Direct quantification of trace amounts of a chronic myeloid leukemia

Oct 1, 2018 - Here, we describe a quantification method for the detection of BCR-ABL targets at very low concentrations (< 10 copies/sample) in the ...
0 downloads 0 Views 871KB Size
Subscriber access provided by Kaohsiung Medical University

Article

Direct quantification of trace amounts of a chronic myeloid leukemia (CML) biomarker using locked nucleic acid capture probes Sourav Mishra, Yoonhee Lee, and Joon Won Park Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b03350 • Publication Date (Web): 01 Oct 2018 Downloaded from http://pubs.acs.org on October 7, 2018

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

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

Page 1 of 25 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

Direct quantification of trace amounts of a chronic myeloid leukemia (CML) biomarker using locked nucleic acid capture probes Sourav Mishra, Yoonhee Lee† and Joon Won Park* Department of Chemistry, Pohang University of Science and Technology, 77 Cheongam-Ro, Nam-Gu, Pohang 37673, Republic of Korea

Abstract Molecular monitoring is indispensable for the clinical management of chronic myeloid leukemia (CML) patients. Real-time quantitative polymerase chain reaction (RT-qPCR) is the gold standard for the quantitative assessment of BCR-ABL transcript levels, which are critical in clinical decision-making. However, the frequent recurrence of the disease after drug discontinuation for 60% of patients has necessitated more sensitive and specific techniques to detect residual BCR-ABL transcripts. Here, we describe a quantification method for the detection of BCR-ABL targets at very low concentrations (< 10 copies/sample) in the presence of a million copies of normal BCR and ABL genes. In this method, a fully modified locked nucleic acid (LNA) and a LNA/DNA chimera were used as capture probes, and the quantitative imaging mode of atomic force microscopy (AFM) was employed. Targets with one of the major breakpoints (found in more than 95% of CML patients), b3a2 and b2a2, were quantified. The BCR-ABL target captured on a miniaturized LNA-probe spot was scanned at nanometric resolution, and the samples containing one to ten copies of the BCR-ABL genes were examined. It was observed that the highest sensitivity, i.e., the detection of a single copy of the target gene, could be achieved through multiple runs, and the observed cluster number was well correlative (adjusted R2 = 0.999) to the target copy number in the sample solution. This observation clearly demonstrates that the LNA-based platform is effective in quantifying BCR-ABL targets with extremely low copy numbers, highlighting the potential applicability of AFM for use in the direct quantification of such targets without amplification or labeling.

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 2 of 25

Introduction Chronic myeloid leukemia (CML) is a malignant clonal disorder of hematopoietic stem cells, characterized by reciprocal translocation between chromosome 9 and chromosome 22, which results in the formation of the Philadelphia (Ph) chromosome.1, 2 The BCR-ABL fusion gene is produced as a molecular consequence of this translocation event. The diagnosis of CML is based on the detection of the BCR-ABL gene, and different fusion proteins which are produced depending on the breakpoint in the BCR gene. Typically, there are three breakpoint cluster regions in the BCR gene, namely major (M-bcr), minor (m-bcr), and micro (µ-bcr).3 A breakpoint in the M-bcr region of the BCR gene is observed in more than 95% of CML patients, and two major breakpoints are b3a2 (e14a2) and b2a2 (e13a2).4 The percentage of CML patients corresponding to a particular breakpoint (either b3a2 or b2a2) varies depending on the nature of the population, and the occurrence of co-expression is rare. Several recent studies point towards the fact that only 40% of CML patients with undetectable BCR-ABL transcripts on tyrosine kinase inhibitor therapy remain disease-free after drug discontinuation,5,

6

whereas 60% of

patients suffer from relapse of the disease. In this regard, the development of quantification techniques with an enhanced detection limits has become necessary to track residual disease and to determine whether the therapy is complete. Furthermore, real-time quantitative polymerase chain reaction (RT-qPCR) has an edge as a sensitive molecular biological tool for diagnostics;7, 8 this technique can precisely quantify nucleic acids by tracking fluorescence intensity, which is exponentially correlated with the amplification cycle. In spite of being an indispensable molecular biological tool, there are limitations for RT-qPCR to follow low-copy number transcripts (fewer than 10 copies) because false positive signals can occur due to errors in amplification, such as the formation of primer dimers.9 The amplification efficiency of RT-qPCR depends on the structure of the nucleic acid, and calibration errors between targets and standards also contribute to the uncertainty in this method. Digital PCR is introduced as an alternative technique to overcome such limitations.10 The target molecules are separated into a large number of partitions or droplets and are amplified in parallel; target concentrations are determined by counting positive partitions after the completion of the PCR. However, apart from the subsampling error (including insufficient reagents in some partitions), correct partition volume is a crucial factor to calculate the number of targets, and potential errors in volume can occur that can create non-trivial bias for the absolute quantification.11 Therefore, new-generation

ACS Paragon Plus Environment

Page 3 of 25 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

technological advancements have focused on lab-on-chip based approaches,12 optofluidic approaches,13 and direct specific detection methods14, 15 of DNA or RNA for use in molecular diagnostics without amplification. Furthermore, recent state-of-the-art technology for the visualization of single molecule, such as single molecule force spectroscopy and single molecule fluorescence microscopy, has been explored. Among the single molecule methods, atomic force microscopy (AFM)-based single molecule force spectroscopy represents a particularly valuable methodology to investigate important biological events, allowing for the probing of intra- and intermolecular forces with high sensitivity in near-physiological conditions and without any labeling. Specifically, AFM is routinely used as a tool to probe DNA hybridization event at the single molecule level.16,

17

Force-distance-curve-based imaging represents another important

aspect of AFM operation, which generates a map of F-D curves over sample surfaces and simultaneously correlates them with the topographic images within a reasonable time-frame.18, 19 It allows the localization of on-surface molecular interactions at nanometric resolution.20,

21

Recently, Lee et al.22 explored the capability of AFM-force mapping to quantify a lower copy number of synthetic BCR-ABL target genes (b2a2) in a DNA-based assay. Although the achieved sensitivity limit was encouraging with this approach, the specificity was poor because of non-negligible interference from the normal genes (BCR and ABL). This outcome certainly necessitates more specific probe designs so that these interferences can be eliminated through enhanced discrimination. Among the available alternative nucleic acids, locked nucleic acid (LNA) has emerged as a potentially better choice for use in DNA probes in the last two decades.23,

24

It contains a

modified ribose moiety in which the 2′-oxygen and 4′-carbon are linked by a methylene bridge, locking the sugar in an N-type conformation. LNA has an exceptionally high affinity and specificity towards its complementary natural nucleic acid analogues (DNA/RNA).23, 24 It was reported that surface-tethered LNA probes could capture PCR amplicons very efficiently with excellent sensitivity and specificity.25 The substitution of LNA into the DNA oligonucleotide capture probes resulted in noticeable enhancement in the discrimination among highly similar (90% sequence identity) mRNAs with concurrent increase in capture sensitivity.26 A miChip using LNA capture probes was also reported for microRNA expression profiling, which overcame several issues attributed to monitoring miRNA expression levels.27 Recently, Mishra et al. explored the applicability of an optimized LNA-based sensing platform for single nucleobase

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

mismatch discrimination, using AFM-based single molecule force spectroscopy (SMFS) approach.28 Herein, we report the quantification of BCR-ABL genes with the major breakpoint (b3a2 or b2a2) at very low concentrations (˂ 10 copies/sample) in the presence of a million copies of normal genes (particularly BCR and ABL, relevant for clinical samples) by force-distance (FD) curve- based atomic force microscopy onto an LNA-based assay. The LNA-modified capture probes were designed in such a way that they can specifically bind to the junctional region of their respective BCR-ABL target with unprecedented specificity among the normal genes, unlike the DNA capture probe. Both synthetic BCR-ABL genes containing b3a2 or b2a2 junction were considered as the target probes, as they are common to more than 95% of CML patients worldwide. To count all the targets captured on the surface directly, the entire surface was scanned. This was ensured by fabricating miniaturized LNA-modified probe spots (~2 µm).

Experimental Section Preparation of LNA-modified capture probes: Two LNA-modified capture probes were designed to recognize specifically the translocated junction (b3a2 and b2a2) of the respective BCR-ABL target. A 12-mer fully modified LNA capture probe was employed to recognize BCRABL target containing a b3a2 junction. In contrast, a 21-mer LNA/DNA chimeric capture probe was designed to recognize the BCR-ABL target with a b2a2 junction. Both LNA-modified capture probes were custom-synthesized (Exiqon, Denmark). In addition, the capture probes were functionalized with an amine group at the 3′-end for substrate immobilization. Preparation of BCR-ABL targets, normal BCR and ABL sequences, and detection DNAs: All the DNA was custom-synthesized (Integrated DNA Technologies Inc.). The BCR-ABL targets of 160-mer and 154-mer consist of sequences of two major BCR-ABL transcripts, b3a2 and b2a2, respectively. Normal BCR and ABL (155-mer) were employed with respect to the b3a2 junction to assess the specificity. Similarly, normal BCR and ABL (148-mer) were employed with respect to the b2a2 junction. The detection DNAs (20-mer) were designed to bind the BCR region of the respective BCR-ABL targets and functionalized with an amine group at the 5′-end for the immobilization on the AFM tip.

ACS Paragon Plus Environment

Page 4 of 25

Page 5 of 25 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 patterned glass slides, capture probe fabrication, target hybridization and AFM tip modification: All these experimental details are provided in the Supporting Information. Quantitative imaging with AFM and data analysis: All the adhesion force mapping experiments were performed in the quantitative imaging (QI) mode using NanoWizard 3 atomic force microscope (JPK Instrument). The detail descriptions are provided in the Supporting Information.

Results Designing LNA-modified capture probes towards enhanced specificity: Capture probe designing towards the quantification of the BCR-ABL gene for its two major breakpoint junctions, with enhanced sensitivity and specificity in the presence of a million copies of normal genes is one of the primary aspects of the present study. We considered differences in the melting temperature for different probe-target duplexes as a measure of specificity for optimal probe designing. We aimed to exploit the maximum possible melting temperature difference between the specific and non-specific probe-target duplexes to attain a higher detection specificity. LNA was the preferred choice in this regard, as the thermal stability of a duplex could be enhanced from +2 °C to +8 °C per substitution of LNA monomers depending on the length and the sequence of the probe.23 However, the effect of LNA substitution decreases with increases in the probe length.29 Here, the probe sequences were designed based on the sequence of the BCR-ABL transcripts for their two major breakpoints: b3a2 and b2a2. The probe length was determined by considering the number of overlapping bases of both BCR and ABL at the junction and its capability of maximizing the melting temperature differences. In the case of b3a2, there is a single base overlap in the junction region (Supplementary Figure S1A). In the case of b2a2, the junction region is comprised by an overlap of six bases (Supplementary Figure S2A). In general, the probe length was kept short to exploit the maximum effect of LNA substitution, shorter length oligonucleotides have a tendency to organize in end-tethered and highly extended configuration,30 and duplex stabilization is best achieved with a short oligonucleotide.31 Additionally, it is worth noting that use of fully modified LNA probes is effective, as long as the probe length is kept within 15 bases.

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

In the case of the b3a2 junction, as there is only a single base overlap in the junction region, we started capture probe designing by putting a complementary LNA base first to that particular base and then sequentially added complementary LNA bases to both directions. A 6-mer fully modified LNA capture probe, fully complementary to the b3a2 junction, was considered to be the capture probe to start with because shorter nucleic acid duplexes (< 4 base pairs) were thermodynamically unstable. The melting temperatures for specific (probe-BCR-ABL duplex) and non-specific (probe-BCR duplex, probe-ABL duplex) duplexes were estimated via the melting temperature prediction tool of Exiqon. The estimated melting temperature of 32 °C ([NaCl]: 115 mM) for the fully complementary specific probe-target duplex turned out to be inappropriate because typical hybridization temperature recommended for the nucleic acid hybridization should be 20 °C below the calculated melting temperature of the duplex to achieve a near-maximum hybridization rate.32, 33 Particularly, in the case of the LNA probes (different lengths), the typical hybridization temperatures recommended should be between 15 °C – 30 °C below the melting temperature of the duplex.34 Hence, we increased the number of LNA bases in the capture probe sequentially and estimated the melting temperature difference between the specific and non-specific duplexes. A 12-mer fully modified LNA capture probe was designed, and it has melting temperature of 83 °C to BCR-ABL gene (12-base pair, b3a2 junction), 42 °C (7-base pair) and 48 °C (6-base pair) to normal BCR gene, and the normal ABL gene, respectively (Supplementary Figure S1C). A similar approach was followed for the capture probe design of the b2a2 junction. As the junction comprised by an overlapping region of six bases, even a fully modified LNA capture probe at its maximum permissible length (15-mer) was proven to be insufficient for our purpose (the melting temperature differences between the specific and non-specific complexes was too close). Longer lengths of fully modified LNA capture probes were not appropriate because of the self-complementarity issue. Hence, we designed a LNA/DNA chimeric capture probe that was suitable to our purpose. To our end up to 60% of the bases of an oligonucleotide sequence were modified with LNA for the probe length of 16-25-mer, and use of more than four consecutive LNA bases was avoided.29, 35 Melting temperatures for the specific and non-specific probe-target complexes were calculated to find an optimal probe. A 21-mer LNA/DNA chimeric capture probe was found, and it has melting temperature of 84 °C to BCR-ABL gene (21-base pair, b2a2

ACS Paragon Plus Environment

Page 6 of 25

Page 7 of 25 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

junction), 67 °C (15-base pair) and 67 °C (12-base pair) towards normal BCR gene and normal ABL gene, respectively (Supplementary Figure S2C). The extent of self-complementarity and the stability of the secondary structures for both capture probes were verified by the LNA oligo optimizer tool provided by Exiqon, and the tool showed that the probes work well at elevated hybridization temperatures (approximately 20 °C below than the melting temperature of the specific probe-target duplex). In addition, for both the capture probe sequences, a hexyl spacer [-(CH2)6-] was introduced at the 3′-end, because it provides flexibility to enable efficient target capture. Probe spot fabrication and the hydrodynamic radius of individual surface-captured target molecules: FluidFM technology was used to fabricate miniaturized capture probe spots, where a microchanneled cantilever equipped with a pyramidal tip of 800 nm aperture was employed.36 The microchannel was filled with the capture probe solution, and the LNA-modified capture probe was spotted onto a photolithographically etched and activated with N-hydroxysuccinimide (NHS) glass slide, at a position with known (x, y) coordinates. The typical spot diameter was in the range of 1.5 µm to 2.2 µm. Force-distance (FD) curves were collected at all pixels of the LNA-modified probe spot, and the AFM tip recognized the surface-captured target molecule (BCR-ABL) only when the tip entered the area defined by the hydrodynamic radius of the target-DNA (Figure 1). Therefore, measurement of the hydrodynamic radius of the surface-captured target molecule is the starting point to determine the optimal pixel size; too small pixel sizes will lead to longer mapping time, while larger pixel sizes will increase the probability of missing the target. For the latter case, each captured target will be shown as a single pixel at best. Additionally, it becomes difficult to distinguish specific pixels (cluster-forming) from randomly distributed non-specific pixels (noncluster-forming). As described earlier, the LNA-modified capture probes (12-mer/21-mer) were designed to specifically bind to BCR-ABL targets at its breakpoint junction (b3a2/b2a2) (Supplementary Figures S1A and S2A). The detection DNA (20-mer), complementary to the BCR region of the respective targets, was tethered to the AFM tips (Supplementary Figures S1A and S2A). When the AFM tip approached the captured target, the detection probe hybridized with the target. Upon retraction of the tip, the target was stretched until the rupture of the relatively weaker duplex occurred (on the AFM tip side). High-resolution adhesion force maps to

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

determine the right pixel size were acquired at each 2-3 nm. Specific F-D curves with relevant molecular stretching were observed in the map, and the most probable adhesion force and stretching distance were calculated. For the target BCR-ABL sequence containing b3a2 junction, the average value of the most probable adhesion force and stretching distance were estimated to be 35.2 ± 2.5 pN and 7.2 ± 1.7 nm, respectively, from three different locations. Histograms of the most probable adhesion force and stretching distance from one representative measurement are presented in Supplementary Figures S3A and S3B. Similarly, the average value of the most probable adhesion force and stretching distance were estimated to be 33.1 ± 4.9 pN and 9.0 ± 0.8 nm, respectively, for the BCR-ABL target with a b2a2 junction (Supplementary Figure S4A). It is apparent that the adhesion force and stretching distance are comparable for both cases. The pixels attributed to the specific adhesion are collected in a circular region implying the motion of a tethered DNA (Figure 1 and Supplementary Figure S4B). To assess the reliability of the measurement, we performed a control experiment by mapping a blank (without exposure to the target) spot first, followed by the hybridization with the target molecules (BCR-ABL having b3a2 junction). We then mapped it again with the quantitative imaging (QI) mode. The absence of characteristic clusters for the blank LNA spot and the occurrence of positive clusters after the hybridization with the target on the same spot confirmed the validity of the approach (Supplementary Figure S5). The hydrodynamic radius of the target molecules was estimated by measuring the cluster radius (Rc) via ellipse fitting.37 For both targets, three to six QI maps were acquired at a position, and three successive maps were superimposed to generate one or two overlaid maps for each position; five overlaid maps were generated for each type of target molecules from three different positions. The average cluster radii of 13.6 ± 1.4 nm and 21.9 ± 1.7 nm were estimated for BCR-ABL targets related to the b3a2 junction and b2a2 junction, respectively. It is worth mentioning that for both the cases (b3a2 and b2a2), the observed cluster radii were larger than those suggested by the observed stretching distance. In particular, the average cluster radius for two different breakpoint specific BCR-ABL targets differed significantly, although the average stretching distances were comparable. This can be accounted by considering two major factors: contribution from the rigid duplex part (generated via capture probe-target hybridization) to the hydrodynamic radius and the flexibility of the secondary structure of the target after capture probe binding at room temperature. A rigid rod like behavior of the duplex part is expected.38

ACS Paragon Plus Environment

Page 8 of 25

Page 9 of 25 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

Thus, a 21-mer duplex, owing to probe-target hybridization, for b2a2 junction, certainly contributes more (~ 6.3 nm) to its hydrodynamic radius compared to a 12-mer duplex (~ 3.4 nm) for the b3a2 junction. The formation of fully modified LNA-DNA duplex (A-type) in the case of b3a2 junction certainly reduced the contribution further. In addition, relatively more compact secondary structure (with a greater number of bases involved in the hairpin formation; see Supplementary Figures S6 and S7 as obtained from the UNAFold program (Integrated DNA Technologies Inc.)) of the target molecule for b3a2 junction, compared to b2a2 junction, results in a smaller cluster radius. Based on these observations, the optimal pixel size for the quantitative imaging of entire LNAmodified probe spot was determined to visualize individual surface-captured target molecules. Examining a spot (2 µm in diameter) at each 5 nm (400 × 400 pixels) takes 65 min (at scan speed of 18 µm/sec). We typically scanned an area slightly larger than the spot size to ensure that we covered the entire area, and for efficiency, a pixel size of 6.0 nm for b3a2 junction and that of 7.0 nm for b2a2 junction were adopted. Optimizing hybridization temperature: Optimization of the hybridization temperature is another key aspect of this study, as we aimed to exploit the melting temperature difference between the specific and non-specific duplexes for the enhanced specificity. An optimal hybridization temperature window was identified in accordance with the melting temperature of a particular probe-target duplex, in which the formation of specific probe-target duplexes can be achieved at a near-maximum hybridization rate with the least possibility of simultaneous nonspecific probe-target hybridization (Supplementary Figure S8). In-depth descriptions are provided in the Supporting Information. Superimposing the quantitative imaging maps and validating the positive cluster: Three consecutive QI maps were collected over a probe spot area and were overlaid to improve the reliability of the measurement.39 In the QI mode, we simultaneously recorded the topography, slope and adhesion map over the sample surface. The recorded data was then converted into three characteristic two-dimensional (2D) images. As reported earlier, by virtue of the height images, we estimated the lateral drift (in terms of pixels) between the first two images and between the last two images with an in-house MATLAB program.22 For both BCR-ABL targets (b3a2 and b2a2), the FD curves correspond to every adhesion map were filtered to sort out those

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

having the appropriate adhesion force value (≤ 50 pN) and specific stretching distance (7-40 nm). The respective 2D images were then generated to visualize the positive pixels. After drift correction, three consecutive specific adhesion maps were then overlaid. In the overlaid specific adhesion map, pixels correspond to a single detection event are colored in green, and pixels with two or three specific detection events are colored in red. Because of the high detection probability at the center of the clusters, red pixels occurred frequently at the center region of a cluster in the overlaid maps for both cases (Figures 2A and 2B). As negative controls, we first exposed 100 fM (100 × 10-15 M, 2.2 × 106 copies) normal BCR gene to the LNA-modified capture probe spot capable of capturing genes with the b3a2 junction. We found that there was no positive cluster for which the radius was close to that which was previously defined (Figure 2C). In contrast, the ABL part of the BCR-ABL fusion gene should be silent during our AFMbased quantification approach (Supplementary Figures S1 and S2). However, the hybridization with ABL should be avoided, because ABL gene can block the capture probe sites. No interference from the ABL gene was verified through the quantitative analysis of the translocated genes (vide infra). In the case of the b2a2 junction, we exposed the LNA-modified capture probe spot to a target solution containing 100 fM of the normal BCR and ABL genes. As with the earlier observation, the occurrence of red pixels was rare, and even the largest cluster was too small to be qualified (Rc = 8.3 nm) (Figure 2D). Based on this observation, we were able to set out sound criteria to assign real positive clusters for both BCR-ABL targets. First, in order to be qualified as a valid cluster in both cases, a cluster must have at least one pixel where repetitive detection events was observed. Second, for the b3a2 junction, the cluster radius must be over 11.0 nm, and for the b2a2 junction, the cluster radius must be over 16.0 nm. The radii were slightly smaller than those of average hydrodynamic radii (13.6 nm and 21.9 nm) obtained from high-resolution maps, but because the pixel size was 5.0 - 6.0 nm and 5.0 - 7.0 nm for b3a2 and b2a2, respectively, this discrepancy was understandable. These selection criteria were consistently followed for the quantitative analysis of all samples. Analysis of samples containing BCR-ABL targets of less than ten copies in the presence of a million copies of normal BCR and ABL genes: The BCR-ABL target (b3a2 and b2a2) solutions of 45 zM (45 × 10-21 M, 40 µL, one copy), 90 zM (40 µL, two copies), 225 zM (40 µL, five copies), and 450 zM (40 µL, ten copies) were prepared by serial dilution, and each solution was exposed to a designated LNA-modified capture probe spot along with 100 fM (2.2 × 106

ACS Paragon Plus Environment

Page 10 of 25

Page 11 of 25 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

copies) of the respective normal BCR and ABL genes (1:1 ratio). As reasoned earlier, all the probe-target hybridization reactions were carried out at 60 °C for 24 h to ensure the effective capture of the particular targets. To account the experimental error related to sample preparation, fluctuations in the capture/detection efficiency, and other factors affecting the quantification, the experiments were repeated several times using independently prepared samples. For the samples containing a single copy or two copies of the target molecule, the experiments were repeated five times, and, for the samples with five copies or ten copies of the target molecule, three replicates were performed (Figures 3A and 3B). For both BCR-ABL targets, valid clusters were counted based on the earlier mentioned criteria, and the average number of clusters was plotted with respect to the target copy number in solution (Supplementary Figures S9A and S9B). For samples with a zero-copy number of target molecules (BCR-ABL), no cluster was quantified for any of the runs. For the samples containing a single copy of the target molecule, a single count of the cluster was detected (Figures 4A and 4B) once during the five replicate experiments irrespective of the type of breakpoint junction. Similarly, for the samples with two copies of the target molecule, one or two clusters were observed during two of the five repeat experiments (Figures 4C and 4D). For the samples containing five or ten copies, positive cluster(s) were always detected in all replicate examinations. The average numbers of the clusters were observed to be 2.33 and 4.67 for the b3a2 junction and 2.0 and 4.33 for the b2a2 junction. In both cases, the observed number of positive clusters was well correlated with the number of target copies in the sample solution (adjusted R2 of 0.999 for both b3a2 and b2a2 junction, linear regression model). In the case of the b3a2 junction, the detection efficiency was 47% for both samples containing five and ten copies of the target molecules. For the b2a2 junction, it was 40% and 43%, respectively, for five and ten copies of the target molecules. For the samples with a single copy of the target molecule, the detection efficiency was 20%, irrespective of the type of breakpoint. The difference in detection efficiency for two different BCR-ABL targets (b3a2 and b2a2 junction) was statistically analyzed using a two-tailed t-test of the data sets. The observed P-value of 0.8520 (P ˃ 0.05) implies that the observed difference in detection efficiency is not statistically significant. In addition, such detection efficiency for all cases is consistent with the previously observed efficiency22 in the absence of normal genes. It is evident that to detect such low copy numbers of BCR-ABL targets specifically in the presence of outnumbered of normal BCR and ABL genes (relevant for clinical samples), multiple runs are essential, and the average

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

number of the cluster from the map is linear to the copy number of the translocated genes in solution.

Discussion A new quantification method for BCR-ABL targets (with b3a2 and b2a2 breakpoints) at very low concentrations (< 10 copies/sample) in the presence of a million copies of normal BCR and ABL genes is examined using the quantitative imaging mode of AFM along with the advent of a specifically designed LNA-based capture probes. Relatively higher melting temperature differences between the specific and non-specific probe-target duplexes owing to LNA substitutions were exploited by carrying out hybridization and washing at an elevated temperature, in which the possibility of competing hybridization was nullified. The hydrodynamic radius of the captured BCR-ABL target molecules on the surface was utilized as a key criterion to sort out the qualified clusters. The most noteworthy aspect of the current approach is the specificity of the detection achieved along with its sensitivity limit. In this study, by virtue of the LNA-modified capture probes, we were able quantify a few copies of the BCR-ABL targets in the presence of a million copies of normal BCR and ABL genes, unlike DNA-based quantification. This is particularly important because quantitative assessment of minimal residual disease (MRD) is critical in the molecular monitoring of CML. The standard techniques used to track MRD count the number of copies of BCR-ABL gene with respect to the normal genes in the sample. As the current approach is direct, devoid of any sorts of modification of the target molecules (including labeling), amplification, it will further reduce the uncertainty in the quantification of low-copy number BCR-ABL transcripts. However, the overall detection efficiency of approximately 50% for the b3a2 junction and 45% for the b2a2 junction (from the slope of supplementary Figures S9A and S9B) and the necessity for multiple runs are still issues. These shortcomings might be improved by optimizing the salt concentration and type of cations present during the hybridization, as it was reported that the formation of the LNA-DNA duplex on the surface is largely impacted by these factors.28,

40

As we are particularly interested in the quantification of very low copy

numbers of BCR-ABL targets, improvements in consistent and enhanced transport (or mixing) seem to be the next step forward. Because apparently lower variation in mapping was observed, optimizing mapping parameters, such as pixel size, scan speed, and the number of repetitions,

ACS Paragon Plus Environment

Page 12 of 25

Page 13 of 25 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

should help to further decrease the coefficient of variation of the measurement. The use of microfluidic devices during the sample preparation and handling could be considered a viable choice to avoid sample loss as well as to reduce uncertainty about the copy numbers due to the serial dilution.41

Conclusion In summary, this study provides us with a platform to quantify BCR-ABL targets (for two major breakpoints, b3a2 and b2a2) with less than ten copies in a clinically relevant environment. The applicability of AFM is particularly encouraging in this regard, as it allows the direct quantification of targets without any modification of the target molecules, amplification, or labeling. LNA in particular certainly holds the promise to be a better choice to surpass the limitation of conventional DNA-based assays towards the molecular monitoring of CML to assess the minimal residual disease (MRD).

Associated Content Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Additional text describing experimental details, LNA-modified capture probes, detection and target DNAs, and normal BCR and ABL sequences used in this study, melting temperature of the respective specific and non-specific probe-target duplexes, distributions of most probable adhesion force and stretching distance related to the localization of individual targets, control experiment, secondary structure of target DNAs, experiment to optimize hybridization temperature, plots showing target cluster correlation (PDF)

Author Information Corresponding Author *[email protected] Present Addresses

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 14 of 25

Department of Electrical Engineering, Columbia University, 500 W. 120th Street, New York

10027, USA Notes The authors declare no competing financial interest.

Acknowledgement This

work was supported

by

the

National

Research

2016K2A9A1A03904662, NRF-2017R1A2B3008478),

and

Foundation of Brain

Research

Korea

(NRF-

Program

of

NRF (2015M3C7A1030964).

References 1. Nowell, P. C.; Hungerford, D. A. A minute chromosome in human chronic granulcytic leukaemia. Science 1960, 132, 1497-1499. 2. Rowley, J. D. A new consistent chromosomal abnormality in chronic myeloid leukemia identified by Quinacrine fluorescence and Giemsa staining. Nature 1973, 243, 290-293. 3. Melo, J. V. The diversity of BCR-ABL fusion proteins and their relationship to leukemia phenotype. Blood 1996, 88, 2375-2384. 4. Heisterkamp, N.; Stam, K.; Groffen, J.; de Klein, A.; Grosveld, G. Structural organization of the BCR gene and its role in the Ph’ translocation. Nature 1985, 315, 758-761. 5. Mahon, F. X.; Rea, D.; Guilhot, J.; Guilhot, F.; Huguet, F.; Nicolini, F.; Legros, L.; Charbonnier, A.; Guerci, A.; Varet, B.; Etienne, G.; Reiffers, J.; Rousselot, P. Intergroupe Francais des Leucemies Myeloides Chroniques: Discontinuation of imatinib in patients with chronic myeloid leukemia who have maintained complete molecular remission for at least 2 years: the prospective, multicentre Stop Imatinib (STIM) trial. Lancet Oncol. 2010, 11, 10291035. 6. Ross, D. M.; Branford, S.; Seymour, J. F.; Schwarer, A. P.; Arthur, C.; Yeung, D. T.; Dang, P.; Goyne, J. M.; Slader, C.; Filshie, R. J.; Mills, A. K.; Melo, J. V.; White, D. L.; Grigg, A. P.; Hughes, T. P. Safety and efficacy of imatinib cessation for CML patients with stable

ACS Paragon Plus Environment

Page 15 of 25 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

undetectable minimal residual disease: results from the TWISTER study. Blood 2013, 122, 515522. 7. Heid, C. A.; Stevens, J.; Livak, K. J.; Williams, P. M. Real time quantitative PCR. Genome Res. 1996, 6, 986-994. 8. VanGuilder, H. D.; Vrana, K. E.; Freeman, W. M. Twenty-five years of quantitative PCR for gene expression analysis. Biotechniques 2008, 44, 619-626. 9. Hughes, T.; Janssen, J. W. G.; Morgan, G.; Martiat, P.; Saglio, G.; Pignon, J. M.; Pignatti, F.; Mills, K.; Keating, A.; Gluckman, E.; Bartram, C. R.; Goldman, J. M. False-positive results with PCR to detect leukaemia-specific transcript. Lancet 1990, 335, 1037-1038. 10. Vogelstein, B.; Kinzler, K. W. Digital PCR. Proc. Natl. Acad. Sci. USA 1999, 96, 9236-9241. 11. Jacobs, B. K. M.; Goetghebeur, E.; Clement, L. Impact of variance components on reliability of absolute quantification using digital PCR. BMC Bioinformatics 2014, 15, 283. 12. Craighead, H. Future lab-on-a-chip technologies for interrogating individual molecules. Nature 2006, 442, 387-393. 13. Psaltis, D.; Quake, S. R.; Yang, C. Developing optofluidic technology through the fusion of microfluidics and optics. Nature 2006, 442, 381-386. 14. Yahiatene, I.; Doose, S.; Huser, T.; Sauer, M. Correlation-matrix analysis of two color coincidence events in single-molecule fluorescence experiments. Anal. Chem. 2012, 84, 27292736. 15. Nam, J. M.; Stoeva, S. I.; Mirkin, C. A. Bio-bar-code-based DNA detection with PCR-like sensitivity. J. Am. Chem. Soc. 2004, 126, 5932-5933. 16. Lee, G. U.; Chrisey, L. A.; Colton, R. J. Direct measurement of the forces between complementary strands of DNA. Science 1994, 266, 771-773. 17. Strunz, T.; Oroszlan, K.; Schafer, R.; Guntherodt, H. J. Dynamic force spectroscopy of single DNA molecules. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 11277-11282.

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

18. Dufrene, Y. F.; Martinez-Martin, D.; Medalsy, I.; Alsteens, D.; Muller, D. J. Multiparametric imaging of biological systems by force-distance curve-based AFM. Nat. Methods 2013, 10, 847854. 19. Kim, Y.; Kim, W.; Park, J. W. Principles and applications of force spectroscopy using atomic force microscopy. Bull. Korean Chem. Soc. 2016, 37, 1895-1907. 20. Jung, Y. J.; Albrecht, J. A.; Kwak, J. W.; Park, J. W. Direct quantitative analysis of HCV RNA by atomic force microscopy without labeling or amplification. Nucleic Acids Res. 2012, 40, 11728-11736. 21. Iyer, S.; Gaikwad, R. M.; Subba-Rao, V.; Woodworth, C. D.; Sokolov, I. AFM detects differences in the surface brush of normal and cancerous cervical cells. Nat. Nanotechnol. 2009, 4, 389-393. 22. Lee, Y.; Kim, Y.; Lee, D.; Roy, D.; Park, J. W. Quantification of fewer than ten copies of a DNA biomarker without amplification or labeling. J. Am. Chem. Soc. 2016, 138, 7075-7081. 23. Singh, S. K.; Nielsen, P.; Koshkin, A. A.; Wengel, J. LNA (locked nucleic acids): synthesis and high-affinity nucleic acid recognition. Chem. Commun. 1998, 4, 455-456. 24. Koshkin, A. A.; Nielsen, P.; Singh, S. K.; Wengel, J. LNA (Locked Nucleic Acid): an RNA mimic forming exceedingly stable LNA:LNA duplexes. J. Am. Chem. Soc. 1998, 120, 1325213253. 25. Orum, H.; Jakobsen, M. H.; Koch, T.; Vuust, J.; Borre, M. B. Detection of the factor V Leiden mutation by direct allelespecific hybridization of PCR amplicons to photoimmobilized locked nucleic acids. Clin. Chem. 1999, 45, 1898-1905. 26. Kauppinen, S.; Nielsen, P. S.; Mouritzen, P.; Nielsen, A. T.; Vissing, H.; Møller, S.; Ramsing, N. B. LNA microarrays in genomics. Pharma Genomics 2003, 3, 24-34. 27. Castoldi, M.; Schmidt, S.; Benes, V.; Hentze, M. W.; Muckenthaler, M. U. miChip: an arraybased method for microRNA expression profiling using locked nucleic acid capture probes. Nat. Protocol. 2008, 3, 321-329.

ACS Paragon Plus Environment

Page 16 of 25

Page 17 of 25 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

28. Mishra, S.; Lahiri, H.; Banerjee, S.; Mukhopadhyay, R. Molecularly resolved label-free sensing of single nucleobase mismatches by interfacial LNA probes. Nucleic Acids Res. 2016, 44, 3739-3749. 29. You, Y.; Moreira, B. G.; Behlke, M. A.; Owczarzy, R. Design of LNA probes that improve mismatch discrimination. Nucleic Acids Res. 2006, 34, e60. 30. Steel, A. B.; Levicky, R. L.; Herne, T. M.; Tarlov, M. J. Immobilization of nucleic acids at solid surfaces: effect of oligonucleotide length on layer assembly. Biophys. J. 2000, 79, 975-981. 31. Springer, T.; S'ıpova, H.; Stepanek, J.; Homola, J. Shielding effect of monovalent and divalent cations on solid-phase DNA hybridization: surface plasmon resonance biosensor study. Nucleic Acids Res. 2010, 38, 7343-7351. 32. Anderson, M. L. M.; Young, B. D. Quantitative filter hybridization. In Nucleic Acid Hybridization: A Practical Approach; Hames, B. D., Higgins, S. J., Eds.; IRL Press, Oxford, 1985; pp. 73-111. 33. Beltz, G. A.; Jacobs, K. A.; Eickbush, T. H.; Cherbas, P. T.; Kafatos, F. Isolation of multigene families and determination of homologies by filter hybridization methods. Methods Enzymol. 1983, 100, 266-285. 34. Fontenete, S.; Guimarães, N.; Leite, M.; Figueiredo, C.; Wengel, J.; Azevedo, N. F. Hybridization-based detection of Helicobacter pylori at human body temperature using advanced locked nucleic acid (LNA) probes. PLoS ONE 2013, 8, e81230. 35. Lundin, K. E.; Højland, T.; Hansen, B. R.; Persson, R.; Bramsen, J. B.; Kjems, J.; Koch, T.; Wengel, J.; Smith, C. I. Biological activity and biotechnological aspects of locked nucleic acids. Adv. Genet. 2013, 82, 47-107.

36. Gruter, R. R.; Voros, J.; Zambelli, T. FluidFM as a lithography tool in liquid: spatially controlled deposition of fluorescent nanoparticles. Nanoscale 2013, 5, 1097-1104. 37. Lee, Y.; Kwon, S. H.; Kim, Y.; Lee, J. B.; Park, J. W. Mapping of surface-immobilized DNA with force-based atomic force microscopy. Anal. Chem. 2013, 85, 4045-4050.

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

38. Baumann, C. G.; Smith, S. B.; Bloomfield, V. A.; Bustamante, C. Ionic effects on the elasticity of single DNA molecules. Proc. Natl. Acad. Sci. U. S. A. 1997, 94, 6185-6190. 39. Pfreundschuh, M.; Alsteens, D.; Wieneke, R.; Zhang, C.; Coughlin, S. R.; Tampe, R.; Kobilka, B. K.; Muller, D. J. Identifying and quantifying two ligand-binding sites while imaging native human membrane receptors by AFM. Nat. Commun. 2015, 6, 8857. 40. Mishra, S.; Ghosh, S.; Mukhopadhyay, R. Maximizing mismatch discrimination by surfacetethered locked nucleic acid probes via ionic tuning. Anal. Chem. 2013, 85, 1615-1623. 41. Yuen, P. K.; Li, G.; Bao, Y.; Muller, U. R. Microfluidic devices for fluidic circulation and mixing improve hybridization signal intensity on DNA arrays. Lab Chip. 2003, 3, 46-50.

ACS Paragon Plus Environment

Page 18 of 25

Page 19 of 25 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 Captions Figure 1. Schematic representation for the direct quantification of trace amounts of BCR-ABL genes in the presence of a million copies of normal BCR and ABL genes onto an LNA-based assay. Localization of an individual BCR-ABL target (b3a2) was achieved with quantitative imaging. Where a DNA-modified AFM tip scanned the surface with 2.0 nm separation and a cluster of positive pixels was observed. The cluster radius was measured by ellipse fitting and the variation of the characteristics of the FD curves within the cluster were evident in accordance to pulling angle. Three consecutive QI maps were recorded over the entire spot and overlaid after correcting for drift. The cluster with at least one repetitive detection event is counted as positive cluster. Figure 2. Superimposing specific adhesion maps. The white broken circles represent the boundary of the spots obtained from the morphology maps. (A) Three successive adhesion maps (400 × 400 pixels, 2.4 × 2.4 µm2) after hybridization with BCR-ABL target sequence containing b3a2 junction (800 zM, 40 µL) were overlaid after correcting for the drift. (B) Drift-compensated overlaid map (360 × 360 pixels, 1.8 × 1.8 µm2) for BCR-ABL target having b2a2 junction (1.0 aM, 40 µL) from three successive adhesion maps. In both cases the pixels correspond to single specific detection event are colored green, and pixels where two or three specific FD curves were recorded are colored red. Due to the drift, some sections are not overlapped, and are colored gray. Locations of the positive clusters, evident through repeated detection, are marked with white circles (Rc ≥ 11.0 nm for b3a2 and Rc ≥ 16.0 nm for b2a2 junction respectively). For each case one of the representative cluster having repetitive detection pixels is shown in an inset of the respective overlaid maps. (C) Overlaid adhesion force map upon exposure to 100 fM normal BCR target related to b3a2 junction. No positive cluster was observed. (D) Overlaid adhesion force map after hybridization with 100 fM normal BCR and ABL targets corresponding to b2a2 junction. The radius of even the largest cluster (yellow square box) is only 8.3 nm. Figure 3. Number of positive clusters quantified in each LNA-modified probe spot for different concentrations of the BCR-ABL target; (A) b3a2 junction, (B) b2a2 junction in the presence of 100 fM of the normal BCR and ABL genes.

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. Quantification of a few copies of BCR-ABL target in presence of million copies of normal BCR and ABL genes. Overlaid (A) height image and (B) adhesion force map (400 × 400 pixels, 2.8 × 2.8 µm2) for BCR-ABL target having b2a2 junction (45 zM, 40 µL). One positive cluster was detected (marked with a white circle). Overlaid (C) height image and (D) adhesion force map (400 × 400 pixels, 2.0 × 2.0 µm2) for BCR-ABL target containing b3a2 junction (90 zM, 40 µL). Two positive clusters were observed (marked with white circles). Multiple examinations were required in order to detect two copies and single of BCR-ABL target irrespective of the junction type.

ACS Paragon Plus Environment

Page 20 of 25

Page 21 of 25

BCR-ABL transcript (Partial sequence)

BCR

ABL

Photolithographically etched glass slide

Junction region comprised by overlapping bases

BCR-ABL cDNA

FluidFM for miniaturized spot fabrication

1 2

46 pN

20 nm (I) 15 nm

0

Stretching at center

(I) (II)

Microchanneled AFM cantilever filled with LNA capture probe solution

Hybridization at 60 °C, 24 h Stringent washing at 70 °C

Adhesion force map

LNA-modified capture probe (bind to the junctional region with unprecedented specificity)

Sample solution

Detection DNA (attached to AFM tip)

20 pN

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

Visualization of individual BCRABL gene

LNA capture probe spot Quantification of the copy number of captured target genes

(II) Rc = 14.4 nm

Ellipsoid fitting result

Boundary

Cluster Radius (Rc)

Specifically captured target genes

Adhesion frequency

0

1

2-3

Figure 1. Schematic representation for the direct quantification of trace amounts of BCR-ABL genes in the presence of a million copies of normal BCR and ABL genes onto an LNA-based assay. Localization of an individual BCR-ABL target (b3a2) was achieved with quantitative imaging. Where a DNA-modified AFM tip scanned the surface with 2.0 nm separation and a cluster of positive pixels was observed. The cluster radius was measured by ellipse fitting and the variation of the characteristics of the FD curves within the cluster were evident in accordance to pulling angle. Three consecutive QI maps were recorded over the entire spot and overlaid after correcting for drift. The cluster with at least one repetitive detection event is counted as positive cluster.

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

(A)

(B)

0

1

(C)

Page 22 of 25

(D)

2-3

Figure 2. Superimposing specific adhesion maps. The white broken circles represent the boundary of the spots obtained from the morphology maps. (A) Three successive adhesion maps (400 × 400 pixels, 2.4 × 2.4 μm2) after hybridization with BCR-ABL target sequence containing b3a2 junction (800 zM, 40 μL) were overlaid after correcting for the drift. (B) Drift-compensated overlaid map (360 × 360 pixels, 1.8 × 1.8 µm2) for BCR-ABL target having b2a2 junction (1.0 aM, 40 µL) from three successive adhesion maps. In both cases the pixels correspond to single specific detection event are colored green, and pixels where two or three specific FD curves were recorded are colored red. Due to the drift, some sections are not overlapped, and are colored gray. Locations of the positive clusters, evident through repeated detection, are marked with white circles (Rc ≥ 11.0 nm for b3a2 and Rc ≥ 16.0 nm for b2a2 junction respectively). For each case one of the representative cluster having repetitive detection pixels is shown in an inset of the respective overlaid maps. (C) Overlaid adhesion force map upon exposure to 100 fM normal BCR target related to b3a2 junction. No positive cluster was observed. (D) Overlaid adhesion force map after hybridization with 100 fM normal BCR and ABL targets corresponding to b2a2 junction. The radius of even the largest cluster (yellow square box) is only 8.3 nm.

ACS Paragon Plus Environment

Page 23 of 25 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

(A)

Analytical Chemistry

(B)

Figure 3. Number of positive clusters quantified in each LNA-modified probe spot for different concentrations of the BCR-ABL target; (A) b3a2 junction, (B) b2a2 junction in the presence of 100 fM of the normal BCR and ABL genes.

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

(A)

(B)

(C)

0

1

2-3

Page 24 of 25

(D)

0

1

2-3

Figure 4. Quantification of a few copies of BCR-ABL target in presence of million copies of normal BCR and ABL genes. Overlaid (A) height image and (B) adhesion force map (400 × 400 pixels, 2.8 × 2.8 µm 2) for BCR-ABL target having b2a2 junction (45 zM, 40 µL). One positive cluster was detected (marked with a white circle). Overlaid (C) height image and (D) adhesion force map (400 × 400 pixels, 2.0 × 2.0 µm 2) for BCR-ABL target containing b3a2 junction (90 zM, 40 µL). Two positive clusters were observed (marked with white circles). Multiple examinations were required in order to detect two copies and single of BCR-ABL target irrespective of the junction type.

ACS Paragon Plus Environment

Page 25 of 25 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

For TOC only

LNA-modified capture probe

pN nm

Specific capture of BCR-ABL gene

Adhesion frequency 2-3 0 1

ACS Paragon Plus Environment