of different starting amounts of synthetic T2:ERG, PCA3, and KLK2 targets. Error bars represent the standard deviation of three technical replicates.



Figure 4. Validation of detection accuracy. ∆IExt at 394 nm determined by the “mix-to-go” Ag colloidal strategy for (A) T2:ERG, (B) PCA3, and (C) KLK2 from 73 urine samples vs. qRT-PCR detection outcomes. The high/low “mix-to-go” measurements on the y-axes show strong agreement with positive/negative qRT-PCR outcomes on the x-axes.


Page 18 of 21

Page 19 of 21 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 5. PCa molecular profiling. (A) Heat map diagram of ∆IExt at 394 nm representing (B) T2:ERG, (C) PCA3, and (D) KLK2 levels in high-risk PCa (n = 19), low-risk PCa (n = 35), and healthy (n = 19) samples. Data are shown as mean ± standard deviation **P < 0.0001; *P < 0.05; NS, not significant. P values were determined by one-way ANOVA.



Figure 6. PCa detection and risk stratification. Based on combined T2:ERG, PCA3, and KLK2 expression in urine samples, ROC curves were used to determine the potential for predicting (A) PCa and (B) high-risk PCa (GS ≥ 7) on biopsy.


Page 20 of 21

Page 21 of 21 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


“Mix-to-Go” Silver Colloidal Strategy for Prostate Cancer Molecular

Recommend Documents