In vivo measurements of blood flow and glial cell function with two-photon laser scanning microscopy

Abstract

Two-photon laser scanning microscopy is an ideal tool for high-resolution fluorescence imaging in intact organs of living animals. With regard to in vivo brain research, this technique provides new opportunities to study hemodynamics in the microvascular system and morphological dynamics and calcium signaling in various glial cell types. These studies benefit from the ongoing developments for in vivo labeling, imaging, and photostimulation. Here, we review recent advances in the application of two-photon microscopy for the study of blood flow and glial cell function in the neocortex. We emphasize the dual role of two-photon imaging as a means to assess function in the normal state as well as a tool to investigate the vascular system and glia under pathological conditions, such as ischemia and microvascular disease. Further, we show how extensions of ultra-fast laser techniques lead to new models of stroke, where individual vessels may be targeted for occlusion with micrometer precision.

Full text

Page 1

C H A P T E RT E N
In Vivo Measurements of
Blood Flow and Glial Cell
Function with Two-Photon
Laser-Scanning Microscopy
Fritjof Helmchen* and David Kleinfeld†
Contents
1. Introduction
2. Two-Photon Microscopy of Fluorescent Labels as a Tool
for Brain Imaging
2.1. Animal preparation
2.2. In vivo staining of blood vessels and glial cells
2.3. Functional imaging
3. Photoprocesses for Targeted Disruption of Vascular Flow
3.1. Photoinduced thrombosis
3.2. Plasma-mediated ablation
4. Outlook
Acknowledgments
References
232
234
235
235
241
247
247
247
248
251
251
Abstract
Two-photon laser scanning microscopy is an ideal tool for high-resolution
fluorescence imaging in intact organs of living animals. With regard to in vivo
brain research, this technique provides new opportunities to study hemody-
namics in the microvascular system and morphological dynamics and calcium
signaling invariousglial cell types. Thesestudies benefit fromtheongoing devel-
opments for in vivo labeling, imaging, and photostimulation. Here, we review
recent advances in the application of two-photon microscopy for the study of
blood flow and glial cell function in the neocortex. We emphasize the dual role of
two-photonimagingasameanstoassessfunctioninthenormalstateaswellasa
tool to investigate the vascular system and glia under pathological conditions,
Methods in Enzymology, Volume 444
ISSN 0076-6879, DOI: 10.1016/S0076-6879(08)02810-3
#2008 Elsevier Inc.
All rights reserved.
* Department of Neurophysiology, Brain Research Institute, University of Zurich, Zurich, Switzerland
{Department of Physics, University of California, San Diego, La Jolla, California
231

Page 2

such as ischemia and microvascular disease. Further, we show how extensions
of ultra-fast laser techniques lead to new models of stroke, where individual
vessels may be targeted for occlusion with micrometer precision.
1. Introduction
All tissues are mixtures of various cellular components that perform
specialized tasks. In the central nervous system (CNS), several cellular
structures exist along with the electrically excitable neuronal cells that
process information and fulfill the primary job of the CNS. All neuronal
networks within the brain are tightly interwoven with networks of glial
cells, andthus the underlying microvasculature. Thethree major subtypesof
glial cells are astrocytes, oligodendrocytes, and microglial cells (Fig. 10.1).
Astrocytes have long been recognized for their predominant role in main-
taining the homeostasis of neurons. They contribute to the regulation of local
blood supply and provide nutrition to the neurons. More recently it has been
recognized that the communication between astrocytes and neurons is much
tighterthanpreviouslythought, for example via therelease of gliotransmitters
from astrocytes and through the ability of astrocytes to sense glutamate
Basal lamina
Interneuron
Astrocyte
Astrocytes
Neuron
Pericyte
Microglia
Capillary
Endothelial
cell
Tight
junction
Figure 10.1 Vascular constituents of the rodent blood^brain barrier. The barrier is
formedbycapillaryendothelialcellsthatformthe capillary wall.These aresurrounded
by basal lamina, pericytes, and astrocytic perivascular endfeet. Astrocytes provide one
cellular link to neurons; the other is supplied by direct connections from inhibitory
interneurons.The figure also shows microglial cells that populate the brain. (Adapted
from Abbott, N. J., R€ onnb€ ack, L., and Hansson, E. (2006). Astrocyte^endothelial inter-
actionsattheblood^brainbarrier.Nat.Rev.Neurosci.7, 41^53.)
232
Fritjof Helmchen and David Kleinfeld

Page 3

(Volterra and Meldolesi, 2005). Astrocytes are also closely linked to the
microvasculature system. Their ‘‘endfeet’’ enwrap the entire vasculature sys-
tem so that they are in continuous communication with the endothelial cells.
In the brain, endothelial cells from a tight blood–brain barrier, which leads to
immunological isolation of the brain. Recent work has probed the in vivo
dynamics of astroglial signaling in the control of blood flow (Takano et al.,
2006) and following sensory stimulation (Wang et al., 2006). Lastly, recent
evidencesuggeststhatastrocyteshaveimportantphysiologicalfunctionsonthe
scale of individual synapses as well as on the level of neural circuits.
Oligodendrocytes are the myelin-producing cells of the CNS and play a
role roughly analogous to Schwann cells in the peripheral nervous system.
Oligodendroctytes form the myelin sheets around a neuronal axon, partic-
ularly in the long-range axonal tracts that form the white matter, and
accelerate the propagation of action potentials. Finally, microglial cells are
the immune-competent cells in the CNS. In contrast to astrocytes and
oligodendrocytes, which are of ectodermal origin, microglial cells derive
from the mesoderm and are part of the mononuclear phagocytic system.
Microglial cells invade the brain during development and become perma-
nent residents. They can be activated by a variety of stimuli, mostly in
response to any kind of tissue injury, and quickly transform into phagocyt-
ing cells. Recent work, detailed below, has probed the in vivo dynamics of
microglia during resting conditions and in response to brain vascular injury
(Davalos et al., 2005; Nimmerjahn et al., 2005).
We now turn to the angioarchitecture of the vascular system per se.
The topology of arteriole networks varies among different brain regions.
Intheneocortex,thetopologyisstereotyped andwell understood(Fig. 10.2).
The surface of cortex is covered by highly interconnected mesh-like
networks of arterioles formed by anastomoses between branches of the
great cerebral arteries (Brozici et al., 2003). The most studied of these is
the network formed by anastomoses of the middle cerebral artery, which
supplies parietal cortex and parts of the striatum. Recent data support the
notion that the surface network functions as an ideal grid, so that interruption
to flow at a single point receives compensation from flow in neighboring
regions (Schafferet al., 2006).The subsurface microvascular also formsa series
of loops, in three rather than two dimensions. While these loops are quite
tortuous, their detailed organization is a topic of ongoing research. Finally,
the penetrating arterioles connect the surface arteriole network to the under-
lying microvessels that course throughout the parenchyma. The penetrating
arterioles are bottlenecks, in that occlusion of a single penetrating vessel leads
to a loss of supply to all microvessels in the territory fed by that arteriole
(Nishimura et al., 2007). Recent work, detailed below, considers the changes
in blood flow dynamics that accompany a single, targeted occlusion
(Nishimura et al., 2006; Schaffer et al., 2006) as a model of microstroke
(Vermeer et al., 2003).
In Vivo Measurements with TPLSM
233

Page 4

2. Two-Photon Microscopy of Fluorescent
Labels as a Tool for Brain Imaging
Historically, experimental methods to investigate glial function in the
intact brain have been limited. Different from neurons, glial cells are
electrically mostly silent, which makes them invisible to standard in vivo
electrophysiological methods, such as extracellular recordings. Similarly, a
fine-scale investigation of blood flow down to the level of individual
capillaries has been difficult, although coarse measurements of spatially
averaged blood flow in the brain are routinely performed, such as by laser
Doppler flowmetry. Advances in imaging technologies that have been
achieved during the past decade now provide a means to investigate blood
flow and glial function in vivo. The key technology is two-photon laser-
scanning fluorescence microscopy (TPLSM) (Denk et al., 1990, 1994),
which can achieve penetration depths of 500 to 1000 mm into tissue
(Helmchen and Denk, 2005). Two-photon microscopy not only provides
micrometer spatial resolution deep within the tissue, but also permits
Surface arteriole
network
Subsurface
microvasculature
Penetrating
arteriole
1mm
Figure 10.2
latureofamousewasfilledwithfluorescentlylabeledagaroseandaregionimagedwith
theall-opticalhistologymethod(seeTsaietal.,2005).Thereconstructedimagewashigh-
lightedinthe vicinityofasinglepenetratingarteriole(yellowhue).(Fromunpublished
dataof P.S.TsaiandP.BlinderintheKleinfeldlaboratory.)
Major vasculartopologiesinthesupplyofbloodtoneocortex.Thevascu-
234
Fritjof Helmchen and David Kleinfeld

Page 5

dynamic measurements over a wide range of time scales, ranging from
milliseconds to years. Besides improvements in microscope technology
(Helmchen and Denk, 2002), the development of various techniques for
fluorescence labeling of specific tissue components in vivo has been of
utmost importance. In particular, many variants of fluorescent proteins
(FPs) have been created as anatomical markers or biosensors. These con-
structs can be specifically expressed in particular subsets of cells and are
revolutionizing the field of neuroscience by opening new possibilities to
study cellular dynamics in the living animal (Garaschuk et al., 2007;
Kleinfeld and Griesbeck, 2005; Miyawaki, 2005). In the following discus-
sion, we restrict our focus to progress that has been made for in vivo labeling
of the components of the vascular and the glial system in mice and rats.
We highlight emerging applications for revealing their functions under both
physiological and pathological conditions.
2.1. Animal preparation
Most measurements of blood flow dynamicsand glial function involve acute
experiments with anesthetized animals. Access to the cortex is through a
craniotomy (Kleinfeld and Delaney, 1996), although with transgenic mice a
thinned-skull preparation may be used with the advantage that the dura may
be kept intact (Frostig et al., 1993; Grutzendler et al., 2002). A metal frame
(Fig. 10.3A)is employed both aspart of achamber abovethe craniotomy, or
thinned skull, and as a support for the animals head in the imaging apparatus
(Fig. 10.3B). We typically mount the animal and all support gear, including
air/gas supplies, on a single plate in our surgical suite and then move the
entire plate to the microscope (Kleinfeld et al., 2008) (Fig. 10. 3C). Anes-
thesia, airway condition, and body temperature are maintained throughout
the experiment and blood gases can be collected, typically every 4 h with
rats, to assess the health of the animals.
2.2. In vivo staining of blood vessels and glial cells
Labeling of the structures of interest with appropriate fluorescent markers is
a prerequisite for dissecting cellular functions with the use of TPLSM. Such
markers label cells or cellular structures as a means to identify cells and target
them for experimental manipulations. Further, markers may also have a
functional dependence, such that their fluorescence properties change
according to changes in particular physiological parameters, such as intra-
cellular calcium concentration. We focus on methods that are available for
staining components of the vascular and glial system in the rodent brain;
reviews of methods for in vivo staining of neurons and neuronal networks
can be found elsewhere (Feng et al., 2000; Garaschuk et al., 2006b, 2007;
Go ¨bel and Helmchen, 2007).
In Vivo Measurements with TPLSM
235