Toward versatile digital bioanalysis Free

Digital bioanalysis has emerged as a transformative tool that plays a pivotal role in advancing various life science domains.^1–3 This innovative approach, which originates from single-molecule biophysics, provides insights into the mechanisms of biomolecules with unprecedented sensitivity and precision.^4–9 Digital bioanalysis has demonstrated significant potential to contribute to disease diagnostics as an invaluable testing platform for infectious diseases and latent disorders. Its innate capacity for high sensitivity and precision enables it to detect even the rarest biomarkers or pathogens, holding promise for applications in early disease diagnosis and continuous disease monitoring.^10–12 With continuing technological evolution, digital bioanalysis can catalyze advances in both life sciences and biomedical applications.

Digital bioanalysis is an exceptionally sensitive approach that can fractionate target biomolecules of interest at the single-molecule level for subsequent detection by coupling them with the fluorogenic reaction of enzymes. The presence of biomolecules fractionated at the single-molecule level is then detected as a binary fluorescent signal (representing the presence as “1” or absence as “0”) based on a defined threshold, giving rise to the term “digital” bioanalysis. Digital bioanalysis detection modalities can be broadly classified into three categories (Fig. 1).

• Direct detection of target enzymes: This involves the detection of target enzymes that catalyze fluorogenic reactions. Functional profiling of these target enzymes at the single-molecule level is possible using multiple fluorescent substrates with the appropriate reaction points [Fig. 1(a)]. This innovative technology has recently been used as a diagnostic platform. For example, it can be used to quantify the number of disease-related enzymes in liquid samples at the isozyme level.^6,10

• Indirect detection of target biomolecules by activation of fluorogenic enzymes: In this approach, fluorogenic enzymes are directly activated via one-to-one binding with target biomolecules. This strategy involves the utilization of nucleic acid cleavage enzymes, such as the CRISPR-Cas protein, which are activated upon binding to a specific nucleic acid sequence. This method has been used to detect genes and their fragments as well as miRNAs and cfDNA [Fig. 1(b)]. Recently, this strategy has garnered significant interest as an amplification-free rapid genetic test to identify viral and bacterial genes at the single-molecule level for rapid detection.^13–19 

• Indirect detection of target biomolecules by one-to-one binding with fluorogenic enzymes via antibodies: This involves one-to-one binding between target biomolecules and fluorogenic enzymes mediated by antibodies. It is akin to an enzyme-linked immunosorbent assay (ELISA) performed at the single-molecule level, referred to as a digital ELISA. This technique significantly enhances the detection sensitivity by approximately a million-fold [Fig. 1(c)].^11,20 Digital ELISA has served as a platform demonstrating high sensitivity for decades and has been employed for detecting protein-based antigens.

While digital bioanalysis offers unparalleled advantages in terms of sensitivity, accuracy, and throughput over conventional bioassays, its widespread adoption has been hampered by the high cost of the associated devices and consumables.^21–23 Despite substantial research and development efforts spanning decades, these devices and consumables have remained inaccessible to many researchers because of their high costs, limiting their use to a relatively small research community. Here, our primary objective is to highlight the critical prerequisites for versatile digital bioanalysis, with a specific focus on the advancement of economically feasible and openly accessible instrumentation. Furthermore, we aim to underscore the favorable prospects emerging in the field of biomedical applications resulting from the acquisition of these versatile capabilities.

In digital bioanalysis, it is necessary to capture each target molecule in a microdroplet for highly sensitive and rapid detection.^- Hence, two types of microdevices are primarily used to form microdroplets in conventional methods: microfluidic devices and microchamber arrays (Fig. 2). In microfluidic devices, oil and aqueous solutions are mixed in microchannels to form water-in-oil droplets via hydrodynamic forces [Fig. 2(a)].²⁴ The capacity to generate microdroplets using a commercially available device has reached ∼1000/s, each with a volume of approximately 100 pL.²⁵ The advantages of reduced droplet volume include shorter detection times. Furthermore, an increased number of microdroplets enhances the detection sensitivity. Therefore, it is essential to increase the efficiency of microdroplet production and further reduce their volume; however, this has remained a technical challenge for decades.²⁴ High cost associated with this technology is another important consideration. The cost of microfluidic devices is expensive, and specialized instruments such as flow controllers are required.²¹ Although microfluidic devices that do not require flow controllers have recently been developed for the mass production of microdroplets,²⁶ a concerted effort to reduce costs is essential to increase accessibility and accelerate widespread adoption.

In microchamber arrays, approximately 1 × 10⁶ microdroplets can be formed simultaneously by filling the chamber with an aqueous solution and sealing the orifices with oil [Fig. 2(b)].^1,3 The volume of the microdroplets can be arbitrarily controlled over a wide range of 100 aL–1 pL because it can be defined by the size of the microchamber.³ Therefore, microchamber arrays can generate small-volume microdroplets in less time than microfluidic devices, rendering them more suitable for highly sensitive and rapid digital bioanalysis. Microchamber arrays can be classified into two main types: those fabricated by the photolithography of fluoropolymers²⁷ and those fabricated by the injection molding of plastics, such as cyclo-olefin polymer (COP)²⁸ and polycarbonate (PC).^14,15 In terms of well-established manufacturing processes, microchamber arrays made of fluoropolymers have been widely used in basic research.³ Through-hole structures are formed on the fluoropolymer film coated on the glass substrate (thickness: 0.13–0.17 mm) using photolithography, which are used as microchambers. Although fluoropolymers exhibit high water repellency, the bottom of the chamber is hydrophilic glass. This property makes it easy to retain microdroplets in the chamber after sealing with oil and facilitates chemical modifications, such as biotinylation or PEGylation,²⁹ thereby enhancing the functionality of digital bioanalysis. In addition, fluorescent signals in the microchambers can be detected with high sensitivity because a thin glass substrate allows the use of high numerical aperture (NA) and high-magnification objectives. However, fluoropolymers are relatively expensive, and their manufacturing throughput is low, resulting in relatively high costs. The global supply of fluoropolymers and fluorinated solvents has recently been reduced because of environmental concerns regarding polyfluoroalkyl substances (PFASs).³⁰ CYTOP has historically played a central role as a component of microchamber arrays;³ however, recent scarcity in the supply chain has necessitated the use of alternative materials to fabricate microchamber arrays.

Various plastic-based microchamber arrays have been developed in recent years,^14,15 including commercial products,²⁸ in cooperation with private companies. The initial cost of the injection molding machine is high; however, the material cost of plastic-based microchamber arrays is low, and the production time is only a few seconds, rendering it suitable for mass production.³¹ For example, our latest microchamber array was fabricated in collaboration with Fujifilm using a compact disk (CD) production machine based on injection molding,^14,15 which enabled low-cost mass production. Microchamber arrays are made of PC or COP, which have low autofluorescence and can stably trap microdroplets; hence, they are suitable for long-term fluorescence observations in digital bioanalysis as well as fluoropolymer-based arrays.^14,15,28 Collectively, microchamber arrays offer a variety of options for different applications; however, some peripherals are required to form microdroplets. The most common, including those that are commercially available, are flow cells mounted on top of the microchamber array to exchange the aqueous solution and oil,^3,28 which are expensive to manufacture and assemble. Therefore, similar to microfluidic devices, further cost reductions are required for their general use.

During microdroplet formation, target molecules are captured stochastically, i.e., most of the target molecules are discarded. Consequently, although digital bioanalysis can detect a target at the single molecule level, there is still room for improvement in detection sensitivity due to the inefficient capture of target molecules into the microdroplets. To address this issue, various methods have been developed for enriching target molecules with microparticles, such as magnetic beads^{11,14,20,23,32,33} (Fig. 3). In these methods, microparticles are introduced into a sample solution to immobilize target molecules on their surface through antibody or biotin–avidin interactions. External forces are then applied to these microparticles, efficiently capturing the target molecules within microdroplets. For example, using high-density microparticles allows gravity to efficiently capture them within microdroplets, significantly enhancing the capture efficiency of target molecules.^11,20 Alternatively, when magnetic beads are employed as microparticles, manipulating a magnet enables higher external force to the microparticles, enhancing the capture efficiency of target molecules more rapidly.^14,23

Multicolor microparticles are also used for the simultaneous detection of multiple types of biomolecules.^32,33 For example, by establishing a one-to-one correspondence between the color and the antibody immobilized on the microparticles, it is possible to arbitrarily control the target molecules captured on each colored microparticle. Therefore, matching the color of the microparticles with the fluorescent signal from the droplet enables multiplexed digital bioanalysis, allowing simultaneous detection and counting of multiple types of target molecules.

Observation of the fluorogenic reaction in a microdroplet is essential for digital bioanalysis, which requires a device for fluorescence detection. Fluorescence microscopy is conventionally used for digital bioanalysis [Fig. 4(a)].^1,3 While it can acquire fluorescence images with high spatial resolution and sensitivity, it exhibits inherent issues. Fluorescence microscopy is expensive, and a long time is required to acquire fluorescence images of a large number of microdroplets owing to its small field of view. Therefore, it is necessary to develop an alternative fluorescence detection device for digital bioanalysis to reduce the initial installation cost and further accelerate detection. As commercial complementary metal–oxide–semiconductor (CMOS) sensors with higher resolution have been developed recently, fluorescence images with ∼3 μm resolution can be acquired almost without magnification by combining them with low aberration and high NA macro lenses.^34,35 Our current device (COWFISH) that utilizes a telecentric macro lens³⁴ is capable of simultaneously acquiring fluorescence images of several hundred thousand microdroplets owing to the large field of view (11.8 mm × 7.9 mm), resulting in a significant reduction in fluorescence image acquisition time [Fig. 4(b)]. The total cost of COWFISH, which consists of low-cost commercial components, is approximately $8000, remarkably lower than that of fluorescence microscopy. The commercial availability of these affordable devices has been anticipated in the future, thereby enhancing their versatility. Furthermore, a device in which a microchamber array was directly integrated on top of a CMOS sensor with a built-in fluorescence imaging system has been developed [Fig. 4(c)].³⁶ This device eliminates the need for an external optical system and saves space. However, both the CMOS sensor and microchamber array are disposable, resulting in higher consumable costs.

Digital bioanalysis requires several solution manipulations such as reaction solution preparation and microdroplet formation. It is essential to develop an automated system to improve the reproducibility and efficiency of solution manipulation. Two types of automated systems have been mainly used for solution manipulation in digital bioanalysis (Fig. 5): an automated flow control system for solution delivery and exchange with a flow cell-type microdevice,^28,37 and automated dispensing systems for solution dropping and aspiration with an open-type microdevice.^14,15 In an automated flow control system, a pump is connected to the inlet or outlet of the flow cell to introduce or exchange solutions. Although the flow cell adds to the cost of the microdevice, it allows for the arbitrary control of flow parameters, such as the flow rate and pressure, rendering the operation highly reproducible. Automated flow control systems for digital bioanalysis are commercially available; however, they are expensive.^28,37

Automated dispensing systems are limited to digital bioanalysis performed using a microchamber array with a simple enclosure designed to prevent the solution from spreading (Fig. 5).^14,15 Solution exchange can be performed by dropping and aspirating solutions inside the enclosure in a highly reproducible manner because the dispensing protocol can be well controlled in an automated system. Recently, open-source automated dispensing systems, including their operating programs, have been developed and are commercially available at relatively low price.³⁸ Using such an open-source dispensing system, it is possible to automate digital bioanalysis while reducing installation costs.

In addition to solution manipulation, fluorescence detection and coupling are required for complete automation. When fluorescence microscopy is used as a fluorescence detection system, it can be easily coupled with an automated dispensing system because both systems are automated in most cases. For example, our automated system (opn-SATORI),^14,15 which consists of a commercially available fluorescence microscope (A1, Nikon) and a dispenser (LX24, Biotec), demonstrates full automation of digital bioanalysis by simple coupling via serial and/or TTL communication (Fig. 5). Despite easy coupling, it is difficult to control automated systems for fluorescence detection and dispensing using the same program due to the absence of publicly accessible control programs, resulting in complex and non-user friendly systems. Consequently, control programs for both the automated fluorescence detection and dispensing systems should be publicly accessible. This would enable integrating them easily, thereby reducing the barriers to achieving fully automated digital bioanalysis.

Although digital bioanalysis demonstrates excellent technological advancements, certain technical challenges require attention. These include addressing the cost and availability of microdevices for the mass production of microdroplets and the development of a low-cost and fully automated device for fluorescence detection. We are optimistic that these issues will be resolved effectively in the future, enabling the creation of a seamlessly integrated and fully automated system that incorporates these fundamental technologies. In pursuit of this goal, it is essential to reduce the expense of consumables and facilitate the dissemination of open-source control programs for a range of automated devices, which can be achieved through collaboration with private companies. Currently, we are actively engaged in the collaborative development of affordable microdevices and customizable fully automated systems with private sector partners. Through these efforts, we anticipate that the practical application of these innovations in the near future will significantly enhance the adaptability of digital bioanalysis, thereby offering researchers from diverse areas of the life sciences the opportunity to use this technology extensively.

The reduced cost of consumables and microdevices provides a significant advantage for the introduction of digital bioanalysis to clinical testing. For example, although digital PCR and ELISA have been commercially available for a long time, their consumable costs are higher than those of conventional PCR and ELISA. This cost disparity is a potential limitation of their use in clinical settings. There is a high likelihood that digital bioanalysis will find extensive applications as a clinical test because of its high sensitivity and throughput, provided that consumable costs can be substantially lowered. We are in the process of developing an amplification-free digital genetic test called SATORI^13,39 and aim for its widespread adoption by reducing consumable costs and enhancing device compactness and automation.

We thank all members of the Watanabe laboratory for their constructive discussions. This work was supported by JST CREST (No. JPMJCR19H5), AMED (Nos. JP22he0422018 and JP22fk0108542), and JSPS Grant-in-Aid for Transformative Research Areas A (No. 20H05931) and Scientific Research A (No. 21H04645) to R.W. and JSPS Grant-in-Aid for Scientific Research B (No. 22H01996) to J.A.

The authors have no conflicts to disclose.

Jun Ando: Writing – original draft (supporting). Rikiya Watanabe: Writing – original draft (equal); Writing – review & editing (equal).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.