7/14/2015

1

Single-cell-level measurements of

transcription heterogeneity of highly mobile identical genes

Enrico Gratton and Paolo Annibale

Laboratory for Florescence Dynamics

University of California, Irvine

AAPM 57th Annual Meeting, Anaheim, 2015

Art by Liz Hinde

• Motivation: Spatial dynamics of proteins in the nucleus of cells

• Approach: Orbital tracking method

• System:

• Tracking a gene on the chromatin

• Detection and localization of gene dynamics

• Discussion/conclusions

Outline of the presentation

Some open questions about chromatin

The elongation rate of Polymerase II (PolII) in eukaryotes varies largely across different cell types and genes There is not yet a consensus whether intrinsic factors such as the position, local mobility or the engagement by an active molecular mechanism of a genetic locus could be the determinants of the observed heterogeneity

Approach Employing high-speed 3D fluorescence nanoimaging we resolve at the single cell level multiple, distinct regions of mRNA synthesis within a labeled transgene array This approach allows measuring the transcription kinetics as a function of time with millisecond resolution for total times of thousands of seconds We can also determine the motion of the transcribed gene and correlate the motion with transcription and with the motion of nearby genes

7/14/2015

2

-0.8 -0.4 0.0 0.4 0.8

-0.8

-0.4

0.0

0.4

0.8

y (

m

)

x (m)

3-D particle tracking in the nucleus The orbital tracking method

Valeria Levi

Valeria Levi, Qiaoqiao Ruan, Matthew Plutz, Andrew S Belmont, and Enrico Gratton. Biophys J. 2005; 89(6): 4275-85.

Department of Biological Chemistry, School of Sciences University of Buenos Aires (UBA),

Instrument setup

0

1

2

3

Inte

nsi

ty

angle

0

1

2

3

Inte

nsi

ty

angle particle

0

1

2

3

Inte

nsi

ty

angle

0

3

laser orbit

Principle of particle tracking using the orbital scanner

• Localization precision depends on the photons collected along the orbit • Temporal resolution depends on the sampling time of the intensity along the orbit • 3D orbits in space allow 3D tracking at very high speed

7/14/2015

3

DNA

lac operator repeat

(32)

= lac repressor-EGFP

Chromatin labeling in vivo (CHO-K1)

C6-14

2 mm

Belmont et al.

cell nucleus

0400

800

1200

50

100

150

200

250

0

400

800

1200

Inte

nsi

ty

y (n

m)

x (nm)

High contrast

No photobleaching

Contrast ratio 3-10

on N=39

-0.8 -0.4 0.0 0.4 0.8-0.8

-0.4

0.0

0.4

0.8

x (m)

y (

m

)

In a fixed cell position

is determined with a

standard deviation of

about 20 nm

Dynamics of the fluorescent spot in the context

of structures in the cell and nucleus

association with nucleolus (54%) or nuclear envelope (32%) (N=59)

corralled motion

jumps between regions of confined motion (N=30/59)

-0.8 -0.4 0.0 0.4 0.8-0.8

-0.4

0.0

0.4

0.8

y (

m

)

x (m)

-0.8 -0.4 0.0 0.4 0.8-0.8

-0.4

0.0

0.4

0.8

y (

m

)

x (m)

Motion of the whole cell/nucleus ?

Bar, 2 m

DHFR-BAC cell line

multiple insertions of Lac-operator repeat

simultaneous tracking of

two particles

laser orbit

particle 1

particle 2

7/14/2015

4

-0.8 -0.4 0.0 0.4 0.8

-0.8

-0.4

0.0

0.4

0.8y (

m

)

x (m)

jumps in 36% of the cells (N=80)

Motion of the whole cell/nucleus ?

Statistical tests: Analysis of direction/velocity

αd

d

vector

vector

J

a = 1 if g ≤ 20o (for at least 5 contiguous points) = 0 if g > 20o

g

High values of J the sequence is moving fast and in a linear path

0 100 200 3000

1

2

3

4

J

time (s)

random motion

-0.4 -0.2 0.0 0.2 0.4

-0.4

-0.2

0.0

0.2

0.4

y (

m

)

x (m)

0 100 200 3000

1

2

3

4

J

time (s)

C6-14 cell

Cumulative J probability identifies correlated

movements at two locations in the same

chromosome

Uncorrelated jumps Correlated jumps

7/14/2015

5

Observation: the trajectory is not due to a “pure” diffusion process.

The MSD is not Gaussian (Levy’s statistics)

-0.6 -0.3 0.0 0.3 0.6

-0.6

-0.3

0.0

0.3

0.6

y (

m

)

x (m)

200 nm

random diffusion

MSD t

0 2 4 60

50

100

MS

D (

m

2)

t (s)

active transport

MSD t

Local random diffusion and jumps

100 200 300 4000.00

0.05

0.10

0.15

0.20

pro

bab

ilit

y

jump size (nm)

0 1 2 3 40.00

0.05

0.10

0.15

0.20

pro

bab

ilit

y

residence time (min)

Statistics of jump size and residence time

Distribution maximum is 120nm Distribution maximum is 40s

Statistical tests: Analysis of instantaneous

velocity (not MSD!)

-0.4 -0.2 0.0 0.2 0.4

-0.4

-0.2

0.0

0.2

0.4

y (

m

)

x (m)

87 % of points in jumps are blue (N=23)

0 100 200 300 400 500 6000

200

400

600

800

Nd

ata

velocity (nm/s)

threshold

7/14/2015

6

Jumps in the trajectories are ATP dependent

the jumps are ATP and

temperature dependent

during the steps the DNA

sequence moves “linearly”

Experimental tests

Strategy

Statistical tests

-0.4 -0.2 0.0 0.2 0.4

-0.4

-0.2

0.0

0.2

0.4

y (

m

)

x (m)

-0.4 -0.2 0.0 0.2 0.4

-0.4

-0.2

0.0

0.2

0.4

y (

m

)

x (m)

control azide + deoxy-glucose

No jumps (N = 30)

ATP depletion

Open questions about chromatin

Variability of elongation of Polymerase II (PolII) across different cell types and genes Intrinsic factors such as the position, local mobility or the engagement by an active molecular mechanism of a genetic locus could be the determinants of the observed variability

Approach High-speed 3D fluorescence nanoimaging can resolve at the single cell level multiple, distinct regions of mRNA synthesis within a transgene array We can measure kinetics of transcription with millisecond resolution for total times of thousands of seconds We can also observe the motion of the transcribed gene and correlate the motion with transcription and with the motion of nearby genes

terminator 1 kb

3D Tracking of a DNA Locus During Transcription

910 nm excitation 488 nm and 561 nm excitation

mRNA (MS2-EGFP)

CFP reporter 1 kb 24x MS2 repeats 1.3 kb

PolII

promoter

256x Lac op 10 kb

U2OS 2-6-3 Cell line is a courtesy of Dr. Robert Singer (Albert Einstein College, New York) and Dr. Xavier Darzacq (ENS,Paris)

Adapted from Janicki et al., Cell 2004

Lac repressor mCherry

MS2-EGFP

orbit

Paolo Annibale

7/14/2015

7

3D Motion of the center of mass of the gene

The resolution of the gene location in 3D is 10 nm, enough to establish distances among genes

Time evolution of the fluorescence intensity in the green channel (MS2) along the orbit

Each loci has a different dynamics even if the same transcript is being produced at the same time. Some loci are persistently more active than others during times of 1600s

Photobleaching recovery of one individual loci of gene expression

7/14/2015

8

12x10-3

10

8

6

4

2

0

-2

AC

F

4 6 8

12 4 6 8

102 4 6

s

-1

1

x1

0-3

tau = 5.8 ± 0.4 s

12x10-3

10

8

6

4

2

0

-2

AC

F

4 6 8

12 4 6 8

102 4 6

s

-0.5

0.5

x1

0-3

tau = 4.2 ± 0.1 s

12x10-3

10

8

6

4

2

0

-2

AC

F

4 6 8

12 4 6 8

102 4 6

s

+Dox +Dox

+Actinomycin D

Cell with a single gene

Induction of expression with different treatments and at different time

The loci respond to

expression regulators

A cell line with a single copy

of the gene has very similar

behavior than the cell with

many genes

one loop= 60 bp = 60 steps with kstep

24 stem loops = 24 steps with keff

kstep

keff

kinit

krelease

80x10-6

60

40

20

0AC

F s

imula

tion

4 6 8

12 4 6 8

102 4 6 8

100s

6x10-3

5

4

3

2

1

0

AC

F th

eory &

exp

erim

ent

kpause

Darzacq 2007 Larson 2011

Comparison of experimental and theoretical expectations according to proposed models

Elongation rate statistics from different loci and different cells

Cell variance of average elongation rate

Petal variance of elongation rate

Effect of activators and inhibitors of transcription

7/14/2015

9

Elongation rate is loci dependent

Color coded average rate of elongation for different petals Black dashed lines: experimental MSD of the petals. Orange line: average of experimental MSD. Red line: simulated MSD for a wormlike chromatin fiber at equilibrium. Most MSD traces display crossover between a t0.75 rigid rod regime for time scales <5-10 s (dotted orange line) to t0.25-0.5 (dashed-dotted blue and green lines) regime characteristic of entangled fibers between approximately 10-100 s. Scatter plot of confinement radius Rmax .

Simulated trace of intensity fluctuations and angular displacement of a petal A peak in the cross-correlation function is present only if there is a predominant directional motion. Binary map of cross-correlation of the fluorescence intensity and the angular displacement of the petals, calculated in 128 s time bins. Each line corresponds to a different experiment, and the sign of the cross-correlation amplitude is arranged in order to display negative values first.

Correlation between angle and intensity

Spatial pair cross-correlation at points along the orbit using a tri-fold trajectory

7/14/2015

10

Model of transcription and mRNA export from multimer copies of the same locus within a transgene array

Conclusions

•We followed a loci of gene transcription with nanometer spatial resolution and millisecond time resolution for 1000’s of seconds

•In a large transgene array only few sites of elongation occur

•A gene with active RNA synthesis has a spatial dynamics correlated to the RNA elongation and release

•The specific rate of transcription varies between identical genes in the same cell and at the same time

•RNA transcripts follow specific trajectories after release which imply the presence of topological barriers

•Our observations point out to a loci dependent regulation of gene expression

Thank you

Valeria Levi Paolo Annibale

lfd

NIH-P41-GM103540 NIH-P50-GM076516 LH