Deep tissue imaging of cell structures

A proper diagnosis of skin cancer can only be established by performing a biopsy, a procedure in which a tissue sample is surgically removed from the suspect lesion for microscopic examination. The technique is invasive, time consuming and also requires skilled personnel. Hence the high level of interest for developing non-invasive techniques that would allow microscopic investigation of cell structures in vivo. The problem, however, is that biological tissue, such as human skin, scatters and absorbs light to an extent significantly limiting the imaging depth of a conventional light microscope. In this context, two-photon laser-scanning fluorescence microscopy (TPM) is emerging as a very promising technique.

TPM is based on non-linear optical processes involving the simultaneous absorption of two photons resulting in fluorescence emission. In conventional fluorescence microscopy, one-photon excitation is achieved using UV or visible light. TPM, however, uses near-infrared light, which penetrates biological tissue far better. The two-photon process requires a high level of light intensity, usually provided by a femtosecond laser, and the fluorescence is only obtained at the focal point. With laser-scanning, a 3D image of the cell structure of biological tissue can be acquired without having to physically extract a sample.

TPM is particularly suitable for studying deep tissue. For example, it has been used to image skin both ex vivo¹ and in vivo.^2,3 Autofluorescence from endogenous fluorophores such as reduced nicotinamide adenine dinucleotide phosphate, melanin and collagen, can be detected from the epidermis and upper dermis.³ However, the histological features that can be visualised with TPM require characterization to develop the diagnostic criteria required for clinical use.

We recently used TPM to study freshly excised non-melanoma skin cancer lesions using an approach similar to in vivo imaging.⁴ Optical tissue sections parallel to the tissue surface were obtained, revealing recognized histopathological criteria in specimens of squamous cell carcinoma in situ and superficial basal cell carcinoma. To further facilitate image interpretation, we also developed an algorithm for the automatic detection of cell nuclei and indicative of abnormal cell distributions, as illustrated in Figure 1.⁵

Figure 1. Two-photon image of in situ epidermal cells of squamous cell carcinoma obtained from optical sectioning ex vivo: (a) original image, (b) image-processed view, and (c) automatically detected cell nuclei. The suspected multinucleated cell is highlighted in red. Field of view is 76 × 76μm. Image courtesy of J. Paoli and C. Ljungblad.

Another attractive application of TPM is that it can be used to investigate the transport of exogenous substances, such as topical drugs, in the skin. Low permeability makes topical drug administration problematic, hence the need for an improved understanding of their uptake and distribution mechanisms. Skin permeability is conventionally studied using diffusion cells, a method that provides no information on how the compounds distribute and are transported in the cellular matrix. In this context, TPM has emerged as a valuable tool for visualising the distribution of fluorescent compounds in the skin. For example, some reports have investigated the effect of a penetration enhancer on the skin barrier⁶ and we have used a similar method to study xenobiotics in human skin in vitro⁷. An example image is shown in Figure 2.

Figure 2. A two-photon laser-scanning fluorescence microscopy 3D image of sulphorhodamine-stained human epidermis in vitro. The dye is located in the intercellular region, leaving the corneocytes as black unstained bricks. Swelling of the corneocytes can be observed. Field of view is 321 × 321×54μm. Image courtesy of C. Simonsson.

To conclude, TPM represents a promising technique to perform optical biopsies for non-invasive skin cancer diagnostics. Classical histopathologic features of squamous cell carcinoma in situ and basal cell carcinoma have both been observed in TPM images. In addition, TPM can also be used to investigate drugs topically applied on human skin. Even though two-photon excitation has the advantage of improved light penetration when compared to conventional optical microscopy, some issues still require further study. Future research will be focused on improving image acquisition and optics so that imaging can be performed even deeper into the tissue, and to better control and quantify the fluorescence signal as a function of tissue depth.

The author thanks J. Paoli, C. Simonsson, C. Ljungblad, and M. Smedh for providing images and valuable discussions. The Centre for Cellular Imaging at G¨oteborg University is also acknowledged for the use of imaging equipment and kind support from the staff.