State University of New York at Buffalo

Two-photon fluorescence confocal microscopy of maize tissue
--Pan, SJ; Shih, A; Liou, WS; Cheng, PC

Confocal laser scanning microscopy (LSM) has contributed significantly in the understanding of biological specimens in 3D. However, due to high attenuation (scattering and absorption) of light and strong autofluorescence, application of this technology to botanical specimens is generally more difficult than to animal specimens. Imaging maize tissue presents no exception. The light attenuation is particularly pronounced in the UV spectrum and becomes significantly lower in the IR (Cheng and Kriete, in: Handbook of Biological Confocal Microscopy, Plenum Press, 1995; Bhawalkar et al., Bioimaging, 1997).

Conventional confocal fluorescence microscopy uses shorter wavelength light to excite fluorophores to emit longer wavelength fluorescence. The high attenuation resulting from scattering and absorption (e.g. pigmented structures) in botanical specimens hinders the penetration of the excitation beam and the detection of fluorescence emitting from the specimen. Therefore, it frequently limits the imaging technique to the very surface of the specimen (Cheng et al., 1993, in Multidimensional Microscopy, Springer-Verlag). In addition, the presence of chlorophyll and other photoactive compounds in the plant cells exhibits strong autofluorescence, which hinders the feasibility of using conventional fluorophores such as FITC and Texas Red in multichannel fluorescence microscopy. It has been shown that botanical specimens (maize specifically) generally exhibit a much lower absorbance in deep red and near infrared light (Bhawalkar et al., Bioimaging, 1997), therefore, imaging the deeper part of the specimen by IR is desirable. However, the use of a longer wavelength results in lower transverse and axial resolution and prevents the use of most of the existing biological fluorescent dyes.

The use of two-photon induced fluorescence in conjunction with laser scanning microscopy (LSM) has provided an alternative method in the study of botanical specimens in three dimensions. The technique is based on simultaneous absorption of two photons, each having at least half the energy of the band gap, to excite a fluorophore and induce fluorescence. Because of the quadratic dependence of two-photon induced fluorescence intensity on the excitation intensity, it is possible to achieve depth discrimination even without a confocal aperture in front of the photo-detector (Denk et al., Science, 1990 73-76). Since the two-photon absorption cross-section is much lower than the linear absorption cross-section, the two-photon process is very inefficient in comparison to the single-photon fluorescence process used in conventional fluorescence microscopy. As a result, the signal strength in two-photon fluorescence microscopy is frequently poor.

We report here the use of one of a group of newly developed highly efficient upconverters (fluorophores), 4-[N-(2-hydroxyethyl)-N-methyl)amino pheny]-4'-(6-hydroxyhexyl sulfonyl)stilbene, abbreviated as APSS, for the imaging of maize stem. Two-photon technique, using this fluorophore, is capable of imaging structures over 250mm deep into the maize stem with a minimum degradation in image intensity and still capable of achieving submicron resolution. Maize stem fixed in 1:3 ethanol/acetic acid was used in this experiment. After fixation, the specimen was washed thoroughly in water and stained with 0.1% APSS in acetone. Following many washes in acetone, the tissue was cleared in methyl salicylate. The APSS has high affinity to cell wall.

For the two-photon confocal microscopy, an Olympus Fluorview laser scanning confocal microscope equipped with a Plan-apo 40X (NA=1.3) objective lens was used. The light source was a diode pumped mode-locked laser (850nm, Flare laser, Clark-MXR Inc.) producing a train of pulses of duration ~150fs each, at a frequency of ~120MHz. The average power in the laser beam was 30mW. A 800nm long-wavelength reflecting dichroic beam splitter (Chroma Technology Co., 725DCSP) was used to separate the illumination and detection paths and a 540nm short pass filter was used in the detecting path. Figure 1a and 1b are two optical sections of a stem vascular bundle stained by APSS. Figure 2 (a and b) are isosurface renderings of a stack of 250 optical sections (at 1mm spacing).

This article is dedicated to Professor Dr. D. B. Walden on the occasion of his retirement after many years of continued inspiration through discussions with his students and colleagues in the filed of genetics. The APSS fluorophore was kindly provided by Dr. P. N. Parsad of the Department of Chemistry, SUNY. This project was supported in part by the Academic Development Fund of SUNY to PCC.

Figure 1 (a and b). Two-photon fluorescence image of maize vascular bundle stained by APSS.
Figure 2 (a and b). Isosurface rendered views of a maize stem showing vascular bundles. The 3D rendering was generated from a stack of 250 optical sections.

Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors

Return to the MNL 71 On-Line Index
Return to the Maize Genome Database Page