specific Are proteins fixed or mobile in the cell? FRAP Fluorescence Recovery After Photobleaching
STUDYING PROTEIN DYNAMICS IN LIVING CELLS (2001) Jennifer Lippincott-Schwartz, Erik Snapp & Anne Kenworthy Nature Reviews Molecular Cell Biology 2, 444-456
1
Are specific proteins fixed or mobile in the cell?
STUDYING PROTEIN DYNAMICS IN LIVING CELLS (2001) Jennifer Lippincott-Schwartz, Erik Snapp & Anne Kenworthy Nature Reviews Molecular Cell Biology 2, 444-456
2
1
Fluorescence Recovery After Photobleaching (FRAP) What does it tell us?
On a membrane a well defined region* is bleached and the recovery monitored. The data is fit to:
Fo
Immediately after bleach
F (t ) # ( $ ! ) n =+ F0 n! n =0
1 % 4 Dt & 1 + n '1 + 2 ( "r * )
and the diffusion coefficient is obtained.
*a Gaussian intensity profile on the membrane is assumed (of course).
The paper on single photon FPR: Axelrod, et al., (1976) Mobility measurement by analysis of fluorescence photobleaching recovery kinetics. Biophys J. 16:1055-1069, 1976.
3
Do we need the photobleaching or can we learn about diffusion without it?
4
2
We previously considered the noise in photon detection
Light detection is the detection of photons, which is a statistical process. If for a given intensity a detector registers on average lets say 100 photons in the detection time interval at a given intensity it will in most measurements not be exactly 100 but may be 93, 106, 102, 89. How large are these statistical fluctuations? We found that the variance is equal to the mean photon count.
2 2 "N = m #"n = m # p = N
!
Microscopy-2
Biophysical Methods
Slide 5
Fluorescence Correlation Spectroscopy (FCS)
I ( r,z) = I0e"2r
2 2 2 2 # r "2z # z
e
!
6
3
Simplest cases a single diffusion coefficient (Normal diffusion) 2D FCS 3D FCS
$# &% 1 &% 2 % & % # "r $ ! &%1+ % & ( ' ' "z ( ! D $ & & & & & (
1/ 2
" 1$ 1 G (! ) = $ Nv $ 1 + ! $! D &
# % % % % '
# 1% 1 G (! ) = % Nv % 1 + ! %! D '
"D =
!r2 4D
D = lateral diffusion time D = diffusion coefcient
G(0) =
1 1 = NV Veff C
Veff = " 3 2# r2# z
!
7
The autocorrelation function is
Autocorrelation Function (ACF) G (" ) =
#F ( t )#F ( t + " )
(#F (t ))
2
, #F ( t ) = F ( t ) $ Fmean , " = lag time
This is calculated from digital data (counts/bin, where the fundamental binsize can be as short as 12 ns). ni is the number of photons collected in the ith bin and ! j*binsize is the lag time (). N is the total number of bins.
N+ j
G( j ) =
"n i"n i+ '2 j
N 1$ & #"n i ) N % i=1 (
N * #"n i * "n i+ j , G( j ) =
i=1
$N '2 (N + j )&#"ni ) % i=1 (
, "n i = n i + n mean
!
8
4
FCS fitting functions in terms of shape, they all look pretty much the same and it can be difficult to discriminate between models.
1 0.8 0.6 0.4 1 0.8 0.6 0.4 0.2 0 0.001 0.01 0.1 1 10
1-comp. diff.
1-comp. diff.
0.2 0 0.001 0.01 0.1 1 10
0.5
1.8 1.6
0.4 0.3 0.2 0.1
0.01 0.1 1
triplet only (NO diffusion)
1.4 1.2 1 0.001
exponential
0.01 0.1 1 10
0 0.001 1
1.5 1 0.5 0 0.001
0.8 0.6 0.4 0.2
1-comp. diff. /1-comp. diff. + exponential
1-comp diff. + triplet
0.01
0.1
1
10
0 0.001
0.01
0.1
1
10
1 0.8 0.6 0.4 0.2 0 0.001
1 0.8 0.6 0.4 0.2 0 0.001 0.1 10 1000
1-comp. diff. / triplet
2-comp. diff.
0.01
0.1
1
10
9
Fluorescence cross-correlation spectroscopy (FCCS): 2 forms: 1. Two-color cross correlation uses two different detectors that collect the emission from two different color fluorophores and then cross-correlates the signal.
G (" ) =
! FA (t ) ! FB (t + " ) FA (t ) FB (t )
2. Single color FCCS two detectors are used but only one color fluorophore. The emission is split (50/50) and send to the two detectors and cross-correlated. This can remove detector artifacts arising from after-pulsing.
10
5
Dual Color Cross-Correlation: prospects for observing protein-protein interactions
If the two labeled species (red and green) are not bound they move independently and do not generate a cross-correlation signal (black trace). If some fraction of the labeled species are bound together that fraction produces a crosscorrelation signal. From the G(0) values of the cross-correlation and autocorrelation the fraction of the total bound can be calculated for either species.
autocorrelation
Fluorescence Burst Analysis (FBA)
It is also possible to record continuous streams of photon counting data which is analyzed to locate the photon bursts that occur as a molecule diffuses through the illumination volume or passes through a focused beam in a nanochannel. You can obtain: 1. The molecular brightness - number of photons per burst. 2. The burst width distribution (diffusional information). 3. The "waiting time distribution" (WTD) - the times between bursts (additional diffusional information). The molecular brightness is a more sensitive indicator of dimerization than diffusion alone. A factor of 2 increase the molecular weight changes the diffusion coefficient 21/3 = 1.2, but doubles the brightness.
(t) G
ab
delay (ms)
cross-correlation
(t) G
ab
delay (ms)
11
12
6