Introduction

Miniaturized microdevices have various advantages in terms of functions compared with their macroscopic counterparts: higher sensitivities in sensors, high efficiencies in reactors, small amount of samples required, fast response, etc. Microelectromechanical systems (MEMS) technologies have been developed by applying semiconductor microfabrication technologies from the late 1980s. MEMS technologies offer the advantages of the batch fabrication of many devices and have the ability to integrate multiple functional units in a small area as well as the ability to fabricate small structures at the submicrometer level. MEMS technologies have successfully been applied to sensors [1, 2], actuators [3], microreactors [4, 5], radiofrequency devices [6], and optical devices such as an optical attenuator [7], an optical mirror array [8], an optical scanner array [9], and optical switches [10–12]. Besides the application of MEMS to industrial products such as accelerometers and movable mirror arrays for a video projection system, applications to biological and biochemical experiments have become a hot topic. MEMS technologies have also enabled the integration of multiple-function units on a small chip, so-called micro total analysis systems (μTAS) or lab on a chip, whose basic idea was proposed by Manz et al. [13] in 1990. μTAS allow rapid biological or chemical analysis with a small amount of sample thanks to the merits of miniaturization. While the popular industrial application is limited to DNA chips, there has been an increase both in the amount of research aimed at biological and biochemical applications and in the number of applications [14, 15].

Single molecule observation and characterization of biomolecules enables us to determine the individual behavior and characteristics of molecules usually hidden in the ensemble-averaging and time-averaging of bulk experiments [16, 17]. The major advantages in single molecule characterization are the ability to measure the variations in molecules, the dynamics on a short time scale, and the mechanical motion of biomolecules. The typical size of an individual biomolecule is on the order of a few nanometers to tens of nanometers. Giant nucleic acids such as a λDNA are longer than 15 μm. Today, we can fabricate microstructures less than 100 nm in size by MEMS technologies. Because their scales have become closer, developing MEMS devices to manipulate a single biomolecule might realize novel experiments or measurements with higher sensitivity that could not have been achieved by a conventional single molecule manipulation technique such as using magnetic tweezers [18, 19], optical tweezers [20, 21], atomic force microscope cantilevers [22], and liquid flow [23]. Some attempts to manipulate a single biomolecule with microstructures or microfabricated devices have been reported. Controlling and on–off switching of actomyosin as a nanoactuator [24] and electrostatic manipulation of DNA in microfabricated structures [25, 26] were successfully achieved. However, these works did not utilize the full ability of MEMS technologies such as high sensitivity or manipulation of an individual molecule.

This review is aimed at introducing some new achievements in MEMS application to single biomolecule manipulation and characterization by our group. The target biomolecules are proteins (a motor protein F1-ATPase—hereafter F1—and an enzyme β-galactosidase—hereafter β-gal) and DNAs. Although many other exciting results exist, four examples are highlighted owing to the limited scope of this review. The first is a microthermodevice for a real-time rapid temperature alternation experiment whose capability was demonstrated by controlling a rotary biomolecular motor, F1 [27]. The rapid control of the temperature enabled us to perform rapid motion control of F1. The experimental results suggest that the torque of F1 increases at higher temperatures. The second example is the application of a femtoliter container whose temperature can be controlled rapidly by a microthermodevice [28]. The catalytic reaction of an enzyme, β-gal, was observed in the microcontainer with rapid temperature alternation. This method realized the enzymatic assay at temperatures higher than that at which enzymes would “normally” be denatured in bulk experiments. The third example is a systematic isolation and immobilization of a single DNA molecule with a microdevice [29]. It consists of a polymer sheet with channels and aluminum electrodes deposited on a cover slip. A DNA molecule passed through a 2-μm-wide channel and both ends of a single DNA molecule were immobilized at integrated electrodes by dielectrophoresis force. The fourth example is the conductivity measurement of DNA bundles; this measurement demonstrates that humidity is the main factor affecting the electrical conduction in a DNA bundle captured between microactuated silicon tweezers [30]. The result suggests that (1) DNA has a quasi-ohmic behavior under constant temperature and humidity conditions and (2) the exponential dependence of conductivity on the relative humidity could be explained by the change of permittivity arising from water adsorption on the DNA backbone. A quantitative model was also proposed to explain the electrical properties of the DNA bundle. For more detailed information, especially on the experimental conditions and fabrication processes, we encourage readers to refer to the original papers. The significance of each microdevice for its biological application will be discussed specifically for each topic. In the conclusion, the general advantages and the future prospects of MEMS devices for the single biomolecule experiments will be discussed.

Motion control of a rotary biomolecular motor by a microfabricated local heating device

Motor proteins [31] are the major targets in single molecule observation because their mechanical motions allow measurement of the energy consumption of a single molecule, which is important information for understanding molecular mechanisms. Enzymes such as kinesin [32], dynein [33], myosin [34], DNA helicase [35], and RNA polymerase [36] act as a linear motor which moves on the rail proteins or DNA. Bacterial flagella protein complex [37] and a subunit of ATP synthase, F1) [17, 38–41], act as a rotary motor. Changing the chemical conditions in bioexperiments cannot realize real-time rapid alternation in experimental conditions [42–44]. One of the most promising ways to measure the response of an individual molecule is rapid temperature alternation because the activity of proteins is known to be temperature dependent from bulk experiments [45, 46]. In this work, Arata et al. proposed a simple and versatile method to control the motion and activity of a single biomolecular motor by rapid temperature alternation.

Conventional devices which control the temperature under a microscope have the following limitations: (1) the response speed is low (on a time scale of minutes), (2) thermal expansion of the stage plate of the microscope causes defocusing during temperature alternation, and (3) the heat causes damage to the objective lenses. These limitations can be overcome by controlling the temperature locally [47, 48] because reducing the volume of the heating target reduces the total amount of heat generated to alter the temperature. Local heating with an infrared laser may damage proteins by exposing the samples to excess energy though it can alter the temperature rapidly in a local area [48]. A microfabricated local heating device can provide a good solution by integrating a microheater and a microthermosensor on a glass plate. The local heating microdevice was fabricated by patterning nickel on the glass plate (Fig. 1a). The microheater was designed with a circular shape so as to minimize the temperature nonuniformity within the heater. The diameter of the circular microheater was 400 μm, and the width was 20 μm. The maximum response speed was 71.5 and 56.9 °C/s for temperature rise and fall, respectively. The microthermosensor was located at the center of the microheater to realize real-time monitoring of the temperature. The temperature distribution in water was measured with high spatial resolution using the fluorescent dye rhodamine B. Since rhodamine B solution exhibits a strong and reversible temperature-dependence variation of its fluorescence intensity, it is useful as a temperature detector. The temperature distribution on a microthermodevice was successfully measured from the fluorescence intensity distribution of rhodamine B solution. This method allows temperature distribution measurement with an accuracy of 3 °C and a spatial resolution of 5.3 μm. The maximum temperature difference within the circular microheater of our microthermodevice was measured to be 3 °C [49].

Fig. 1
figure 1

a Ni-patterned glass plate, which is placed on a pitch conversion board and connected to lead wires (left). Microscope view of an integrated microheater and microthermoresistive sensor (right). Reducing the size of the thermodevice can reduce the total amount of heat generated to alter the temperature. This enables us to avoid drawbacks in conventional devices and realizes high response speed. b Microscope view of two microbeads being rotated by a single F1-ATPase (F1) molecule. The angular velocity is 1.5 rps. The time course of the rotation of F1 while alternating the temperature. The angular velocity changed from 1.2 to 4.4 rps and back to 1.7 rps. (Reprinted with permission from Arata et al. [27]. Copyright 2006, American Institute of Physics.)

Full size image

An F1 molecule was attached to a glass surface by infusing F1 contained in buffer in the flow cell. Subsequently, microbeads were attached to the rotary axis of F1 by infusing them into the flow cell. Microbeads are necessary because the size of a single F1 molecule is too small, around 8 nm in diameter and 14 nm in height, to be observed with an optical microscope. The angular velocity change of F1 rotation was successfully observed by altering the temperature. Figure 1b shows the time course of accumulative revolution of F1 while the temperature was switched from room temperature to 50 °C and back to room temperature. Its angular velocity changed from 1.2 to 4.4 rps and back to 1.7 rps. The drawbacks in conventional temperature controlling devices under the microscope such as defocusing caused by thermal expansion of the stage plate were not observed. When the temperature was increased gradually, the velocity of F1 increased gradually but stopped at a temperature lower than 50 °C because the enzyme cannot survive at high temperatures for a long time. To measure the velocity at higher temperatures, the temperature should be altered rapidly, because the enzyme can survive at higher temperatures for a short period, even at high temperatures that would denature it in a steady state. This rapid temperature altering system enabled us to measure the activity of an F1 molecule at temperatures higher than 50 °C. The torque of F1 was calculated by subtracting the effect of water viscosity that causes the drag force on the microbead attached to the rotating axis of F1. Consequently, the temperature dependence of the torque of F1 could be calculated over a wide temperature range: from room temperature to 70 °C. As a result, the torque seems to increase at higher temperatures, increasing at the rate of 4% per 10 °C.

MEMS technology enables us to integrate a microheater and a microthermosensor in a small area with ideal design. Also, the microheater and the microthermosensor can be easily electrically connected to a computer and sampled or controlled. Additionally, MEMS technology has mass-production capability that is suitable for biochemical research which requires a large number of experiments for statistical data analysis. This MEMS-based experimental tool is versatile in other scientific experiments which require rapid temperature alternation under the microscope.

Rapid temperature alternation by an on-chip thermodevice for enzymatic measurement at high temperatures

Even though the activity of an enzyme decreases with excess temperature applied over a long time period, we expect the enzyme to stay active by limiting the exposure time [45, 46] at temperatures higher than the molecular destruction temperature in a steady state. Enzymatic activity at a high temperature may be measured by momentary heating. However, to raise and reduce the temperature could not be realized by conventional technology in biochemical experiments. There are two problems to solve. One is to change the temperature rapidly and the other is to trap a certain number of molecules within the locally heated space. Arata et al. have developed an on-chip microthermodevice with an integrated microheater and microthermosensor on a glass plate as described in the previous section. Here, the chip was also equipped with a microcontainer array that traps enzymes in the temperature-controlled area (Fig. 2). Polydimethylsiloxane (PDMS) was used to form the microchamber because PDMS has the following advantageous characteristics [50]: it has no effect on biomaterials such as proteins and cells [51]; it can be easily patterned into minute structures; it is transparent and allows easy observation of the target inside the structured PDMS; it is compatible with mass production; its adhesion to the flat surface is strong and it is easy to seal with a flat substrate. Cylindrical microcontainers with a height of 1.5–2.2 μm, a diameter of 1–8 μm, and a volume of 1–110 fL were manufactured. They were first manufactured to measure the chemical activities of enzymes at the single molecule level by restricting the diffusion of products [5, 52]. Here, these containers are utilized for trapping a certain number of molecules within a small volume that is heated.

Fig. 2
figure 2

a Scanning electron microscope image of a patterned polydimethylsiloxane (PDMS) sheet. The 56.0-fL chambers had a standard deviation of 1.63 in volume. b Optical microscope view of an integrated microheater and microthermosensor on a glass plate. c The temperature-controlled microchamber array. d Activity controlled by temperature pulses with a duty ratio of 1:3 (four pulses in 48 s). The fluorescence intensity increased from 0.28 to 0.56. The lines follow the intensities of four individual microchambers. Activity was successfully increased, and this indicates that the enzymes remain active and increase their activity over a short period, even at temperatures that cause damage after prolonged exposure. (Reprinted with permission of the American Chemical Society from Arata et al. [28].)

Full size image

β-gal is widely used as a reporter in molecular biology [53]. β-gal is the enzyme which catalyzes hydrolysis of fluorescein di-β-d-galactopyranoside (FDG). A buffer solution containing a fluorogenic substrate of β-gal, FDG, was hydrolyzed to fluorescein. The fluorescence intensity represents the amount of the product produced by β-gal [5]; therefore, we can measure the activity of β-gal from the time course of the fluorescence intensity. β-gal solution was mixed with 400 μM FDG in 100 mM phosphate buffer pH 7.5 and contained in a PDMS microchamber. In this experiment, the volume of the microcontainer was 60 fL; at a β-gal concentration of 37 nM, each container contained around 1,300 enzymes. Applying 60 °C heat pulses of 4 s, we successfully increased the activity of the enzyme (Fig. 2d). This indicates that the enzymes remained active and increased their activity over a short period, even at temperatures that cause damage after prolonged exposure; the activity of β-gal versus temperature shows a peak at around 45–50 °C, a decrease with temperature after the peak and complete cessation above 60 °C (data shown in Arata et al. [28]). The activity of the enzyme is also controllable by changing the duty ratio of the heating pulses.

This method gives new kinetic information such as the temperature stability of biomolecules and also makes it possible to perform highly sensitive experiments in biochemistry, for example, the Arrhienius plot at higher temperatures which does not include the effect of denatured enzymes. As for industrial application, this method is suitable for realizing highly sensitive biosensors. This device can be directly applied to other proteins, cells, and biomaterials for respective heating experiments without any modifications of the device and of the microscope.

Single DNA molecule isolation and trapping in a microfluidic device

DNA has been widely investigated as it contains genetic information, critical in the understanding of the human body and the future advancement of medical science. In view of this, many researchers have attempted to study DNA through the observation of immobilized DNA molecules. As an example, Harada et al. [54] observed the rotation of single DNA molecule during transcription by RNA polymerase which was immobilized on a cover slip. Immobilization is critical as the first step in realizing a biophysical system used for the manipulation and characterization of biomolecules which move randomly in their natural state. At present, immobilization methods of DNA molecules include (1) chemical binding [55–58], (2) physical stretching and binding [59, 60], and (3) electrical stretching and binding [25, 26, 61]. Mixing DNA molecules with a thiol group caused the modified DNA molecules to chemically adhere onto a gold substrate [58]. As an example of a physical binding method, Guan and Lee [60] presented a unique technique for highly ordered arrays of stretched DNA. By using a PDMS sheet with aligned 5-μm-diameter chambers to cover a DNA solution droplet, and then peeling the sheet off slowly, they were able to stretch and immobilize DNA molecules on the peripheral edges of a circular chamber as a result of surface tension and flow pressure. Finally, for electrical binding and stretching, Washizu and Kurosawa [25] proposed an effective way of DNA trapping by dielectrophoresis. In this case, DNA molecules were fully stretched along the electric flux line and immobilized on an aluminum electrode. Microfluidic devices and microfabricated chambers are effective approaches for single molecule experiments. The channel sizes usually used for μ-TAS are in the range from several tens of nanometers to several hundred micrometers. Enclosing molecules in these small areas provides precise handling and good observation for researchers.

Kumemura et al. [29] designed and fabricated a microfluidic device for isolation and immobilization of a single DNA molecule (Fig. 3) which was composed of a PDMS film having wide and narrow channels, and electrodes deposited on a cover slip. The channels were 300 μm wide and 4 μm deep for the wide channel and 1 μm wide and 4 μm deep for the narrow channel, respectively. In this microdevice, the balance between electrophoresis and electroosmosis in the channel was effectively exploited for DNA molecule isolation. By applying a DC voltage, DNA molecules, which are negatively charged, flow from the anode to the cathode in a microchannel. On the other hand, because the surface of PDMS and the cover slip are negatively charged (silanol group), cations in water which are positioned close to channel walls move from the cathode to the anode. Therefore, electroosmotic flow appeared in the opposite direction to DNA electrophoresis in this case. In the 300-μm-wide channel, the electrophoretic flow dominated over the effect of electroosmosis. In the narrow channel, the electroosmosis produced significant flow at the entrance of the 1-μm-wide microchannel. DNA molecules that reached the entrance of the microchannel were faced with electroosmotic flow in the reverse direction and many of them were prevented from passing through the microchannel (Fig. 4). This efficiency of the molecule separation was estimated by measuring the molecular flux (number of molecules per second) at given locations in the device. The relationship between the molecular flux in the narrow channel and the applied voltage of electrophoresis was examined and the optimized voltage was obtained for molecule separation. This technique leads to successful single molecule immobilization between electrodes, which is the next step. For immobilization of DNA molecules, electrostatic force (dielectrophoresis) was employed. By applying 900-kHz AC voltage between electrodes positioned at the exit of the narrow channel, a single DNA molecule stretched along the lines of electric force, and was attracted to both edges of the electrodes (Fig. 4). Because the DNA binding on aluminum is nonspecific and permanent [25, 26], once the DNA molecule is immobilized on the aluminum electrodes, the stretched molecule provides a stable appropriate template for further biomolecular assay.

Fig. 3
figure 3

a A fabricated microdevice. A and B are electrodes for electrophoresis. C and D are electrodes for dielectrophoresis. b Blowup of the microdevice. E is an access hole, F is a wide channel (300 μm wide), G is a narrow channel (2 μm wide), and H is an electrode for immobilization. (Reproduced with permission from Kumemura et al. [29]. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

Full size image
Fig. 4
figure 4

a The balance between electrophoresis and electroosmosis at the entrance of the narrow channel. The thick solid line shows electrophoresis-induced force line, the dotted line shows electroosmotic flow, and thin solid lines show DNA trajectories. The electrophoresis drift pushes DNA towards the channel entrance. The electroosmotic flows is in the reverse direction and spreads out when exiting the microchannel. This diverging electroosmosis pushes the molecule away from the entrance (case 2). Only the molecule moving in the channel axis can enter (case 1) b The fluorescence view of an immobilized single DNA molecule between fabricated electrodes. (Reproduced with permission from Kumemura et al. [29]. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

Full size image

Humidity dependence of charge transport through DNA revealed by silicon microtweezers

DNA has been considered as a possible building block, such as an electric nanowire, for molecular electronic devices. For such an application to nanoelectronics, it is necessary to know the electrical properties of DNA and the physical origin of DNA conductivity. Tran et al. [62] studied the influence of temperature on DNA conductivity and also noted the effect of water. Other groups later discussed the major influence of humidity on DNA conductivity [63–66]. However, the relation between DNA conductivity caused by humidity and the influence of other dimensional parameters such as DNA length or the bundle diameter remained to be determined.

In order to address the question of DNA conductivity and the influence of humidity, Yamahata et al. [67] handled DNA molecules with MEMS-based microtweezers; the device enables the straightforward trapping of DNA between aluminum-coated silicon nanotips by AC dielectrophoresis [68, 69]. The microtweezers were designed to be used for simultaneous mechanical stretching and electrical characterization of the captured DNA bundle (Fig. 5a) [70–72]. With these microtweezers, Yamahata et al. measured the variation of the current I as a function of the applied voltage U (two points measurement method) on DNA bundles (various diameters, φ) in air under a relative humidity of 50–60% at 25 °C. The diameter of the dried bundles was estimated afterwards by field emission scanning electron microscopy (Fig. 5b). The current I measured for different bundles approximately represents the quasi-linear dependence on U. At a constant voltage, the current I shows the quadratic dependence on the diameter φ of dried bundles, I ∝ φ 2. Yamahata et al. also stretched a thin DNA bundle and simultaneously measured its conductivity (Fig. 6). The relative variation of the current, ΔI/I, as a function of the relative elongation of the bundle, Δl/l, shows that ΔI/I = −Δl/l, indicating that I ∝ 1/l (Fig. 6). These results indicate that the DNA bundle has a quasi-ohmic behavior, I = σSU/l, where S ∝ φ 2 is the section of the bundle and σ is its conductivity. With an assumption that electron charges q are carried by mobile ions of density n, and the kinetics of charge exchanges between ions results in a mobility μ, the conductivity of a DNA bundle can be described as σ = qnμ.

Fig. 5
figure 5

3D view of the microelectromechanical system (MEMS) silicon device and micrograph of a DNA bundle. a The MEMS nanotweezers were fabricated using silicon on insulator technology. The electrostatic actuator, the differential capacitive sensor, and the nanotips are electrically isolated. The displacement, Δl, of the moving tip results in a change of the variable capacitances C 1 and C 2 used for sensing. b A DNA bundle captured between silicon tips coated with aluminum. The bundle has a mean diameter of approximately380 nm. (Reprinted with permission of the Biophysical Society from Yamahata et al. [30].)

Full size image
Fig. 6
figure 6

Effect of the elongation of a DNA bundle on its conductivity. The measurements were performed for a DC voltage of U = 10 V. The data were recorded at 25 °C for a relative humidity of 50–60% (typical currents of few tens of picoamperes). The function y = −x is a perfect match with the least-squares linear regression curve calculated over 50 measurements (y = −1.008x, R 2 = 0.91). (Reprinted with permission of the Biophysical Society from Yamahata et al. [30].)

Full size image

Following these experiments, to know the origin of the conductivity of the DNA bundle, Yamahata et al. measured the transient current through a DNA bundle in different humidity conditions. The relative humidity was decreased slowly from 75 to 45% in 6 h, at a constant temperature of 21 °C. Figure 7a shows the transient current measured for different values of the relative humidity. In the transient current measurement, three main phenomena resulting in the current decrease can be identified. The authors attributed the first rapid decrease of I with a typical time constant of τ 1≈0.3 s to electrical capacitance charging in the device, the second one (τ 2≈3 s) to the electrical double layer (EDL) charging that takes place at the interface between the electrodes and the moisturized DNA bundle, and the long-term current decrease (τ 3≈600 s) to the consumption of carriers by electrolysis [66, 67], respectively. The time constant τ 3 is rather large and it is assumed that τ 3 >> τ 1, τ 2, so that the current can be approximated by an exponential decay with a time constant of τ 3 = l 2/μU after a certain time t >> τ 1, τ 2 [73, 74]. With such a timescale in Fig. 7a (0 s ≤ t ≤ 120 s), the combined effects of τ 2 (which varies with the relative humidity) and τ 3 can be seen: For lower humidity levels, because of the reduced concentration of charge carriers, the time required to establish the EDL is longer, resulting in an increase of τ 2 (this affects the slope of the curve for t < 60 s). For higher humidity levels, mainly the slope due to τ 3 can be observed. With the assumption that the decrease in carrier charge was always negligible, the stabilized current (recorded at t = 60 s in Fig. 7a) at different values of the relative humidity is plotted in Fig. 7b. As previously observed by Kleine-Ostmann et al. [66], the conductivity of the DNA bundle increased exponentially with the relative humidity. The values of τ 3, which were extracted from transient measurements recorded over several minutes, did not vary significantly with the relative humidity, indicating that the carrier mobility μ remains constant with the relative humidity since τ 3 = l 2/μU. Therefore, Yamahata et al. concluded that the carrier concentration n is responsible for the current increase with humidity and proposed the following relation for the conductivity:\(\sigma \propto \sqrt {N_{\text{w}} } \,\exp \left( {{{ - q^2 } \mathord{\left/ {\vphantom {{ - q^2 } 2}} \right. \kern-\nulldelimiterspace} 2}\varepsilon dRT} \right)\), where N w is the number of water molecules adsorbed per nucleotide, R is the gas constant, T is the absolute temperature, d is the equilibrium separation distance of the charges in the neutral species, and ɛ is the permittivity. In this equation, the expression exp(−q 2/2ɛdRT) takes into account the energy needed to separate the charge species during the ion-generation mechanism [63, 75].

Fig. 7
figure 7

Effect of the humidity on the transient current and on the conductivity of a DNA bundle. Data recorded at 21 °C (±1 °C overall fluctuation) for different humidity levels (relative humidity ±0.2% for each curve). The relative humidity was decreased from 75 to 45% in 6 h. a Transient current through a DNA bundle for different humidity conditions. The recording were obtained for a 5-V step. The dotted lines are guides for the eye. b Exponential decrease of the current with decreasing humidity. The curve is an exponential fit. (Reprinted with permission of the Biophysical Society from Yamahata et al. [30].)

Full size image

The increase of σ is mainly caused by the increase of permittivity ɛ with the relative humidity [75, 76]. Several authors attempted to model the charge transfer in the DNA–water complex and considered this system as a heterogeneous dielectric medium consisting of different regions with different dielectric permittivities [77–79]. These authors proposed that DNA in solution has different dielectric zones [79]: the nucleobase, the bases and sugar–phosphate backbone, the bound-water zone (3 Å thick) adjacent to the surface of the DNA fragment (ɛ 1), and the bulk water zone (ɛ 2). Yamahata et al. considered that there is space (definitely much more than 3 Å) between each DNA strand in the bundle: the bundle is a spongelike material rather than a compact assembly of DNA molecules, and the water molecules can penetrate within the bundle. This is consistent with the fact that the current varies linearly with the cross section. Indeed, this dependence shows that the conductivity does not only occur at the surface of the DNA bundle, but also comes from the contribution of each DNA strand inside the bundle. Hence, Yamahata et al. concluded that in their case the global dielectric constant was a function of ɛ 1 and ɛ 2, which both depend on the relative humidity.

Conclusions

In summary, MEMS-based microsystems have a wide range of possibilities to serve as powerful tools in manipulating and analyzing biomaterials at the molecular level. The miniaturization of experimental tools achieved a fast response speed of a thermodevice which realized rapid temperature alternation in biological experiments. Trapping DNA in a microchannel realized isolation and immobilization on electrodes at the single DNA molecule level. Silicon-based microtweezers enabled us to analyze the humidity dependence of charge transport through DNA. These microdevices with MEMS technologies have additional advantages such as batch process for mass production by a batch process and on-chip integration of multifunctions. This may increase the efficiency of bioexperiments which requires a large number of data and experiments. Furthermore, this approach requires no biological mutations or chemical treatments to the biomaterial itself; hence, it is widely applicable and the compact devices also allow samples to be analyzed at the point of need. We strongly believe that the approach to design and manufacture MEMS devices based on requirements in single molecule measurements might provide breakthroughs and powerful tools in molecular biology.