The truth about the movement of NO across cell membranes

Jeffrey Garvin

Hypertension and Vascular Research DivisionDepartment of Internal Medicine

Henry Ford Hospital

Acetylcholine-induced EDRF release

NE

Ach -8 -7 -6 w

2 g

5 min

intact

rubbed

Ach -8 -7

-6 wNE

Furchgott et al., Nature 1980

NO synthesis

COOHHCNH2

CH2

CH2

CH2

NH CNH NH2

COOHHCNH2

CH2

CH2

CH2

NH CNOH

NH2

COOHHCNH2

CH2

CH2

CH2

NHCO

NH2

NO+

L-arginine L-citrulline

Why do we care about NO?

NO is involved in:

1. CNS function and cognition2. Cardiac contractility

3. Peripheral vascular resistance4. Respiration

5. Gut motility and ion absorption6. Renal perfusion and transport

7. Reproduction

Properties of NO

1. It is small.

2. It is non-polar.

3. It is RELATIVELY lipophilic with a partition coefficient of about 5.

4. It is a gas.

5. Its reactive (different from O2 and CO2).

soluble guanylatecyclase

NO

How many think NO diffuses through two bilayers

NOS 3

endothelial cell vascular smooth muscle cell

distance0

5

10

Energy profile of NO with distance based on partition coefficient

mem

bran

e

mem

bran

e

Arb

itra

ry e

ner

gy

un

its

guan

ylyl

cyc

lase

NONO

NO

NO

NO

NO

NONO

NO

NONO

NO

soluble guanylatecyclase

NOS 3NONO

NO

NO

NO

NO

NONO

NO

NONO

NO

NO

A slightly more “realistic” model of NO diffusion through bilayers

endothelial cell vascular smooth muscle cell

There have been no direct measurements of the NO permeability of any cell membrane!!!

There has been one calculation which is widely cited. This value of 76 cm/s was calculated based onsteady-state measurements of NO within an artificial membrane using 2 mM NO.

Free diffusion creates several problems:1. Free diffusion is relatively slow;2. The amount of NO trapped in the membrane is relatively large;3. If NO is only around transiently, the

membrane could act as a trap;4. There is no control over where NO goes; 5. There is no way to regulate NO release;6. There is little control over NO entry.

As you have heard today “gas channels” including aquaporin-1(AQP-1) has been

shown to transport CO2 and other gases.

Organs where AQP-1 and NO synthase are expressed

45

30

20

Gly-AQP-1

AQP-1

EC VSMC(15 g) (5 g)

AQP-1 expression by aortic EC and VSMCisolated from CD1 mice

soluble guanylatecyclase

NONOS 3

Hypothesis

AQP-1AQP-1

vascular smooth muscle cell

endothelial cell

If our hypothesis is correct:1. NO permeability (PNO) should correlate with water permeability (Pf).

2. Increasing AQP-1expression should increase NO flux.

3. Inhibitors of AQP-1 should reduce NO flux.

4. NO flux should be saturable.

5. Purified AQP-1 should transport NO.

inlet

outlet

Camera & Image Analysis

ArcLamp

objective lens

dichroic mirror

mirror

excitation 488 nm

emission535±50 nm

Measuring NO with DAF in cultured cells

Correlation of PNO and Pf in stably transfected CHO cells

1. NO permeability (PNO) correlates with water permeability (Pf).

2. Increasing AQP-1 expression should increase NO flux.

3. Inhibitors of AQP-1 should reduce NO flux.

4. NO flux should be saturable.

5. Purified AQP-1 should transport NO.

Effect of transiently transfecting CHO cells with aquaporin-1 (AQP-1) on NO influx

NO gradient by SPM (5 M NO)

1. NO permeability (PNO) correlates with water permeability (Pf).

2. Increasing AQP-1 expression increases NO flux.

3. Inhibitors of AQP-1 should reduce NO flux.

4. NO flux should be saturable.

5. Purified AQP-1 should transport NO.

Effect of DMSO, an AQP-1 inhibitor, on NO influx into transiently transfected CHO cells

NO gradient by SPM (5 M NO)

0

5

10

15

20

25

30

C AQP-1 C AQP-1

HgCl2

NO

influ

x [fl

uore

scen

ce u

nits

/sec

] p < 0.005

Effect of 20 M HgCl2, an AQP-1 inhibitor, on NO influx into transfected CHO cells

NO gradient by SPM (5 M NO)

0

10

20

30

40

50

C AQP-1 C AQP-1

DMSO

NO

influ

x[fl

uore

scen

ce u

nits

/sec

] p < 0.03

NO gradient by gas (5 M NO)

Effect of DMSO on NO influx into transiently transfected CHO cells

1. NO permeability (PNO) correlates with water permeability (Pf).

2. Increasing AQP-1 expression increases NO flux.

3. Inhibitors of AQP-1 reduce NO flux.

4. NO flux should be saturable.

5. Purified AQP-1 should transport NO.

0 1 2 3 4 5 60

10

20

30

40

50N

O in

flux

[fluo

resc

ence

uni

ts/s

ec]

NO [M]

K1/2= 0.54M

Concentration-dependent NO flux using NO gas

1. NO permeability (PNO) correlates with water permeability (Pf).

2. Increasing AQP-1 expression increases NO flux.

3. Inhibitors of AQP-1 reduce NO flux.

4. NO flux is saturable.

5. Purified AQP-1 should transport NO.

1 M NO

AQP-1

control

NO flux into proteolipisomes made with purified AQP-1

1. NO permeability (PNO) correlates with water permeability (Pf).

2. Increasing AQP-1 expression increases NO flux.

3. Inhibitors of AQP-1 reduce NO flux.

4. NO flux is saturable.

5. Purified AQP-1 increases NO transport.

Do other aquaporins transport NO?

Partial aquaporin family tree

AQP-0

AQP-5

AQP-4AQP-6

AQP-3

AQP-9

AQP-1

AQP-8

AQP-10

AQP-2

AQP-7

AQP-Z

Adapted from Agre et al. J Physiol 542:3-16, 2002

MOCK AQP-3 AQP-4

NO

influ

x [f

luor

esce

nce

units

/sec

]

0

1

2

3

4

5

6

Effect of transiently transfecting CHO cells with AQP-3 on NO influx

MOCK AQP-3 AQP-4

NO

influ

x [f

luor

esce

nce

units

/sec

]

0

1

2

3

4

5

6

Effect of transiently transfecting CHO cells with AQP-4 on NO influx

AQP-3 and AQP-4 may transport NO. More data are required.

How does NO transport by AQP-1 compare to diffusion through the bilayer in “real” cells?

Is it physiologically significant?

Aortic ring preparation

bath solution

aorticring

forcetransducer

outletinlet

gas

0

20

40

60

80

100

-10 -9 -8 -7 -6 -5

Log [Ach] concentration

% c

on

tra

ctio

n t

o P

E

WT

AQP-1 -/-

p < 0.0001

Acetylcholine-dependent relaxation of aortic rings from wild type and AQP-1 -/- mice

The reduction in Ach-induced relaxation in AQP-1 -/-mice is NOT due to:

1.Less NOS 3. There is more in AQP-1 -/- mice than WT.

2. Defective signaling down-stream of NO. Donors that release NO inside VSMCs and cGMP relax rings from AQP-1 -/- mice the same as WT.

The reduction in Ach-induced relaxation in AQP-1 -/-mice could be due to:

1. Reduced NO efflux out of endothelial cells; and/or

2. Reduced NO influx into vascular smooth muscle cells.

Stimulus: 1 mM Ach

0.0

0.5

1.0

1.5

2.0

2.5

3.0

GAPDH AQP-1 siRNA siRNA

NO

rel

ease

[pA

mps

/g]

p < 0.05

NO

rel

ease

(pA

mps

/g)

NO

rel

ease

(pA

mps

/g)

Effect of inhibiting AQP-1 on NO release by pancreatic endothelial cells

NO release by cultured aortic endothelial cells from wild type and AQP-1 -/- mice

p < 0.0001

Relaxation of denuded aortic rings to spermine NONOate, an NO donor that releases NO into the bathing media

NO influx into isolated aortic vascular smooth muscle cells from wild type and AQP-1 -/- mice

0.00

0.05

0.10

0.15

0.20

WT AQP-1

NO

influ

x [fl

uore

scen

ce u

nits

/sec

]

p < 0.002

NO

flux

[flu

ore

sce

nce

un

its/s

ec]

NO

flux

[flu

ore

sce

nce

un

its/s

ec]

We are trying to show that the reduction in NO transport by AQP-1 is physiologically relevant in vivo by showing that total peripheral resistance does not decrease in response to acetylcholine in these mice as much as wild type mice

BUT

it seems that these mice have compensation mechanism including increased prostaglandin production and NOS expression that has frustrated our attempts thus far.

Conclusion1. AQP-1 transports NO.

2. Transport of NO by AQP-1 occurs faster than by diffusion through the bilayer by about a factor of 2.

3. Transport of NO by AQP-1 appears to be physiologically significant.

4. Reduced Ach-dependent relaxation of aortic rings from AQP-1 -/- mice is due to both reduced efflux out of endothelial cells and reduced influx into vascular smooth muscle cells.