A close look at brain dynamics: Cells and vessels seen by in vivo two-photon microscopy

Progress in Neurobiology 121 (2014) 36–54

Contents lists available at ScienceDirect

Progress in Neurobiology journal homepage: www.elsevier.com/locate/pneurobio

A close look at brain dynamics: Cells and vessels seen by in vivo two-photon microscopy Stefano Fumagalli a,b, Fabrizio Ortolano a,b, Maria-Grazia De Simoni a,* a b

IRCCS – Istituto di Ricerche Farmacologiche Mario Negri, Department of Neuroscience, via La Masa 19, 20156 Milan, Italy Neurosurgical Intensive Care Unit, Fondazione IRCCS Ca’ Granda/Ospedale Maggiore Policlinico, Via Francesco Sforza 28, 20122 Milan, Italy

A R T I C L E I N F O

A B S T R A C T

Article history: Received 6 November 2013 Received in revised form 17 June 2014 Accepted 29 June 2014 Available online 8 July 2014

The cerebral vasculature has a unique role in providing a constant supply of oxygen and nutrients to ensure normal brain functions. Blood vessels that feed the brain are far from being simply channels for passive transportation of fluids. They form complex structures made up of different cell types. These structures regulate blood supply, local concentrations of O2 and CO2, transport of small molecules, trafficking of plasma cells and fine cerebral functions in normal and diseased brains. Until few years ago, analysis of these functions has been typically based on post mortem techniques, whose interpretation is limited by the need for tissue processing at specific times. For a reliable and effective picture of the dynamic processes in the central nervous system, real-time information in vivo is required. There are now few in vivo systems, among which two-photon microscopy (2-PM) is a truly innovative tool for studying the brain. 2-PM has been used to dissect specific aspects of vascular and immune cell dynamics in the context of neurological diseases, providing exciting results that could not have been obtained with conventional methods. This review summarizes the latest findings on vascular and immune system action in the brain, with particular focus on the dynamic responses after ischemic brain injury. 2-PM has helped define the hierarchical architecture of the brain vasculature, the dynamic interaction between the vasculature and immune cells recruited to lesion sites, the effects of blood flow on neuronal and microglial activity and the ability of cells of the neurovascular unit to regulate blood flow. ß 2014 Elsevier Ltd. All rights reserved.

Keywords: Two-photon microscopy Live imaging Central nervous system Neuroscience Acute brain injury Neurodegenerative diseases

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Blood vessel dynamics by 2-PM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Blood flow dynamics after acute brain injury: 2-PM quantitative analysis . . . . . . . . . . . . . . . . . . . . . 2.1. Hierarchical structure of the cerebrovascular network: mechanisms for blood flow maintenance in 2.2. Blood flow regulation of dendritic spine structural plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. O2 imaging and blood flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Vascular regulation of immune cell activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Route of infiltration and imaging T-cells infiltrating in the CNS (meningitis and stroke) . . . . . . . . . . 3.1. Resident immune cells (microglia) and imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. 3.3. Relevant indices for immune cell active/functional state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.......... .......... .......... penetrating .......... .......... .......... .......... .......... ..........

......... ......... ......... arterioles . ......... ......... ......... ......... ......... .........

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

37 37 39 39 40 41 41 42 43 44

Abbreviations: 2-PM, two-photon microscopy; APC, antigen presenting cell; BBB, blood–brain barrier; ChABC, bacterial enzyme chondroitinase ABC; CNS, central nervous system; EAE, experimental autoimmune encephalopathy; EATs, erythrocyte-associated transients; ECM, extracellular matrix; FRET, fluorescence resonance energy transfer; GFP, green fluorescent protein; IFNg, interferon-g; IL, interleukin; LCMV, lymphocytic-choriomeningitis virus; LPS, lipopolysaccharide; MCAo, middle cerebral artery occlusion; MMP, matrix metalloproteinases; MRI, magnetic resonance imaging; NFAT, nuclear factor of activated T-cells; PCR, polymerase chain reaction; RBC, red blood cell; ROS, reactive oxygen species; SNR, signal to noise ratio; TGFb, tumor growth factorb. * Corresponding author at: Laboratory of Inflammation and Nervous System Diseases, IRCCS – Istituto di Ricerche Farmacologiche Mario Negri, via Giuseppe La Masa 19, 20156 Milan, Italy. Tel.: +39 02 390 14 505; fax: +39 02 390 01 916. E-mail addresses: [email protected] (S. Fumagalli), [email protected] (F. Ortolano), [email protected] (M.-G. De Simoni). http://dx.doi.org/10.1016/j.pneurobio.2014.06.005 0301-0082/ß 2014 Elsevier Ltd. All rights reserved.

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

4.

5.

6.

Regulation of blood flow by cells of the neurovascular unit . . . . . . . . . . Imaging of astrocytes enveloping the endothelium . . . . . . . . . . . 4.1. Imaging of pericytes’ actions on brain capillaries . . . . . . . . . . . . . 4.2. Boosting 2-PM forward: (nearly) available technological improvements Speeding up the system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Increasing the resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction The cerebral vasculature has a unique role in providing a constant supply of oxygen and nutrients to the central nervous system (CNS) to ensure normal brain functions. Blood vessels that feed the brain are not only channels for passive transportation of fluids, but are complex structures made of a number of different cell types (Abbott, 2002; Lok et al., 2007; Nedergaard et al., 2003). These structures regulate blood supply, local concentrations of O2 and CO2, transportation of small molecules, trafficking of plasma cells and fine cerebral functions in both normal and diseased brain (Begley and Brightman, 2003; Konsman et al., 2007). The cerebrovascular structure is generally referred to as the blood– brain barrier (BBB), and forms a physical barrier between the brain parenchyma and systemic blood circulation (Begley and Brightman, 2003; Konsman et al., 2007). The main components of the BBB are endothelial cells, astrocytes, pericytes, perivascular macrophages and microglia, all closely linked to neurons (Berezowski et al., 2004; Ramsauer et al., 2002; Schiera et al., 2003; Zenker et al., 2003). This structure as a whole is defined as the neurovascular unit. In the case of a cerebral pathology, the neurovascular unit responds with local redistribution of the blood flow, changes in the concentrations of blood gases, or increased permeability to specific molecules (Konsman et al., 2007). Other events happen at the endothelial–brain interface and are pathological hallmarks of several CNS conditions, often involving activation of the immune response. The brain immune response can develop with the direct contribution of the neurovascular unit. Mannose-binding lectin deposition on endothelial cells has been described in brain ischemia and has deleterious effects on account of activation of the complement system cascade (Gesuete et al., 2009; Orsini et al., 2012). Moreover, in many diseased states, inflammatory cells from the periphery are recruited to the brain, and their entry into the brain parenchyma is strictly regulated by the neurovascular unit (Davoust et al., 2008; Scholz et al., 2007). The neurovascular unit is thus involved in metabolic and inflammatory pathways, being important in generating and sustaining many pathophysiological cascades that precede or follow a diseased condition of the brain. The inflammatory cascade, in particular, offers a potential, promising therapeutic target, with its role in injury progression in most – virtually all – CNS diseases and arises early in the pathology, with long-lasting effects (Magnus et al., 2012; Dirnagl et al., 1999). The interplay between the immune system and blood vasculature seems fundamental in the pathobiology of CNS diseases, and has stimulated increasing interest in recent years. The dynamic nature of cerebrovascular events and immune cell trafficking, as they evolve over time, requires adequate analysis methods. To date, our knowledge of vascular and immune responses in the diseased brain has been typically based on post mortem techniques, i.e. immunohistochemistry and flow cytometry (Gelderblom et al., 2009). The interpretation of these data is limited by the need for tissue processing at specific time points, but, to get a reliable and effective picture of the dynamic processes involved in vascular and immune events, we need real-time

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

37

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

46 46 46 47 47 48 49 50

information in vivo. Various approaches are available for in vivo imaging, and two-photon microscopy (2-PM) offers a major qualitative advance. This technique allows high-resolution visualization of individual blood vessels and cells and maps of cellular connectivity within a three-dimensional space (Oh et al., 2014; Ortolano et al., 2010; Svoboda and Yasuda, 2006), overcoming the main limitations in spatial resolution and three-dimensional acquisition associated with other in vivo imaging approaches (e.g. MRI, intravital microscopy). Furthermore, 2-PM can provide time-lapse recordings of dynamic events, with timescales range from tens of millimeters per second (blood flow) to days (modification of vascular geometry during development, Fig. 1). The frame acquisition rate must therefore be set appropriately (Table 1). 2-PM has been successfully used to dissect specific aspects of vascular and immune cell dynamics in neurological diseases (Cordiglieri et al., 2010; Fumagalli et al., 2011; Schaffer et al., 2006; Shih et al., 2012) yielding results that could not be obtained with standard methods. This review gathers the latest 2-PM findings on vascular and immune system action in the brain, contributing to the picture of their dynamic interactions. 2. Blood vessel dynamics by 2-PM Blood flow in the brain has been typically studied and measured in vivo using approaches such as laser Doppler flowmetry or MRI (Nakase et al., 1997). These instruments focus on an area measuring hundreds of microns, yielding a mean value obtained from the blood vessels pertinent to that area. MRI allows the acquisition of three-dimensional images which can be processed to yield structural information, including gross vascular architecture, tissue perfusion, BBB integrity, onset of hemorrhage and immune cell infiltration (Klohs et al., 2014). To image vascular architecture and measure perfusion, MRI exploits the different relaxation times between tissue and blood in response to the application of a magnetic field pulse. This is briefly turned on so to align proton spins. As the pulse is turned off, the tissue recovers its initial proton alignment, while the intravascular bed is replenished with fresh proton spins due to blood flowing, allowing to distinguish tissue vs. blood signals. Poor perfusion results in delayed fresh spin replenishment and decrease in oxygen content, allowing to detect disease-related perfusion impairments (Santosh et al., 2008). Blood-associated signal can be enhanced using contrast agents that can be exploited to study the occurrence of hemorrhage (Strbian et al., 2008). Due to their noninvasive nature, imaging solutions based on MRI are readily translatable to clinical use. However, MRI in vivo cannot provide sufficient spatial resolution to proper study small vessels, individual vascular events or interactions between cell and vessels. Moreover, MRI in vivo is limited by the long delays between sequential images, hampering the acquisition of very fast dynamic events (Cho et al., 2011). With the introduction of 2-PM, the possibility of imaging at high spatial resolution and acquisition rate meant that single vessels could be visualized to get parameters associated with vessel architecture (diameter, length, number of

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

38

Fig. 1. Timescales of different dynamic events in the CNS, measured by two-photon microscopy. Each dynamic event in the CNS, studied by in vivo 2-PM, has its own specific time frame of movement, ranging from tens of millimeters per second (blood flow) to days (modification of vascular geometry during development).

branches, topology of vessels), blood flow velocity, changes in O2 concentrations, and to repeat the same measures on the same individual vessel over time. This allowed to collect data on vasculature before (baseline) and after a given pathological event, thus enabling longitudinal analysis (Denk et al., 1990; Svoboda and Yasuda, 2006). In vascular pathologies such as brain ischemia, most studies have focused on the hemodynamic changes in different conditions modeling stroke. Stroke induces a variety of vascular changes,

including blood flow redistribution and BBB leakage. These events trigger pathophysiological cascades that evolve over time and space, leading to expansion of the brain injury and offering potential targets for therapy (Iadecola and Anrather, 2011). 2-PM has also been applied to visualize neuronal activation in response to functional stimuli and the effects of neuronal stimulation on the cerebral micro-circulation. Activated neurons can change the geometry of microvessels and generate both dilatatory or constrictory events to support new metabolic needs (Lindvere et al., 2013).

Table 1 Frame acquisition rates to image specific dynamic events. The best options are reported for frame acquisition rates for different dynamic events visualized by time-lapse 2-PM, as set in previous papers. Cell type

Visualized event

Frames per minute

Study

Endothelium

Blood flow velocity (RBC speed)

12 000 000 (2.6 kHz) 96 000 (1.6 kHz) 78 000 (1.3 kHz) 60 000 (1 kHz) 49 800 (0.83 kHz) 76 0.5 3 0.00007 (1 per day)

Kim et al. (2012) Schaffer et al. (2006) Nishimura et al. (2006) Chaigneau et al. (2003) Fumagalli et al. (2013) Lindvere et al. (2013) Ferna´ndez-Klett et al. (2010) Cho et al. (2011) Harb et al. (2013)

60

Ding et al. (2009)

600–1200 60–240

Nimmerjahn et al. (2009) Nimmerjahn et al. (2009)

49 800 (0.83 kHz) 20 4 2 2 1 2

Fumagalli et al. (2011) Ortolano et al. (2010) Mues et al. (2013) Fumagalli et al. (2011) Bartholoma¨us et al. (2009) Kim et al. (2009) Bartholoma¨us et al. (2009)

Microglia contacts with presynaptic structures

2 0.5 0.3 1 0.2 0.1

Davalos et al. (2005) Davalos et al. (2012) Masuda et al. (2011) Grinberg et al. (2011) Marker et al. (2010) Wake et al. (2009)

Dendritic spine loss after stroke

0.004

Zhang and Murphy (2007)

Vascular dilatation/constriction Vascular leakage upon BBB opening Vascular plasticity Astrocytes

Ca2+ waves in astrocytes (fluorescence increases by DF/F0) Burst axial extents and 3D volumes

Lymphocytes

Intravascular flowing T-cells T-cells in brain slice Extravascular T-cells

Microglia

Microglia ramification extension/withdrawal

Intravascular crawling T-cells

Microglia cell body displacement

Neurons

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

39

Fig. 2. Method for RBC speed calculation. RBC appear as round-shaped black objects in the blood stream, as they do not take up the fluorescent dye. Linescans at 830–1600 Hz are acquired and piled up to yield an image (A) whose axis correspond to distance (mm) and time (s). RBC, by flowing within the vessel following blood flow direction, form diagonal streaks. The slopes of the streaks are calculated to obtain RBC speed as mm/s (B).

2.1. Blood flow dynamics after acute brain injury: 2-PM quantitative analysis Most 2-PM studies on hemodynamics rely on the possibility to label blood vessels with a fluorescent dye (e.g. rhodamine–dextran, Q-dots) and measure blood flow velocity and direction in individual vessels (Chaigneau et al., 2003). Briefly, to calculate blood flow velocity, this technique assesses the time needed by a red blood cell (RBC) to travel a certain distance in a blood vessel (RBC velocity, Fig. 2). As RBCs do not take up the fluorescent dye used to label blood vessels in vivo, they are detectable as nonfluorescent shadows in the binary images. To acquire sufficient snapshots of linearly moving RBCs, the acquisition rate is enhanced by using line-scan acquisitions (rates ranging from 830 to 1600 Hz), with line-scans placed at the center of the vessel where blood flow velocity is highest. A typical line-scan image is depicted in Fig. 2A, in which black (non-fluorescent) streaks resulting from RBC passage are clearly observable. The quantifiable blood flow velocity assessed by this method ranges from 0 (stalled) to about 30 mm/s. Recent advances in the line-scanning acquisition protocol (Kim et al., 2012) have widened the application to fastflowing big arteries (500 mm/s, Amirbekian et al., 2009; Huo et al., 2008) or to pathological states associated with high flow, such as atherosclerosis and vascular anomalies. Measures of the vascular network architecture, such as diameter, length and branching (Cho et al., 2011; Lindvere et al., 2013) are also feasible, as 2-PM allows three-dimensional imaging in stacks 250–800 mm thick from intact tissues. Assuming that vessel flow is laminar, RBC velocity can be combined with the diameter of the lumen to calculate the average volume flux as an additional parameter associated with blood flow (Shih et al., 2009). These measures have been employed to study vascular rearrangements in rodent models of stroke. In the stroke setting, 2-PM experiments produce a wide range of output data that would not have been obtained relying on single indicators, thus allowing to draw a comprehensive picture of vascular changes (Pinard et al., 2002; Shih et al., 2009; Tasdemiroglu et al., 1992). Middle cerebral artery occlusion (MCAo) has been used to model cerebral ischemia in 2-PM studies, and imaging has been done over a broad infarct area that is also associated with functional deficits (Fumagalli et al., 2013, 2011; Sigler and Murphy, 2010). Besides this model, local vessel occlusions have been generated, using laser illumination of specific vessels (Schaffer, 2010). Since 2-PM is easy to focus on

tissue portions measuring only a few microns, its laser beam can be directed exclusively to specific vessels, causing a laser-induced cortical thrombosis (Nimmagadda et al., 2008). This model of thrombosis has been used to occlude specific vessels and to study the changes in selected vessels downstream or upstream to the occlusion (Helmchen and Kleinfeld, 2008; Nishimura et al., 2006). On account of the limitation of laser penetration into intact tissues (800 mm), these studies focused on the response of the cortical vasculature (Fig. 3), revealing intriguing aspects of vascular dynamics after ischemia. 2.2. Hierarchical structure of the cerebrovascular network: mechanisms for blood flow maintenance in penetrating arterioles The complex and often redundant network of cerebral vessels help confer protection from an ischemic event, given the rapid reorganization of blood supply after an occlusion. Schaffer and colleagues (Schaffer et al., 2006) described how the surface network of communicating arterioles appears to be well protected against single-point occlusions of surface arterioles by virtue of its architecture alone. Blood flow downstream of a targeted occlusion, as in MCAo ischemic models, is not interrupted, but it is

Fig. 3. Structure of rodent cortical vasculature. Two-photon in vivo microscopy is usually applied to the brain cortex, to a maximum depth of 800 mm. The typical vasculature architecture in the volume imaged has big vessels lying parallel to the brain surface (pial vasculature) and smaller arterioles penetrating the parenchyma and branching into small capillaries. Thick are sized 50 mm.

40

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

re-established by reversal in the direction of flow, fed by one of the downstream branches (Schaffer et al., 2006). Occlusion of a single vessel is also associated with a change in blood flow velocity which depends on the topological relationship of the clotted vessel with surrounding vessels: RBC speed is reduced by less than half in the first downstream branch, while it remains close to baseline values in parallel and far-downstream vessels (Schaffer et al., 2006). The rapid re-establishment of the new pattern of flow, about 1 s after clot formation, ensures there is no major change in vessel diameter (Schaffer et al., 2006). Thus cortical surface arterioles show persistent perfusion with stability in vessel diameter, without reaching the deficit in blood flow that could cause irreversible neuropathology (Baron, 2001; Hossmann, 1994; Zhao et al., 1997). Indeed no downstream pathology is observed after surface arteriole occlusion, whereas occlusion of a penetrating arteriole has been associated with downstream neuropathology. Penetrating arterioles can contribute significantly to blood supply over a cortical area up to 350 mm radius and, in addition, collateral flow from neighboring penetrating arterioles is limited (Nishimura et al., 2007). After occlusion of a penetrating arteriole, RBC velocity is dramatically reduced up to ten branches distal to the target vessel, dropping to approximately 0.1–0.3 times baseline values (Nishimura et al., 2007). Similar topological differences in blood flow redistribution have been observed for venules. While the occlusion of surface venules does not cause a permanent drop of RBC, thanks to the redundancy of surface vessels, occlusion of cortical ascending venules reduces blood flow by 20%, an effect that may cause cognitive dysfunction when protracted in time (Cho et al., 2011). In particular, when a cortical ascending venule is occluded and there is no collateral venule, capillaries one to four branches upstream to the target venule show dramatic reduction in RBC velocity, reversal of flow direction and increased diameter (Cho et al., 2011). In the cortex, venules outnumber arterioles and although they share a similar pattern of capillary branching their density is higher. For this reason occlusion of a venule affects RBC velocity up to a distance of four branchings, a far shorter distance than that caused by occlusion of an arteriole (up to ten downstream branchings). Overall, 2-PM studies on brain cortical vasculature clearly show the hierarchical organization of the cerebral blood vessel network, based mainly on vascular geometry. Occlusions at different levels of the hierarchy cause different degrees of ischemic damage, since vascular geometry has a fundamental role in redistributing blood flow. Surface vessels communicate through frequent connections and form a highly redundant network. In case of an occlusion, preocclusion blood flow is rapidly re-established by virtue of the vessel geometry itself, preventing neuropathology. This may also serve to prevent downstream penetrating arterioles from excessive drops in blood flow. The network formed by penetrating vessels is less redundant and, as a consequence of their architecture, penetrating vessels individually feed wide areas. After occlusion, flow reversal and an alternative blood supply are limited in penetrating arterioles, resulting in wide hypoperfused areas and potential tissue damage (Shih et al., 2009). The network of vascular connections must be kept stable, as this – as discussed above – is an important defence against drops in blood supply. The organization of the mature vascular architecture is indeed highly stable, though the adult brain keeps the ability to increase vascular density in response to specific needs (e.g. pathological status) or fluctuations in the energy demand. Harb and colleagues have characterized the long-term dynamics of vascular remodeling by in vivo time-lapse 2-PM in mice at different stages of life (Harb et al., 2013). They described distinct patterns of microvascular remodeling, ranging from the extensive expansion of the microvascular network (microvascular sprouting, endothelial proliferation and vessel elongation) in the early postnatal brain,

to the decline of remodeling and the acquisition of vascular stability in the mature brain. Interestingly, newly formed sprouts in the developing brain received restricted plasma entry and were more susceptible to pruning than connected vessels. This may imply the importance of vessel connections for proper blood distribution and anastomosis formation. Hypoxia (10% oxygen) induced new and lasting vessel formation in young adult mice, while 4–5-month-old mice almost completely lost the ability to remodel the vasculature, thus failing to counterbalance the metabolic mismatch between energy demand and blood supply (Harb et al., 2013). This further supports the idea that the vasculature plays a pivotal role in both initiating and counteracting ischemic damage. 2.3. Blood flow regulation of dendritic spine structural plasticity 2-PM gives enough spatial resolution to allow investigation of structures finer than the cerebral vasculature, such as neuronal dendrites and dendritic spines. In vivo imaging of these structures is limited by the intrinsic ability of photons in the infrared spectrum to penetrate intact tissue and by the slight shifts of the structures of interest caused by heart/breathing rates and vasomotion that may reduce resolution (Sigler and Murphy, 2010). Most 2-PM studies on dendritic plasticity have therefore been conducted close to the pial surface where dendrites originating from neurons located in layers 5, 2 and 3 bend and run almost parallel to pial vessels (Sigler and Murphy, 2010). It should be pointed out, however, that efforts were made to enhance resolution by increasing the penetration of excitation and by limiting motion artifacts with improved surgery (see Section 5.2). The main advantage of using 2-PM to visualize dendritic activity is its ability to image neurons in a longitudinal manner, e.g. before and after a given stimulus. The investigators altered the animals’ sensory experience under 2-PM imaging by whisker trimming or odorous stimuli and recorded the effects on the structural neuronal wiring. Dendritic spines showed a rapid turnover with enduring motor memories (Xu et al., 2009; Yang et al., 2009). Besides monitoring of physiological synaptic responsiveness, 2PM has been applied to measure structural plasticity of dendritic spines in different mouse models of pathological conditions, such as Alzheimer’s disease (Bittner et al., 2012), Rett syndrome (Landi et al., 2011), Huntington’s disease (Murmu et al., 2013), the syndrome caused by methyl-CpG-binding protein-2 duplication (Jiang et al., 2013), or brain ischemia (Sigler and Murphy, 2010). As for the latter, dendritic plasticity has been evaluated in relation to microcirculation during acute stroke. 2-PM provided simultaneous visualization of the cerebral vasculature and dendrites, revealing the effects of a vessel occlusion on the turnover of dendritic spines, so the relationship between blood perfusion and dendritic structure could be defined. Under normal conditions, the neuronal network is hard-wired with minimum variations over time. Severe ischemia induced by the vasoconstrictor endothelin, reducing blood supply to 10% of normoperfused vessels, caused a rapid loss of dendritic spines as early as 10 min from occlusion (Tsai et al., 2009; Zhang and Murphy, 2007). If perfusion was re-established within 50 min, the dendritic and spine structures were largely restored (Zhang et al., 2005). The mean distance of dendritic spines from the closest blood vessel is 13 mm. Ischemia, as expected, induces the appearance of nonperfused areas and, in the absence of flowing vessels, the dendritic structures could be maintained up to 80 mm from the nearest perfused vessel (Zhang and Murphy, 2007). In general, differences in vessel perfusion cause a sharp transition between intact and damaged dendritic structures, with vessels able to support dendritic viability up to distance of 80 mm.

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

In peri-infarct areas, where blood flow supply is maintained over critical values and redundant vascular architecture helps preserve perfusion, dendritic structures showed the highest plasticity, especially in the later stages of stroke pathogenesis. One to two weeks after ischemia, the dendritic structures in peri-infarct areas reached their maximal plasticity, indicated by a 4-fold increase in spinogenesis, still detectable at six weeks (Brown et al., 2009, 2007). These changes selectively involved the peri-infarct area neurons, since spine turnover rates at locations more than 1.5 mm from the lesion were not affected. Hemorrhagic events of different severity are common complications following ischemic/reperfusion injury. Bleedings into the ischemic area are observed frequently in ischemic patients and occur also as unwanted effects of the thrombolytic therapy (Jauch et al., 2013). After brain ischemia two waves of hemorrhagic events occur: early events (<24 h), promoted by leukocyte-derived matrix metalloproteinase (MMP) 2 and MMP9, and delayed hemorrhages (>24 h), promoted by brain-derived MMP2, MMP9, MMP3 and endogenous tissue plasminogen activator (Jickling et al., 2014). Bleedings, in turn, induce the influx of proteases into the brain. One of the main effects of this proteolytic activity is the digestion of the extracellular matrix (ECM) with subsequent loosening of the ECM structures. Recent hypothesis attributed a significant role to ECM in controlling synaptic plasticity (Orlando et al., 2012). During brain development, the juvenile form of ECM is permissive to dendritic plasticity and synaptic wiring. In the adult brain ECM acquires its mature form, becomes more dense, fills peri-synaptic spaces and stabilizes dendritic spines, thus helping maintain brain homeostasis (Frischknecht and Gundelfinger, 2012). Treatments digesting ECM restore synaptic plasticity (Berardi et al., 2004; Dityatev et al., 2010; Pizzorusso et al., 2006). In particular, in a 2-PM in vivo experiment, the digestion by the bacterial enzyme chondroitinase ABC (ChABC) of the chondroitin sulfate proteoglycans, a key component of ECM, increased the motility and the plasticity of cortical spines (De Vivo et al., 2013). ChABC is not present in the brain, but the endogenous proteases that penetrate the brain through the ischemic damaged BBB may have similar effects. Proteases indeed target the chondroitin sulfate proteoglycans, thus causing loosening of ECM structure. Whether this effect is beneficial or toxic in response to ischemia is not known. It is conceivable that, at early phases after ischemia, the action of proteases, disrupting the supportive role of ECM, mediates mainly neuronal death and apoptosis (Lu et al., 2008; Yang et al., 2007). At delayed phases, however, ECM degradation may be needed for initiating the reparative processes favoring neurogenesis and angiogenesis (Baeten and Akassoglou, 2011; Kim and Joh, 2012). 2.4. O2 imaging and blood flow For a complete understanding of brain metabolism, measures of the vascular parameters reported above should be paralleled by quantification of the amounts of oxygen that can be diffused to surrounding tissue (Raichle and Mintun, 2006). Classical PO2 measurements with electrodes (Lecoq et al., 2009; Ndubuizu and LaManna, 2007) are limited by their invasivity and lack of intravascular PO2 sensitivity (Lecoq et al., 2009). 2-PM offered a solution to these limitations. A phosphorescent probe, PtP-C343 (Finikova et al., 2007), was combined with 2-PM to measure PO2 and blood flow simultaneously in cerebral vessels up to a depth of 300 mm in rat brain (Lecoq et al., 2011). This probe proved sensitive enough to PO2, its phosphorescence lifetime decreasing as PO2 increased. This technique allows very-high resolution mapping of oxygen levels in the brain, measuring vascular PO2 between two flowing RBC. The

41

authors reported a local increase of PO2 associated with erythrocytes in microvessels, defined as erythrocyte-associated transients (EATs, Golub and Pittman, 2005), the probe phosphorescence lifetime being significantly shorter in the immediate vicinity of RBC than in the bulk plasma (Lecoq et al., 2011). EATs depend on multiple microvascular features, such as blood flow, density and RBC orientation, as well as on oxygen consumption by the neuropil (Wang and Popel, 1993). In basal conditions, mean capillary PO2 increases with RBC flow velocity, while PO2 between two flowing RBC (inter-RBC) does not change. Inter-RBC is in equilibrium with nearby tissue PO2 so it can be used as an indicator of local tissue PO2 (Parpaleix et al., 2013). This technique can be applied to report tissue PO2 and, as neuronal activity is associated with changes in oxygen demand, it may help in defining local maps of neuronal activity (Parpaleix et al., 2013). Imaging of oxygen microvascular dynamics has been done in living mice after odor stimulation to induce changes in neuronal activity (Lecoq et al., 2011; Parpaleix et al., 2013). Odor triggered a dip in tissue PO2 followed by an increase of vascular PO2 (Lecoq et al., 2011). 3. Vascular regulation of immune cell activity Activation of the immune response is a pathogenic mechanism associated with cerebral ischemia (Iadecola and Anrather, 2011). After the insult different immune populations belonging to the innate and adaptive responses are activated/recruited to the site of injury and promote inflammatory signaling. The effect of individual cellular components of the inflammatory cascade can be either detrimental or beneficial, depending on the stage of tissue injury, the magnitude of the response and whether the inflammatory component also activates neuroprotective pathways (Nawashiro et al., 2000; Zhang et al., 2000). Recruitment/activation of immune cells is time-dependent and involves many cell types. Within the first hours after ischemia, resident microglia are activated, undergo morphological changes and begin to produce inflammatory cytokines and chemokines. After the early activation of this microglial response, immune cell types from the peripheral immune system, such as neutrophils, macrophages, dendritic cells and T-lymphocytes, are recruited to the site of injury (Garcia et al., 1994) following a defined time course (Gelderblom et al., 2009). Macrophages, dendritic cells and T-lymphocytes accumulate soon after microglial activation, reaching their peak concentration in 3– 4 days after injury. The early accumulation of these immune cells favors the subsequent influx of neutrophils through upregulation of cell adhesion molecules including intercellular adhesion molecule-1, vascular cell adhesion molecule-1 and E-selectin, promoting neutrophil accumulation and migration into the brain parenchyma (Gelderblom et al., 2009; Huang et al., 2006). Upon transmigration through the activated cerebral endothelium, neutrophils acquire a neurotoxic phenotype involving the release of toxic proteases associated with de-condensed DNA (Allen et al., 2012). To assess the dynamic aspects of immune cell behavior after ischemic injury, 2-PM has been used in a number of studies, employing vital fluorescent dyes to label lymphocytes, and transgenic mouse models bearing the gene for the green fluorescent protein (GFP) under the promoter of genes selectively expressed in T-cells (hCD2, De Boer et al., 2003; Ortolano et al., 2010) or in microglia (cx3cr1, Davalos et al., 2005). Thus, in recent years, most attention has focused on to inflammatory mechanisms driven by T-cells and microglia/macrophages after ischemia, and selective strategies interfering with their function (De´nes et al., 2008; Gee et al., 2008; Hurn et al., 2007; Liesz et al., 2009; Shichita et al., 2009; Yilmaz et al., 2006). In the context of brain immunity, the cerebral vasculature has a fundamental role in regulating cell trafficking, recruitment and

42

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

activity. While this may be obvious when related to cell infiltration, its role in the fate of T-cells infiltrated into the brain or in instructing resident microglia is still unexplored. 2-PM studies have shown that BBB components can instruct circulating immune cells prior to infiltration and have indicated the mechanisms of vascular alterations capable of activating resident microglia. These results have depicted a new scenario in the vascular–immune cell interaction within the CNS. 3.1. Route of infiltration and imaging T-cells infiltrating in the CNS (meningitis and stroke) As reported above, immune system cells are recruited to the injured brain in a time-dependent manner (Gelderblom et al., 2009). T-cells have attracted interest as they have pivotal roles in the progression of many pathological brain conditions, thus offering a potential therapeutic target. Their access to the brain is through the vasculature which is normally strictly closed to immune cell passage by the BBB, but which may permit cell infiltration after an injury. T-cells move along the blood vessels’ inner surface searching for a suitable place to enter into the tissue. Once the right spot is found, extravasation takes 10–20 min (Bartholoma¨us et al., 2009), and involves integrins and G-coupled proteins expressed on endothelial cells that allow cellular migration (Bauer et al., 2009; Engelhardt, 2006).

After brain ischemia, BBB opening is a very early event occurring within minutes after vessel occlusion (Fig. 4). This event may further favor T-cell recruitment to lesion site. In a clinically relevant model of brain ischemia (Zanier et al., 2013) we analyzed the opening of the BBB by 2-PM and quantified it on the basis of the extravasated fluorescent signal associated with fluorescent dextran, injected before to ischemia to label blood vessels. We focused on the brain microvasculature, and individually analyzed small vessels, exploiting the high resolution achievable with 2-PM. The vessel edges became markedly irregular during ischemia indicating that the microvascular structure was rapidly affected. In shamoperated mice vascular borders kept their regular even shape during the whole imaging session and this was regarded as an indicator of a healthy state even after exposure to the laser beam (Fig. 4). The rapid onset of extravasation was also described by MRI, an in vivo technique providing information over wider tissue portions (Strbian et al., 2008). BBB opening is a first step for T-cell infiltration which may be passively recruited to site of injury at early phases after ischemia. MRI showed that after an initial early opening (25 min after ischemic onset, Strbian et al., 2008), the BBB subsequently opens in a time dependent manner, showing biphasic maximal leakage 4–8 h and 12–16 h after ischemic onset (Klohs et al., 2009). The T-cell infiltration peak was reported three days after ischemia (Gelderblom et al., 2009) and continued over a

Fig. 4. Extravasation after experimental brain ischemia. Extravasation became evident as early as 20 min after transient occlusion of the middle cerebral artery (tMCAo). At baseline (A) the Texas-red fluorescent dextran (70 kDa) signal was confined to the vessels, with minimum background noise widespread in the parenchyma. During tMCAo (B) the fluorescent signal was also clearly visible outside the vessels indicating extravasation. A differential image of the red insert is obtained by subtracting image A from image B. The resulting image (C) shows the parenchymal accumulation of dextran. In sham-operated mice (D, E and F) no extravasated fluorescent signal is detectable, and the apparent signal in the differential image in (F) is due to slight misalignment of the vessels. Bars = 100 mm. (G) Integrated density ratio of extra- to intra-vascular dextran. In ischemic mice extravasation is significantly greater as early as 20 min after occlusion onset. Data are expressed as mean + SD, n = 24 vessels for ischemic mice and n = 22 vessels for sham-operated animals. Two-way ANOVA (p < 0.0001) followed by post hoc Bonferroni test: ***p < 0.001 vs. sham. (H) At baseline (pre), vessel borders appear straight and regular, while during tMCAo and 1 h after ischemic onset, they became markedly irregular possibly due to the incipient BBB damage (white arrows). In a shamoperated mouse, there was no change in the vessel morphology at any time (H). Scale bar = 25 mm.

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

period not necessarily associated with BBB opening peaks. Conceivably therefore, the BBB, once opened, not only allows a passive transfer of immune cells from the bloodstream to the brain, but participates in their recruitment actively. Owens and colleagues (Owens et al., 2008) elegantly described how extravasated T-cells are likely to move in the perivascular space rather than in the brain parenchyma, and possibly come into contact with the different cell types making up the BBB. Direct demonstration of this has only recently become available as in different models of brain diseases 2-PM observations clearly showed T-cells patrolling the perivascular space of CNS blood vessels prior to tissue invasion. In a model of viral meningitis, naive GFP+ CD8+T-cells expressing a receptor able to recognize the lymphocytic-choriomeningitis virus (LCMV) were visualized in the brain after craniotomy. The number of T-cells increased significantly in sixday symptomatic mice. T-cells moved in the perivascular areas with preserved blood vessel integrity and their motility was confined to the meningeal space. There, they were able to recruit myelomonocytic cells which in turn caused vascular injury with subsequent BBB leakage, associated with the pathology (Kim et al., 2009). Although stroke causes BBB leakage, with an increased expression of adhesion molecules, on the whole facilitating T-cell infiltration, there are far fewer T-cells in the brain after stroke than in viral or autoimmune diseases (Flu¨gel et al., 2007). For this reason, time-lapse in vivo imaging of T-cell dynamics in a model of stroke poses a technical challenge. In adult male CBA/CaXC57F7BL/ 10 hCD2-GFP transgenic mice that express GFP specifically in Tcells under the control of the hCD2 promoter, we previously studied T-cell dynamics in a model of permanent focal ischemia (Fumagalli et al., 2011). The transgenic model has the great advantage that it does not require the injection of fluorescent Tcells from a donor mouse, thus leaving the physiological concentration of circulating T-cells unaltered. Furthermore, the transgenic model allows visualization of the whole T-cell population involved in the ischemic response, not just the labeled fraction as for the transfer technique. However it is not possible to track a specific population of T-cells in vivo, as in transgenic mice GFP expression is not restricted to a specific population of T-cells so one cannot distinguish the T-cell subsets. Three days after ischemia, similarly to what was reported in meningitis, many T-cells crawl along the abluminal space of blood vessels where they interact physically with BBB components such as perivascular macrophages, astrocytes and endothelial cells. Two main T-cell motility classes exist concomitantly in an ischemic area: one stationary and one moving fast and over long distances, often along the abluminal space of vessels. This might reflect the evolution of the T-cell response from a naive population to effector/memory cells and it could be related to the behavior of Tcells which, after encountering an antigen, have an altered profile of adhesion molecules and chemokine receptors. Real-time imaging in the lymph node showed that, in contrast to CD4 Tcells, whose migration is slower (Rush et al., 2009), memory CD8 Tcells move faster than naive cells (Chtanova et al., 2009), suggesting that faster migration could be a defining characteristic of CD8 as opposed to CD4 memory T-cells. It is not clear whether the T-cell driven immune response after ischemia needs complete priming of the T-cells, since we still have no direct evidence of autoreactive T-cells attacking CNS (Iadecola and Anrather, 2011). At least very early after ischemia, when BBB damage occurs, it is likely that T-cells enter the brain passively, without priming. The mechanisms through which T-cells participate in ischemic damage are largely unknown and might be either deleterious or beneficial. In acute phases, unprimed T-cells contribute to damage in the context of innate, antigen-independent immunity, by secreting pro-inflammatory cytokines such as IFNg, or reactive

43

oxygen species (Arumugam et al., 2005; Brait et al., 2010; Leys et al., 2005). In later phases, autoreactive CD4+ and CD8+ T-cells develop, with detrimental effects as well (Leys et al., 2005), but concomitantly regulatory T-cells develop with the production of protective factors such as IL-10 and TGFb (Gee et al., 2009). IL-4 and low doses of IFNg that are produced locally, and whose main source are T-helper cells, push microglia activation toward a protective phenotype (Butovsky et al., 2007) and can secrete neurotrophic factors (Hallenbeck et al., 2006). Since T-cells are the main suppliers of both these cytokines in the ischemic brain, this event has been defined as protective autoimmunity. The multiple functions of T-cells yield a puzzling picture. It seems that their early detrimental effects are not related to adaptive immunity whilst their participation in protective functions may rely on T-cell priming and the development of regulatory T-cells. These latter cells have in fact been suggested as having a role in maintaining self-tolerance and in downregulating various immune responses (Gelderblom et al., 2009). Since T-cells are present in the ischemic brain 24–48 h after insult (Kleinschnitz et al., 2010), and peak at three days (Gelderblom et al., 2009), they may anticipate the establishment of an adaptive response that usually requires 7–10 days from antigen presentation to clonal expansion of autoreactive T-cells and target organ attack. T-cells may escape their classical activation steps, and be activated by other cell types that specifically reside in the brain. The BBB plays a vital role in adaptive immunity. As T-cells enter the brain, part of them spend time crawling in the perivascular space (Fumagalli et al., 2011) where the immunologic synapse could be formed with astrocytic end feet, perivascular phagocytes, or even microglia, all of which have the capacity to act as antigen presenting cells (Kivisa¨kk et al., 2009) to prime T-cells. The primed T-cells can have different downstream effects, ranging from support of the inflammatory response to the education of microglia to protective functions. 3.2. Resident immune cells (microglia) and imaging One of the possible target effector cells of primed T-cells may be resident microglia (Butovsky et al., 2007). Microglia rapidly sense injury with their uniform presence throughout the cerebral tissue and their continuous sampling of the surrounding microenvironment, observable also in physiological conditions (Davalos et al., 2005; Hanisch and Kettenmann, 2007). Microglia can play either a detrimental or a beneficial role as they can acquire a variety of phenotypes covering a wide range of functional polarization. After ischemia, activated microglia may acquire phenotypes belonging to the ‘classical’ (M1) or the ‘alternative’ (M2) active status (David and Kroner, 2011; Michelucci et al., 2009; Perego et al., 2011). M1 activation is generally referred to as a pro-inflammatory and cytotoxic phenotype, with production of nitric oxide, ROS and proinflammatory cytokines, while the M2 phenotype is an alternative activation state, associated with resolution of inflammation, scavenging of debris, promotion of angiogenesis, tissue remodeling and repair. Specific environmental signals, including effector molecules, timing of activation and degree of injury can induce these different polarization states (Porta et al., 2009; Zanier et al., 2011). As a direct consequence of their commitment, microglia can worsen the final outcome or promote tissue recovery. Specific functional states of microglia are associated with changes in cell morphology or cell displacement (Soltys et al., 2005), so specific shapes or motility rates are strongly indicative of microglial activity. Analysis of cell dynamics over time can provide information on microglial activity in an inflamed territory. In the last few years, microglial dynamics has been widely studied by in vivo 2-PM using a transgenic mouse model expressing GFP under the control of the fractalkine receptor (cx3cr1), which is

44

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

constitutively expressed by resident microglia (Davalos et al., 2005; Fumagalli et al., 2013; Liang et al., 2009; Masuda et al., 2011; Nimmerjahn et al., 2005; Ohsawa and Kohsaka, 2011). These studies indicated the high dynamic nature of microglia within the CNS. In physiological conditions, microglia extend and retrieve their processes to sample the surrounding microenvironment (Davalos et al., 2005) and make direct contact with other cell types in the CNS, such as neurons (Wake et al., 2009) and perivascular astrocytes (Mathiisen et al., 2010). When an injury occurs, microglia activate, become hypertrophic and retract their processes, acquiring an ameboid morphology (Fumagalli et al., 2013). This is triggered by sensing ATP (Davalos et al., 2005) through the adenosine receptor A2A whose expression is increased in microglial cells soon after injury. In contrast, the expression of P2Y12R decreases (Haynes et al., 2006; Orr et al., 2009). This pathway is also responsible for the microglia displacement induced by LPS stimulation (Gyoneva et al., 2014; Ohsawa and Kohsaka, 2011). The ability of microglia to travel over brain tissue seems to depend on the kind of danger signal received. In a model of spreading depression in organotypic slice cultures that causes loss of synaptic activity, microglia showed displacement by Le´vy flightlike movements in response to diminished synaptic activity (Grinberg et al., 2011). In contrast, in models such as traumatic brain injury (induced by laser ablation, Davalos et al., 2005), the photothrombotic stroke model and global or focal ischemia (Davalos et al., 2005; Fumagalli et al., 2013; Masuda et al., 2011), microglia did not show any body movement in response to damage. The lack of microglial displacement after ischemia might be related to the rapid decline of ATP levels due to vessel occlusion, ATP being needed for the cytoskeletal modifications exploited by microglia to migrate within tissue (Davalos et al., 2005; Grinberg et al., 2011). The ability of microglia to rapidly modify their shape in response to an ischemic insult may need less ATP, but is dramatically affected by the residual capillary blood flow around their soma (Masuda et al., 2011). After complete loss of blood flow in their territory, microglia become stalled, unable to withdraw ramifications. Acquisition of the ameboid morphology requires minimum blood flow and ATP supply from surrounding capillaries (Masuda et al., 2011). The resilience of microglia is a key feature of this population, as they recover process activity once blood flow is re-established, even in areas with severe ischemia (Masuda et al., 2011). Our group has reported that the ameboid evolution of microglial morphology in response to brain ischemia is impaired in CX3CR1-deficent mice (Fumagalli et al., 2013). In the absence of CX3CR1, microglia do not withdraw their processes or acquire their typical activated hypertrophic morphology. This effect is paralleled by a significant change in the inflammatory environment where the expression of CD11b, the lysosomal marker CD68 (Ramprasad et al., 1996) and the M1 polarization marker iNOS is reduced, while that of the M2 polarization marker Ym1 increases at the lesion site; all these events contribute to protection from ischemic injury at 24 h (Fumagalli et al., 2013). Besides the changes that mirror the microglial functional response to a pathological treat, the changes in morphology and motility of microglia also happen in an age-related manner. As for morphology, microglial cell soma volume increases while mean ramification length decreases with age. Dynamic changes in aged mice (27–28 month old) include quicker soma movements and decreased baseline process motility compared to young mice (3 month old, Hefendehl et al., 2013). Being these observations done in an unperturbed environment, it should be considered that the surveying microglia, which contribute to the maintenance of tissue homeostasis, progressively narrow the area sampled by individual cells in aged mice. To partially compensate this in aged mice, microglial cells increase their number by 14% in cortex, and are

more likely to form clusters of three or more cells within small brain regions (extending within a 40 mm radius). Although the morphological features of microglia in aging brains are reminiscent of the active state, it is not clear whether the aged microglia are truly activated. It is likely that the aged microglia enter in a dysfunctional state, which may have remarkable effects on the integrity of other structures microglia come into contact with such as neuronal or vascular networks. Possible consequences of the dysfunctional behavior of aged microglia regard a dysregulated response to injuries, changes in neuroprotective functions and an increase in the toxic responses (Hefendehl et al., 2013). Cerebral vessels can activate microglia because of bleeding. Cortical microhemorrhages in aging brains, while not causing acute stroke symptoms (Cullen et al., 2005; Farrall and Wardlaw, 2009), do activate microglial cells up to 200 mm from the origin of the bleeding (Rosidi et al., 2011). The area of activation after microhemorrhages (400 mm diameter) correlates with the amount of extravasated blood plasma, suggesting that plasma components may trigger microglial activity (Rosidi et al., 2011). Since BBB leakage occurs early after ischemia (Strbian et al., 2008), plasma components may be among the first activating signals that elicit a microglial response. A mechanism of microglial activation by plasma components was demonstrated in a model of EAE by in vivo 2-PM (Davalos et al., 2012). The plasma protein fibrinogen leaks out during the development of EAE-induced BBB damage, and precedes the neurological signs. Extravasated fibrinogen binds to its receptor CD11b/CD18, expressed on the microglial surface, and promotes microglial mobilization similarly to the ATP/P2Y12R pathway (Davalos et al., 2012). Soluble fibrinogen is cleaved by thrombin, a serine proteases that leaks out during the opening of the BBB, to form insoluble fibrin deposits at site of hemorrhages. Fibrinogen induces clustering of microglia at sites of fibrin deposition, mainly close to the leaked vasculature. Clustered microglia release ROS and promote demyelination, thus contributing to early axonal damage (Davalos et al., 2012). Proteases play a prominent role in the opening of the BBB and in turn, as a consequence of bleeding, accumulate in the CNS exerting neurotoxic effects (see Section 2.3). A direct toxic effect on synaptic plasticity was described for thrombin that, upon its accumulation, saturates the NMDA-dependent neuronal potentiation, thus preventing further synaptic responsiveness (Maggio et al., 2013). Other proteases that accumulate in the CNS during pathogenetic events include MMPs. The influx of MMPs in the diseased CNS is a consequence of inflammatory processes and involves microglial activation. Active microglia and endothelial cells express the extracellular MMP inducer (CD147) which is a key molecule in the process of leukocyte infiltration (Agrawal et al., 2013). During the development of EAE, the sensing of CD147 induces leukocytes to secrete MMP2 and MMP9 that, in turn, selectively cleave b-dystroglycan at the parenchymal basement membrane to allow cell transmigration across capillaries (Agrawal et al., 2006). This event is accompanied by remodeling of ECM, whose effects on brain homeostasis are discussed in Section 2.3. 3.3. Relevant indices for immune cell active/functional state As documented above, 2-PM introduced different measures of the dynamic behavior of immune cells within the CNS. However, differently from the molecular markers, a given parameter of cellular dynamics may not be linked to a specific cellular active/ functional state in a straightforward manner. The increasing literature exploring the morphology and the motility of immune cells in physiological or diseased conditions may help define specific dynamic indices of a given active/ functional state, in the view of understanding the immune cell

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

function through the measure of its movements. This may be possible by integrating the 2-PM data with conventional molecular analysis – immunohistology, real time PCR, western blots – in different pathophysiological conditions. We now discuss a number of dynamic parameters that are associated to a function of the immune cells within the CNS, whose pertinence depends mainly upon cell type, e.g. infiltrating immune cells or resident microglia. Among the infiltrating immune cell types, T-cells have been the most studied by time-lapse 2-PM. T-cells do not undergo any sharp morphological changes, thus the shape descriptors do not apply well to the definition of T-cell activation which is better represented by cell motility. We document here the parameters of cellular dynamics that regard T-cells, although some of these measurements may be successfully applied to other cell types too. T-cells need to penetrate into an inflamed tissue to exert their function. Once found the right spot, they infiltrate, target an antigen presenting cell (APC), form the immunological synapsis and become activated toward a specific antigen. From a dynamic point of view this paradigm is realized by intra-vascular flowing, perivascular crawling, parenchymal migration, arrest in proximity of an APC and reprise of migration. For this reason one can describe T-cell functionality by assessing their velocity, trajectory, meandering behavior, location relative to the vascular structures, coefficient of arrest and displacement (Cordiglieri et al., 2010; Fumagalli et al., 2011; Gibson et al., 2012; Kawakami and Flu¨gel, 2010; Odoardi et al., 2007; Zenaro et al., 2013). Unprimed T-cells flow with the blood and show velocities in the order of few millimeters per second realizing linear trajectories. Upon an inflammatory challenge T-cells start crawling along the perivascular space, reduce their speed (to about ten micrometers per minute) and move along tracks characterized by linear stretches. Once infiltrated, T-cells keep migrating at ten micrometers per minute, but they significantly increase their meandering behavior (Bartholoma¨us et al., 2009; Fumagalli et al., 2011). The motility of T-cells is further affected by their ability to recognize a specific antigen. When immunized for ovalbumin, T-cells slowed down after administration of ovalbumin (from an average of 8.0 mm min 1 to 2.8 mm min 1), and many of them stopped when in close contact with meningeal and perivascular phagocytes, labeled with tetramethylrhodamine–dextran (Odoardi et al., 2007). In viral meningitis model, T-cell velocity was affected by their interaction with LCMV specific antigen. When this interaction was prevented, T-cells moved at an average speed of 5.2 mm min 1, far faster than their usual average 3.4 mm min 1 in the context of antigen presentation (Kim et al., 2009). Similarly, the fact that CD4+ T-cells can slow down upon specific antigen encounter has been described (Miller et al., 2002; Zinselmeyer et al., 2005). When primed and tolerized, CD4+ T-cells reduced their speed from 10 mm min 1 to 4–5 mm min 1. 2-PM thus demonstrated clearly that T-cells slow down when the immune synapse is formed and they are primed/tolerized. This biological event can be visualized by 2-PM exploiting the association between immune synapse and the variation of intracellular Ca2+ concentration. T-cells use Ca2+ as second messenger for signal transduction cascades. Intracellular Ca2+ peaks in T-cells are related, though not exclusively, to antigen recognition. Upon recognition of exogenous antigen in peripheral lymph nodes, migratory T-cells come to arrest, immediately raising intracellular Ca2+ (Mempel et al., 2004; Shakhar et al., 2005). Mues and colleagues have developed a genetically encoded Ca2+ indicator, based on a FRET event, to follow in vivo transient T-cell Ca2+ peaks; in the CNS, T-cells have intracellular Ca2+ transients during contact with perivascular phagocytes and intracellular Ca2+ levels remain high during parenchymal

45

migration (Mues et al., 2013). This may be due to the presentation of small amounts of local myelin autoantigen by perivascular phagocytes, sufficient to activate and direct T-cells toward their final destination, but too weak to induce extended periods of T-cell arrest. Although long, stable contacts with APC are needed to drive Tcell activation in the peripheral lymph nodes (Cahalan and Parker, 2008; Mempel et al., 2004), in the CNS the first T-cell activating signal depends on short transient contacts with perivascular phagocytes. This has been shown with in vivo 2-PM in combination with a fluorescent sensor for the nuclear factor of activated T-cells (NFAT, Lodygin et al., 2013). NFAT nuclear translocation, revealed by colocalization of the NFAT sensor with histone protein H2B sensor, takes place shortly after T-cell receptor activation and can be used as an indicator of this activation (Lodygin et al., 2013). Most of the previously cited dynamic parameters were described for the first time in EAE and then translated to other brain diseases such as cerebral ischemia. In this context, data from EAE studies served as paradigm, relying on the general assumption that T-cell behavior in various CNS pathologies complies with similar mechanisms. However this assumption still needs confirmation. Due to maximum 2-PM penetration depths, T-cells in EAE are studied within the white matter of the spinal cord (Bartholoma¨us et al., 2009; Odoardi et al., 2007), while in meningitis or stroke models T-cells are studied in the gray matter (Fumagalli et al., 2011; Kim et al., 2009). It should be pointed out that important differences exist between gray and white matter, which may possibly be even larger than those between different disease models. These differences include vascular architecture and myelin composition, both possibly deeply influencing T-cell behavior. Neutrophils have been less studied in the CNS by in vivo timelapse 2-PM, thus few parameters of functional or active state are available. Neutrophils share with T-cells some common features such as the ability to travel over distances in the tissue they invade and the perivascular location after ischemia (Enzmann et al., 2013), suggesting that some of the parameters used for T-cell movements may also apply to neutrophils. A recent work described that neutrophils locomote in a tidy fashion toward an inflamed site showing an ameboid-like migration mediated by MMP and using the dense collagen network of the inflamed tissue as a guidance (Lerchenberger et al., 2013). Inhibition of MMP significantly impaired the migration of neutrophils by decreasing their directionality, velocity and distance traveled, suggesting these as important features of the full neutrophil functionality. Furthermore, the directional migration required a morphological polarization that can be measured by the shape descriptor eccentricity (the cell long axis divided by the cell short axis), which decreased by application of the MMP inhibitor (Lerchenberger et al., 2013). Although this study applies 2-PM to a peripheral inflamed mouse organ – the cremaster muscle – it is conceivable that these motility descriptors would be useful to track neutrophil functional state in the CNS. Shape descriptors are particularly well applicable to the analysis of microglial activation. Microglia do change their morphology when they activate, express novel surface antigens and release cytokines and chemokines in response to a treat in the CNS (Hanisch and Kettenmann, 2007; Perego et al., 2011). The ameboid appearance and the thickening of ramifications and cell soma are key morphological features accounting for the activation of microglia due to a damage (Cunningham et al., 2013; Fumagalli et al., 2013; Kozlowski and Weimer, 2012) or to aging (Hefendehl et al., 2013). The ameboid morphology has been associated to the innate activation of microglia, characterized by tissue clearance through phagocytosis, while the ramified shape seems to belong to

46

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

a more protective behavior of microglia (Fumagalli et al., 2013; Vinet et al., 2012). In the very early phases after an insult, a morphological polarization of microglia preceding the ameboid shape has been described. This change is caused by the extension of ramifications toward the lesion site so as to properly sense the challenge (Davalos et al., 2005). Microglia displace over very limited distances, however significant differences in displacement because of aging (Hefendehl et al., 2013) or certain diseased conditions (Davalos et al., 2012; Grinberg et al., 2011) have been documented. In particular microglia do clusterize in damaged areas where significant amounts of fibrinogen accumulate (Davalos et al., 2012). The continuous extension and withdrawal of ramifications is a basal dynamic behavior exploited by microglia to explore the surrounding micro-environment (Davalos et al., 2005). After injury, microglia modify their branches either by extension toward a site of injury or withdrawal to become ameboid. This is largely due to the specific timing of activation and the kind of signal received, e.g. ATP and ADP for extension, CX3CL1 for withdrawal (Fumagalli et al., 2013; Ohsawa and Kohsaka, 2011). While the number and the length of ramifications account for the striking morphological differences caused by microglial activation, the rate of extension/withdrawal per ramification does not reveal an active state. There is evidence that the average speed of extension/ retraction (micrometers per minute) declines during a vascular occlusion when ATP is not available, but it soon recovers to the baseline values after new perfusion is allowed (Masuda et al., 2011). This implies that in the activated microglia after ischemia the intrinsic process dynamics remains unaltered. Similar results have been obtained after the administration of LPS, as the motility of processes did not change either at two or twenty-eight days after LPS (Kondo et al., 2011). It should be hypothesized that microglia do quickly rearrange the length and number of branches without loosing the exploring behavior. 4. Regulation of blood flow by cells of the neurovascular unit Changes in blood supply, BBB damage or microhemorrhages are all events capable of regulating the activity of both resident and recruited CNS cells. Neurons, astrocytes, pericytes and, to some extent, lymphocytes too can regulate blood flow, completing a dynamic picture of cellular–vascular interplay. Local increases in blood flow are needed to sustain neuronal activity, this being defined as neurovascular coupling. Local dilatation of capillaries takes place in response to whisker stimulation, giving the metabolic support needed for neuronal activity (Stefanovic et al., 2008). The vascular tone also changes in pathological conditions such as ischemia or Alzheimer’s disease (Zlokovic, 2005). The local blood flow supply is controlled by glial and neuronal cells proximal to the vasculature and depends on different mechanisms for vascular autoregulation, a homeostatic mechanism needed to ensure constant blood flow in the brain. 4.1. Imaging of astrocytes enveloping the endothelium Local constriction or dilatation of blood vessels depends on the activity of the smooth muscle cells associated with the endothelium. Smooth muscle cells are simply effector cells that regulate blood vessel diameter in response to signals received from astrocytic endfeet. Astrocytes are ideally placed to sense neuronal stimulation and transfer the signal to smooth muscle cells, since astrocytes surround synapses and envelop blood vessels with their endfeet (Attwell et al., 2010). When astrocytes are activated on sensing neuron-released glutamate, they release metabolites of arachidonic acid (prostaglandins and epoxy-eicosatrienoic acids) causing relaxation of smooth muscle cells (Gordon et al., 2008;

Metea and Newman, 2006; Peng et al., 2004, 2002; Zonta et al., 2003). Interestingly, in contrast with these findings, a 2-PM study on brain slices from rats and mice provided evidence of vasoconstriction by astrocytes (Mulligan and MacVicar, 2004). The authors investigated the effects of selective waves of increased Ca2+ concentrations on the astrocytic endfeet that were stimulated by localized flash photolysis of caged Ca2+ with 2-PM (Brown et al., 1999; Soeller and Cannell, 1999). In response to the increased [Ca2+] with propagation to the endfeet, astrocytes were able to cause vasoconstriction. The peak [Ca2+] increase preceded vasoconstriction by 2.7  0.5 s, enough time for the synthesis and release of second messengers diffusing from astrocytic endfeet to smooth muscle cells in the vascular walls. The authors identified the Ca2+-sensitive photolypase A2 as responsible for the vasoconstriction (Mulligan and MacVicar, 2004). Studies on brain slices gave contrasting results on vascular control by astrocytes. Zonta and colleagues proposed a role in vasorelaxation (Zonta et al., 2003) and Mulligan and MacVicar a role in vasoconstriction (Mulligan and MacVicar, 2004). These studies are limited by the lack of blood perfusion in organotypic slice cultures, so there may be differences from intact animals. In vivo 2-PM imaging of astrocytes typically relies on the use of Ca2+sensitive indicator dyes, combined with either labeling of astrocytes by the fluorescent dye sulforhodamine 101 (Nimmerjahn et al., 2004) or transgenic mice with reporter fluorescent genes. Takano and colleagues, using in vivo 2-PM in transgenic mice expressing GFP under the control of GFAP, found that photolysis of caged Ca2+ in astrocytic endfeet caused vasodilatation (Takano et al., 2007). They also showed that, in the early stages of Alzheimer’s disease astrocytes have abnormal activity, with spontaneous increases in Ca2+ signaling causing repetitive cycles of vascular relaxation/constriction. This astrocytic dysfunction and the subsequent abnormal microcirculation precede amyloid deposition and neuronal loss. Nimmerjahn and colleagues demonstrated that astrocytes, which display a non-uniform brain distribution, showed specific dynamics depending on brain areas and layers they populate. In particular Bergmann glia had a different Ca2+ excitability compared to protoplasmatic astrocytes (Nimmerjahn et al., 2009). Moreover, Bergmann glia exhibited three forms of Ca2+ excitation, namely sparkles, bursts and flares, all involving the networking of many fine fibers emerging from a few tens up to hundreds of cells. Interestingly, while sparkles and bursts were observed in awake animals at rest, flares were only elicited during active locomotion, with similar raise times as blood perfusion level increases. This might indicate that flares are responsible of astrocyte-induced local functional hyperemia in response to increased neuronal activity (Nimmerjahn et al., 2009). Brain ischemia can also cause changes in [Ca2+] in astrocytes. In a model of photothrombosis, repetitive and transient Ca2+ signals were detected starting 20 min after clot formation (Ding et al., 2009). Photothrombosis increased the amplitude and frequency of Ca2+ transients for up to 3 h after occlusion onset. Antagonists for mGluR5 and GABABR attenuated astrocytic Ca2+ signaling, suggesting a role of the two receptors in the elevation of Ca2+ stores (Ding et al., 2009). In vivo 2-PM studies have given fundamental information on the vasodilatatory function of astrocytes, and have suggested a functional role in the early development of neurodegenerative conditions. 4.2. Imaging of pericytes’ actions on brain capillaries Studies of the control of vascular tone by astrocytes, mostly by post mortem analysis or in vitro (Attwell et al., 2010), have suggested for years that the neurovascular coupling was mediated

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

exclusively by smooth muscle cell activation. Recently, a similar ability has been proposed for pericytes, whose presence and distribution along capillaries suggested their ability to regulate blood flow at that level (Hamilton et al., 2010; Puro, 2007; Shepro and Morel, 1993). A few studies found that cultured pericytes contracted in response to arachidonic acid and neurotransmitters, suggesting a role in blood flow regulation (Peppiatt et al., 2006; Puro, 2007; Shepro and Morel, 1993). Direct evidence of pericyte control of blood flow was provided by a 2-PM study using GFP fluorescent mice (Ferna´ndez-Klett et al., 2010). Under conditions of tissue hypoperfusion caused by topical administration of an agonist of TBXA2, a mediator of vasoconstriction, local capillary constrictions were evident and more pronounced at pericyte bodies (Ferna´ndez-Klett et al., 2010). However, pericytes were not involved in the increase in capillary diameter after bicucullin, a GABAA receptor antagonist, which induces recurring bursts of neuronal spike activity. Thus, in neurovascular coupling, capillaries dilate passively in response to increased pressure in upstream arterioles, and this does not require pericyte functionality (Ferna´ndez-Klett et al., 2010). Pericyte contractile features are needed for vasoconstriction in response to an altered perfusion state such as ischemia. During ischemia, pericytes constrict capillaries smaller than 5 mm in diameter, and the effects may be enhanced by blood cell trafficking. Red and white cells have to deform considerably to flow through a 5 mm-diameter vessel (Attwell et al., 2010) and, since ischemia significantly increases the number of circulating blood cells, collateral capillary clots may form (no-reflow phenomenon, Del Zoppo and Mabuchi, 2003; Del Zoppo, 1997) giving rise to new small non-perfused areas. 5. Boosting 2-PM forward: (nearly) available technological improvements The number of publications that exploit 2-PM has increased significantly over the last years, as many labs have been supplied with this technology. A huge body of scientific data has been now obtained by the application of 2-PM to different biological systems, ending up with the availability of a wide range of protocols for 2PM imaging. New protocols include tasks for preparing animals to in vivo imaging (Hefendehl et al., 2012; Pai et al., 2012), use of fluorescent dye indicators (Albertazzi et al., 2013; Hirase et al., 2004; Lecoq et al., 2011), scan methods for specific biological events (Marker et al., 2010) and quantitative analysis of acquired images (Jungblut et al., 2012; Kozlowski and Weimer, 2012; Tomek et al., 2013). The selection of the best imaging protocol should consider: (1) the spatial location of the event, e.g. deep in parenchyma or at pial surface; (2) the features of the object to be visualized and quantified, e.g. motion and size; (3) the expected outcome, e.g. change in cellular motility, morphology or ion transcellular trafficking and (4) whether acute or chronic imaging is required. Good surgical practice by skillful operators is required to prepare animals for in vivo imaging. Clean craniotomy preparations will allow to obtain high quality images because of preservation of the underneath brain structures and limitation of signal aberrations. The appropriate surgical technique, e.g. open window or skull thinning, should be chosen according to the imaging purposes. As discussed later on (Section 5.2) openwindow guarantees higher depths of excitation and best fluorescent signal collection compared to skull thinning, the latter being less invasive and weakly associated with inflammation (Isshiki and Okabe, 2013; Xu et al., 2007; Yang et al., 2010). Thus open-window is usually applied to acute imaging of brain vasculature (Schaffer, 2010), while skull thinning is used for studying dendritic spines, immune cells or in chronic conditions (Marker et al., 2010).

47

Notably, after the first protocol for 2-PM imaging in awake animals was introduced (Dombeck et al., 2007), a recent broad interest for functional imaging raised. New protocols have been published to allow the application of in vivo 2-PM to behaving animals (Andermann et al., 2013; Chen et al., 2013; Kaifosh et al., 2013; Scott et al., 2013), in view of collecting data on synaptic functional plasticity. Functional 2-PM imaging typically provides the measure of Ca2+ transients in vivo without the confounding factors associated with the anesthetic regimen. The parameters for image acquisition should be set to fit the best dimensional and temporal features of the object of interest. Digital sampling should adhere to the general rule that the digital unit (voxel) must be sized below the half size of the smallest object of interest, e.g. a single cell or a dendritic spine. This is associated with the microscope intrinsic limits of resolution that will be discussed in Section 5.2. As for temporal sampling, the trajectory of a moving object should be represented by time points frequent enough not to leave wide gaps between two consecutive spatial positions. In this regard Table 1 offers a list of previously published protocols that may help setting up the system for appropriate acquisition frame rate. The appropriate choice of the parameters for image acquisition ensures that the information contained in the image has sufficient quantity and quality to be correctly quantified. Quantitative analysis of digital images is the art of transforming a visual sensation into its numerical form. This enables to describe, classify and mathematically interpret spatial and temporal components of the image itself. The numerical descriptors that are typically applied to 2-PM image quantification include fluorescence intensity and indices of morphology (volume, surface, sphericity, cell complexity by Sholl analysis) and motility (track velocity, displacement, path length, meandering, clusterization rate, ramification elongation/withdrawal). Computational techniques play a central role in the process of extracting meaningful quantitative data from images. In recent years, joint efforts between biologists and informaticians have been dedicated to develop new tools for reliable image quantification. The science of bioimage-informatics has rapidly become crucial for the evolution of microscopy and for the full exploitation of sophisticated optical methods, such as 2-PM. The issue of bioimage-informatics, which has been detailed in a recent publication (The quest for quantitative microscopy, 2012), goes beyond the purposes of this review. Besides the definition of the best imaging options to properly exploit the potential of up-to-dated microscopes, a big effort has been also put into the development of more efficient hardware set-ups, with particular focus on the advancements that allow to increase the speed of acquisition or the quality of the final image. 5.1. Speeding up the system A recent article by Ducros and colleagues settled a new limit for 2-PM temporal resolution, enabling real-time imaging of events that used to range out from the previous scan speed boundaries (Ducros et al., 2013). As reported above (refer to Fig. 1), 2-PM must be set to an appropriate frame acquisition rate to follow the dynamic events within intact tissues. Surely this associates with the need for sufficient hardware scan speed. Different strategies may apply to improve the scan speed of acquisition without any major hardware upgrade such as setting of a very small region of interest and decreasing of the pixel resolution. Unfortunately this applies well only to a limited number of purposes, including the measurement of blood flow velocity through RBC speed, being this based on the

48

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

acquisition of single linescans sized few microns (see Section 2.1, Chaigneau et al., 2003; Schaffer et al., 2006). In those cases in which the use of small-sized regions of interest (e.g. by acquiring a single linescan, or reducing the final image pixel resolution) may not be applicable, 2-PM fails to provide sufficient scan speed to properly visualize – and quantify – dynamic events. This limitation arises especially when 2-PM interrogates simultaneous quick events such as fluorescence transients due to multiple Ca2+ signals or RBC passage in an array of brain capillaries. Commonly used 2-PM protocols only allow acquisition of quick, isolated events, with no chance to provide recordings of cell or vascular networks at one time. This is mainly due to intrinsically slow scanning mechanisms. Some attempts have been done to overcome this limitation by reducing the background scanning time (Grewe et al., 2010; Lillis et al., 2008; Otsu et al., 2008; Reddy and Saggau, 2005; Sacconi et al., 2008), however none has been able to meet an acceptable compromise between signal-to-noise ratio and fast temporal resolution. A clear, well contrasted signal is achieved by increasing either the excitation laser intensity or the pixel dwell time, with the risk of pushing the laser power above the threshold of photodamage or further slow down the acquisition frame rate. The work by Ducros et al. introduced the encoded multisite two-photon microscopy (eMS2-PM, Ducros et al., 2013). This technique promises to obtain a clear-cut signal-to-noise ratio without jeopardizing the temporal resolution. The eMS2-PM exploits the Hadarmard transform, on the basis of which illuminating and collecting multiple sites at once will enhance the signal-to-noise ratio (Diaspro, 2013). To generate multiple excitation beams, the excitation laser beam is split into sub-beams of significant milliwatt power by using multiple reflection beam splitters or spatial light modulators (Niesner et al., 2007; Nikolenko et al., 2008). The multiple sub-beams are then used to create multisite, simultaneous excitation points organized in defined patterns and the resulting emitted signals are conveyed to a single photomultiplier. To achieve this goal, Ducros et al. modified the microscope setup by adding a spatial light modulator and a digital mirror device. To test the system, they firstly run a proof-of-concept experiment using fluorescent microbeads quickly sliding under the objective. The experiment revealed a 0.72 ms temporal resolution, with satisfactory fluorescent signal intensity and signal-to-noise ratio even in the presence of a thick layer of scattering medium, simulating the effect of brain tissue scattering, between the objective and the specimen. The application of eMS2-PM in vivo allowed the authors to measure simultaneously the RBC speed in twelve regions of interest along cortical microvessels, with a temporal resolution of 0.96 ms (Ducros et al., 2013).eMS2-PM introduces the advantage of imaging the fluorescence changes within few milliseconds, simultaneously and in presence of scattering structures, e.g. the brain cortex. Imaging with eMS2PM can be done over a 300 mm depth in living animals, still with the ability of discriminating signals coming from close structures such as RBC distanced by 7 mm. On the other hand some limitations apply to this method: the simultaneous imaging is only doable in xy planes – no rapid three-dimensional acquisitions – and the absence of phototoxicity needs to be confirmed in vivo. 5.2. Increasing the resolution The power of resolution is a striking feature of the 2-PM. As already discussed, none of the other available in vivo imaging approaches reaches the same level of resolution at the depths explored by 2-PM (Shih et al., 2012). A good practice of 2-PM should aim at obtaining images as close as possible to the best resolution the system can provide. However,

due to the physical properties of the specimens, the theoretical maximum resolution of the microscope cannot be fully reached. Resolution declines increasing imaging depth, since factors such as light absorption and scattering strongly combine leading to blurred signal and poor final image quality (Theer and Denk, 2006). The point spread function (PSF) describes the ability of an optic system to image a point-like light source sized below the resolution limit. This optical propriety sets the maximal resolution and is adversely affected by several factors associated with the biological nature of the sample (Chaigneau et al., 2011). Biological tissues have inhomogeneous optical proprieties, so that the shape of PSF changes with depth, being enlarged (loss of axial and lateral resolution) in response to optical aberrations, collection of out-offocus signal and reduction of the signal to noise ratio (SNR) (Chaigneau et al., 2011; Ji et al., 2010; Theer and Denk, 2006). Another factor that strongly influences the shape of PSF is the excitation wavelength used. In intact tissues it has been demonstrated that the depth-dependent deterioration of spatial resolution is lower at lexc = 1110 nm than at lexc = 920 nm (Herz et al., 2010). Notably, also the craniotomy procedure used to expose the area of interest influences the PSF, the open skull window providing higher image quality at points deeper than 50 mm compared to skull thinning (Isshiki and Okabe, 2013). The best application of in vivo 2-PM provides images extending over thick tissue specimens. This implies that all the factors that erode resolution progressively sum up and settle a limiting depth for satisfactory in vivo 2-PM imaging. While generally considered to range between 800 mm and 1 mm, maximum depth may also be lower depending on the object to visualize. The very small structures such as dendritic spines or thin microglial ramifications, that are sized close to the microscope resolution limit, require high resolution to be properly quantified (Mancuso et al., 2013). Work dealing with fine measures of dendritic spines or microglia morphology is usually done within a range of 200–250 mm beneath the outer cerebral surface, where proper resolution can be achieved (Chaigneau et al., 2011; Herz et al., 2010). Many areas involved in neurodegenerative processes, e.g. hippocampus or caudatum–putamen, remain inaccessible, thus limiting the potential applications of in vivo 2-PM. A few novel hardware set-ups have been recently proposed to better exploit the theoretical maximum resolution of 2-PM. These advancements ground on different means to correct the signal alterations associated with the excitation or the light collection paths. A spatial filter in 4p geometry has been developed to obtain sub-femto-liter excitation volumes (Dilipkumar et al., 2011). This produces a highly localized excitation, limiting the generation of out-of-focus signal and improving the three-dimensional resolution. Excitation may be also improved by using an optical parametric oscillator that further extends the excitation wavelength to the infrared (Herz et al., 2010). Longer excitation wavelengths allow to limit photodamage and to reach deeper tissue penetration due to reduced scattering of the excitation photons. The use of optical parametric oscillator has been also applied to obtain simultaneous three-color fluorescence imaging (Mahou et al., 2012), so to provide better excitation performance for deep-tissue imaging and to increase detection efficiency due to absent overlap between excitation and emission spectra. The implementation of deformable membrane mirrors in the excitation path has been proposed to correct brain tissue-induced wavefront distortion (Chaigneau et al., 2011). This solution has been applied to in vivo 2-PM of the mouse neocortex, and has proved to decrease the size of the PSF and to increase the SNR, thus enhancing the quality of images. The artifacts arising from the mechanical positioning of the objective lens along the z axis can be minimized by the use of

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

inertia-free focus control, as recently demonstrated (Dal Maschio et al., 2011). Adaptive optics, based on the segmentation of the rear pupil of objective lens, can correct optical inhomogeneities and have been demonstrated to improve signal resolution to 400 mm depths in fixed brain slice (Ji et al., 2010). Good resolution coupled with high scanning speeds at similar depths have been obtained in vivo by the use of acousto-optic scanning technology (Katona et al., 2012). 6. Conclusions Biological events occur in a physical space within a specific time frame and require the action of solid objects such as blood vessels and cells. These objects exert their specific function in intact tissues, where there is a complex network of interactions. Most of the information associated with spatial motility, time-dependent dynamics and tissue integrity (e.g. simultaneous presence of all the cell types involved in the biological effect) is lost with conventional biochemical, biomolecular or histological techniques. In neurobiology 2-PM offered a new way to tackle previously unexplored mechanisms in physiological or pathological conditions, thanks to its ability to provide a three-dimensional, high-resolution representation of the CNS over time in living animals.

49

We have reviewed how 2-PM has provided previously unavailable information on dynamic pathophysiological events after CNS injury. We focused on recent findings on vascular dynamics after brain ischemia and the interaction between cerebral vasculature and cells that populate or are recruited to the CNS. Most of 2-PM studies have been conducted in vivo, combining fluorescent labeling of vessels with genetically encoded fluorescent reporters. They have revealed, on the one hand, a passive role of the vasculature in the pathophysiology of brain ischemia, since vascular geometry itself can lead to damage. On the other hand, the vasculature can have an active role in the inflammatory response, or may in turn be regulated (e.g. local hyperemia or vasoconstriction) by cells of the neurovascular unit, with final impact on the pathogenesis of stroke (Fig. 5). Since these events all involve intrinsic dynamic movements, 2PM analysis is clearly the first choice. 2-PM offers a useful tool to confirm mechanisms previously suggested using conventional means, and to investigate unexplored events. Some obvious technical limitations apply to in vivo 2-PM, such as the depth of acquisition up to about 800 mm, the small regions of interest (e.g. limited tissue sampling) and the need for skillful surgical preparation of animals. A recent correlative approach combined 2-PM with post mortem confocal light sheet microscopy

Fig. 5. Dynamic events at the neurovascular unit after ischemic injury. During physiological homeostasis blood flow is maintained at constant levels, with transient local increases in response to specific neuronal metabolic demand. Cells such as RBC, lymphocytes and monocytes circulate without extravasating. Occasionally, lymphocytes can travel along the perivascular wall. Astrocytes and microglia make contact with endothelial cells as part of the blood–brain barrier. Neurons are located at a mean distance of 13 mm from the closest vessel. After an injury such as brain ischemia several events take place, breaking the physiological homeostasis. Time-lapse 2-PM imaging provided the chance to visualize many dynamic events which are associated with pathology. (1) The architecture of the vasculature is redundant: when occlusion occurs, blood is redistributed through the pial vasculature (yellow arrows in inserts) to provide an alternative supply to downstream penetrating arterioles. (2) There is a local increase of PO2 in the proximity of RBC bodies. In stroke research, non-fluorescent RBCs are widely used for the measurement of blood flow velocity. (3) Neurons settled in hypoperfused areas and more than 80 mm from a flowing vessel lose dendritic spines and may suffer irreversible synaptic dysfunction. (4) Lymphocytes need to travel along the vessels (abluminal side) before completing extravasation, possibly to receive priming signals from the components of the neurovascular unit. Once extravasated, they can secrete activatory signals capable of driving the microglia/macrophages polarization choice. (5) Only where there is residual or re-established blood flow, can microglia be supplied with sufficient ATP for activation, changing their shape and acquiring a specific polarization profile. (6) Pericytes cause local vasoconstriction after ischemia, possibly favoring the formation of new clots in microvessels because of reduction in diameter which blocks the normal flow of blood cells.

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

50

(Silvestri et al., 2012), so to counter-balance the intrinsic spatial limitations of 2-PM with a technique able to visualize the entire brain, although with a time limitation, in turn compensated by 2PM (Silvestri et al., 2013). Further technical developments and new transgenic mouse lines with fluorescent reporters will help improve 2-PM. As for new technical approaches, the recent development of the eMS2-PM (Ducros et al., 2013) and the line-scanning particle image velocimetry (Kim et al., 2012) introduced a significant advancement to 2-PM that in future may be usefully applied to speed up the process of image acquisition (see Section 5.1). Other advancements have been proposed to increase the image resolution at significant depths, where conventional 2-PM systems fail to provide sufficient details (see Section 5.2). This will extend the use of 2-PM in vivo imaging to deep brain areas that may be important in the study of neurodegenerative processes. Regarding the animal models, new transgenic lines already exist with specific imaging purposes such as the brainbow mice (Livet et al., 2007) for visualizing neurons in different color ranges or the thy1GFP-M mice, in which neurons and motile dendritic cells are fluorescently labeled (Laperchia et al., 2013). A significant effort is also dedicated to the improvement of the existing transgenic lines (Cai et al., 2013) and it is hoped that all these tools will be soon applied to the study of normal and pathological conditions. However, it should be noted that conventional 2-PM systems do not allow the simultaneous excitation of more than two distinct fluorescence colors, thus limiting the in vivo application of 2-PM to transgenic lines with multiple color labeling (e.g. brainbow mice). Waiting for the radiation and the stable use of the newest upgrades, 2-PM however already stands as an advanced imaging technique, successfully and largely exploited by researchers in the field of brain injury. Future improvements in models of brain diseases will greatly depend on technological progress to produce reliable and solid data. Wider use of 2-PM, with its ability to investigate dynamic events, combined with conventional post-mortem techniques, more closely linked to molecular information, will play a fundamental role in that technological improvement. A comprehensive representation of a pathological event will help raise the standards of preclinical research, allowing the development of new, more effective therapeutic strategies.

Conflict of interest None.

References Abbott, N.J., 2002. Astrocyte–endothelial interactions and blood–brain barrier permeability. J. Anat. 200, 629–638. Agrawal, S., Anderson, P., Durbeej, M., van Rooijen, N., Ivars, F., Opdenakker, G., Sorokin, L.M., 2006. Dystroglycan is selectively cleaved at the parenchymal basement membrane at sites of leukocyte extravasation in experimental autoimmune encephalomyelitis. J. Exp. Med. 203, 1007–1019. Agrawal, S.M., Williamson, J., Sharma, R., Kebir, H., Patel, K., Prat, A., Yong, V.W., 2013. Extracellular matrix metalloproteinase inducer shows active perivascular cuffs in multiple sclerosis. Brain 136, 1760–1777. Albertazzi, L., Storti, B., Brondi, M., Sulis Sato, S., Ratto, G.M., Signore, G., Beltram, F., 2013. Synthesis, cellular delivery and in vivo application of dendrimer-based pH sensors. J. Vis. Exp.. Allen, C., Thornton, P., Denes, A., McColl, B.W., Pierozynski, A., Monestier, M., Pinteaux, E., Rothwell, N.J., Allan, S.M., 2012. Neutrophil cerebrovascular transmigration triggers rapid neurotoxicity through release of proteases associated with decondensed DNA. J. Immunol. 189, 381–392. Amirbekian, S., Long Jr., R.C., Consolini, M.A., Suo, J., Willett, N.J., Fielden, S.W., Giddens, D.P., Taylor, W.R., Oshinski, J.N., 2009. In vivo assessment of blood flow patterns in abdominal aorta of mice with MRI: implications for AAA localization. Am. J. Physiol. Heart Circ. Physiol. 297, H1290–H1295.

Andermann, M.L., Gilfoy, N.B., Goldey, G.J., Sachdev, R.N.S., Wo¨lfel, M., McCormick, D.A., Reid, R.C., Levene, M.J., 2013. Chronic cellular imaging of entire cortical columns in awake mice using microprisms. Neuron 80, 900–913. Arumugam, T.V., Granger, D.N., Mattson, M.P., 2005. Stroke and T-cells. Neuromolecular Med. 7, 229–242. Attwell, D., Buchan, A.M., Charpak, S., Lauritzen, M., Macvicar, B.A., Newman, E.A., 2010. Glial and neuronal control of brain blood flow. Nature 468, 232–243. Baeten, K.M., Akassoglou, K., 2011. Extracellular matrix and matrix receptors in blood–brain barrier formation and stroke. Dev. Neurobiol. 71, 1018–1039, http://dx.doi.org/10.1002/dneu.20954. Baron, J.C., 2001. Perfusion thresholds in human cerebral ischemia: historical perspective and therapeutic implications. Cerebrovasc. Dis. 11 (Suppl 1), 2–8. Bartholoma¨us, I., Kawakami, N., Odoardi, F., Schla¨ger, C., Miljkovic, D., Ellwart, J.W., Klinkert, W.E.F., Flu¨gel-Koch, C., Issekutz, T.B., Wekerle, H., Flu¨gel, A., 2009. Effector T cell interactions with meningeal vascular structures in nascent autoimmune CNS lesions. Nature 462, 94–98. Bauer, M., Brakebusch, C., Coisne, C., Sixt, M., Wekerle, H., Engelhardt, B., Fa¨ssler, R., 2009. Beta1 integrins differentially control extravasation of inflammatory cell subsets into the CNS during autoimmunity. Proc. Natl. Acad. Sci. U. S. A. 106, 1920–1925. Begley, D.J., Brightman, M.W., 2003. Structural and functional aspects of the blood– brain barrier. Prog. Drug Res. 61, 39–78. Berardi, N., Pizzorusso, T., Maffei, L., 2004. Extracellular matrix and visual cortical plasticity: freeing the synapse. Neuron 44, 905–908. Berezowski, V., Landry, C., Dehouck, M.-P., Cecchelli, R., Fenart, L., 2004. Contribution of glial cells and pericytes to the mRNA profiles of P-glycoprotein and multidrug resistance-associated proteins in an in vitro model of the blood– brain barrier. Brain Res. 1018, 1–9. Bittner, T., Burgold, S., Dorostkar, M.M., Fuhrmann, M., Wegenast-Braun, B.M., Schmidt, B., Kretzschmar, H., Herms, J., 2012. Amyloid plaque formation precedes dendritic spine loss. Acta Neuropathol. 124, 797–807. Brait, V.H., Jackman, K.A., Walduck, A.K., Selemidis, S., Diep, H., Mast, A.E., Guida, E., Broughton, B.R.S., Drummond, G.R., Sobey, C.G., 2010. Mechanisms contributing to cerebral infarct size after stroke: gender, reperfusion, T lymphocytes, and Nox2-derived superoxide. J. Cereb. Blood Flow Metab. 30, 1306–1317. Brown, C.E., Aminoltejari, K., Erb, H., Winship, I.R., Murphy, T.H., 2009. In vivo voltage-sensitive dye imaging in adult mice reveals that somatosensory maps lost to stroke are replaced over weeks by new structural and functional circuits with prolonged modes of activation within both the peri-infarct zone and distant sites. J. Neurosci. 29, 1719–1734. Brown, C.E., Li, P., Boyd, J.D., Delaney, K.R., Murphy, T.H., 2007. Extensive turnover of dendritic spines and vascular remodeling in cortical tissues recovering from stroke. J. Neurosci. 27, 4101–4109. Brown, E.B., Shear, J.B., Adams, S.R., Tsien, R.Y., Webb, W.W., 1999. Photolysis of caged calcium in femtoliter volumes using two-photon excitation. Biophys. J. 76, 489–499. Butovsky, O., Bukshpan, S., Kunis, G., Jung, S., Schwartz, M., 2007. Microglia can be induced by IFN-gamma or IL-4 to express neural or dendritic-like markers. Mol. Cell. Neurosci. 35, 490–500. Cahalan, M.D., Parker, I., 2008. Choreography of cell motility and interaction dynamics imaged by two-photon microscopy in lymphoid organs. Annu. Rev. Immunol. 26, 585–626. Cai, D., Cohen, K.B., Luo, T., Lichtman, J.W., Sanes, J.R., 2013. Improved tools for the Brainbow toolbox. Nat. Methods 10, 540–547. Chaigneau, E., Oheim, M., Audinat, E., Charpak, S., 2003. Two-photon imaging of capillary blood flow in olfactory bulb glomeruli. Proc. Natl. Acad. Sci. U. S. A. 100, 13081–13086. Chaigneau, E., Wright, A.J., Poland, S.P., Girkin, J.M., Silver, R.A., 2011. Impact of wavefront distortion and scattering on 2-photon microscopy in mammalian brain tissue. Opt. Express 19, 22755–22774. Chen, J.L., Andermann, M.L., Keck, T., Xu, N.-L., Ziv, Y., 2013. Imaging neuronal populations in behaving rodents: paradigms for studying neural circuits underlying behavior in the mammalian cortex. J. Neurosci. 33, 17631–17640. Cho, E.E., Drazic, J., Ganguly, M., Stefanovic, B., Hynynen, K., 2011. Two-photon fluorescence microscopy study of cerebrovascular dynamics in ultrasound-induced blood–brain barrier opening. J. Cereb. Blood Flow Metab. 31, 1852–1862. Chtanova, T., Han, S.-J., Schaeffer, M., van Dooren, G.G., Herzmark, P., Striepen, B., Robey, E.A., 2009. Dynamics of T cell, antigen-presenting cell, and pathogen interactions during recall responses in the lymph node. Immunity 31, 342–355. Cordiglieri, C., Odoardi, F., Zhang, B., Nebel, M., Kawakami, N., Klinkert, W.E.F., Lodygin, D., Lu¨hder, F., Breunig, E., Schild, D., Ulaganathan, V.K., Dornmair, K., Dammermann, W., Potter, B.V.L., Guse, A.H., Flu¨gel, A., 2010. Nicotinic acid adenine dinucleotide phosphate-mediated calcium signalling in effector T cells regulates autoimmunity of the central nervous system. Brain 133, 1930–1943. Cullen, K.M., Ko´csi, Z., Stone, J., 2005. Pericapillary haem-rich deposits: evidence for microhaemorrhages in aging human cerebral cortex. J. Cereb. Blood Flow Metab. 25, 1656–1667. ˜ o, V., Noctor, S.C., 2013. Microglia regulate the Cunningham, C.L., Martı´nez-Cerden number of neural precursor cells in the developing cerebral cortex. J. Neurosci. 33, 4216–4233. Dal Maschio, M., De Stasi, A.M., Benfenati, F., Fellin, T., 2011. Three-dimensional in vivo scanning microscopy with inertia-free focus control. Opt. Lett. 36, 3503–3505.

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54 Davalos, D., Grutzendler, J., Yang, G., Kim, J.V., Zuo, Y., Jung, S., Littman, D.R., Dustin, M.L., Gan, W.-B., 2005. ATP mediates rapid microglial response to local brain injury in vivo. Nat. Neurosci. 8, 752–758. Davalos, D., Ryu, J.K., Merlini, M., Baeten, K.M., Le Moan, N., Petersen, M.A., Deerinck, T.J., Smirnoff, D.S., Bedard, C., Hakozaki, H., Gonias Murray, S., Ling, J.B., Lassmann, H., Degen, J.L., Ellisman, M.H., Akassoglou, K., 2012. Fibrinogen-induced perivascular microglial clustering is required for the development of axonal damage in neuroinflammation. Nat. Commun. 3, 1227. David, S., Kroner, A., 2011. Repertoire of microglial and macrophage responses after spinal cord injury. Nat. Rev. Neurosci. 12, 388–399. Davoust, N., Vuaillat, C., Androdias, G., Nataf, S., 2008. From bone marrow to microglia: barriers and avenues. Trends Immunol. 29, 227–234. De Boer, J., Williams, A., Skavdis, G., Harker, N., Coles, M., Tolaini, M., Norton, T., Williams, K., Roderick, K., Potocnik, A.J., Kioussis, D., 2003. Transgenic mice with hematopoietic and lymphoid specific expression of Cre. Eur. J. Immunol. 33, 314–325. De Vivo, L., Landi, S., Panniello, M., Baroncelli, L., Chierzi, S., Mariotti, L., Spolidoro, M., Pizzorusso, T., Maffei, L., Ratto, G.M., 2013. Extracellular matrix inhibits structural and functional plasticity of dendritic spines in the adult visual cortex. Nat. Commun. 4, 1484. Del Zoppo, G.J., 1997. Microvascular responses to cerebral ischemia/inflammation. Ann. N. Y. Acad. Sci. 823, 132–147. Del Zoppo, G.J., Mabuchi, T., 2003. Cerebral microvessel responses to focal ischemia. J. Cereb. Blood Flow Metab. 23, 879–894. De´nes, A., Ferenczi, S., Hala´sz, J., Ko¨rnyei, Z., Kova´cs, K.J., 2008. Role of CX3CR1 (fractalkine receptor) in brain damage and inflammation induced by focal cerebral ischemia in mouse. J. Cereb. Blood Flow Metab. 28, 1707–1721. Denk, W., Strickler, J.H., Webb, W.W., 1990. Two-photon laser scanning fluorescence microscopy. Science 248, 73–76. Diaspro, A., 2013. Taking three-dimensional two-photon excitation microscopy further: encoding the light for decoding the brain. Microsc. Res. Tech.. Dilipkumar, S., Diaspro, A., Mondal, P.P., 2011. Spatial filter based 3D resolution improvement and polarization properties of multiphoton multiple-excitationspot-optical microscopy. Rev. Sci. Instrum. 82, 063705. Ding, S., Wang, T., Cui, W., Haydon, P.G., 2009. Photothrombosis ischemia stimulates a sustained astrocytic Ca2+ signaling in vivo. Glia 57, 767–776. Dirnagl, U., Iadecola, C., Moskowitz, M.A., 1999. Pathobiology of ischaemic stroke: an integrated view. Trends Neurosci. 22, 391–397. Dityatev, A., Schachner, M., Sonderegger, P., 2010. The dual role of the extracellular matrix in synaptic plasticity and homeostasis. Nat. Rev. Neurosci. 11, 735–746. Dombeck, D.A., Khabbaz, A.N., Collman, F., Adelman, T.L., Tank, D.W., 2007. Imaging large-scale neural activity with cellular resolution in awake, mobile mice. Neuron 56, 43–57. Ducros, M., Goulam Houssen, Y., Bradley, J., de Sars, V., Charpak, S., 2013. Encoded multisite two-photon microscopy. Proc. Natl. Acad. Sci. U. S. A. 110, 13138–13143. Engelhardt, B., 2006. Molecular mechanisms involved in T cell migration across the blood–brain barrier. J. Neural Transm. 113, 477–485. Enzmann, G., Mysiorek, C., Gorina, R., Cheng, Y.-J., Ghavampour, S., Hannocks, M.-J., Prinz, V., Dirnagl, U., Endres, M., Prinz, M., Beschorner, R., Harter, P.N., Mittelbronn, M., Engelhardt, B., Sorokin, L., 2013. The neurovascular unit as a selective barrier to polymorphonuclear granulocyte (PMN) infiltration into the brain after ischemic injury. Acta Neuropathol. 125, 395–412. Farrall, A.J., Wardlaw, J.M., 2009. Blood–brain barrier: ageing and microvascular disease – systematic review and meta-analysis. Neurobiol. Aging 30, 337–352. Ferna´ndez-Klett, F., Offenhauser, N., Dirnagl, U., Priller, J., Lindauer, U., 2010. Pericytes in capillaries are contractile in vivo, but arterioles mediate functional hyperemia in the mouse brain. Proc. Natl. Acad. Sci. U. S. A. 107, 22290–22295. Finikova, O.S., Troxler, T., Senes, A., DeGrado, W.F., Hochstrasser, R.M., Vinogradov, S.A., 2007. Energy and electron transfer in enhanced two-photon-absorbing systems with triplet cores. J. Phys. Chem. A 111, 6977–6990. Flu¨gel, A., Odoardi, F., Nosov, M., Kawakami, N., 2007. Autoaggressive effector T cells in the course of experimental autoimmune encephalomyelitis visualized in the light of two-photon microscopy. J. Neuroimmunol. 191, 86–97. Frischknecht, R., Gundelfinger, E.D., 2012. The brain’s extracellular matrix and its role in synaptic plasticity. Adv. Exp. Med. Biol. 970, 153–171. Fumagalli, S., Coles, J.A., Ejlerskov, P., Ortolano, F., Bushell, T.J., Brewer, J.M., De Simoni, M.-G., Dever, G., Garside, P., Maffia, P., Carswell, H.V., 2011. In vivo realtime multiphoton imaging of T lymphocytes in the mouse brain after experimental stroke. Stroke 42, 1429–1436. Fumagalli, S., Perego, C., Ortolano, F., De Simoni, M.-G., 2013. CX3CR1 deficiency induces an early protective inflammatory environment in ischemic mice. Glia 61, 827–842. Garcia, J.H., Liu, K.F., Yoshida, Y., Lian, J., Chen, S., del Zoppo, G.J., 1994. Influx of leukocytes and platelets in an evolving brain infarct (Wistar rat). Am. J. Pathol. 144, 188–199. Gee, J.M., Kalil, A., Thullbery, M., Becker, K.J., 2008. Induction of immunologic tolerance to myelin basic protein prevents central nervous system autoimmunity and improves outcome after stroke. Stroke 39, 1575–1582. Gee, J.M., Zierath, D., Hadwin, J., Savos, A., Kalil, A., Thullbery, M., Becker, K.J., 2009. Long term immunologic consequences of experimental stroke and mucosal tolerance. Exp. Transl. Stroke Med. 1, 3. Gelderblom, M., Leypoldt, F., Steinbach, K., Behrens, D., Choe, C.-U., Siler, D.A., Arumugam, T.V., Orthey, E., Gerloff, C., Tolosa, E., Magnus, T., 2009. Temporal and spatial dynamics of cerebral immune cell accumulation in stroke. Stroke 40, 1849–1857.

51

Gesuete, R., Storini, C., Fantin, A., Stravalaci, M., Zanier, E.R., Orsini, F., Vietsch, H., Mannesse, M.L.M., Ziere, B., Gobbi, M., De Simoni, M.-G., 2009. Recombinant C1 inhibitor in brain ischemic injury. Ann. Neurol. 66, 332–342. Gibson, V.B., Benson, R.A., Bryson, K.J., McInnes, I.B., Rush, C.M., Grassia, G., Maffia, P., Jenkinson, E.J., White, A.J., Anderson, G., Brewer, J.M., Garside, P., 2012. A novel method to allow noninvasive, longitudinal imaging of the murine immune system in vivo. Blood 119, 2545–2551. Golub, A.S., Pittman, R.N., 2005. Erythrocyte-associated transients in PO 2 revealed in capillaries of rat mesentery. Am. J. Physiol. Heart Circ. Physiol. 288, H2735–H2743. Gordon, G.R.J., Choi, H.B., Rungta, R.L., Ellis-Davies, G.C.R., MacVicar, B.A., 2008. Brain metabolism dictates the polarity of astrocyte control over arterioles. Nature 456, 745–749. Grewe, B.F., Langer, D., Kasper, H., Kampa, B.M., Helmchen, F., 2010. High-speed in vivo calcium imaging reveals neuronal network activity with near-millisecond precision. Nat. Methods 7, 399–405. Grinberg, Y.Y., Milton, J.G., Kraig, R.P., 2011. Spreading depression sends microglia on Le´vy flights. PLoS ONE 6, e19294. Gyoneva, S., Davalos, D., Biswas, D., Swanger, S.A., Garnier-Amblard, E., Loth, F., Akassoglou, K., Traynelis, S.F., 2014. Systemic inflammation regulates microglial responses to tissue damage in vivo. Glia, http://dx.doi.org/10.1002/glia.22686. Hallenbeck, J., Del Zoppo, G., Jacobs, T., Hakim, A., Goldman, S., Utz, U., Hasan, A., 2006. Immunomodulation strategies for preventing vascular disease of the brain and heart: workshop summary. Stroke 37, 3035–3042. Hamilton, N.B., Attwell, D., Hall, C.N., 2010. Pericyte-mediated regulation of capillary diameter: a component of neurovascular coupling in health and disease. Front. Neuroenergetics 2 . Hanisch, U.-K., Kettenmann, H., 2007. Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat. Neurosci. 10, 1387–1394. Harb, R., Whiteus, C., Freitas, C., Grutzendler, J., 2013. In vivo imaging of cerebral microvascular plasticity from birth to death. J. Cereb. Blood Flow Metab. 33, 146–156. Haynes, S.E., Hollopeter, G., Yang, G., Kurpius, D., Dailey, M.E., Gan, W.-B., Julius, D., 2006. The P2Y12 receptor regulates microglial activation by extracellular nucleotides. Nat. Neurosci. 9, 1512–1519. Hefendehl, J.K., Milford, D., Eicke, D., Wegenast-Braun, B.M., Calhoun, M.E., Grathwohl, S.A., Jucker, M., Liebig, C., 2012. Repeatable target localization for long-term in vivo imaging of mice with 2-photon microscopy. J. Neurosci. Methods 205, 357–363. Hefendehl, J.K., Neher, J.J., Su¨hs, R.B., Kohsaka, S., Skodras, A., Jucker, M., 2013. Homeostatic and injury-induced microglia behavior in the aging brain. Aging Cell. Helmchen, F., Kleinfeld, D., 2008. Chapter 10. In vivo measurements of blood flow and glial cell function with two-photon laser-scanning microscopy. Methods Enzymol. 444, 231–254. Herz, J., Siffrin, V., Hauser, A.E., Brandt, A.U., Leuenberger, T., Radbruch, H., Zipp, F., Niesner, R.A., 2010. Expanding two-photon intravital microscopy to the infrared by means of optical parametric oscillator. Biophys. J. 98, 715–723. Hirase, H., Creso, J., Singleton, M., Bartho´, P., Buzsa´ki, G., 2004. Two-photon imaging of brain pericytes in vivo using dextran-conjugated dyes. Glia 46, 95–100, http://dx.doi.org/10.1002/glia.10295. Hossmann, K.A., 1994. Viability thresholds and the penumbra of focal ischemia. Ann. Neurol. 36, 557–565. Huang, J., Upadhyay, U.M., Tamargo, R.J., 2006. Inflammation in stroke and focal cerebral ischemia. Surg. Neurol. 66, 232–245. Huo, Y., Guo, X., Kassab, G.S., 2008. The flow field along the entire length of mouse aorta and primary branches. Ann. Biomed. Eng. 36, 685–699. Hurn, P.D., Subramanian, S., Parker, S.M., Afentoulis, M.E., Kaler, L.J., Vandenbark, A.A., Offner, H., 2007. T- and B-cell-deficient mice with experimental stroke have reduced lesion size and inflammation. J. Cereb. Blood Flow Metab. 27, 1798–1805. Iadecola, C., Anrather, J., 2011. The immunology of stroke: from mechanisms to translation. Nat. Med. 17, 796–808. Isshiki, M., Okabe, S., 2013. Evaluation of cranial window types for in vivo twophoton imaging of brain microstructures. Microscopy (Oxf.), http://dx.doi.org/ 10.1093/jmicro/dft043. Jauch, E.C., Saver, J.L., Adams Jr., H.P., Bruno, A., Connors, J.J.B., Demaerschalk, B.M., Khatri, P., McMullan Jr., P.W., Qureshi, A.I., Rosenfield, K., Scott, P.A., Summers, D.R., Wang, D.Z., Wintermark, M., Yonas, H., American Heart Association Stroke Council, Council on Cardiovascular Nursing, Council on Peripheral Vascular Disease, Council on Clinical Cardiology, 2013. Guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 44, 870–947. Ji, N., Milkie, D.E., Betzig, E., 2010. Adaptive optics via pupil segmentation for highresolution imaging in biological tissues. Nat. Methods 7, 141–147. Jiang, M., Ash, R.T., Baker, S.A., Suter, B., Ferguson, A., Park, J., Rudy, J., Torsky, S.P., Chao, H.-T., Zoghbi, H.Y., Smirnakis, S.M., 2013. Dendritic arborization and spine dynamics are abnormal in the mouse model of MECP2 duplication syndrome. J. Neurosci. 33, 19518–19533. Jickling, G.C., Liu, D., Stamova, B., Ander, B.P., Zhan, X., Lu, A., Sharp, F.R., 2014. Hemorrhagic transformation after ischemic stroke in animals and humans. J. Cereb. Blood Flow Metab. 34, 185–199. Jungblut, D., Vlachos, A., Schuldt, G., Zahn, N., Deller, T., Wittum, G., 2012. SpineLab: tool for three-dimensional reconstruction of neuronal cell morphology. J. Biomed. Opt. 17, 076007.

52

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

Kaifosh, P., Lovett-Barron, M., Turi, G.F., Reardon, T.R., Losonczy, A., 2013. Septohippocampal GABAergic signaling across multiple modalities in awake mice. Nat. Neurosci. 16, 1182–1184. Katona, G., Szalay, G., Maa´k, P., Kasza´s, A., Veress, M., Hillier, D., Chiovini, B., Vizi, E.S., Roska, B., Ro´zsa, B., 2012. Fast two-photon in vivo imaging with threedimensional random-access scanning in large tissue volumes. Nat. Methods 9, 201–208. Kawakami, N., Flu¨gel, A., 2010. Knocking at the brain’s door: intravital two-photon imaging of autoreactive T cell interactions with CNS structures. Semin. Immunopathol. 32, 275–287. Kim, J.V., Kang, S.S., Dustin, M.L., McGavern, D.B., 2009. Myelomonocytic cell recruitment causes fatal CNS vascular injury during acute viral meningitis. Nature 457, 191–195. Kim, T.N., Goodwill, P.W., Chen, Y., Conolly, S.M., Schaffer, C.B., Liepmann, D., Wang, R.A., 2012. Line-scanning particle image velocimetry: an optical approach for quantifying a wide range of blood flow speeds in live animals. PLOS ONE 7, e38590. Kim, Y.-S., Joh, T.H., 2012. Matrix metalloproteinases, new insights into the understanding of neurodegenerative disorders. Biomol. Ther. (Seoul) 20, 133–143. Kivisa¨kk, P., Imitola, J., Rasmussen, S., Elyaman, W., Zhu, B., Ransohoff, R.M., Khoury, S.J., 2009. Localizing central nervous system immune surveillance: meningeal antigen-presenting cells activate T cells during experimental autoimmune encephalomyelitis. Ann. Neurol. 65, 457–469. Kleinschnitz, C., Schwab, N., Kraft, P., Hagedorn, I., Dreykluft, A., Schwarz, T., Austinat, M., Nieswandt, B., Wiendl, H., Stoll, G., 2010. Early detrimental T-cell effects in experimental cerebral ischemia are neither related to adaptive immunity nor thrombus formation. Blood 115, 3835–3842. Klohs, J., Rudin, M., Shimshek, D.R., Beckmann, N., 2014. Imaging of cerebrovascular pathology in animal models of Alzheimer’s disease. Front. Aging Neurosci. 6, 32. Klohs, J., Steinbrink, J., Bourayou, R., Mueller, S., Cordell, R., Licha, K., Schirner, M., Dirnagl, U., Lindauer, U., Wunder, A., 2009. Near-infrared fluorescence imaging with fluorescently labeled albumin: a novel method for non-invasive optical imaging of blood–brain barrier impairment after focal cerebral ischemia in mice. J. Neurosci. Methods 180, 126–132. Kondo, S., Kohsaka, S., Okabe, S., 2011. Long-term changes of spine dynamics and microglia after transient peripheral immune response triggered by LPS in vivo. Mol. Brain 4, 27. Konsman, J.P., Drukarch, B., Van Dam, A.-M., 2007. (Peri)vascular production and action of pro-inflammatory cytokines in brain pathology. Clin. Sci. 112, 1–25. Kozlowski, C., Weimer, R.M., 2012. An automated method to quantify microglia morphology and application to monitor activation state longitudinally in vivo. PLOS ONE 7, e31814. Landi, S., Putignano, E., Boggio, E.M., Giustetto, M., Pizzorusso, T., Ratto, G.M., 2011. The short-time structural plasticity of dendritic spines is altered in a model of Rett syndrome. Sci. Rep. 1, 45. Laperchia, C., Allegra Mascaro, A.L., Sacconi, L., Andrioli, A., Matte`, A., De Franceschi, L., Grassi-Zucconi, G., Bentivoglio, M., Buffelli, M., Pavone, F.S., 2013. Twophoton microscopy imaging of thy1GFP-M transgenic mice: a novel animal model to investigate brain dendritic cell subsets in vivo. PLOS ONE 8, e56144. Lecoq, J., Parpaleix, A., Roussakis, E., Ducros, M., Houssen, Y.G., Vinogradov, S.A., Charpak, S., 2011. Simultaneous two-photon imaging of oxygen and blood flow in deep cerebral vessels. Nat. Med. 17, 893–898. Lecoq, J., Tiret, P., Najac, M., Shepherd, G.M., Greer, C.A., Charpak, S., 2009. Odorevoked oxygen consumption by action potential and synaptic transmission in the olfactory bulb. J. Neurosci. 29, 1424–1433. Lerchenberger, M., Uhl, B., Stark, K., Zuchtriegel, G., Eckart, A., Miller, M., PuhrWesterheide, D., Praetner, M., Rehberg, M., Khandoga, A.G., Lauber, K., Massberg, S., Krombach, F., Reichel, C.A., 2013. Matrix metalloproteinases modulate ameboid-like migration of neutrophils through inflamed interstitial tissue. Blood 122, 770–780. Leys, D., He´non, H., Mackowiak-Cordoliani, M.-A., Pasquier, F., 2005. Poststroke dementia. Lancet Neurol. 4, 752–759. Liang, K.J., Lee, J.E., Wang, Y.D., Ma, W., Fontainhas, A.M., Fariss, R.N., Wong, W.T., 2009. Regulation of dynamic behavior of retinal microglia by CX3CR1 signaling. Invest. Ophthalmol. Vis. Sci. 50, 4444–4451. Liesz, A., Suri-Payer, E., Veltkamp, C., Doerr, H., Sommer, C., Rivest, S., Giese, T., Veltkamp, R., 2009. Regulatory T cells are key cerebroprotective immunomodulators in acute experimental stroke. Nat. Med. 15, 192–199. Lillis, K.P., Eng, A., White, J.A., Mertz, J., 2008. Two-photon imaging of spatially extended neuronal network dynamics with high temporal resolution. J. Neurosci. Methods 172, 178–184. Lindvere, L., Janik, R., Dorr, A., Chartash, D., Sahota, B., Sled, J.G., Stefanovic, B., 2013. Cerebral microvascular network geometry changes in response to functional stimulation. Neuroimage 71, 248–259. Livet, J., Weissman, T.A., Kang, H., Draft, R.W., Lu, J., Bennis, R.A., Sanes, J.R., Lichtman, J.W., 2007. Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature 450, 56–62. Lodygin, D., Odoardi, F., Schla¨ger, C., Ko¨rner, H., Kitz, A., Nosov, M., van den Brandt, J., Reichardt, H.M., Haberl, M., Flu¨gel, A., 2013. A combination of fluorescent NFAT and H2B sensors uncovers dynamics of T cell activation in real time during CNS autoimmunity. Nat. Med. 19, 784–790. Lok, J., Gupta, P., Guo, S., Kim, W.J., Whalen, M.J., van Leyen, K., Lo, E.H., 2007. Cell– cell signaling in the neurovascular unit. Neurochem. Res. 32, 2032–2045. Lu, A., Clark, J.F., Broderick, J.P., Pyne-Geithman, G.J., Wagner, K.R., Ran, R., Khatri, P., Tomsick, T., Sharp, F.R., 2008. Reperfusion activates metalloproteinases that contribute to neurovascular injury. Exp. Neurol. 210, 549–559.

Maggio, N., Itsekson, Z., Dominissini, D., Blatt, I., Amariglio, N., Rechavi, G., Tanne, D., Chapman, J., 2013. Thrombin regulation of synaptic plasticity: implications for physiology and pathology. Exp. Neurol. 247, 595–604. Magnus, T., Wiendl, H., Kleinschnitz, C., 2012. Immune mechanisms of stroke. Curr. Opin. Neurol. 25, 334–340. Mahou, P., Zimmerley, M., Loulier, K., Matho, K.S., Labroille, G., Morin, X., Supatto, W., Livet, J., De´barre, D., Beaurepaire, E., 2012. Multicolor two-photon tissue imaging by wavelength mixing. Nat. Methods 9, 815–818. Mancuso, J.J., Chen, Y., Li, X., Xue, Z., Wong, S.T.C., 2013. Methods of dendritic spine detection: from Golgi to high-resolution optical imaging. Neuroscience 251, 129–140. Marker, D.F., Tremblay, M.-E., Lu, S.-M., Majewska, A.K., Gelbard, H.A., 2010. A thinskull window technique for chronic two-photon in vivo imaging of murine microglia in models of neuroinflammation. J. Vis. Exp.. Masuda, T., Croom, D., Hida, H., Kirov, S.A., 2011. Capillary blood flow around microglial somata determines dynamics of microglial processes in ischemic conditions. Glia 59, 1744–1753. Mathiisen, T.M., Lehre, K.P., Danbolt, N.C., Ottersen, O.P., 2010. The perivascular astroglial sheath provides a complete covering of the brain microvessels: an electron microscopic 3D reconstruction. Glia 58, 1094–1103. Mempel, T.R., Henrickson, S.E., Von Andrian, U.H., 2004. T-cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature 427, 154–159. Metea, M.R., Newman, E.A., 2006. Glial cells dilate and constrict blood vessels: a mechanism of neurovascular coupling. J. Neurosci. 26, 2862–2870. Michelucci, A., Heurtaux, T., Grandbarbe, L., Morga, E., Heuschling, P., 2009. Characterization of the microglial phenotype under specific pro-inflammatory and anti-inflammatory conditions: effects of oligomeric and fibrillar amyloid-beta. J. Neuroimmunol. 210, 3–12. Miller, M.J., Wei, S.H., Parker, I., Cahalan, M.D., 2002. Two-photon imaging of lymphocyte motility and antigen response in intact lymph node. Science 296, 1869–1873. Mues, M., Bartholoma¨us, I., Thestrup, T., Griesbeck, O., Wekerle, H., Kawakami, N., Krishnamoorthy, G., 2013. Real-time in vivo analysis of T cell activation in the central nervous system using a genetically encoded calcium indicator. Nat. Med. 19 (6), 778–783. Mulligan, S.J., MacVicar, B.A., 2004. Calcium transients in astrocyte endfeet cause cerebrovascular constrictions. Nature 431, 195–199. Murmu, R.P., Li, W., Holtmaat, A., Li, J.-Y., 2013. Dendritic spine instability leads to progressive neocortical spine loss in a mouse model of Huntington’s disease. J. Neurosci. 33, 12997–13009. Nakase, H., Kempski, O.S., Heimann, A., Takeshima, T., Tintera, J., 1997. Microcirculation after cerebral venous occlusions as assessed by laser Doppler scanning. J. Neurosurg. 87, 307–314. Nawashiro, H., Brenner, M., Fukui, S., Shima, K., Hallenbeck, J.M., 2000. High susceptibility to cerebral ischemia in GFAP-null mice. J. Cereb. Blood Flow Metab. 20, 1040–1044. Ndubuizu, O., LaManna, J.C., 2007. Brain tissue oxygen concentration measurements. Antioxid. Redox Signal. 9, 1207–1219. Nedergaard, M., Ransom, B., Goldman, S.A., 2003. New roles for astrocytes: redefining the functional architecture of the brain. Trends Neurosci. 26, 523–530. Niesner, R., Andresen, V., Neumann, J., Spiecker, H., Gunzer, M., 2007. The power of single and multibeam two-photon microscopy for high-resolution and highspeed deep tissue and intravital imaging. Biophys. J. 93, 2519–2529. Nikolenko, V., Watson, B.O., Araya, R., Woodruff, A., Peterka, D.S., Yuste, R., 2008. SLM microscopy: scanless two-photon imaging and photostimulation with spatial light modulators. Front. Neural Circuits 2, 5. Nimmagadda, A., Park, H.-P., Prado, R., Ginsberg, M.D., 2008. Albumin therapy improves local vascular dynamics in a rat model of primary microvascular thrombosis: a two-photon laser-scanning microscopy study. Stroke 39, 198–204. Nimmerjahn, A., Kirchhoff, F., Helmchen, F., 2005. Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science 308, 1314–1318. Nimmerjahn, A., Kirchhoff, F., Kerr, J.N.D., Helmchen, F., 2004. Sulforhodamine 101 as a specific marker of astroglia in the neocortex in vivo. Nat. Methods 1, 31–37. Nimmerjahn, A., Mukamel, E.A., Schnitzer, M.J., 2009. Motor behavior activates Bergmann glial networks. Neuron 62, 400–412. Nishimura, N., Schaffer, C.B., Friedman, B., Lyden, P.D., Kleinfeld, D., 2007. Penetrating arterioles are a bottleneck in the perfusion of neocortex. Proc. Natl. Acad. Sci. U. S. A. 104, 365–370. Nishimura, N., Schaffer, C.B., Friedman, B., Tsai, P.S., Lyden, P.D., Kleinfeld, D., 2006. Targeted insult to subsurface cortical blood vessels using ultrashort laser pulses: three models of stroke. Nat. Methods 3, 99–108. Odoardi, F., Kawakami, N., Klinkert, W.E.F., Wekerle, H., Flu¨gel, A., 2007. Bloodborne soluble protein antigen intensifies T cell activation in autoimmune CNS lesions and exacerbates clinical disease. Proc. Natl. Acad. Sci. U. S. A. 104, 18625–18630. Oh, S.W., Harris, J.A., Ng, L., Winslow, B., Cain, N., Mihalas, S., Wang, Q., Lau, C., Kuan, L., Henry, A.M., Mortrud, M.T., Ouellette, B., Nguyen, T.N., Sorensen, S.A., Slaughterbeck, C.R., Wakeman, W., Li, Y., Feng, D., Ho, A., Nicholas, E., Hirokawa, K.E., Bohn, P., Joines, K.M., Peng, H., Hawrylycz, M.J., Phillips, J.W., Hohmann, J.G., Wohnoutka, P., Gerfen, C.R., Koch, C., Bernard, A., Dang, C., Jones, A.R., Zeng, H., 2014. A mesoscale connectome of the mouse brain. Nature 508, 207–214. Ohsawa, K., Kohsaka, S., 2011. Dynamic motility of microglia: purinergic modulation of microglial movement in the normal and pathological brain. Glia 59, 1793–1799.

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54 Orlando, C., Ster, J., Gerber, U., Fawcett, J.W., Raineteau, O., 2012. Perisynaptic chondroitin sulfate proteoglycans restrict structural plasticity in an integrindependent manner. J. Neurosci. 32, 18009–18017, 18017a. Orr, A.G., Orr, A.L., Li, X.-J., Gross, R.E., Traynelis, S.F., 2009. Adenosine A(2A) receptor mediates microglial process retraction. Nat. Neurosci. 12, 872–878. Orsini, F., Villa, P., Parrella, S., Zangari, R., Zanier, E.R., Gesuete, R., Stravalaci, M., Fumagalli, S., Ottria, R., Reina, J.J., Paladini, A., Micotti, E., Ribeiro-Viana, R., Rojo, J., Pavlov, V.I., Stahl, G.L., Bernardi, A., Gobbi, M., De Simoni, M.-G., 2012. Targeting mannose binding lectin confers long lasting protection with a surprisingly wide therapeutic window in cerebral ischemia. Circulation. Ortolano, F., Maffia, P., Dever, G., Rodolico, G., Millington, O.R., De Simoni, M.G., Brewer, J.M., Bushell, T.J., Garside, P., Carswell, H.V., 2010. Advances in imaging of new targets for pharmacological intervention in stroke: real-time tracking of T-cells in the ischaemic brain. Br. J. Pharmacol. 159, 808–811. Otsu, Y., Bormuth, V., Wong, J., Mathieu, B., Dugue´, G.P., Feltz, A., Dieudonne´, S., 2008. Optical monitoring of neuronal activity at high frame rate with a digital random-access multiphoton (RAMP) microscope. J. Neurosci. Methods 173, 259–270. Owens, T., Bechmann, I., Engelhardt, B., 2008. Perivascular spaces and the two steps to neuroinflammation. J. Neuropathol. Exp. Neurol. 67, 1113–1121. Pai, S., Danne, K.J., Qin, J., Cavanagh, L.L., Smith, A., Hickey, M.J., Weninger, W., 2012. Visualizing leukocyte trafficking in the living brain with 2-photon intravital microscopy. Front. Cell Neurosci. 6, 67. Parpaleix, A., Houssen, Y.G., Charpak, S., 2013. Imaging local neuronal activity by monitoring PO2 transients in capillaries. Nat. Med. 19, 241–246. Peng, X., Carhuapoma, J.R., Bhardwaj, A., Alkayed, N.J., Falck, J.R., Harder, D.R., Traystman, R.J., Koehler, R.C., 2002. Suppression of cortical functional hyperemia to vibrissal stimulation in the rat by epoxygenase inhibitors. Am. J. Physiol. Heart Circ. Physiol. 283, H2029–H2037. Peng, X., Zhang, C., Alkayed, N.J., Harder, D.R., Koehler, R.C., 2004. Dependency of cortical functional hyperemia to forepaw stimulation on epoxygenase and nitric oxide synthase activities in rats. J. Cereb. Blood Flow Metab. 24, 509–517. Peppiatt, C.M., Howarth, C., Mobbs, P., Attwell, D., 2006. Bidirectional control of CNS capillary diameter by pericytes. Nature 443, 700–704. Perego, C., Fumagalli, S., De Simoni, M.-G., 2011. Temporal pattern of expression and colocalization of microglia/macrophage phenotype markers following brain ischemic injury in mice. J. Neuroinflammation 8, 174. Pinard, E., Nallet, H., MacKenzie, E.T., Seylaz, J., Roussel, S., 2002. Penumbral microcirculatory changes associated with peri-infarct depolarizations in the rat. Stroke 33, 606–612. Pizzorusso, T., Medini, P., Landi, S., Baldini, S., Berardi, N., Maffei, L., 2006. Structural and functional recovery from early monocular deprivation in adult rats. Proc. Natl. Acad. Sci. U. S. A. 103, 8517–8522. Porta, C., Rimoldi, M., Raes, G., Brys, L., Ghezzi, P., Di Liberto, D., Dieli, F., Ghisletti, S., Natoli, G., De Baetselier, P., Mantovani, A., Sica, A., 2009. Tolerance and M2 (alternative) macrophage polarization are related processes orchestrated by p50 nuclear factor kappaB. Proc. Natl. Acad. Sci. U. S. A. 106, 14978–14983. Puro, D.G., 2007. Physiology and pathobiology of the pericyte-containing retinal microvasculature: new developments. Microcirculation 14, 1–10. Raichle, M.E., Mintun, M.A., 2006. Brain work and brain imaging. Annu. Rev. Neurosci. 29, 449–476. Ramprasad, M.P., Terpstra, V., Kondratenko, N., Quehenberger, O., Steinberg, D., 1996. Cell surface expression of mouse macrosialin and human CD68 and their role as macrophage receptors for oxidized low density lipoprotein. Proc. Natl. Acad. Sci. U. S. A. 93, 14833–14838. Ramsauer, M., Krause, D., Dermietzel, R., 2002. Angiogenesis of the blood– brain barrier in vitro and the function of cerebral pericytes. FASEB J. 16, 1274–1276. Reddy, G.D., Saggau, P., 2005. Fast three-dimensional laser scanning scheme using acousto-optic deflectors. J. Biomed. Opt. 10, 064038. Rosidi, N.L., Zhou, J., Pattanaik, S., Wang, P., Jin, W., Brophy, M., Olbricht, W.L., Nishimura, N., Schaffer, C.B., 2011. Cortical microhemorrhages cause local inflammation but do not trigger widespread dendrite degeneration. PLoS ONE 6, e26612. Rush, C.M., Millington, O.R., Hutchison, S., Bryson, K., Brewer, J.M., Garside, P., 2009. Characterization of CD4+ T-cell-dendritic cell interactions during secondary antigen exposure in tolerance and priming. Immunology 128, 463–471. Sacconi, L., Mapelli, J., Gandolfi, D., Lotti, J., O’Connor, R.P., D’Angelo, E., Pavone, F.S., 2008. Optical recording of electrical activity in intact neuronal networks with random access second-harmonic generation microscopy. Opt. Express 16, 14910–14921. Santosh, C., Brennan, D., McCabe, C., Macrae, I.M., Holmes, W.M., Graham, D.I., Gallagher, L., Condon, B., Hadley, D.M., Muir, K.W., Gsell, W., 2008. Potential use of oxygen as a metabolic biosensor in combination with T2*-weighted MRI to define the ischemic penumbra. J. Cereb. Blood Flow Metab. 28, 1742–1753. Schaffer, C.B., 2010. Optical tools to produce and study small strokes in animal models. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2010, 3377–3378. Schaffer, C.B., Friedman, B., Nishimura, N., Schroeder, L.F., Tsai, P.S., Ebner, F.F., Lyden, P.D., Kleinfeld, D., 2006. Two-photon imaging of cortical surface microvessels reveals a robust redistribution in blood flow after vascular occlusion. PLoS Biol. 4, e22. Schiera, G., Bono, E., Raffa, M.P., Gallo, A., Pitarresi, G.L., Di Liegro, I., Savettieri, G., 2003. Synergistic effects of neurons and astrocytes on the differentiation of brain capillary endothelial cells in culture. J. Cell. Mol. Med. 7, 165–170. Scholz, M., Cinatl, J., Scha¨del-Ho¨pfner, M., Windolf, J., 2007. Neutrophils and the blood–brain barrier dysfunction after trauma. Med. Res. Rev. 27, 401–416.

53

Scott, B.B., Brody, C.D., Tank, D.W., 2013. Cellular resolution functional imaging in behaving rats using voluntary head restraint. Neuron 80, 371–384. Shakhar, G., Lindquist, R.L., Skokos, D., Dudziak, D., Huang, J.H., Nussenzweig, M.C., Dustin, M.L., 2005. Stable T cell-dendritic cell interactions precede the development of both tolerance and immunity in vivo. Nat. Immunol. 6, 707–714. Shepro, D., Morel, N.M., 1993. Pericyte physiology. FASEB J. 7, 1031–1038. Shichita, T., Sugiyama, Y., Ooboshi, H., Sugimori, H., Nakagawa, R., Takada, I., Iwaki, T., Okada, Y., Iida, M., Cua, D.J., Iwakura, Y., Yoshimura, A., 2009. Pivotal role of cerebral interleukin-17-producing gammadeltaT cells in the delayed phase of ischemic brain injury. Nat. Med. 15, 946–950. Shih, A.Y., Driscoll, J.D., Drew, P.J., Nishimura, N., Schaffer, C.B., Kleinfeld, D., 2012. Two-photon microscopy as a tool to study blood flow and neurovascular coupling in the rodent brain. J. Cereb. Blood Flow Metab.. Shih, A.Y., Friedman, B., Drew, P.J., Tsai, P.S., Lyden, P.D., Kleinfeld, D., 2009. Active dilation of penetrating arterioles restores red blood cell flux to penumbral neocortex after focal stroke. J. Cereb. Blood Flow Metab. 29, 738–751. Sigler, A., Murphy, T.H., 2010. In vivo 2-photon imaging of fine structure in the rodent brain: before, during, and after stroke. Stroke 41, S117–S123. Silvestri, L., Allegra Mascaro, A.L., Costantini, I., Sacconi, L., Pavone, F.S., 2013. Correlative two-photon and light sheet microscopy. Methods 66 (2), 268–272. Silvestri, L., Bria, A., Sacconi, L., Iannello, G., Pavone, F.S., 2012. Confocal light sheet microscopy: micron-scale neuroanatomy of the entire mouse brain. Opt. Express 20, 20582–20598. Soeller, C., Cannell, M.B., 1999. Two-photon microscopy: imaging in scattering samples and three-dimensionally resolved flash photolysis. Microsc. Res. Tech. 47, 182–195. Soltys, Z., Orzylowska-Sliwinska, O., Zaremba, M., Orlowski, D., Piechota, M., Fiedorowicz, A., Janeczko, K., Oderfeld-Nowak, B., 2005. Quantitative morphological study of microglial cells in the ischemic rat brain using principal component analysis. J. Neurosci. Methods 146, 50–60. Stefanovic, B., Hutchinson, E., Yakovleva, V., Schram, V., Russell, J.T., Belluscio, L., Koretsky, A.P., Silva, A.C., 2008. Functional reactivity of cerebral capillaries. J. Cereb. Blood Flow Metab. 28, 961–972. Strbian, D., Durukan, A., Pitkonen, M., Marinkovic, I., Tatlisumak, E., Pedrono, E., Abo-Ramadan, U., Tatlisumak, T., 2008. The blood–brain barrier is continuously open for several weeks following transient focal cerebral ischemia. Neuroscience 153, 175–181. Svoboda, K., Yasuda, R., 2006. Principles of two-photon excitation microscopy and its applications to neuroscience. Neuron 50, 823–839. Takano, T., Han, X., Deane, R., Zlokovic, B., Nedergaard, M., 2007. Two-photon imaging of astrocytic Ca2+ signaling and the microvasculature in experimental mice models of Alzheimer’s disease. Ann. N. Y. Acad. Sci. 1097, 40–50. Tasdemiroglu, E., Macfarlane, R., Wei, E.P., Kontos, H.A., Moskowitz, M.A., 1992. Pial vessel caliber and cerebral blood flow become dissociated during ischemiareperfusion in cats. Am. J. Physiol. 263, H533–H536. The quest for quantitative microscopy, 2012. Nat. Methods 9, 627. Theer, P., Denk, W., 2006. On the fundamental imaging-depth limit in two-photon microscopy. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 23, 3139–3149. Tomek, J., Novak, O., Syka, J., 2013. Two-photon processor and SeNeCA: a freely available software package to process data from two-photon calcium imaging at speeds down to several milliseconds per frame. J. Neurophysiol. 110, 243–256. Tsai, P.S., Kaufhold, J.P., Blinder, P., Friedman, B., Drew, P.J., Karten, H.J., Lyden, P.D., Kleinfeld, D., 2009. Correlations of neuronal and microvascular densities in murine cortex revealed by direct counting and colocalization of nuclei and vessels. J. Neurosci. 29, 14553–14570. Vinet, J., Weering, H.R.J., van Heinrich, A., Ka¨lin, R.E., Wegner, A., Brouwer, N., Heppner, F.L., Rooijen, Nv., van Boddeke, H.W., Biber, K., 2012. Neuroprotective function for ramified microglia in hippocampal excitotoxicity. J. Neuroinflammation 9, 27. Wake, H., Moorhouse, A.J., Jinno, S., Kohsaka, S., Nabekura, J., 2009. Resting microglia directly monitor the functional state of synapses in vivo and determine the fate of ischemic terminals. J. Neurosci. 29, 3974–3980. Wang, C.H., Popel, A.S., 1993. Effect of red blood cell shape on oxygen transport in capillaries. Math. Biosci. 116, 89–110. Xu, H.-T., Pan, F., Yang, G., Gan, W.-B., 2007. Choice of cranial window type for in vivo imaging affects dendritic spine turnover in the cortex. Nat. Neurosci. 10, 549–551. Xu, T., Yu, X., Perlik, A.J., Tobin, W.F., Zweig, J.A., Tennant, K., Jones, T., Zuo, Y., 2009. Rapid formation and selective stabilization of synapses for enduring motor memories. Nature 462, 915–919. Yang, G., Pan, F., Gan, W.-B., 2009. Stably maintained dendritic spines are associated with lifelong memories. Nature 462, 920–924. Yang, G., Pan, F., Parkhurst, C.N., Grutzendler, J., Gan, W.-B., 2010. Thinned-skull cranial window technique for long-term imaging of the cortex in live mice. Nat. Protoc. 5, 201–208. Yang, Y., Estrada, E.Y., Thompson, J.F., Liu, W., Rosenberg, G.A., 2007. Matrix metalloproteinase-mediated disruption of tight junction proteins in cerebral vessels is reversed by synthetic matrix metalloproteinase inhibitor in focal ischemia in rat. J. Cereb. Blood Flow Metab. 27, 697–709. Yilmaz, G., Arumugam, T.V., Stokes, K.Y., Granger, D.N., 2006. Role of T lymphocytes and interferon-gamma in ischemic stroke. Circulation 113, 2105–2112. Zanier, E.R., Montinaro, M., Vigano, M., Villa, P., Fumagalli, S., Pischiutta, F., Longhi, L., Leoni, M.L., Rebulla, P., Stocchetti, N., Lazzari, L., De Simoni, M.-G., 2011. Human umbilical cord blood mesenchymal stem cells protect mice brain after trauma. Crit. Care Med. 39, 2501–2510.

54

S. Fumagalli et al. / Progress in Neurobiology 121 (2014) 36–54

Zanier, E.R., Pischiutta, F., Villa, P., Paladini, A., Montinaro, M., Micotti, E., Orru`, A., Cervo, L., De Simoni, M.G., 2013. Six-month ischemic mice show sensorimotor and cognitive deficits associated with brain atrophy and axonal disorganization. CNS Neurosci. Ther. 19, 695–704. Zenaro, E., Rossi, B., Angiari, S., Constantin, G., 2013. Use of imaging to study leukocyte trafficking in the central nervous system. Immunol. Cell Biol. 91, 271–280. Zenker, D., Begley, D., Bratzke, H., Ru¨bsamen-Waigmann, H., von Briesen, H., 2003. Human blood-derived macrophages enhance barrier function of cultured primary bovine and human brain capillary endothelial cells. J. Physiol. (Lond.) 551, 1023–1032. Zhang, S., Boyd, J., Delaney, K., Murphy, T.H., 2005. Rapid reversible changes in dendritic spine structure in vivo gated by the degree of ischemia. J. Neurosci. 25, 5333–5338. Zhang, S., Murphy, T.H., 2007. Imaging the impact of cortical microcirculation on synaptic structure and sensory-evoked hemodynamic responses in vivo. PLoS Biol. 5, e119.

Zhang, Z.G., Zhang, L., Jiang, Q., Zhang, R., Davies, K., Powers, C., Bruggen, N.V., Chopp, M., 2000. VEGF enhances angiogenesis and promotes blood–brain barrier leakage in the ischemic brain. J. Clin. Invest. 106, 829–838. Zhao, W., Belayev, L., Ginsberg, M.D., 1997. Transient middle cerebral artery occlusion by intraluminal suture: II. Neurological deficits, and pixel-based correlation of histopathology with local blood flow and glucose utilization. J. Cereb. Blood Flow Metab. 17, 1281–1290. Zinselmeyer, B.H., Dempster, J., Gurney, A.M., Wokosin, D., Miller, M., Ho, H., Millington, O.R., Smith, K.M., Rush, C.M., Parker, I., Cahalan, M., Brewer, J.M., Garside, P., 2005. In situ characterization of CD4+ T cell behavior in mucosal and systemic lymphoid tissues during the induction of oral priming and tolerance. J. Exp. Med. 201, 1815–1823. Zlokovic, B.V., 2005. Neurovascular mechanisms of Alzheimer’s neurodegeneration. Trends Neurosci. 28, 202–208. Zonta, M., Angulo, M.C., Gobbo, S., Rosengarten, B., Hossmann, K.-A., Pozzan, T., Carmignoto, G., 2003. Neuron-to-astrocyte signaling is central to the dynamic control of brain microcirculation. Nat. Neurosci. 6, 43–50.