Upgrading to five-color imaging

We’ve recently been working on upgrading one of our microscopes to do five-color imaging, using a DAPI / FITC / Cy3 / Cy5 / Cy7 filter set. To do so, we replaced our Sutter Lambda XL with a Sutter Lambda LS xenon arc lamp, containing a special 1100 nm cold mirror (thanks Sutter!) so that more of the infrared emission makes it to the output. The spectra for both lamps, measured at the liquid light guide output, are shown below.new LS

As you can see, there is substantially more output in the near-IR from the LS than the XL. The sharp fall-off at 785 nm is due to the IR blocking filter (a Semrock 785nm short-pass) we use to keep the Nikon Perfect Focus System (PFS), which operates at 870 nm, happy.  IR light at 870 nm from the lamp would interfere with the PFS, so we cut off the longer wavelengths from reaching the microscope. The gradual fall-off in the IR is due in part to the fact that these spectra aren’t corrected for the spectral sensitivity of my Ocean Optics spectrophotometer, which falls off rapidly in the IR. The fall off in the UV is due in part to the 380 series liquid light guide we’re using to connect the lamp to the microscope.

As you can see, choosing a lamp, liquid light guide, and filter combination that can span the 380 – 750 nm excitation range while blocking at 870 nm is a bit tricky and requires some trade-offs. Nevertheless, the excitation with the LS extends out about 120 nm further into the IR and is substantially brighter, allowing use of the five-band filter set.

Illumination Power (mW)

ChannelIR-optimized LSXLSpectra-X
DAPI19 (387/11)2.9 (402/15)13 (387/11)
FITC64 (485/20)11 (490/20)21 (485/20)
Cy387 (560/25)12 (555/25)77 (560/25)
Cy556 (650/13)3.8 (645/30)11 (650/13)
Cy748 (740/13)

This table gives the power, in mW, for three different microscopes with three different light sources in the NIC. In each case the power is measured at the output of a 10x/0.45 Plan Apo objective. This is not exactly a fair comparison, because the bulbs are different ages, the filters are different, and the lamps are connected to the microscopes differently; the Spectra-X is attached through the Nikon epi-illuminator; the liquid light guides from the other lamps are directly mounted to the microscope with a Sutter LG-N29 adapter. This typically yields a ~2-fold increase in brightness compared to the Nikon epi-illuminator arm. Finally, the bulb in the XL is old and I suspect that is why the intensity is so low. That said, the IR-optimized LS is very bright and provides plenty of light across all wavelengths. It definitely justifies having to change bulbs, at least until we can get an LED unit with even more wavelengths in it.

This is an easy way to add an additional channel for fluorescence imaging. Extending further into the UV or IR to add additional channels is challenging owing to a lack of good dyes in those spectral regions, poor camera sensitivity in the IR, and low transmission of optics in the UV.  To get more than five channels probably requires using long-Stokes shift dyes, spectral unmixing, or the use of more exotic dyes with narrower emissions such as quantum dots or lanthanides.

2 thoughts on “Upgrading to five-color imaging

  1. An increasingly common configuration of our Spectra-X Light Engine is with a 740nm Source offering Cy7 excitation, by swapping out our Teal Source for this NIR Source. This should help labs who desire this breadth, while offering the stellar advancements of Solid State Lighting!

    Great Post as Always Kurt!

  2. It turns out that the liquid light guide on our XL had gone bad, which is why the XL intensities are so low. With a replacement light guide installed, the intensities are 2-5x larger.

Comments are closed.