Investigation of laser-induced cell lysis using time-resolved imaging

Using time-resolved imaging, we investigated the lysis of conﬂuent PtK2 cell cultures by pulsed laser microbeam irradiation. Images obtained at time delays of 0.5 ns to 50 m s demonstrate lysis to be mediated by laser-induced plasma formation resulting in pressure wave propagation and cavitation bubble formation. Image analysis enabled quantitative characterization of the pressure wave and cavitation bubble dynamics. The zone of cell damage exceeded the plasma size and serves to implicate cavitation bubble expansion as the primary agent of cell injury. © 2004 American Institute of Physics. @ DOI: 10.1063/1.1705728 #

Highly focused laser microbeams with femtosecond to nanosecond pulse durations are used increasingly for targeted cell lysis, 1 microsurgery of cellular structures, 2 sampling of cellular contents for biochemical analysis, 3,4 and molecular delivery via transient permeabilization of the plasma cell membrane. [5][6][7] Recently, Venugopalan and coworkers examined optical breakdown ͑plasma formation͒ in pure water produced by 6 ns laser pulses at ϭ532 and 1064 nm when focused at high numerical aperture ͑NA͒. 8 They showed that the laser parameters used to produce cell lysis as well as transient cell membrane permeabilization are sufficient to produce optical breakdown. 5 Based on these results, they concluded that laser-induced plasma formation is the mechanism driving a broad range of laser-based cellular micromanipulation techniques.
In this study, we examined the physical mechanisms governing laser-induced cell lysis by imaging potorous rat kidney epithelial ͑PtK2͒ cell cultures following the delivery of a 6 ns pulse ͑FWHM͒ from a ϭ532 nm Nd:YAG laser ͑Brilliant B, Quantel Inc.͒ focused at high NA. The resulting plasma formation, pressure wave propagation, and cavitation bubble dynamics were imaged at delay times spanning five orders of magnitude ͑0.5 ns-50 s͒. The laser output was introduced via the epifluorescence port of an inverted microscope ͑Zeiss Axiovert S100͒ into the back aperture of the microscope objective. A bright-field ͑40ϫ; 0.8 NA, Zeiss Achroplan͒ or phase-contrast ͑40ϫ, 0.65 NA, ph2, Zeiss A-Plan͒ objective was used to both visualize the cells and deliver the laser radiation. PtK2 cells were cultured in phenol red free medium transparent to 532 nm radiation. Cells were grown in Petri dishes with a glass cover slide window for microscope examination.
Time resolved images were acquired by a gated intensi-fied CCD camera ͑PI-MAX, Roper Scientific͒ triggered at specific delay times following the arrival of the Nd:YAG laser pulse. A filter ͑LP 570, Edmund Optics͒ was used to prevent scattered laser radiation from reaching the camera. Imaging was enabled by transilluminating the sample at the desired time delay. For delays Ͻ1 s, we used a dye cell pumped by a split-off portion of the 532 nm laser pulse. The dye fluorescence ͑LDS 698, Exciton Inc͒ was 15 ns in duration and coupled into a multimode optical fiber whose length determined the time delay for illumination. For delays Ͼ1 s, we used an electronically triggered flash lamp ͑Nanolite KL-L, High-Speed Photo Systeme͒ with a 40 ns duration. Images acquired at delays Ͻ1 s were taken using a camera gate width of 0.5 ns while images taken at larger delay times employed a 200 ns gate width. The short gate width provided clear images without averaging and enabled measurement of pressure wave and cavitation bubble dynamics with improved accuracy compared to previous studies. 8,9 While optical breakdown in water 9,10 and selective targeting of cells loaded with light-absorbing micro-and nano-particles 11 have been imaged, this letter provides time-resolved visualization and analysis of cell lysis produced by laser-induced plasma formation.
The threshold for plasma formation in the cell culture, i.e., the pulse energy providing a 50% probability for optical breakdown, was 8 J. The observation of plasma luminescence was taken as evidence of optical breakdown. We examined cell lysis at pulse energies corresponding to 1ϫ and 3ϫ threshold. No cellular effects were observed in the absence of plasma formation. In an earlier study, we measured a plasma threshold of 1.9 J using a cuvette that allowed the immersion of a 0.9 NA objective in distilled water. The higher threshold found here is due to the use of a lower NA objective combined with optical aberrations introduced by epi-fluorescence lamp conditioning optics, a dichroic beamsplitter, and a glass coverslip that lie in the path of the laser a͒ Author to whom correspondence should be addressed; electronic mail: vvenugop@uci.edu APPLIED PHYSICS LETTERS VOLUME 84, NUMBER 15 12 APRIL 2004 beam when using a conventional microscope system. 12 Figure 1 shows typical time-resolved images using bright-field ͓Figs. 1͑a͒ and 1͑c͔͒ and phase-contrast ͑Figs. 1͑b͒ and 1͑d͔͒ objectives. The maximum plasma size is reached at 20 ns and the pressure wave resulting from the plasma expansion is clearly seen in Figs. 1͑a͒ and 1͑b͒. At 67 ns ͓Figs. 1͑c͒ and 1͑d͔͒, the cavitation bubble resulting from the plasma expansion is visible and the pressure wave has propagated outwards. In Fig. 1͑c͒, the outer portions of the cavitation bubble appear dark due to the reflection of the illumination by the curved bubble surface. However, light can pass through the bubble center and cellular damage is clearly visible in this region. While phase-contrast imaging provides better visualization of the PtK2 cells and the shock wave, the region beneath the cavitation bubble surface is obscured. Figure 2͑a͒ provides the radial size of the pressure wave and bubble rim versus time at threshold. Each data point represents the average and standard deviation of three images. Figure 2͑b͒ presents the speeds of the pressure wave u s (t) and bubble rim u B (t) as determined by the derivative of the data fits shown in Fig. 2͑a͒. At early times the pressure wave is supersonic with maximum speeds approaching 2000 m/s for both 1ϫ and 3ϫ ͑not shown͒ threshold; i.e., it is a shock wave. These shock waves decay to sonic speed within 100 ns. The bubble wall speeds are plotted once the bubble rim is clearly distinguished from the shock wave and possess maximum values of 940 and 1070 m/s at 1ϫ and 3ϫ threshold, respectively.
The shock wave amplitudes p s were obtained from the pressure wave speeds using the following relation: 9 where 0 ϭ998 kg/m 3 is the density of water, c 0 ϭ1483 m/s is the speed of sound in water, C 1 ϭ5190 m/s, and C 2 ϭ25 306 m/s. These results are shown in Fig. 3 and reveal maximum pressures of 420-460 MPa which decay rapidly as the shock wave propagates. Figure 4͑a͒ is an image taken 5 s after laser irradiation at 3ϫ threshold showing the large cavitation bubble produced by the breakdown process. The maximum bubble diameter D B was determined from the images and used to calculate the bubble energy E B using the following relation: 9 where p ϱ is atmospheric pressure and p v is the vapor pressure within the expanded bubble ͑2330 Pa at 20°C͒. Table I provides values for E B and D B , along with the laser pulse energy E p , the diameter of the cell lysis region D lys , and the maximum plasma size D pl . The data show that a significant fraction of the laser pulse energy is converted to bubble energy. The cell lysis region, as shown in Fig. 4͑b͒, is defined as the area denuded of cells surrounding the irradiation site and is much smaller than the maximum cavitation bubble size. Specifically, D lys ϭ44 and 80 m for 1ϫ and 3ϫ threshold, respectively, while the corresponding maximum  Fig. 4͑a͒ shows that the later stages of bubble expansion produce considerable deformation of the cells surrounding lysis region as evidenced by their ''bulging'' away from the irradiation site. However, fluorescent dye assays evaluating acridine orange uptake and propidium iodide exclusion immediately after irradiation and further examination after 24 h revealed that nearly all of these cells have intact membranes and are viable. Thus, the viable PtK2 cells neighboring the denuded area survive the deformation produced by shear stress associated with the fluid flow around the cavitation bubble as well as exposure to shock wave pressures of 50-120 MPa. A study of cell deformation induced by ultrasound contrast agent bubbles came to similar conclusions regarding the ability of cells to survive cavitation bubble dynamics. 13 However, a study on corneal endothelial cells showed that cell death was produced in a relatively large region around the denuded area. 14 Close examination of our images reveals the lysis region to be fully developed within 1 s, i.e., during the bubble expansion. This is consistent with a recent theoretical analysis of cell deformation resulting from shock wave passage and cavitation bubble dynamics in extracorporeal shockwave lithotripsy. 15 Application of this analysis to our situation reveals that the hydrodynamic stresses associated with the fluid flow driven by the expanding cavitation bubble is primarily responsible for cell lysis. Moreover, we found that the border of the cell lysis region is specified by the location at which the product of the transient shear stress and the characteristic bubble expansion time falls below a critical value; implying that cell lysis is sensitive to both the stress amplitude and its duration.