Biology Reference
In-Depth Information
Segments that show large spikes or drifts in fluorescence intensity (due to cell
movement) are excluded from the analysis. Since a constant intensity light source
produces a photon count distribution that follows Poisson statistics, as fluorescent
molecules enter and diffuse through the nonhomogenously illuminated
observation volume, the fluctuations in fluorescence intensity result in a
broadening of the Poisson distribution. This super-Poisson characteristic is
observed in the tail of PCH curve.
Sample histograms are shown in
Fig. 10.3
C. The shape of the histogram is a function
of the number of fluorescent molecules and their molecular brightness. To generate a
histogram, each 10 s fluorescence intensity trace is broken down into 1 million 10
m
s
intervals or bins. The number of 10
m
s bins is plotted on the
y
-axis and photon
counts on the
x
-axis, revealing the number of bins that register 1,2,3,
,
n
photon
counts during one 10 s observation period. Both histograms show an average of
1.25 counts per 10
...
s bin time, which equals 125,000 counts per second.
Dividing by the average number of molecules (calculated from the amplitude of
the autocorrelation curves in
Fig. 10.3
B using Eq.
(10.2)
with a 3D model, where
N
m
7) yields an average molecular brightness of 17,857 CPSM. Histograms for mo-
nomeric YFP (not shown) yield a molecular brightness of 8750 CPSM. In this man-
ner, comparison of molecular brightness values with known controls can be used to
determine receptor-receptor interactions. Even though the 2D FCS and 3D PCH
models yielded different numerical values, both predicted a dimeric structure for
the beta2-adrenergic receptor.
The residuals of the curve fit (
Fig. 10.3
D) provide a measure of how well the data
fit the selected model. In the example provided, the data were fit to a one-component
model for a homogenous population of fluorescence-tagged receptors and
yielded reduced chi-square values equal to unity. The residuals of the curve fit
are
¼
2 standard deviations and are randomly distributed about zero, indicating that
the data are a good fit for the selected model. When the residuals are systematic,
non-uniformly distributed about zero, the data are not adequately described by the
selected model.
PCH provides an estimate of the average molecular brightness of all fluorescent
species present in the sample (
M¨ ller et al., 2000
). Thus, if the receptor being stud-
ied exists as a mixture of species (monomers/dimers/tetramers), then the observed
molecular brightness value would be an average based on the monomer/dimer/tet-
ramer composition of the sample. Initially, PCH data should be fit to a one-
component model with concentration and molecular brightness set as free and
the first-order correction fixed at zero. Reduced chi-square values
<
3 (for runs
3-10 where photobleaching should be minimal) indicate a good fit to the selected
model. If greater than three, two-component and three-component analyses are per-
formed. In this case, molecular brightness values are fixed in multiples of estab-
lished control values for monomers/dimers. Inspection of the residuals of the
curve fit and reduced chi-square analysis are used to determine the goodness of
fit to the selected model.
<
