LIDo banner

Apply now

Find out more about the different routes to entry and our eligibility criteria

Thomas A Phillips: A method for reproducible high-resolution imaging of 3D cancer cell spheroids

kke
Multicellular tumour cell spheroids embedded within three-dimensional (3D) hydrogels or extracellular matrices (ECM) are widely used as models to study cancer growth and invasion.

Standard methods to embed spheroids in 3D matrices result in random placement in space which limits the use of inverted fluorescence microscopy techniques, and thus the resolution that can be achieved to image molecular detail within the intact spheroid.

Here, we leverage UV photolithography to microfabricate PDMS (polydimethylsiloxane) stamps that allow for generation of high-content, reproducible well-like structures in multiple different imaging chambers. Addition of multicellular tumour spheroids into stamped collagen structures allows for precise positioning of spheroids in 3D space for reproducible high-/super-resolution imaging. Embedded spheroids can be imaged live or fixed and are amenable to immunostaining, allowing for greater flexibility of experimental approaches. We describe the use of these spheroid imaging chambers to analyse cell invasion, cell–ECM interaction, ECM alignment, force-dependent intracellular protein dynamics and extension of fine actin-based protrusions with a variety of commonly used inverted microscope platforms.

This method enables reproducible, high-/super-resolution live imaging of multiple tumour spheroids, that can be potentially extended to visualise organoids and other more complex 3D in vitro systems.

INTRODUCTION

Optical microscopy is the most widely used imaging technique due to general ease of use and relatively inexpensive equipment for the more basic instruments.1 Fluorescence microscopy, where specific proteins on the sample are labelled with molecules (fluorophores) that absorb light at one wavelength and emit at another, is highly popular. This is due to its ability to image the distribution of specific proteins, and because the output is a simple image of the sample as a signal on a dark background (in contrast to methods such as phase contrast, where the image generation process is complex). In fluorescence microscopy, samples are often imaged on inverted setups, where both the illumination and collection occur from below the samples. Such geometry is ideally suited to imaging optically transparent 2D samples such as single cells on a coverslip.

However, many biological samples are considerably thicker and 3D requiring alternative solutions that might involve lowering spatial and or temporal resolution, using phototoxic reagents or molecules, adding complexity to the sample preparation and so on. Several commonly employed imaging techniques share pros and cons in their use for imaging of 3D biological samples.

Epifluorescence microscopy allows for excitation of fluorophores with an LED, mercury or xenon light source of a particular wavelength, resulting in emission of red-shifted photons, with light collection via charge-couple device (CCD) or scientific CMOS (sCMOS) camera.2 The simplicity and relatively low cost of epifluorescence instruments is coupled with comparatively inferior resolution and a lack of sectioning capability compared to more advanced imaging systems, rendering this technique unsuitable for thicker samples. The most common approach to achieve optical sectioning is point-scanning confocal microscopy,3 which is an alternative approach wherein laser illumination improves illumination uniformity and a variable diameter pinhole proximal to the detector rejects out-of-focus light. This creates a sectioned image, allowing for high-resolution 3D reconstructions of complex samples. However, the point-scanning nature of this modality results in slow imaging with increased photo-toxicity and bleaching which is particularly important when imaging live samples.

Spinning disk confocal microscopy is perhaps the most common approach to overcome the slow speed of point-scanning confocal microscopy.4 Instead of scanning the sample with a single focused point, an array of points is created using disks with regularly spaced pinholes (and, in a lower set of disks, micro-lenses) in a spiral pattern (Nipkow disks). By sweeping an array of points over the sample, imaging is faster than point-scanning confocal and less photo-toxic, while still achieving sectioning. All the methods described here are limited by the Abbe limit, which quantifies how resolution is limited by the diffractive properties of light, for fluorescence microscopy in practice to around 200 nm.5

Read full publication here