1Neural Development Group, Division of Cell and Developmental Biology, College of Life Sciences, University of Dundee, Dundee, UK, 2Wellcome Trust Centre for Gene Regulation and Expression, College of Life Sciences, University of Dundee, Dundee, UK
Das, R. M., Wilcock, A. C., Swedlow, J. R., Storey, K. G. High-resolution Live Imaging of Cell Behavior in the Developing Neuroepithelium. J. Vis. Exp. (62), e3920, doi:10.3791/3920 (2012).
The embryonic spinal cord consists of cycling neural progenitor cells that give rise to a large percentage of the neuronal and glial cells of the central nervous system (CNS). Although much is known about the molecular mechanisms that pattern the spinal cord and elicit neuronal differentiation1, 2, we lack a deep understanding of these early events at the level of cell behavior. It is thus critical to study the behavior of neural progenitors in real time as they undergo neurogenesis.
In the past, real-time imaging of early embryonic tissue has been limited by cell/tissue viability in culture as well as the phototoxic effects of fluorescent imaging. Here we present a novel assay for imaging such tissue for long periods of time, utilizing a novel ex vivo slice culture protocol and wide-field fluorescence microscopy (Fig. 1). This approach achieves long-term time-lapse monitoring of chick embryonic spinal cord progenitor cells with high spatial and temporal resolution.
This assay may be modified to image a range of embryonic tissues3, 4 In addition to the observation of cellular and sub-cellular behaviors, the development of novel and highly sensitive reporters for gene activity (for example, Notch signaling5) makes this assay a powerful tool with which to understand how signaling regulates cell behavior during embryonic development.
1. Dish Preparation
2. Embryo Electroporation
3. Collagen and Slice Culture Medium Preparation
4. Slice Culture
5. Imaging Embryo Slices
6. Representative Results
An example time-lapse sequence of a spinal cord progenitor cell is shown in Fig. 2a and corresponding Movie 1. Imaging was started on a spinal cord slice from a two-day-old (HH stage 12) embryo. This cell was transfected with a construct expressing GFP-αTubulin. During this early stage, neural progenitor cells undergo predominantly progenitor-progenitor mode divisions during which the cells divide to generate two further cycling progenitor cells. Fig. 2b and corresponding Movie 2 shows a cell transfected with GFP-GPI (GPI anchored GFP), marking the cell membrane. This cell undergoes a division during which the basal process splits into two and is equally inherited by the daughter cells.
Figure 1. Slice culture protocol for time-lapse imaging.
Figure 2. Cell behavior during normal spinal cord development. (a) Selected frames from Movie 1. Cell expressing GFP-tubulin divides with a cleavage plane that is perpendicular to the apical surface (2 h 20 min), generating two daughter cells that divide once again (24 h 23 min and 25 h 54 min). (b) Selected frames from Movie 2. Cell transfected with GFP-GPI undergoes cell division during which the basal process is split (white arrows) and equally inherited by the daughter cells.
Movie 1. Click here to view movie.
Movie 2. Click here to view movie.
We present here a novel time-lapse imaging assay to monitor cell behavior in chicken embryonic slice culture. This assay enables high resolution imaging of living tissue for up to 70 hours, although time frames between 24-48 hours are easier to capture. The use of a high NA oil objective and relatively short intervals between time-points enables image acquisition at high resolution, as well as allowing us to monitor cell behavior that often occurs rapidly, and can be easily missed during conventional time-lapse imaging using confocal microscopy.
The major advantage of this assay is the ability to image cells over long periods of time. The use of wide-field6 instead of confocal laser scanning7 microscopy is critical for this. Confocal microscopy has traditionally offered several advantages over wide-field microscopy, including the ability to take optical sections and eliminate out of focus information directly7, but the use of a pinhole leads to a loss of light to the detector8, excluding a significant amount of information, and necessitating longer exposure times. Wide-field microscopy, on the other hand, uses full field illumination and all the light passing through the objective is sent to the detector. This coupled with a high quantum efficiency cooled coupled device (CCD) camera ensures imaging with a high signal to noise ratio compared to confocal microscopy9, with very fast exposure times. In our application, we must image for long periods, record 3D stacks and ultimately resolve small—near-diffraction limited structures. Although confocal microscopy prevents out-of-focus light from ever reaching the detector and thus significantly reduces background in thick, densely-labeled samples7, 8, it only achieves a usable signal-to-noise ratio with much larger light input than wide-field microscopy10. For very light-sensitive sparsely-labeled samples like our electroporated spinal cord slices, therefore, wide-field microscopy performs better than confocal microscopy. When combined with image restoration by deconvolution, which removes out of focus information and improves contrast, wide-field is then particularly good for detection of small, dim objects11. While not appropriate for every tissue imaging experiment, this approach has been very effective for our live cell imaging application.
The choice of fluorescent protein is an important factor that can influence cell survival. We find that the best results are obtained from constructs that use green fluorescent protein (GFP)12 as a marker. We are currently assessing a range of red fluorescent proteins that can be effectively used in combination with GFP for dual channel time-lapses. There are a variety of such proteins available13. Although many proteins are stable as fusions with a fluorescent protein, some proteins may be rendered unstable by fusion, resulting in reduced cell viability. In such cases, the use of constructs containing an internal ribosome entry site (IRES) is useful to separate the protein of interest from the fluorescent protein.
When imaging spinal cord slices, care must be taken to minimize exposure to fluorescent light. The microscope eyepieces must be used only with transmitted light (brightfield) to locate and position slices. Fluorescent images should always be acquired using the microscope software using minimal exposure times. These should be kept in the range of 5-50 ms and the use of neutral density filters should be experimented with. Although longer exposure times may initially produce clearer images, the phototoxic effects of fluorescent light exposure may lead to cell death. The first 5-10 μm of tissue that is closest to the coverslip should not be imaged as this region contains tissue that may have been damaged during slicing.
Although the examples presented here are of embryos electroporated at HH stage 10 and imaged 6-7 hours later, this method can be used for embryos up to HH stage 18. At later stages, the spinal cord tissue is much larger and so is difficult to slice by hand. In these cases, embedding the embryos in low melting point agarose and sectioning with a vibratome may yield better results. Even though we present this assay as a method for imaging cells in the developing spinal cord, this approach has also been modified to image other embryonic tissues, including the sensory placodes3.
No conflicts of interest declared.