Type‐Independent 3D Writing and Nano‐Patterning of Confined Biopolymers

Abstract Biopolymers are essential building blocks that constitute cells and tissues with well‐defined molecular structures and diverse biological functions. Their three‐dimensional (3D) complex architectures are used to analyze, control, and mimic various cells and their ensembles. However, the free‐form and high‐resolution structuring of various biopolymers remain challenging because their structural and rheological control depend critically on their polymeric types at the submicron scale. Here, direct 3D writing of intact biopolymers is demonstrated using a systemic combination of nanoscale confinement, evaporation, and solidification of a biopolymer‐containing solution. A femtoliter solution is confined in an ultra‐shallow liquid interface between a fine‐tuned nanopipette and a chosen substrate surface to achieve directional growth of biopolymer nanowires via solvent‐exclusive evaporation and concurrent solution supply. The evaporation‐dependent printing is biopolymer type‐independent, therefore, the 3D motor‐operated precise nanopipette positioning allows in situ printing of nucleic acids, polysaccharides, and proteins with submicron resolution. By controlling concentrations and molecular weights, several different biopolymers are reproducibly patterned with desired size and geometry, and their 3D architectures are biologically active in various solvents with no structural deformation. Notably, protein‐based nanowire patterns exhibit pin‐point localization of spatiotemporal biofunctions, including target recognition and catalytic peroxidation, indicating their application potential in organ‐on‐chips and micro‐tissue engineering.


DOI: 10.1002/advs.202207403
control of cells and tissues for in vitro and in vivo applications. [1] Biological polymers, such as nucleic acids, proteins, and glycans, which are essential building blocks of biological systems, can be well organized to build complex nanostructures. [2,3] The replication of their structural properties and molecular functions at the cell and tissue levels is applicable in a wide range of biomedical fields, including regenerative medicine, [4] organ manufacturing, [5] and microfluidic tissue engineering. [6] Furthermore, their micropatterned substrates or 3D scaffolds serve as physical or chemical cues to induce cell signaling for the modulation of cellular behaviors, such as cell adhesion, migration, and proliferation. For instance, aligned actin fibers underneath certain cells guide cellular mechanotransduction, thereby leading to directional elongation and migration of fibroblasts, [7] or selective differentiation of stem cells. [8] Additionally, the insertion of submicronized needles or tubes through plasma membranes facilitates the delivery of exogenous biomolecules into living cells without inducing critical cellular damage. [9] The Food and Drug Administration regulates unethical clinical trials, thus, shedding a spotlight on human organ-on-chips that mimic cell microenvironments and simultaneously maintain tissue-specific functions. [10,11] In drug development and pathological studies, organ-mimicking models are now recognized as an innovative approach to in vitro testing, [12] thereby confirming the technical need for 3D micro-and nano-printing of biopolymers with controlled structures, alignments, properties, and even polymeric types.
The 3D multiscale fabrication of biopolymer architectures, however, has been eager to be type-dependent, which is attributed to the diverse rheological and structural properties of the constituents. To date, many different biopolymers have been printed at the desired positions, mainly via nozzle-based extrusion [13] and light-assisted stereolithography. [14,15] In particular, in monolithic 3D nano-printing, their dynamic rheological behaviors become more critical for controlled shaping with nanometer precision. [16,17] For example, unlike fibrous proteins, globular proteins do not retain submicron shapes without physical entanglement in nozzle-based extrusion, [18] thereby limiting the printing resolution and shape fidelity. [19] Certain additives, such as natural fibers (e.g., alginate and collagen) or reinforcement materials (e.g., carbon nanotubes and silver nanoparticles), can be included during extrusion, [20,21] however, printed nanocomposites often experience unfavorable phase separation with the formation of uneven nanostructures, [22,23] and functional loss of the original biopolymers. For uniform compositions, biopolymers are chemically interconnected during simultaneous 3D printing. However, bio-inks inevitably contain crosslinking agents that are inherently toxic to biological entities, and chemical entanglement poses a substantial risk that deteriorates the functional integrity of the printed biopolymers. [24,25] Notably, practical applications (e.g., needle patches, [26] tissue scaffolds, [27] and vascular networks [28] ) demand the construction of freestanding 3D architectures that are not influenced by substrates and their geometry. To expand the repertoire of printable biopolymers at the submicron scale, developing bioprinting techniques that are type-independent and facilitate full control over composition and shape remains exceptionally challenging.
Therefore, in this study, we demonstrate a type-independent 3D writing technique using a systemic combination of nanoscale confinement, evaporation, and solidification of a biopolymer solution. When a femtoliter solution is confined in an ultra-shallow liquid interface between the tip of the fine-tuned nanopipette and the printing point of the substrate surface, solvent-exclusive evaporation and concurrent solution supply occurs, resulting in the directional growth of biopolymeric nanowires. As the choice of biopolymer type did not strongly influence the synchronization of solvent evaporation and solution flow, the simple control over concentrations and molecular weights of biopolymers readily assured 3D complex printing with submicron resolution. Owing to the 3D motor-operated precise nanopipette positioning, various nanowire patterns could be fabricated with different structural complexities and biological functionalities; with a broad range of diameters (80 nm-10 μm) and lengths (≥1 μm), pillar or arch-shaped architectures were built using numerous different biopolymers, even including catalytically-active enzymes. Importantly, the exclusive surface modification of pre-printed nanowires could successfully secure the structural integrity of the 3D nanowire array in diverse solvents with no functional loss of biopolymers. Particularly, pin-point compartmentalization of protein functions can be achieved, thereby realizing the potential to analyze, support, or even stimulate spatially complex cells and tissues.

Fabrication and Characterization of Biopolymer Nanowires
Our direct 3D writing method facilitates the formation of 3D patterns consisting of individually size-controlled nanowires formed using various biopolymers, including nucleic acids, polysaccharides, and many different proteins (Figure 1a). The in situ growth of the biopolymer nanowires relies on two important procedures: nanoscale solution confinement and evaporation-induced solidification. A femtoliter biopolymer solution is readily confined at the tip of the nanopipette, therefore, an ultra-shallow liquid interface can be created by bridging the end of the glass nanopipette and the desired point of the substrate surface. The liquid nanobridge allows air-driven interfacial evaporation, without the escape of solutes, while ensuring a continuous solution flow from the tip (yellow arrow). Precipitation and accumulation of biopolymers can be induced simultaneously at the base of the quasi-liquid, thereby leading to the 3D growth of biopolymer nanowires (white arrow). When the quasi-liquid nanobridge is preserved with no shrinkage, directional pulling-up of the nanopipette can guide the diameter of the nanowire (d) to be equal to that of the nanopipette tip (d t ), and the length can be eventually determined by the pulling-up length.
Real-time observation of our 3D writing process enabled precise positioning and controlling of nanowire growth, as implemented through a customized setup of a 3-axes motorized stage and two imaging cameras at right (90°) angles to the same focal point ( Figure 1b). As an example, a freestanding nanowire composed of pure DNAs was fabricated on a silicon surface with a predesigned diameter (d ≈ 500 nm) and length (l = 40 μm) ( Figure 1c and Movie S1, Supporting Information). When filled with a DNA solution (4 mg mL −1 in deionized water), a nanopipette with a tip diameter of d t = 500 nm was pulled down to the substrate (t 1 ), and its close proximity to the surface enabled the formation of a quasiliquid nanobridge in between (t 2 ). Subsequently, the nanopipette was pulled upward to a designed height (h = 40 μm), resulting in continuous growth of a pure DNA nanowire (t 3 ). A rapid pullingoff movement resulted in a freestanding DNA nanowire with a diameter and length of ≈500 nm and 40 μm, respectively, at the desired location of the substrate (t 4 ). The growth rate of the nanowires was generally >1 μm s −1 .
Without additives, pure biopolymers can form homogeneous nanowires through 3D writing. To investigate i) molecular arrangement and ii) spatial element distribution of biopolymers in the grown nanowires, i) transmission electron microscopy (TEM) and ii) high angle annular dark field-scanning transmission electron microscopy (HAADF-STEM) combined with energy dispersive X-ray spectroscopy (EDS) characterization were performed on lambda DNA nanowires directly fabricated on a lacey carbon film supported copper (Cu) grid ( Figure 1d and Figure   and Figure S2, Supporting Information). Except for Cu EDS signal originated from the grid, DNA constituents (C, O, N, and P) were mainly found in the EDS spectrum, confirming the exceptional purity of the DNA nanowires ( Figure 1g).
Further simple variation of nanowire length permitted accurate quantification even at the sub-femtomole scale by real-time polymerase chain reaction ( Figure S3, Supporting Information), after inspecting homogeneous elemental spatial distribution revealed by the STEM-EDS of DNA nanowires. In sharp contrast to common stereolithography with harmful UV light, [29,30] our direct 3D writing maintains a mild processing environment, such as room temperature and ambient air, thus, resulting in no crit- , is marked by the vertical boundary line from confined growth (gray) to non-confined growth (white). d) Dependence of on the concentration and the length of DNA. value is directly proportional to the concentration of DNA and inversely proportional to the square root of the molecular weight (∝length) of DNA. e) All the normalized critical speeds v c / over various C, M, and d t were well fitted to a single simulated line (black) equal to 1/d t .
ical damage to the printed biopolymers. Irrespective of the type of printed DNA molecules, a single band was visualized in each lane of gel electrophoresis, thereby indicating full preservation of molecular integrity ( Figure S4, Supporting Information).

Growth Behavior of Biopolymer Nanowire
Regarding the uniformity of the nanowire diameter, the pullingup speed of the nanopipette (v) determines two different modes of nanowire growth: Confined and non-confined growth (Figure 2a). A sufficiently low v prevents an abrupt shrinkage of the quasi-liquid interface between the tip of the nanopipette and the top of the nanowire, thereby causing the nanowire diameter (d) to be confined by the tip diameter of the nanopipette (d t ) (as confined growth). In contrast, a sufficiently high v stretches the nanobridge interface to form a conical meniscus, thereby gradually decreasing the nanowire diameter (as non-confined growth). This suggests the existence of a critical pulling-up speed (v c ) at which the growth mode changes from confined to non-confined growth. While the nanowire diameter remains constant at d = d t in the confined growth mode (v ≤ v c ), it can also be decreased with v under the non-confined growth mode (v > v c ). We derived the critical pulling-up speed of the pipette (v c ) for the transition point from confined to non-confined growth. In the confined growth mode, rapid solvent evaporation induces a continuous solution flow from the pipette tip. For a dilute solution, the solution flow rate is practically equal to the solvent evaporation rate at the quasi-liquid nanobridge. Then, www.advancedsciencenews.com www.advancedscience.com where u, x, and E is the solution flow velocity, bridge length, and solvent evaporation flux, respectively. Here, the volume conservation of the solute in the bridge provides the following boundary condition: [31] u ≈̇g − 0 0 (2) wherėg is the growth rate of the nanowire, and and 0 are the volume fractions of the solute in the nanowire and pipette solutions, respectively. By combining Equations (1) and (2), we obtain: where the experimental coefficient ( ), Notably, if the pulling-up speed of pipette, v, is faster than the growth rate,̇g, the liquid nanobridge cannot preserve its diameter and becomes stretched, thus, crossing over to the nonconfined growth mode. This indicates that thėg obtained from Equation (3) is the maximum speed for confined growth, that is, v c =̇g. Therefore, Notably, v c correlates with the concentration (C) and molecular weight (M) of the biopolymer solute in the solution. As >> 0 in dilute solutions, from Equations (3) and (5), respectively, we obtain: Because the nanobridge length (x) is roughly correlated with the viscosity ( ) of the solution, [31] x ∝ 1 √ and the viscosity is correlated with the molecular weight, (n: Mark-Houwink parameter). [32] Thus, from Equations (3), (5), (7), and (8), respectively, we obtain: where n depends on the polymer-solvent system. Further, from Equations (3)-(9), we obtain: where is a coefficient determined by the type of biopolymer and environmental factors.
As the pulling-up speed increased, the growth mode crossed over from confined to non-confined mode (Figure 2b and Figure  S5, Supporting Information). When the DNA nanowire growth for v ≤ v c was investigated with a given nanopipette (d t = 0.82 μm), the nanowire diameter remained constant with v as d = d t , thus, indicating the confined growth (Figure 2b, gray). However, for v > v c , the nanowire diameter decreased gradually with v (d < d t ), thus, indicating non-confined growth (Figure 2b, white). The red line in the figure represents the simulated critical speed (v c ), which is inversely proportional to d t (based on Equation (5)). Although the length of the liquid nanobridge (x) in ( = 4xE 0 − 0 in Equation (4)) may be affected by the tip diameter of the nanopipette (d t ), the dependency between and d t was almost negligible in the implemented range under 2 μm. Practically, upon measuring the diameter of DNA nanowires grown with different tip diameters (from d t = 0.45-1.36 μm), the transition of the growth mode was observed at the red line (see the inset in Figure 2b), which indicates that is independent of d t .
The estimation of depending on experimental conditions, such as concentration, molecular weight, and type of biopolymer, facilitates precise control over the nanowire growth behavior. For example, we measured for DNA nanowires grown at various concentrations (1, 2, and 4 mg mL −1 ) and base lengths (30, 60, and 120 nt for single-stranded DNA [ssDNA]), where the normalized diameter (d/d t ) was plotted depending on vd t (Figure 2c). Here, [ = v c d t by Equation (5)] is marked by the vertical boundary line from confined growth (gray) to non-confined growth (white). Interestingly, the experimentally measured values (Figure 2d, open circles) were proportional to C and 1/M n/2 , respectively, and matched with those calculated using Equation (10) (Figure 2d, dashed lines).
The Mark-Houwink parameter, n, was fitted to 1, which is consistent with a previous study on linear DNA, [33] thus, indicating the high reliability of determination.
From v c identification and value, we can easily predict nanowire growth modes, using Equation (10). When the DNA nanowires were grown at various ranges of v (1-20 μm s −1 ), C (1, 2, 4 mg mL −1 ), M (30, 60, and 120 nt for ssDNA), and d t (450 nm-2 μm), the nanowire growth mode changed at the calculated v c , and all normalized critical speeds (v c / ) over various C, M, and d t fitted well to a single simulated line equal to 1/d t (Figure 2e). To confirm the universality of Equation (10), we elucidated the values for different types of biopolymers ( Figure S6a, Supporting Information). We performed 3D writing with different biopolymers, such as bovine serum albumin (BSA) protein and dextran. Irrespective of the type of biopolymer used, the transition of the growth mode appeared at the critical speed v c , which could be normalized as v c / (black line, Figure S6b, Supporting Information). Our observations suggest that regardless of the concentration (C), molecular weight (M), and type of biopolymer ( ) applied, the identification of v c enables rational design of the growth of biopolymer nanowires with uniform diameters.

Multiplexed Biopolymer Nanowire Arrays Fabricated by 3D Nano-Patterning
Based on 3D writing in confined growth mode, we successfully fabricated diverse 3D nanowire patterns with various di-www.advancedsciencenews.com www.advancedscience.com mensions, geometries, and materials. Homogeneous freestanding nanowires were reproducibly and site-specifically grown with a uniform diameter and length (Figure 3a). On a Si substrate, seven different character patterns were constructed using double-stranded DNA (120 bp) nanowires of d = 850 nm and l = 10 μm, with each nanowire located 10 μm apart (see inset). As different biopolymers could be readily used to form uniform nanowires, three distinct dye-conjugated biopolymers [DNA-Texas Red (red), Dextran-FITC (green), and BSA-7-DCCA (blue)] were demonstrated to form nanowires with individually controlled length (6-18 μm) and diameter (0.5-2 μm) by steering the fine-tuned nanopipettes (Figure 3b). The precise control of nanopipette movement guided the direction of nanowire growth, thereby allowing wiring construction through the interconnection of two different points. Accordingly, we manufactured a complex 3D pattern that combines curved arches and freestanding pillars at the designated positions on a glass substrate (Figure 3c). The curved arches with specific widths of 10-20 μm were placed at regular intervals of 10 μm using DNA-Texas red (red), and the pillars consisting of DNA-FAM (green), with a uniform length of 10 μm, were printed along the edge of the DNA arches.
Even after 3D nano-patterning, the freestanding biopolymer nanowires retained their inherent molecular functions, such as molecular recognition and bio-catalysis, whereas their mechanical stability and structural integrity could be further enhanced via simple post-modification (Movies S3 and S4, Supporting Information); when the 3D nano-patterns were simply immersed in the organic solvent that contained crosslinkers, the surficial biopolymers of the nanowires could be exclusively interconnected to retain their structural shapes even in water. When two different protein nanowires, comprising pure streptavidin (Figure 3d, left) and BSA (right), respectively, were developed to maintain the freestanding structures in an aqueous solution, their unique target binding and antifouling abilities were well conserved. In the presence of 100 nM FITC-biotin, the positions of streptavidin nanowires were fully overlapped with those of green fluorescence signals due to the specific streptavidin-biotin non-covalent interaction, [34] whereas nothing was bound to the nanowires of BSA, which is one of the most effective antifouling proteins in the blood. [35] Importantly, the protein nanowires could retain their original shapes and functionalities in various solvent environments. When immersed in deionized water, high-salt buffer (10 × phosphate-buffered saline), physiological solution (fetal bovine serum (FBS)), and frequently used organic solvents (ethanol, dimethyl sulfoxide (DMSO), and dimethylformamide (DMF)), the streptavidin nanowires underwent no structural deformation (Figure 3e, top), thus, enabling the localization of FITC-biotin (bottom). Notably, the fluorescence intensities varied depending on the solvent environment, presumably proportional to the quantum yield of FITC in each solvent ( Figure S7, Supporting Information).
At the submicron scale, elaborate patterning of proteins, such as receptors, antibodies, and enzymes, would be of great interest due to the requirement of site-specific target recognition or metabolic regulation in controlling biological systems. [36] As site-specific localized chemical cues have application potential in biological studies, such as cell signaling and organelle mimicking, [37] we prepared an enzymatic nanowire pattern based on horseradish peroxidase (HRP) (Figure 3f). To confirm the HRP-accelerated peroxidation at local points, we used tyramide reagents; unlike easily diffusible reagents, such as 2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) [38] and luciferin, [39] they can be immediately deposited at the location of peroxidation, thereby allowing real-time reporting of catalytically-active sites with high precision. [40] Enzymatic and non-enzymatic nanowires were densely printed with HRP and streptavidin on a glass substrate, and Alexa Fluor-488 tyramide and H 2 O 2 were introduced to initiate pin-point peroxidation reactions (Figure 3g). Green fluorescence was observed only on the surface of the enzymatic nanowires (yellow circles), and the fluorescence signals increased logarithmically with time and saturated at 30 s ( Figure S8, Supporting Information), probably because of the complete occupation of the tyramide deposition sites. However, the fluorescence of the adjacent non-enzymatic nanowires was negligible, indicating the availability of enzymatically active 3D-shaped satellites in one batch reaction. Ultimately, such protein-based nanopatterns enabled spatiotemporal control of chemical and physical stimuli with submicron resolution, thus, potentially serving as artificial cellular organelles such as peroxisomes underneath live cells and tissues. [41]

Conclusion
By exploiting the nanoscale confinement of the biopolymer solution in the quasi-liquid nanobridge, we achieved direct 3D writing and nano-patterning of biopolymers. Additionally, through precise control over nanowire growth, we reproducibly fabricated freestanding nanowires and arch-type wiring. The diameter of the nanowires was adjusted in the range 80 nm-10 μm, which could be further extended by fine-tuning the glass pipettes. [42] Based on the minimum distance of 2 μm between the nanowires ( Figure S9a, Supporting Information), their length can be even shortened to be 1 μm ( Figure S9b, Supporting Information) or elongated to more than a few millimeters ( Figure  S10, Supporting Information); although the gradual increments of 250 nm in distance and length were achievable by the spatial resolution of 3-axis motor stage, the maximum distance and length were limited by its z-axis travel range. As the composition of the nanowire is simply determined by the choice of biopolymer solution, different biopolymers of a single or several mixed components could be readily printed at the desired position on a chosen substrate and even on the top of pre-printed wires ( Figure S11, Supporting Information). Currently, the 3D writing only produces one nanowire at a time and writing speed under the confined growth mode is quite slow (<10 μm s −1 ); however, parallel printing of biopolymer nanowires might improve throughput with the development of well-aligned multi-channel nanopipettes. [43] While the process of 3D writing is exceptionally simple in ambient air, the molecular orientation of biopolymers has not been controlled, causing nanowires to possess amorphous characteristics. However, the recent advancement of self-assembly techniques using external stimuli (such as ultrasound [44,45] and pulsed electric field [46] ) may be synergistically combined with our 3D writing of confined biopolymers to facilitate full control over the crystallinity of the printed architectures by varying the mechanical and biochemical properties.
As proteins are involved in most cellular functions and phenotypes in all living species, their spatially controlled 3D immobilization is crucial for sensing, signaling, and transporting small biological entities in microscopic environments. Existing techniques, such as laser-assisted printing [47] or micromolding, [48] frequently require full crosslinking of internal structures for structural miniaturization and shaping of desired architectures, but high-level modification inevitably reduces the functional integrity of embedded biopolymers. [25,49] In contrast, our method enables free-form structuring of different biopolymers and exclusive modification of the surface of the pre-built nanostructures at mild conditions, thereby reducing the risk of functional loss, but enhancing the mechanical stability and environmental tolerance. Fully intact biopolymers can be compartmentalized by crosslinked surface networks that serve as size-exclusive membranes for the materials approaching the inside of the nanowire, thereby realizing the potential for mimicking microvascular channels with capillary exchange function. [50] Moreover, depending on the type of crosslinker, photodegradable or biodegradable biopolymer architectures can be created, which would be essential in the production of nanoneedle patches for intracellular drug delivery. Given the technical availability of high-throughput 3D nano-patterning and hierarchical nanostructuring (e.g., core-shell formation), our type-independent biopolymer printing technique can help fabricate different 3D biopolymer ensembles with application potential in several technologies, such as organ-on-chips.
3D Writing: 3D writing of biopolymers was performed using a customized setup consisting of a print head with a solution-filled nanopipette and a building platform with glass or silicon substrate. For a nanopipette preparation, borosilicate glass capillaries (BF-100-50-10, Sutter Instrument) were tapered by flame-brushing process where the diameter of pipette tip was precisely controlled by programmed heating and pulling conditions (P-97 micropipette puller, Sutter Instrument). To fabricate a nanowire array, the position of the nanopipette and the substrate were spatially controlled by 3-axis motor stage with an accuracy of 250 nm (XA07A and ZA07A, Kohzu Precision). The fabricated nanopipette was filled with a biopolymer solution of which concentration was ranging from 0.5 to 4 mg mL −1 in deionized water, and then it was subsequently pulled down closely to the substrate, thereby forming a quasi-liquid nanobridge in-between. Thereafter, the nanopipette was pulled with programmed speed (>0.5 μm s −1 ) and direction to produce a biopolymeric nanowire of desired shape, and the pipette was rapidly separated to liberate the nanowire on the substrate (Movie S1, Supporting Information). The nanowire fabrication process was monitored in real-time with the customized optical imaging system consisting of CCD cameras (INFINITY 1-2C, Lumenera Camera), objective lens (100x Plan Apo Infinity Corrected Objective, Mitutoyo), and 590 nm LED light source (Precision LED spotlight, Mightex).
Characterization of Biopolymer Nanowires: The crystallinity of biopolymer molecules within the fabricated nanowire was inspected by using a TEM (JEM-2100F, JEOL) operated at 80 kV. The spatial elemental distribution in nanowires was investigated by EDS-mapping operated at 80 kV of a fifth-order aberration(C s )-corrected STEM (JEM-ARM200F, JEOL). For the TEM sample preparation, the biopolymer nanowires (d = 89 ± 5 nm (mean ± s.d.)) were directly fabricated in the lateral direction on a lacey carbon film supported copper TEM grid ( Figure S1 and Movie S2, Supporting Information). For the investigation of nanowire growth behavior, the diameters of nanowires grown at different printing speeds were measured with field emission scanning electron microscopy (FE-SEM, S-3400N, Hitachi). The appearances of nanowire patterns were characterized by using a highresolution FE-SEM (JSM-7800F Prime, JEOL). The fluorescent imaging of the fabricated nanopatterns was performed using a confocal microscope (STELLARIS 5, Leica).
Real-Time Polymerase Chain Reaction (RT-PCR) Analysis: Using nanopipettes filled with the ssDNA-120 solution (25 μM in water), nanowires were grown on tapered optical fiber tips with the constant pulling speed of 1 μm s −1 . The tapered optical fibers were prepared by using a laser-based puller (P-2000, Sutter instrument). To identify the proportionality between the volume of nanowire and the number of DNA molecules in the nanowire, four different nanowires at varying lengths (5, 10, 20, and 40 μm) ( Figure S3a, Supporting Information) were prepared, and then, they were dissolved in water (2 μL). Thereafter, the number of DNA molecules were quantified by RT-PCR (LightCycler 480, Roche). Each sample for RT-PCR reaction (20 μL) contained 0.2 μL of F-primer (50 μM), 0.2 μL of R-primer (50 μM), 2 μL of DNA-dissolved water, 10 μL of LightCycler 480 SYBR green I master mix, and 7.6 of nuclease-free water (T&I). Prepared samples were subjected to PCR cycles (20 s denaturation at 94°C, 20 s annealing at 57°C, and 20 s extension at 72°C).