Fellow imagers,

I'm writing a review paper on Ca imaging in fungi and am considering some
of the pros and cons of "normal" fluorescence, confocal and two-photon
microscopy. I have some questions that stretch beyond the easily-reached
limits of my mathematical skills.

The questions relate to one particular measurement artefact, and the
differences in impact it is likely to have when using these three forms of
microscopy. I dont think I can expect anyone to help with the question
unless I give a detailed background. The artefact was first brought to my
attention several years ago in a paper by Angus Silver et al, 1992, Eur. J.
Physiol. 420, 595-602. They consider a ratio-dye-loaded spherical cell
imaged by "normal"  fluorescence microscopy. To the "eye" of the microscope
the thickness of the cell varies from maximum in the cell centre to a
minimum at the periphery, and the absolute values of the fluorescent
intensities of the two wavelengths also follow the same pattern. Of course,
ratio imaging allows one to account for these effects of varying (apparent)
cell thickness.

But they point out that there are various ways in which the imaging process
may lead to a constant value being either added to subtracted to all the
absolute values, for example by detector saturation, certain forms of
background subtraction, or by a threshold for detection somewhere in the
system. These alterations of the absolute values distort the ratios away
from their true values, and can either generate an apparent gradient across
a cell which doesnt have one, or flatten out a gradient in a cell that
does. This matter is of obvious imporance, because cell biologists are very
often comparing ion levels at the cell periphery with those of more
interior parts of the cell, and for most cells, their periphery is, to the
microscope's eye at least, considerably thinner than their interior.

Although not addressed in the article, it is obvious that the extent of the
havoc wreaked by any set of constant values depends partly on the size of
the constants, and partly on the extent of the difference (for the
ion-insensitive wavelength) between the absolute fluorescence intensities
at the cell centre and periphery. It is the factors that affect this second
aspect I want to concentrate on now. I only have the background to consider
them in a qualitative manner. I am particularly hopeful that someone with
more mathematical/physics expertise may be able to provide some more
quantitative estimates of the impacts of these factors- even if they are
just "ballpark" attempts.

Apart from actual cytoplasmic inhomogeneities, which I dont want to
consider here, there are two physical factors that contribute to the
difference in absolute ion-insensitive intensities at the cell centre and
periphery. In "normal" fluorescence microscopy both would apply. Firstly,
and in "normal" microscopy only, light both excites, and is collected from,
a volume of the cell above and below the plane of focus. Since the height
of the volume in question varies with apparent cell thickness, so must also
the absolute intensity of fluorescence. I have data that suggests that the
height of the volume of excitation/collection can have a dramatic impact on
the intensity of fluorescence: the ion-insensitive fluorescence of
cylindrical fungal cells of 10microns diameter is 33% greater than that of
cells 8 microns in diameter. But my first actual question is: can anyone
suggest a mathematical relationship for how the relative differences in
intensity at the centre and periphery of a spherical cell will vary with
the diameter of the regions of interest, as a proportion of the total cell
diameter? (Assume for the sake of simplicity that there is no light
scattering by the cytoplasm).

Confocal microscopy of course eliminates some of the intensity differences
due to cell thickness, since light collection occurs only from the focal
plane. But there is one component not eliminated by the confocal set-up.
Since the excitation volume is not restricted in the same fashion as the
collection volume, extending above and below the focal plane, a fraction of
the fluorescence excited in that part of the excitation volume outside the
collection volume will stray into the collection volume and be collected
along with its own light.  Can anyone suggest a mathematical relationship
to estimate just what that fraction will be, again assuming no
cytoplasmic.light scattering? And how the fraction would differ between
regions of interest at the cell periphery and cell centre? My guess is that
this component is not too large, because the ion-insensitive fluorescence
of cylindrical fungal cells measured with confocal microscopy shows no
(startling) diameter-dependence.

Two-photon microscopy of course eliminates both contributions to intensity
differences due to cell thickness, because the volume of excitation is the
same as the volume of collection.

As well as any replies to my specific questions, I would of course welcome
any more general comments. Especially if you know of any other sources of
interfering "constants" in addition to the ones I have mentioned, and what
sort of overall  impact you feel they may have on accuracy of measurements.

Cheers,
Geoff Hyde.


Dr Geoff Hyde
Lecturer
School of Biological Science
University of New South Wales
Fax: 612 9385 1558
Ph:  612 9385 1648
Email: [log in to unmask]