Last week, with a sufficiently large—I hope—collection of root images cached, I indulges in a few extras. I entertained a visit from a friend (Mike Blatt) and his wife (Jane), washed out reagent bottles, and fitted in two experiments. Well, not really experiments, more like explorations.
One afternoon, I imaged celery xylem stained either with congo red or with fast scarlet. Long time Lab Fab fans might remember that I have used to celery xylem to check the performance of the liquid crystal system. But this exploration aimed to compare the dyes. I imaged a dozen fields of each and—someday—I will quantify the anisotropy and orientation of the ribs. Just by looking tho, I sense that congo red is more reliable; that’s good because that’s the dye I used for the project.
On another day, I checked out yet a fourth fluorescent cellulose stain: CBM3-488. While the name is unpronounceable, the dye is interesting. I had requested a sample from my colleague and friend, Daniel Cosgrove, whose lab had synthesized it for their studies of cellulose. Actually, my request was back in early spring just when I thought my project was launched; the 8 µL Dan sent me has sat shrouded in aluminum foil in the fridge ever since. I did get as far as determining that the buffer, sodium acetate, used by the Cosgrove lab for staining, is toxic to roots. Dan told me that I could use MES instead, which is root compatible. I really wanted to try the staining with stuff before leaving.
While fast scarlet, congo red, and even Calcofluor, are relatively small fluorescent molecules that happen to bind cellulose strongly and reasonably specifically, CBM3-488 is a different story. The CBM part stands for cellulose binding domain, taken from a protein that evolved a specific binding site for cellulose. This makes it more specific than chemical dyes which tend to bind other molecules in the cell wall to varying extents. The number 3 in the name refers to this being the third such domain engineered for this kind of thing, there being many cellulose-binding proteins to chose among. And finally 488 refers to the fluorescent molecule, Alexa-488, that was coupled to the protein to make it fluorescent. Quite the package.
On the morning of the day in question, I had to calibrate the confocal for excitation at 488 nm (that is what those numbers mean, Alexa comes in almost as many excitation wavelengths as Baskin & Robins has flavors). Using the quad-slide from Rudolf the calibration went smoothly, confirming that this little tool is the bee’s knees. Of course, calibration required a couple of hours. In truth, I was a bit too pooped to tune it up to the nth. But I got it good enough for exploration.
Later that day, I imaged stained roots. I did not do any concomitant growth imaging, I was just having a look at the cellulose staining. I noticed that the dreaded dark zone 4 was more or less absent. This is good and suggests that CBM3’s affinity for cellulose is higher than that of dyes. Not a surprise and confirms my belief that the dark zone’s darkness results from lack of access not lack of cellulose.
But the really surprising thing was that there was no anisotropy. As I cycled thru the four excitation states (each polarized at a different angle), the image hardly changed. Therefore, when I calculated the net orientation of the dye molecules, they were essentially random (Figure 1).
Figure 1. Staining with CBM3-488. Left: Average fluorescence. Note the faint fibers that appear in some cell walls, running obliquely to the cell’s long axis. Right: Orientation image. This image is calculated from the four input images with excitation polarized at specific angles. The salt-and-pepper of cell walls is similar to that of the background, indicating the absence of a net orientation among the CBM3-488 molecules.
Random orientation contrasts strikingly with images from fast scarlet or congo red in which most cells have a strong net orientation. Also, with those dyes, longitudinal cell walls viewed in cross section always appear to be oriented longitudinally because only cellulose microfibrils that are perfectly perpendicular fail to contribute to a longitudinal orientation signal. Not so much with CBM3-488.
Hmm. I can explain this by postulating that the Alexa-488 molecule is bound to the CBM3 domain via a bond that has rotational freedom. In that that case, the molecule would be spinning with a period of nanoseconds, maybe microseconds, but anyway orders of magnitude faster than the few seconds needed for my images. An alternative explanation is that the Alexa bit binds the protein at several places; maybe CBM3 bristles with cysteins, the group used to forge the link.
If either explanation holds up (I am hoping that Daniel knows something about the chemistry), then the day’s work will provide a negative control. I don’t think a negative control is needed; that cellulose is well oriented in plant cell walls is a given, as is oriented binding to cellulose by a small molecule. But still, one never knows. The cell walls stained with CBM3-488 look like those stained with fast scarlet or congo red in average fluorescence, only the orientation signal is missing. Curious if nothing else. Of course should it prove to be the case that Alexa-488 is stapled on to the CBM3 with no wiggling allowed…