Nondestructive testing in industry often relies on holography, which is the reconstruction of optical waves to make a three-dimensional image. In combination with microscopy, digital holography can also provide contact-free, marker-free, quantitative, phase-contrast imaging. Here, we describe a system with high-magnification optics that can detect fast processes and provide long-term measurements, even if they require focus tracking.
Figure 1 shows our system for digital-holographic microscopy. This is called an off-axis system because of the large angle between the illuminating and reference waves. Placing the microscope beneath the specimen enables investigation of transparent samples, including living cells in culture medium. Then, a nondiffractive method reconstructs the digitally captured holograms.1,2 Moreover, the applied algorithm avoids two common problems: twin image, which blurs the results, and zero order, which puts a spot on an image. The sharply focused image of the sample aids helps to align the experimental setup.
Figure 1. This system provides inverse, off-axis, digital-holographic microscopy and uses a Nd-YAG laser (λ = 532nm). Nd-YAG: Neodymium-doped yttrium aluminum garnet.
To measure the integral refractive index of cells, we covered them with a capping glass. It gets pressed onto the sample so that the distance to the carrier glass is nearly constant. Calculating the integral cellular refractive index depends on several variables, including the changes in the holographic optical-path length caused by the cellular sample and air near the specimen, as well as the refractive index of the cell culture medium.
We studied two human pancreatic cancer cell lines: PaTu 8988S, a highly differentiated adenocarcinoma cell line, and PaTu 8988T, a poorly differentiated adenocarcinoma with high metastatic potential. We retrovirally transduced the PaTu 8988T cells with an E-cadherin protein expression construct containing an E-cadherin complementary DNA in the expression vector pLXIN from Clontech (Palo Alto, CA). Research suggests that E-cadherin functions as a tumor-suppressor protein.
The cells were seeded on glass slides or tissue culture plates and analyzed at room temperature and normal atmosphere. For all pancreatic tumor cells, we obtained an average refractive index value ncell = 1.38 ± 0.015. This information was used to determine the thickness of adherently grown PaTu 8988T cells and PaTu 8988T pLXIN E-cadherin cells. In Figure 2, panel (a) shows the reconstructed holographic amplitude image of a PaTu 8988T cell, and panel (d) shows an image for a PaTu 8988T pLXIN E-cadherin cell. Panels (b) and (e) show the digital-holographic, phase-contrast images with a gray-level legend for the phase (Δφ) and the corresponding cell thickness (d). Panels (c) and (f) show pseudo-three-dimensional representations of the cell thickness, and those results agree with scanning-electron micrographs of cells from the same lines.2 The cell types can be differentiated by the phase data and the calculated thickness.
Figure 2. (a, d) Digital-holographic microscopy was used to generate images of PaTu 8988T and PaTu 8988T pLXIN E-cadherin cells. (b,e) Gray-level legends reveal the phase and cell thickness from digital-holographic, phase-contrast images for both cells. (c, f) Approximately three-dimensional renderings show the thickness of the cells.
We also used digital-holographic microscopy to study the dynamics of morphological changes in a PaTu 8988S cell after adding a 10μM solution of a marine cell toxin called latrunculin B from Calbiochem (San Diego, CA). This toxin destroys the cytoskeleton. The upper row of Figure 3 shows the time-dependent, wrapped phase distributions (modulo 2π). With increasing time after adding the toxin, the phase distributions indicate that the cell collapses due to destruction of the cytoskeleton. The lower row of Figure 3 shows the corresponding unwrapped phase data—along the arrow-marked cross-sections in the upper row—for the phase distribution and the calculated cell thickness. Phase and cell thickness decrease locally up to about 50%, and the crinkling of the shrunken cell is clearly visible.
Our results show that digital-holographic microscopy can be used to investigate living cells under conventional laboratory conditions. This opens new possibilities for minimally invasive ways to quantitatively measure the dynamic processes of living cellular systems.
Figure 3. The marine toxin latrunculin B changes the morphology of PaTu 8988S, as shown in series of phase-contrast images (modulo 2π) at the top. Collecting unwrapped phase data along the cross-sections indicated by the arrows provides the cell thickness depicted below.
The German Federal Ministry of Education and Research (BMBF) and the Deutsche Forschungsgemeinschaft (DFG) supported this research.