Dual Mode Endomicroscopy for Assessing Gene Transfection

Add One

Fibre bundle endomicroscopes usually operate in fluorescence mode: the tissue is stained with a fluorescent dye which, when illuminated at a certain wavelength, emits light at a longer wavelength. Collecting this fluorescent emission tends to produce clear, high contrast images, and also allows back-reflections from the fibre bundle to be removed using wavelength selective filters. Reflectance mode endomicroscopes, which create an image from the light back-scattered from the tissue, have been demonstrated several times, but have found little practical application. Now, Cha et al. from John Hopkins University have developed a dual-mode device that simultaneously collects both fluorescence and reflectance images. They have used this device to measure the efficacy of gene transfection – the deliberate insertion of genes into cancerous cells.

The paper, published in Biomedical Optics Express1, describes an instrument which, in principle, is suitable for use in vivo,  although it was only demonstrated for cells in vitro. The device is very similar to existing fibre bundle endomicroscopes that have been reported in the literature. The main difference is the addition of a 50/50 beamsplitter to divide light returning from the tissue into two parts. One is sent directly to a photo-detector, whilst the other is passed through a dichroic filter and an emission filter, leaving only the fluorescence signal to reach a second detector.

Placing the beamsplitter between the dichroic mirror and the probe is a rather counter-intuitive way of doing things, as it means that the ‘reflectance’ image actually contains both the reflectance and the fluorescence components. If the beamsplitter was instead placed between the laser and the dichroic mirror, a ‘pure’ reflectance image would be obtained, and the efficiency of the fluorescence imaging channel would be increased by a factor of two. Why the authors chose their approach is not explained in the paper, but in any case probably doesn’t have a great deal of practical significance since the reflectance signal will tend to dominate over the fluorescence.

A couple of other things are worth mentioning. The device has a frame rate of only 1 Hz, which is slow in comparison to endomicroscopes designed for in vivo use. However, this limitation is due to the use of a pair of galvo scanners and a slow data acquisition card; there is no reason in principle while the dual-mode approach could not be used with a high speed system.  A problem with reflectance mode imaging is the need to remove back-reflections from both ends of the fibre bundle. Solutions usually involve either index matching gels, polarisation selection, or software subtraction. In this case the problem was lessened by the fact that there was no lens on the end of the fibre bundle – it was placed directly into the assay, reducing reflection from the far end considerably. Reflections from the near end were then removed by software subtraction.

Gene Transfection

The application chosen for demonstration of the instrument was gene transfection. The authors cultured HeLa cells and used four different transfection agents to introduce CFP (a fluorescent protein) into the cells. After incubation for 24 hours, the samples were fixed with paraformaldehyde prior to imaging both with the newly developed device and a standard bench-top microscope.

The general idea was that all cells would show up on the reflectance image, while only those which have been transfected with CFP would appear on the fluorescence image. The authors developed software in Matlab to automatically count the number of cells in each image, and so to calculate the efficiency of the transfection. This then allowed them to compare the different transfection agents.

The main novelty here was in building a flexible probe that could potentially be used in vivo. The key comparison was therefore between the flexible probe and the conventional microscope. The images appear broadly similar, but the flexible probe has poor resolution (3.5 µm) due to the spacing of the cores in the fibre bundle, and so couldn’t resolve sub-cellular features. The fluorescence collection efficiency was also lower, which appeared to lead to a slightly smaller measured transfection efficiency that the bench-top microscope, although the difference wasn’t statistically significant.

The next step for the authors will be to repeat these experiments in vivo. This will probably necessitate a higher speed imaging system and a lens-based probe, both of which are relatively simple to implement.


  1. J. Cha, J. Zhang, S. Gurbani, G. Cheon, M. Li, and J. Kang, “Gene transfection efficacy assessment of human cervical cancer cells using dual-mode fluorescence microendoscopy,” Biomed. Opt. Express  4, 151-159 (2013)

2 Comments add one

  1. Simon says:

    Other then Glavos what would they use? Resonant?

    • Mike Hughes says:

      That wasn’t worded very well – I meant to say that they used two ‘conventional’ galvo mirrors as opposed to a resonant scanner + conventional galvo combination.

      Another other option is to do line scanning rather than point scanning, sacrificing some depth sectioning for higher speed. Spinning disks and structured illumination systems have also been shown to work with fibre bundles – as I’m sure you know!

Leave a Comment

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>