Research highlights

Lensless imaging of moving objects

Advances in high pixel density image sensor chips and image processing methods have led to the development of microscopy systems that can operate without the use of lenses, termed ‘lensless’ or ‘lensfree’ imaging.^1–3 By removing the lenses, the size, mass and cost of the imaging systems are dramatically reduced, and the inherent optical limitations of lenses (aberrations and chromaticity) are also minimized. The field of view (FOV) is also enlarged, because lensless imaging can utilize the whole area of the image sensor. However, the image sensor can only capture the shadow image of the object, with a resolution of at most two times the pixel size (∼2 μm).⁴

To overcome this problem, algorithms have been developed that combine several lower resolution shadow images, shifted with respect to each other by a length much smaller than the physical pixel size to create a single high resolution image.^4,5 The subpixel shifted shadow images can be obtained by raster scanning a light source, with typical FOV values of ∼20 mm² and a resolution of ∼0.6 μm. However, imaging of moving objects is not possible since raster scanning is difficult to apply to motile objects. Yang and colleagues have addressed this challenge by utilizing the inherent motion of the motile microorganisms instead of the raster scanning light source to create shifted shadow images.

Lee et al.⁶ used their previously developed imaging platform (ePetri dish),⁴ replaced the raster scanning light source with a single LED and successfully imaged motile protozoa Euglena gracilis (E. gracilis). They filled a poly(dimethylsiloxane) (PDMS) well with culture medium containing E. gracilis and placed it on a CMOS image sensor (Fig. 1A). The image resolution is known to suffer with increasing distance between the object and the image sensor, so the microorganisms were held at a distance of less than 100 μm. Evaporation of medium was prevented by covering it with a droplet of oil (Fig. 1B). A Peltier module combined with a heat sink was placed underneath the CMOS image sensor to dissipate the heat generated from the image sensor board.

Fig. 1 Lensless imaging platform and captured images: (A), (B) schematic of the lensless imaging platform; (C) traces of individual E. gracilis over 1 s; (D) reconstructed high resolution images of E. gracilis. Figure reprinted with permission from the Royal Society of Chemistry from Lee et al.⁶

Using the pixel super-resolution algorithm, high resolution images of the motile microorganisms were reconstructed numerically (Fig. 1C, D). The resolution was 0.95 μm, comparable to a 0.3 NA 10× objective lens with a FOV of 6 mm × 4 mm. In addition, the researchers demonstrated long-term culture of E. gracilis on their lensless imaging platform and conducted automatic cell counting, motion and shape analysis.

One drawback of the current system is the blurring of images due to the perpendicular rotation of the microorganisms with respect to their translational axis. An algorithm capable of addressing this deficiency still needs to be developed. However, the results of this study demonstrate that a miniaturised and cheap imaging system, which rivals conventional optical microscopes in resolution and offers a greatly increased FOV, is within reach. These imaging systems can be readily integrated with conventional cell culture tools and have the potential to greatly enhance long-term biological experiments. They could also enable high temporal-resolution imaging of cell motion and proliferation over several days.

Microdroplets as biofactories

Nanoscale colloidal crystals have been utilized in numerous bioimaging, therapeutic and optoelectronic applications due to their unique physiochemical properties. Previously employed synthesis approaches often involve high costs and exposure to environmental pollutants and other toxic chemicals. More recently, novel strategies have been developed towards biological synthesis of such nanoparticles. These microbial based methods allow for better control of the particle size during biosynthesis, but are limited due to cytotoxicity of some of the materials.⁷

To address this challenge, Park and colleagues⁸ have recently developed an alternative biosynthesis approach that relied on artificial cellular bioreactors (Fig. 2). In this process each reactor consisted of a single hydrogel microdroplet encapsulating a solution of recombinant E. coli extracts and diffused metal precursors. These droplets were generated by the microfluidic double-flow focusing method (two core aqueous phases enveloped by a lipid sheath phase), which allowed for immediate encapsulation of the cellular extracts in the hydrogel precursor solution. As a consequence, specific metal binding proteins could immediately react with the diffusing metal ion precursors and enable consistent nanoparticle synthesis within the bioreactor droplets.

Fig. 2 Schematic of the microdroplet generation method. (a) Microdroplets were produced using co-flow of cell extracts and NIPAAM monomers in a microfluidic device. (b) The polymerized NIPAAM gel served as an artificial membrane. (c) Different types of precursor solutions were dispersed in the artificial cellular bioreactors. (d) The precursors were transferred into the cells, and nanoparticles were formed in the bioreactors. Figure reprinted with permission from the American Chemical Society from Lee et al.⁸

The microengineered hydrogels protected the cellular components from losing bioactivity and helped control the mass transfer of metal precursors. The inner droplet size was directly correlated with the concentration of metal binding proteins, which controlled the biosynthesis of various types of metal nanoparticles. To increase the utility of their approach, the researchers cross-linked the hydrogel prepolymer chemically, avoiding UV light that could potentially damage cellular components.

In the proof-of-principle experiment, magnetic nanoparticles were generated inside the bioreactor droplets using Fe precursors. Cd/Se quantum dots emitting a strong red fluorescent signal were also synthesized. These two examples confirmed the utility of the bioreactor for synthesizing metal alloy nanoparticles. In addition, the size of synthesized Au nanoparticles was controlled by varying the concentrations of the precursors. It was shown that the Au nanoparticle size increased with increasing Au precursor concentration.

By combining the advantages of previous synthesis methods coupled with hydrogel microdroplets, the authors presented a unique and highly versatile approach for biosynthesis of various metal nanoparticles which have a broad range of biomedical applications. The generated bioreactors provided stable kinetics for metal nanoparticle formation and the synthesized products had a homogenous morphology with little variability from reactor to reactor. This approach has the potential to open new doors to understanding the mechanisms behind biosynthetic reactions and metal growth machinery requirements.

Acoustic tweezers for particle manipulation

Capabilities to trap and manipulate small objects (e.g., beads, cells) are needed in many fields, evidenced by the wide use of optical tweezers in biology, physics and chemistry.⁹ Laser manipulation of particles, however, requires materials to have certain optical properties and can incur heating damage on the samples caused by the focused light.¹⁰

Inspired by a dearth of gentle, non-destructive particle trapping methods, Huang and colleagues have adopted acoustic tweezers to trap and manipulate micrometer-size beads, cells, and even millimeter-size C. elegans worms using standing surface acoustic waves (SAW).¹¹ While acoustic waves have been widely used in microfluidics to manipulate cells and particles,¹² most studies on this topic have previously been on particle mixing, separation, and focusing.¹³ In contrast, Ding et al. have demonstrated the application of a novel acoustic method in manipulating various particles, including live objects, inside a flow cell.¹¹

Two orthogonal pairs of chirp interdigitated transducers (IDT) were fabricated by a metal deposition and lift-off process on top of a lithium niobate (LiNbO₃) piezoelectric substrate. A 2.5 × 2.5 mm² PDMS channel was bonded to the substrate asymmetrically between these two pairs of IDTs (Fig. 3A). Each IDT pair was independently biased with a radio frequency signal to generate SAWs. The interference between these SAWs established a differential pressure field in the fluid, such that particles could navigate toward it and be trapped at pressure nodes. By increasing the input power of acoustic waves from 11 dBm to 27 dBm, the velocity of a 10 μm polystyrene bead moving in the fluid increased by a factor of 50 to 1600 μm s⁻¹. The location of the pressure nodes (and therefore the trapped particles) was controlled by changing the applied signal frequency that generated SAWs. In order to operate at a wide range of frequencies (18.5–37 MHz), the IDTs were designed to have a linear gradient in the spacing between their interdigitated electrodes. Two orthogonal electrode pairs were sufficient to control the movement of particles in x and y directions, as shown in the stacked images in Fig. 3B (a single bovine red blood cell following arbitrary patterns).

Fig. 3 (A) Schematic illustrating a microfluidic device with orthogonal pairs of chirped IDT for generating standing SAW. (B) Stacked images showing dynamic control of a bovine red blood cell to trace the letters “PSU”. (C) Stacked images showing single C. elegans trapped, moved in y direction, and then in x direction. Figure reprinted with permission from the National Academy of Science from Ding et al.¹¹

Compared to optical tweezers, the acoustic approach requires lower power density to move an object and so causes less damage to the target species. For example, the power density required by the acoustic tweezers to move a 10 μm polystyrene bead at a velocity of 30 μm s⁻¹ is about 0.5 nW μm⁻², 10⁷ times less than what is needed by optical tweezers. Using HeLa cells as a model system, the impact of acoustic tweezers on cell viability and proliferation was investigated. The authors observed only a small decrease in metabolic activity and no significant change in DNA synthesis after exposing cells for 10 min to a high power (25 dBm) acoustic field. At this power, the temperature of the system increased by only 6 degrees to 31 °C. The acoustic tweezers have also been used to trap, stretch and move a C. elegans worm (Fig. 3C).

This work is the first demonstration of an acoustic-based method to manipulate particles in 2D. It stands out due to its capability to control whole organisms, e.g. worms, its compatibility with a range of particles regardless of their optical, magnetic, and electric properties, and its potential to handle multiple particles in parallel. In the future, an acoustic equivalent of the holographic optical tweezers may be sought, enabling the selective trapping and manipulation of particles and organisms in 3D.

References

1. S. B. Kim, H. Bae, K. I. Koo, M. R. Dokmeci, A. Ozcan and A. Khademhosseini, J Lab Autom, 2012, 17, 43–49.
2. G. Zheng, J. Biophotonics, 2012, 5, 639–649.
3. A. Ozcan, S. Isikman, O. Mudanyali, D. Tseng and I. Sencan, SPIE Newsroom, July 7th 2011, 2011, pp. 002947.
4. G. Zheng, S. A. Lee, Y. Antebi, M. B. Elowitz and C. Yang, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 16889–16894.
5. W. Bishara, T. W. Su, A. F. Coskun and A. Ozcan, Opt. Express, 2010, 18, 11181–11191.
6. S. A. Lee, G. Zheng, N. Mukherjee and C. Yang, Lab Chip, 2012, 12, 2385–2390.
7. S. H. Kang, K. N. Bozhilov, N. V. Myung, A. Mulchandani and W. Chen, Angew. Chem., Int. Ed., 2008, 47, 5186–5189.
8. K. G. Lee, J. Hong, K. W. Wang, N. S. Heo, D. H. Kim, S. Y. Lee, S. J. Lee and T. J. Park, ACS Nano, 2012, 6(8), 6998–7008.
9. D. G. Grier, Nature, 2003, 424, 810–816.
10. K. C. Neuman and A. Nagy, Nat. Methods, 2008, 5, 491–505.
11. X. Ding, S.-C. S. Lin, B. Kiraly, H. Yue, S. Li, I.-K. Chiang, J. Shi, S. J. Benkovic and T. J. Huang, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 11105–11109.
12. H. Bruus, J. Dual, J. Hawkes, M. Hill, T. Laurell, J. Nilsson, S. Radel, S. Sadhal and M. Wiklund, Lab Chip, 2011, 11, 3579–3580.
13. M. Gedge and M. Hill, Lab Chip, 2012, 12, 2998–3007.

---