Pictured above is a basic schematic of the time-shared/holographic optical trapping apparatus we designed. Two-dimensional (2D) trap arrays are formed using a Zeiss Achroplan high NA objective (100x/1.25NA) commercial Axiovert 200M optical microscope in conjunction with either one of two diffractive elements: either two acousto-optic deflectors (AODs), one for X and one for Y, or a spatial light modulator(SLM).
After positioning cells, it is possible to photopolymerize a solution of PEGDA around them, fixing them in place. Polymerization is possible by UV illumination either through the condenser (as shown) or epifluorescently.
I wrote LabView code to control the AODs, allowing for an intuitive user interface for the optical trapping setup. Cells can be positioned by dragging-and-dropping them into preformed traps with a “shepherd” beam. As pictured in the movie on the left with E. Coli, with our optical trapping setup we are able to position cells quickly and with sub-micron accuracy.
With the cells position maintained by polymerization of the hydrogel, we can remove them from the trapping microscope, and perform other experiments or image them in other ways. Pictured in the movie to the left is a 3D reconstruction of a confocal image series of a 5×5 array of bacteria. The data was taken on a Leica SP2 confocal microscope using a 63X objective, with the bacteria fluorescently stained using SYTO 9 (Molecular Probes). We believe the bacteria are all aligned along the Z-axis due to torque generated by the trapping beam.
We even have control over z-positioning with our apparatus. By using the SLM as a Fresnel lens in conjunction with the AODs, we can form an array of optical traps at different focal planes. Demonstrating this on the left is a movie of a 3x3x3 cube of bacteria. The bacteria have been false colored to correspond with the Z-plane they are on.We can also configure the SLM as a holographic array of traps, rather than a Fresnel lens, and do away with the AODs all together, but appreciate the flexibility and speed that the AODs offer.This work is published in “Laser-Guided Assembly of Heterotypic Three-Dimensional Living Cell Microarrays” (PMID: 16891375).
To optimize viability of trapped cells, we use time sharing of the traps – that is, we rapidly flicker the beam between the array of traps. This meant that the effective power felt by the cells was minimized, improving the viability of the cells. The lower the power delivered, the higher the viability of the cells. We also probed the best wavelength to use for trapping E. Coli, and found that the optimal wavelength seemed to be 900nm. Optimizing the wavelength will probably be necessary for different cell types, as different cells may have different absorbance spectra. We have published this work in Physical Review E, “Optimal Optical Trap for Bacterial Viability” (PMID: 18850868)
Looking towards the future, we are working on assembling systems of eukaryotic cells. With approproiately designed traps, it is easy to shift the larger, eukaryotic cells with optical traps, as pictured to the right.