Not a new start as much as a new vehicle: getting off the tram and onto the bus. Or to put it more accurately, getting out of the car with smoke spewing from the radiator and hopping on a bicycle. Will I make my destination? Dunno but at least I will be moving.
At long last, the piece I had 3-D printed was finished (Fig. 1). Way back in February, I wrote about the idea of changing microscopes (the metaphorical vehicle in the previous paragraph). The system I brought over here was designed and built to work on a confocal fluorescence microscope made by Zeiss (the LSM 780). In January, I tried imaging roots (simply; no polarization business) on another kind of fluorescent microscope: a multiphoton microscope, in this case built by Olympus (aka Evident). This microscope imaged the roots wonderfully.
Figure 1. Cell phone photo of the multiphoton microscope (a part thereof). The Petri dish with the seedlings is not present. The 3-D printed piece is the white object above the objective lens and below the main body of the ‘scope.
In my Lab Fab episode about this from January, I attempted to explain the optics that distinguishes confocal from multiphoton microscopy. I won’t repeat that here. But despite the repetition (rerun!), I am going to explain why I am excited about using the multiphoton. Anyway, my excitement is part of the current story.
I see three advantages to the multiphoton.
1: I can image roots the cellulose of which is stained with Calcofluor white. This dye stains the root cell walls more evenly than does either fast scarlet or congo red, which are the two cellulose stains available for the confocal (and the liquid crystals).
2: I can image roots while they remain in the agar and in the Petri dish used for their growth. Because the multiphoton is built around an upright microscope (Fig. 1), the Petri dish with the seedlings can simply be placed on the stage (with the lid removed) and the roots imaged. Even tho the roots are inside the agar medium within the dish, the Multiphoton is able to image them through the agar. By contrast, for the confocal, the roots need to be taken out from the agar, stained for an hour or so in solution, and then mounted on a microscope slide—inserting time and disturbance between imaging them for growth and for cell wall alignment.
3: I get a better discrimination of orientation. OK, this one is theoretical but as a scientist, I like theories. For imaging the polarization of fluorescence, with multiphoton imaging the sensitivity is theorized to go as the fourth power of the sample’s orientation; whereas with confocal, the sensitivity goes as the square. This means that the multiphoton should be better able to resolve the orientation of the dye molecules—and hence of the cellulose—which is the great glorious grail of my project.
Naturally, there is a disadvantage. I brought over the liquid crystals in a carrier (a “slider”) designed for a Zeiss confocal; there is no straightforward way to get that slider into the multiphoton’s beam. Sigh.
Nevertheless, goaded by those advantages, I began a chase, pursuing a crooked-forward approach. This involved contacting someone at Evident, contacting another user who had done a custom modification, ordering a couple of bits, but most of all working with Maninder Dayal, a personable Ph.D. student who helps run a maker space in the University of Birmingham’s Engineering Department. With my input, Maninder designed and printed a bit that (after half-a-dozen trials) does the job (Fig. 1). The piece he printed is the white thing above the objective. It has a flange above to slide onto the microscope and a negative-flange below for the microscope’s lens holder to slide on (with the lens) to it. Finally, the piece has a gap on each side to allow inserting the slider with the liquid crystals (that is the long horizontal bar with wires at one end).
On Tuesday, we found out that the whole contraption works. Wow! That of course then descended me into my favorite thing: calibration. Not surprisingly, there were problems. But I will say that I accomplished a crude calibration and took an image of a root (Fig. 2). And again, this might be (certainly is!) my optimism oozing thru, but it looks better than many images from the confocal. Let’s see how far I can push this approach.
Figure 2. Images of a root showing results of the first session on the multiphoton equipped with the gear for polarized fluorescence. Left-hand panel: Average fluorescence from the five input images (each with a distinct state of linear polarization). A cell in the upper region of the image has its outer facing cell wall more or less in the imaging plane (note the more or less uniform fluorescence across the face of the cell). Right-hand panel: for a region of the image, the calculated, net orientation of the fluorescent molecules is shown as a short red line. Over the well stained cell, the lines run nearly parallel to the cell’s long axis, an orientation consistent with that expected for cellulose in this mature region of the root. Where a line passes over a side wall, whether longitudinal or transverse, the line parallels the wall, again showing the orientation expected for cellulose. Because the roots are inside the agar, root hairs emerge no farther than the small bump present on the left-end of the well stained cell.