International Journal of Neuroscience, 2015; 125(2): 91–99 Copyright © 2014 Informa Healthcare USA, Inc. ISSN: 0020-7454 print / 1543-5245 online DOI: 10.3109/00207454.2014.914511
REVIEW
Neurolight – astonishing advances in brain imaging 1 Ewa Rojczyk-Golebiewska, Artur Palasz,1 John J. Worthington,2 Grzegorz Markowski,3 and Ryszard Wiaderkiewicz1
Department of Histology, Medical University of Silesia, 18 Medyk´ow Street, Katowice, Poland; 2 Manchester Immunology Group, University of Manchester, United Kingdom; 3 Department of Glottodidactic and Distance Learning, University of Silesia, Katowice, Poland
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
1
In recent years, significant advances in basic neuroanatomical studies have taken place. Moreover, such classical, clinically-oriented human brain imaging methods such as MRI, PET and DTI have been applied to small laboratory animals allowing improvement in current experimental neuroscience. Contemporary structural neurobiology also uses various technologies based on fluorescent proteins. One of these is optogenetics, which integrates physics, genetics and bioengineering to enable temporal precise control of electrical activity of specific neurons. Another important challenge in the field is the accurate imaging of complicated neural networks. To address this problem, three-dimensional reconstruction techniques and retrograde labeling with modified viruses has been developed. However, a revolutionary step was the invention of the “Brainbow” system, utilizing gene constructs including the sequences of fluorescent proteins and the usage of Cre recombinase to create dozens of colour combinations, enabling visualization of neurons and their connections in extremely high resolution. Furthermore, the newly- introduced CLARITY method should make it possible to visualize threedimensionally the structure of translucent brain tissue using the hydrogel polymeric network. This original technique is a big advance in neuroscience creating novel viewpoints completely different than standard glass slide immunostaining. KEYWORDS: brainbow, clarity, fluorescence, imaging
Introduction Neuroanatomy is often considered to be a classical descriptive branch of science with limited scope to develop. Indeed, the vast majority of brain structures were precisely and correctly described in the 19th century by the “fathers” of microscopical neuroanatomy such as Santiago Ramon y Cajal (1852–1934) and Camillo Golgi (1843–1926). They invented new techniques of brain tissue impregnation with silver, chrome and gold salts, which allowed visualization of the spatial architecture of dendrites, axons as well as glial cells. An outstanding, German anatomist, Otto Karl Deiters (1834–1863) – is also worth mentioning as he was the author of extremely accurate brain cytoarchitectonic atlases, before his untimely death at the age of 29. Contemporary studies focus not only on the morphological aspects of brain structure, but also try to investigate functions of particReceived 22 December 2013; revised 27 March 2014; accepted 9 April 2014 Correspondence: Ewa Rojczyk-Golebiewska, Department of Histology, ´ Street, 40-752 Katowice, Poland. Medical University of Silesia, 18 Medykow Tel: +48-697-888-440. E-mail: [email protected]
ular neuronal formations as precisely as possible. For this reason in vivo experiments based on animal models currently play an essential role in neuroscience and modern functional neuroanatomy seeks to utilize recent advances in biochemistry, biophysics and biotechnology.
Classical methods of brain imaging The most common methods of brain imaging are nuclear magnetic resonance technique (called magnetic resonance imaging – MRI) and positron emission tomography (PET). Currently, MRI and functional MRI –(fMRI) are standard diagnostic methods used in clinical practice. Classical fMRI is based on detection of brain hemodynamic changes (related to the activity of certain brain regions), blood oxygenation level-dependent (BOLD) changes and cerebral blood flow (CBF) / cerebral blood volume (CBV) fluctuations in brain structures [1]. BOLD imaging, which detects increasing oxygen consumption by neurons is imprecise, as it does not take electrophysiological background into consideration. 91
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
92
E. Rojczyk-Golebiewska et al.
Thus, modification of resonance imaging has been developed, called manganese enhanced MRI (MEMRI), which takes advantage of ion channel dependent neuronal accumulation of manganese indicating bioelectrical cell activity. Moreover, transsynaptic translocation of the manganese gradient allows the creation of a map of neuronal connections [2,3]. PET images are formed by means of radionuclide labeled substances with ability to emit positrons. As a result of non-invasive scanning of fading photons and tomographic technique application, three-dimensional images of isotope distribution are obtained. An important advantage of PET is the possibility to completely and precisely determine the probe distribution in time and space in high resolution – even 1 μm [4]. In recent years, thanks to device miniaturization and the increase in magnetic field intensity (up to 7–15 Tesla) it became possible to use fMRI in studies on small laboratory animals. Similarly to fMRI, PET technique is also used not only in clinical trials, but also in experiments on animal models, especially in studies of reparation changes in CNS, neuroplasticity, receptor physiology, enzymatic activity and in gene expression analysis. Another interesting method of brain structure visualization is diffusion tensor imaging (DTI). It measures water molecules diffusion, which is usually faster along axonal tracts of white matter in comparison to other directions. It is a simple consequence of brain structure anisotropy. Information from DTI can be digitally processed using so called tractography technique, which enables the user to obtain spatial, colour-coded brain maps showing three-dimensional neuronal fibre architectonics [4].
Fluorescent techniques in neurobiology Discovery of green fluorescent protein (GFP) derived from the jellyfish Aequorea victoria and progressive formation of its various spectral variants enabled a permanent cell labeling by expressing fluorescent proteins (FPs) in specific cell types. During gene transfer procedures, DNA sequences encoding FPs were usually attached to other DNA sequences in order to create fusion proteins after translation. Consequently, GFP has become a most common marker of gene expression and a tool for protein targeting [5,6]. However, there are also many GFP-like proteins derived from different organisms as well as other proteins emitting light or responding to it [5].
plication of fluorescent light is optogenetics. It integrates optics, genetics and bioengineering to enable temporalprecise control and modifications of neuronal activity [7]. In this method, rhodopsins (transmembrane proteins bound to retinal, found in various microorganisms) act as molecular sensitizers responding to light when isolated from microbial species and expressed in neurons, they respond to light in different manners, dependent on the type of protein. Halorhodopsins mediate neuron silencing by hyperpolarization in response to yellow light, whereas channelrhodopsins have opposing effects after stimulation with blue light, as they activate neurons by membrane depolarization. Finally, archearhodopsins have the ability to respond to green and yellow light, causing powerful neuron silencing [7–9]. In classical optogenetic technology, a rhodopsinencoding gene construct is inserted into specific cells. After viral vector delivery into animal brain, rhodopsin expression has to be checked with the use of a GFP marker. If cell transduction is successful, the last step of the procedure can be performed – optical electrode attachment (to a certain brain region) and light delivery [8].
Three-dimensional reconstruction The three-dimensional (3D) visualization of labeled structures within a large biological specimen is a huge challenge in neuroimaging. It requires both the progress in microscopic techniques and advances in tissue processing. Development of two-photon microscopy (in combination with fluorescent activity reporters) was a big step forward, as it uses relatively long wavelengths that are less sensitive to scattering. Moreover, in twophoton microscopy signal-to-noise ratio of the emitted fluorescent signal is higher in comparison to singlephoton excitation [10]. Sometimes it is convenient to combine optical fluorescence with electron microscopy in order to reconstruct brain molecular architecture in high resolution. To address this problem, array tomography has been developed which enables simultaneous imaging of serial ultrathin tissue sections collected on a single glass slide. In this approach, immunostaining has to be performed before data processing, so there is a possibility to detect specific antigens and to determine their spatial distribution in the large volume of tissue [5,11]. Even better resolution is provided with the use of serial block-face scanning electron microscopy – as it enables tracing of the thinnest axons and identification of synapses [5,12].
Optogenetics
Retrograde labeling
One of the interesting approaches in neurobiological studies, which require genetic modification and the ap-
Another approach used to visualize complicated neuronal network is to take advantage of labeling International Journal of Neuroscience
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Neurolight
substances, such as fluorescent dyes or proteins, which are absorbed by neurons and transported through axons. In case of retrograde transport, it is possible to label neurons projecting their axons to the region of former substance delivery. The combination of two retrograde labels (for example dyes absorbing different fluorescent wavelenghts) is called double retrograde labeling and is particularly useful for studying brain structures possessing many projections to different brain regions. This technique enables a precise insight into afferent and efferent connection structure [13]. Apart from conventional chemical labels, viruses containing marker proteins can be used to visualize neuronal connections. They have the ability to move between neurons in the synaptic way and to replicate inside infected cells, which considerably increases the signal. Sometimes, in order to obtain more detailed images of complex neural pathways, a combination of double retrograde and viral labeling is performed. In this method, two strains of recombinant viruses are used, each of them expressing unique marker protein. Consequently, neurons belonging to both neuronal loops show co-expression of the labels [13]. Among the numerous neurotropic viruses used for transsynaptic labeling, rabies virus (RV) is particularly worth mentioning. It can be transported only in retrograde manner and its propagation is restricted to neurons directly connected with the infected cell (without affecting neighbouring neurons) [14].
“Brainbow” The general anatomical structure and physiology of the brain is already well known, mainly due to the development of the aforementioned techniques utilizing fluorescent proteins. However, the true structure of this organ is even more complex when considering that it is connected with a multitude of synapses that are often not stable in time and space. This fact gave rise to connectomics – a brand new branch of neurobiology focusing on synapses and their high resolution imaging.
The beginnings of the “Brainbow” One of the first scientists that understood the need to create a time variable, exact map of neural wiring was Professor Jeff Lichtman from the Center for Brain Science at Harvard University [15,16]. He was interested in synaptic plasticity, which was proven by the discovery of phenomena called synapse elimination and synaptic takeover, that are common during vertebrate nervous system development. There is a reduction of an initially huge number of axonal connections formed in the im C
2014 Informa Healthcare USA, Inc.
93
mature organism with some being replaced by others in a kind of competition between axons. This process is most evident and particularly well studied at neuromuscular junctions, so in the mid 1980s Lichtman’s group decided to make the first attempts to visualize synaptic plasticity exactly on this part of the nervous system. They described a technique called “activity-mediated uptake of fluorescent probes” to label terminals of individual axons in developing twitch muscle fibers. After electrical stimulation of axons, different fluorescent probes were taken up in a classical way similar to neurotransmitter reuptake in synapses. As a result, each neuron was labeled by a traceable and unique colour. However, this technique required the complex process of simultaneous stimulation of all neurons. [16,17]. Firstly, Lichtman generated lines of transgenic mice expressing three different fluorescent proteins (RFP, GFP and YFP, together termed XFP) in order to visualize synapse formation by motoneurons. Neuro-specific elements of the thy-1 gene were chosen as a promoter site, as they were proven to generate the strongest reporter gene expression in the cells of interest. The results of these attempts were surprising, because patterns of transgene expression in 25 independently generated transgenic lines (incorporating identical regulatory elements) were diverse. Meaning that some mouse strains expressed XFP in almost all neural cells, whereas others showed fluorescence in only a small percentage of cells. The next step was to cross the animal lines with one another, which resulted in the new strains expressing two or three XFPs in different subsets of neurons. Some cells showed co-expression of fluorescent proteins causing formation of several new colours [16,18]. The possibility to cross different transgenic mouse lines gave opportunity to create specific progeny expressing particular XFPs in particular neural cell types. This kind of manipulation enables a complete reconstruction of synaptic connections in the chosen brain regions. Importantly, XFP expression in cells turned out to be stable which facilitated tracing of individual cell fate over time. Thus, in the first years of the 21st century Lichtman’s lab tried to express different fluorescent proteins in small subsets of motor axons in order to image axons withdrawal and expansion at postsynaptic sites. These studies revealed that synaptic competition is a random process (without obvious spatial bias), quite changeable in time and that not all postsynaptic sites are re-occupied after axon withdrawal [19,20]. The following years provided further advances in time-lapse imaging of neural cells using XFPs. High resolution confocal laser scanning microscopy enabled the visualization of small structures like retinal ganglion cells (RGCs) in detail, showing not only cell bodies, but also axons and whole dendritic arbors. This has a potential
94
E. Rojczyk-Golebiewska et al.
application in studies of normal retinal development and in eye disease modeling [21].
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Creating connectomic map To make an authentic, detailed map of neuronal circuitry, there was a need to use more than several basic colours for fluorescent labeling of neurons. Initially, scientists tried to achieve it by creating chimeric mice, by crossing mouse lines or initiating crossing over process. Nevertheless, these methods were time consuming, cumbersome and generated only a limited number of fluorescent shades [15]. Jean Livet – a member of Lichtman’s lab since 2002 – invented a new technique to solve this problem [16]. It was based on well-known DNA-engineering system called Cre/lox recombination which is often used to create knockout animals and to carry out deletions, insertions, translocations and inversions at specific sites of DNA. Cre is a recombinase, that performs DNA excision or inversion at the sites flanked by lox sequences. Its activation is usually obtained either by tamoxifen administration (in case of so called tamoxifen-inducible CreER recombinases) or by crossing the mouse possessing lox sites with a mouse showing constitutive expression of the Cre enzyme. Livet’s general idea was to create genetic constructs with several XFP genes flanked by multiple lox sites. In such constructs only genes directly following the promoter can be expressed. Then, activated Cre recombinase excises/inverts DNA between one pair of lox sequences. Importantly, there are multiple incompatibile lox site variants (for example loxP, loxN, lox2272). If they are inserted into one construct, Cre recombinase is somehow forced to choose between two mutually exclusive excision events, so only one possibility of recombination occurs in one cell. As a choice of excision/inversion site is made randomly in each cell, this manipulation will result in neural cells labeled in different shades (Figure 1). The more different XFP genes and lox site variants are used, the more colours can be potentially obtained [15,16]. After examination of the first generation of these transgenic mice, it turned out, that some mouse lines had incorporated more than one construct copy. It enabled independent recombination of each copy (so called combinatorial XFP expression), that resulted in multitude of colours obtained. That is why Prof. J.Lichtman named this interesting and simple technique “Brainbow” [16]. In case of three transgene copies, it is possible to obtain ten distinct colours (Figure 2) and the addition of further copies generates an enormous number of shades – upwards of a hundred. As a result, using fluorescent microscopy, we can generate colourful images of different brain regions (Figure 3). In order
to control the number of “Brainbow” construct tandem repeats, researchers used another recombination system (analogous to Cre/lox) called Flp/FRT. Insertion of FRT sites into appropriate place in the DNA sequence allowed excision of excessive transgene copies after Flp recombinase activation [15].
Examples of “Brainbow” applications After invention of “Brainbow”, many efforts focused on generation of multiple transgenic mice lines in order to test this system for different applications. Sometimes it is convenient to restrict recombination process to specific cell types or brain regions. Another interesting idea is to use “Brainbow” expressing mice as a tool for mapping neural circuits. Researchers succeeded in visualizing granular cells of the inner granular layer (IGL) of the cerebellum together with the mossy fibres innervating these cells. Confocal microscopy accompanied by computer-assisted methods enabled us to distinguish and trace individual axons forming the exact map of connections. This study revealed that cerebellar granular cells are polyneuronally innervated by mossy fibres [15]. In conclusion, “Brainbow” represents an enormously promising tool for the visualization of the nervous system in nanoscale and is considered as a true milestone on the way to create the first full “connectomic” map of the brain.
CLARITY Classical methods of brain imaging enable visualization of small volumes of tissue, typically a few micrometerthin microscopic sections. Detailed reconstruction of three-dimensional brain structure was so far possible only by means of computer data consolidation methods, but it was quite complicated and not applicable to immunohistochemistry because of limited antibody penetration capacity.
Optical clearing methods Three-dimensional fluorescent brain imaging requires the sufficient transparency of the whole organ. The gross anatomical CNS structures are opaque, therefore chemical clearing is a useful technique to dissolve this problem. The classical, commonly applied clearing medium contains benzyl alcohol and benzyl benzoate (BABB). [22]. BABB is a hydrophobic agent and thus can only be used after brain dehydration. Unfortunately exposure to ethanol and BABB may cause a significant decline or even disappearance of GFP fluorescence. It has been recently suggested that tetrahydrofuran (THF) does not International Journal of Neuroscience
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Neurolight
95
Figure 1. The application of the Cre/lox system to create a colourful map of neural circuitry.
produce this effect and seems to be a better dehydrating agent than ethanol [23]. Moreover, combined dibenzyl ether (DBE) and THF provide more beneficial chemical clearing of brain tissue and preserve fluorescence [23, 24]. Importantly, the clearing process with DBE is also faster than with BABB [25]. The THF-BABBbased clearing technique makes the unsectioned brainstem and spinal cord transparent and appropriate to 3D deep fluorescence analysis. For instance, this promising method enables the precise visualization of the trajectories of regenerating axons. It also quantitatively determines the injury-related glial reaction [26].
CLARITY for creating translucent brain In recent months, brand new clearing methods have been described by scientists in the USA. It is called CLARITY, which was initially an acronym to describe Clear Lipid-exchanged Acrylamide-hybridized Rigid Imaging/Immunostaining/In situ hybridizationcompatible Tissue hYdrogel. Put simply, it creates a C
2014 Informa Healthcare USA, Inc.
translucent brain tissue that is permeable for macromolecules using hydrogel-tissue hybridization and subsequent electrophoretic clearing. This method is so mild, that it enables tissue molecular structure preservation, making it possible to visualize the whole intact, unsectioned brain using microscopes with long working distance objectives [27]. The simple procedure takes about eight-nine days and can be divided into three steps (Figure 4). Firstly, formaldehyde (crosslinker), hydrogel monomers and thermally triggered polymerization initiators are administered to the tissue which causes not only tissue crosslinking, but also covalent linking of biological elements (for example proteins and nucleic acids) to monomers. The temperature is then increased up to 37 degrees Celsius and thermally triggered polymerization takes place resulting in hydrogel-tissue hybrid formation. This gelatinous meshwork is still covalently linked to native biological molecules of the tissue apart from lipids, which lack the necessary reactive groups.
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
96
E. Rojczyk-Golebiewska et al.
The image of mouse dentate gyrus obtained with the use of “Brainbow” system (left) –courtesy of Dr Jeff Lichtman, Harvard University, and combinatorial XFP expression resulting from tandem transgene copies integration (right).
Figure 2.
The last step is electrophoresis in ionic detergent (for example in SDS) called electrophoretic tissue clearing (ETC), in which micelles are actively transported into the tissue, whereas membrane lipids (in these micelles) are extracted. During ETC tissue is placed in the chamber supplied with temperature-controlled buffer circu-
lator and electrophoresed by applying 20–60 volts to the electrodes. As a result, the brain sample becomes transparent, but all structures and biomolecules are well preserved (Figure 5). It makes structural and molecular information accessible for visualization and analysis (27,28).
Figure 3. The “Brainbow” imaging of the rat hippocampus neuronal population - courtesy of Dr Jeff Lichtman, Harvard University.
International Journal of Neuroscience
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Neurolight
Figure 4.
97
Three steps to make the brain translucent–overview of the CLARITY method.
Whole brain imaging and molecular phenotyping with CLARITY
Three-dimensional view of the rat hippocampus showing eYFP (green), parvalbumin (red) and GFAP (blue)expressing neurons with CLARITY method - courtesy of Dr Karl Deisseroth, Stanford University.
Figure 5.
C
2014 Informa Healthcare USA, Inc.
As CLARITY is compatible with most fluorescence microscopy techniques (like confocal, single-photon, multiphoton imaging), it is possible to obtain images of the whole brain expressing fluorescent proteins. Directly before imaging, it is recommended to perform an immersion in refractive-index-specific solutions matching the CLARITY hybrid (for example in 85% glycerol) in order to decrease light scattering occurring due to heterogeneously distributed protein and nucleic acid complexes in the hybrid. It makes the brain uniformly transparent and facilitates high-resolution visualization [27, 28]. CLARITY enables the immunostaining of entire intact tissue on practical timescales of several weeks. It
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
98
E. Rojczyk-Golebiewska et al.
was proven, that electrophoretic clearing of hydrogeltissue hybrids causes only 8% protein loss, a good result in comparison with classical methods of tissue clearing. Moreover, elution of antibodies in multi-round immunostaining (using the same conditions as in ETC) does not cause tissue damage and the diffusion rate of molecular probes deep into intact tissue is quite fast. For example, 2 weeks-long incubation for each antibody is enough to stain 5 mm-thick adult mouse brains. Together with modern microscopic techniques it is possible to obtain multi-colour images presenting specific protein localization in big tissue volumes [27, 28]. In summary, CLARITY provides access to structural and molecular information in large volumes of tissue without need to perform sophisticated computer data consolidation. It may be applied in preclinical comparative studies and disease animal modeling to obtain broad insight into brain structure and molecular features in different conditions.
2.
3.
4.
5. 6. 7.
8.
9.
Conclusions Development of neurobiology is based to a large degree on superseeding imaging methods. Classical techniques are still being improved and used for new purposes, including basic studies performed on small animals and optogenetics, which opens new possibilities in manipulating the bioelectrical activity of single neurons. In recent years, a large emphasis has been placed on recognizing nervous system microstructure, taking synaptic connections architecture under special consideration. That is why traditional brain microstructure imaging methods are gradually displaced by modern methods (like “Brainbow” and CLARITY), which use current achievements of biophysics, biotechnology and molecular biology. Thanks to these advances, it has became possible to accurately visualize nervous system microstructure, which is the first step leading to the true investigation of the relationship between structure and function of particular groups of neurons.
10.
11.
12.
13.
14.
15.
16. 17.
18.
Declaration of Interest
19.
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
20.
21.
References 22. 1. Van der Linden A, Van Camp N, Ramos-Cabrer P, Hoehn M. Current status of functional MRI on small animals: application
to physiology, pathophysiology and cognition. NMR Biomed. 2007;20(5):522–45. Aoki I, Wu YJ, Silva AC, et al. In vivo detection of neuroarchitecture in the rodent brain using manganese-enhanced MRI. NeuroImage 2004;22(3):1046–59. Lizarbe B, Benitez A, Pel´aez Brioso GA, et al. Hypothalamic metabolic compartmentation during appetite regulationas revealed by magnetic resonance imaging and spectroscopy methods. Front Neuroenergetics 2013;13;5:6 doi: 10.3389/fnene.2013.00006. eCollection 2013 Jacobs RE, Cherry SR. Complementary emerging techniques: high-resolution PET and MRI. Curr Opin Neurobiol. 2001;11(5):621–9. Miyawaki A. Fluorescence imaging in the last two decades. Microscopy (Oxf). 2013;62(1):63–8. Tsien RY. The green fluorescent protein. Annu Rev Biochem. 1998;67:509–544 Bernstein JG, Boyden ES. Optogenetic tools for analyzing the neural circuits of behavior. Trends Cogn Sci. 2011;15(12):592–600. Bernstein JG, Garrity PA, Boyden ES. Optogenetics and thermogenetucs: technologies for controlling the activity of targeted cells within intact neural circuits. Curr Opin Neurobiol. 2012;22(1):61–71. Yao JP, Hou WS, Yin ZQ. Optocenetics: a novel optical manipulation tool for medical investigation. Int J Ophthalmol. 2012;5(4):517–22. Bovetti S, Moretti C, Fellin T. Mapping brain circuit function in vivo using two-photon fluorescence microscopy. Microsc Res Tech. 2014 (epub ahead of print). Micheva KD, Smith SJ. Array tomography: a new tool for imaging the molecular architecture and ultrastructure of neural circuits. Neuron. 2007;55(1):25–36 Denk W, Horstmann H. Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. PLoS Biol. 2004;2(11):e329. Ohara S, Inoue K, Witter MP, Iijima T. Untangling neural networks with dual retrograde transsynaptic viral infection. Frontiers Neurosci. 2009;3(3):344–9. Wickersham IR, Lyon DC, Barnard RJ, et al. Monosynaptic Restriction of Transsynaptic Tracing from Single, Genetically Targeted Neurons. Neuron 2007; 53(5):639–47. Livet J, Weissman TA, Kang H, et al. Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature 2007;450(7166):56–62. Lehrer J. Neuroscience: making connections. Nature 2009;457(29):524–7. Lichtman JW, Wilkinson RS, Rich MM. Multiple innervation of tonic endplates revealed by activity-dependent uptake of fluorescent probes. Nature 1985;314(6009):357–9. Feng G, Mellor RH, Bernstein M, et al. Imaging Neuronal Subsets in Transgenic Mice Expressing Multiple Spectral Variants of GFP. Neuron 2000;28(1):41–51. Walsh MK, Lichtman JW. In Vivo Time-Lapse Imaging of Synaptic Takeover Associated with Naturally Occurring Synapse Elimination. Neuron 2003;37(1):67–73. Keller-Peck CR, Walsh MK, Gan WB, et al. Asynchronous Synapse Elimination in Neonatal Motor Units: Studies Using GFP Transgenic Mice. Neuron 2001;31(3):381–94. Walsh MK, Quigley HA. In vivo time-lapse fluorescence imaging of individual retinal ganglion cells in mice. J Neurosci Methods. 2008;169(1):214–21. Becker K, J¨ahrling N, Kramer ER, et al. Ultramicroscopy: 3D reconstruction of large microscopical specimens. J Biophotonics. 2008;1(1):36–42
International Journal of Neuroscience
Neurolight
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
23. Becker K, Jahrling N, Saghafi S, et al. Chemical clearing and dehydratation of GFP expressing mouse brains. PLoS One. 2012;7(3):e33916. 24. Dodt HU, Leischner U, Schierloh A, et al. Ultramicroscopy: three-dimensional visualization of neuronal networks in the whole mouse brain. Nat Methods. 2007;4(4):331–6 25. Viswanath DS, Ghosh T, Prasad D, et al. Viscosity of liquids: theory, estimation, experiment and data. Dordrecht: Springer; c2007
C
2014 Informa Healthcare USA, Inc.
99
26. Erturk A, Mauch CP, Hellal F, et al. Three-dimensional imaging of the unsectioned adult spinal cord to assess axon regeneration and glial responses after injury. Nat Med 2011;18(1):166–71. 27. Chung K, Wallace J, Kim SY, et al. Structural and molecular interrogation of intact biological systems. Nature 2012;497:332–7. 28. Chung K, Deisseroth K. CLARITY for mapping the nervous system. Nat Methods 2013;10(6):508–13.
REVIEW
Neurolight – astonishing advances in brain imaging 1 Ewa Rojczyk-Golebiewska, Artur Palasz,1 John J. Worthington,2 Grzegorz Markowski,3 and Ryszard Wiaderkiewicz1
Department of Histology, Medical University of Silesia, 18 Medyk´ow Street, Katowice, Poland; 2 Manchester Immunology Group, University of Manchester, United Kingdom; 3 Department of Glottodidactic and Distance Learning, University of Silesia, Katowice, Poland
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
1
In recent years, significant advances in basic neuroanatomical studies have taken place. Moreover, such classical, clinically-oriented human brain imaging methods such as MRI, PET and DTI have been applied to small laboratory animals allowing improvement in current experimental neuroscience. Contemporary structural neurobiology also uses various technologies based on fluorescent proteins. One of these is optogenetics, which integrates physics, genetics and bioengineering to enable temporal precise control of electrical activity of specific neurons. Another important challenge in the field is the accurate imaging of complicated neural networks. To address this problem, three-dimensional reconstruction techniques and retrograde labeling with modified viruses has been developed. However, a revolutionary step was the invention of the “Brainbow” system, utilizing gene constructs including the sequences of fluorescent proteins and the usage of Cre recombinase to create dozens of colour combinations, enabling visualization of neurons and their connections in extremely high resolution. Furthermore, the newly- introduced CLARITY method should make it possible to visualize threedimensionally the structure of translucent brain tissue using the hydrogel polymeric network. This original technique is a big advance in neuroscience creating novel viewpoints completely different than standard glass slide immunostaining. KEYWORDS: brainbow, clarity, fluorescence, imaging
Introduction Neuroanatomy is often considered to be a classical descriptive branch of science with limited scope to develop. Indeed, the vast majority of brain structures were precisely and correctly described in the 19th century by the “fathers” of microscopical neuroanatomy such as Santiago Ramon y Cajal (1852–1934) and Camillo Golgi (1843–1926). They invented new techniques of brain tissue impregnation with silver, chrome and gold salts, which allowed visualization of the spatial architecture of dendrites, axons as well as glial cells. An outstanding, German anatomist, Otto Karl Deiters (1834–1863) – is also worth mentioning as he was the author of extremely accurate brain cytoarchitectonic atlases, before his untimely death at the age of 29. Contemporary studies focus not only on the morphological aspects of brain structure, but also try to investigate functions of particReceived 22 December 2013; revised 27 March 2014; accepted 9 April 2014 Correspondence: Ewa Rojczyk-Golebiewska, Department of Histology, ´ Street, 40-752 Katowice, Poland. Medical University of Silesia, 18 Medykow Tel: +48-697-888-440. E-mail: [email protected]
ular neuronal formations as precisely as possible. For this reason in vivo experiments based on animal models currently play an essential role in neuroscience and modern functional neuroanatomy seeks to utilize recent advances in biochemistry, biophysics and biotechnology.
Classical methods of brain imaging The most common methods of brain imaging are nuclear magnetic resonance technique (called magnetic resonance imaging – MRI) and positron emission tomography (PET). Currently, MRI and functional MRI –(fMRI) are standard diagnostic methods used in clinical practice. Classical fMRI is based on detection of brain hemodynamic changes (related to the activity of certain brain regions), blood oxygenation level-dependent (BOLD) changes and cerebral blood flow (CBF) / cerebral blood volume (CBV) fluctuations in brain structures [1]. BOLD imaging, which detects increasing oxygen consumption by neurons is imprecise, as it does not take electrophysiological background into consideration. 91
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
92
E. Rojczyk-Golebiewska et al.
Thus, modification of resonance imaging has been developed, called manganese enhanced MRI (MEMRI), which takes advantage of ion channel dependent neuronal accumulation of manganese indicating bioelectrical cell activity. Moreover, transsynaptic translocation of the manganese gradient allows the creation of a map of neuronal connections [2,3]. PET images are formed by means of radionuclide labeled substances with ability to emit positrons. As a result of non-invasive scanning of fading photons and tomographic technique application, three-dimensional images of isotope distribution are obtained. An important advantage of PET is the possibility to completely and precisely determine the probe distribution in time and space in high resolution – even 1 μm [4]. In recent years, thanks to device miniaturization and the increase in magnetic field intensity (up to 7–15 Tesla) it became possible to use fMRI in studies on small laboratory animals. Similarly to fMRI, PET technique is also used not only in clinical trials, but also in experiments on animal models, especially in studies of reparation changes in CNS, neuroplasticity, receptor physiology, enzymatic activity and in gene expression analysis. Another interesting method of brain structure visualization is diffusion tensor imaging (DTI). It measures water molecules diffusion, which is usually faster along axonal tracts of white matter in comparison to other directions. It is a simple consequence of brain structure anisotropy. Information from DTI can be digitally processed using so called tractography technique, which enables the user to obtain spatial, colour-coded brain maps showing three-dimensional neuronal fibre architectonics [4].
Fluorescent techniques in neurobiology Discovery of green fluorescent protein (GFP) derived from the jellyfish Aequorea victoria and progressive formation of its various spectral variants enabled a permanent cell labeling by expressing fluorescent proteins (FPs) in specific cell types. During gene transfer procedures, DNA sequences encoding FPs were usually attached to other DNA sequences in order to create fusion proteins after translation. Consequently, GFP has become a most common marker of gene expression and a tool for protein targeting [5,6]. However, there are also many GFP-like proteins derived from different organisms as well as other proteins emitting light or responding to it [5].
plication of fluorescent light is optogenetics. It integrates optics, genetics and bioengineering to enable temporalprecise control and modifications of neuronal activity [7]. In this method, rhodopsins (transmembrane proteins bound to retinal, found in various microorganisms) act as molecular sensitizers responding to light when isolated from microbial species and expressed in neurons, they respond to light in different manners, dependent on the type of protein. Halorhodopsins mediate neuron silencing by hyperpolarization in response to yellow light, whereas channelrhodopsins have opposing effects after stimulation with blue light, as they activate neurons by membrane depolarization. Finally, archearhodopsins have the ability to respond to green and yellow light, causing powerful neuron silencing [7–9]. In classical optogenetic technology, a rhodopsinencoding gene construct is inserted into specific cells. After viral vector delivery into animal brain, rhodopsin expression has to be checked with the use of a GFP marker. If cell transduction is successful, the last step of the procedure can be performed – optical electrode attachment (to a certain brain region) and light delivery [8].
Three-dimensional reconstruction The three-dimensional (3D) visualization of labeled structures within a large biological specimen is a huge challenge in neuroimaging. It requires both the progress in microscopic techniques and advances in tissue processing. Development of two-photon microscopy (in combination with fluorescent activity reporters) was a big step forward, as it uses relatively long wavelengths that are less sensitive to scattering. Moreover, in twophoton microscopy signal-to-noise ratio of the emitted fluorescent signal is higher in comparison to singlephoton excitation [10]. Sometimes it is convenient to combine optical fluorescence with electron microscopy in order to reconstruct brain molecular architecture in high resolution. To address this problem, array tomography has been developed which enables simultaneous imaging of serial ultrathin tissue sections collected on a single glass slide. In this approach, immunostaining has to be performed before data processing, so there is a possibility to detect specific antigens and to determine their spatial distribution in the large volume of tissue [5,11]. Even better resolution is provided with the use of serial block-face scanning electron microscopy – as it enables tracing of the thinnest axons and identification of synapses [5,12].
Optogenetics
Retrograde labeling
One of the interesting approaches in neurobiological studies, which require genetic modification and the ap-
Another approach used to visualize complicated neuronal network is to take advantage of labeling International Journal of Neuroscience
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Neurolight
substances, such as fluorescent dyes or proteins, which are absorbed by neurons and transported through axons. In case of retrograde transport, it is possible to label neurons projecting their axons to the region of former substance delivery. The combination of two retrograde labels (for example dyes absorbing different fluorescent wavelenghts) is called double retrograde labeling and is particularly useful for studying brain structures possessing many projections to different brain regions. This technique enables a precise insight into afferent and efferent connection structure [13]. Apart from conventional chemical labels, viruses containing marker proteins can be used to visualize neuronal connections. They have the ability to move between neurons in the synaptic way and to replicate inside infected cells, which considerably increases the signal. Sometimes, in order to obtain more detailed images of complex neural pathways, a combination of double retrograde and viral labeling is performed. In this method, two strains of recombinant viruses are used, each of them expressing unique marker protein. Consequently, neurons belonging to both neuronal loops show co-expression of the labels [13]. Among the numerous neurotropic viruses used for transsynaptic labeling, rabies virus (RV) is particularly worth mentioning. It can be transported only in retrograde manner and its propagation is restricted to neurons directly connected with the infected cell (without affecting neighbouring neurons) [14].
“Brainbow” The general anatomical structure and physiology of the brain is already well known, mainly due to the development of the aforementioned techniques utilizing fluorescent proteins. However, the true structure of this organ is even more complex when considering that it is connected with a multitude of synapses that are often not stable in time and space. This fact gave rise to connectomics – a brand new branch of neurobiology focusing on synapses and their high resolution imaging.
The beginnings of the “Brainbow” One of the first scientists that understood the need to create a time variable, exact map of neural wiring was Professor Jeff Lichtman from the Center for Brain Science at Harvard University [15,16]. He was interested in synaptic plasticity, which was proven by the discovery of phenomena called synapse elimination and synaptic takeover, that are common during vertebrate nervous system development. There is a reduction of an initially huge number of axonal connections formed in the im C
2014 Informa Healthcare USA, Inc.
93
mature organism with some being replaced by others in a kind of competition between axons. This process is most evident and particularly well studied at neuromuscular junctions, so in the mid 1980s Lichtman’s group decided to make the first attempts to visualize synaptic plasticity exactly on this part of the nervous system. They described a technique called “activity-mediated uptake of fluorescent probes” to label terminals of individual axons in developing twitch muscle fibers. After electrical stimulation of axons, different fluorescent probes were taken up in a classical way similar to neurotransmitter reuptake in synapses. As a result, each neuron was labeled by a traceable and unique colour. However, this technique required the complex process of simultaneous stimulation of all neurons. [16,17]. Firstly, Lichtman generated lines of transgenic mice expressing three different fluorescent proteins (RFP, GFP and YFP, together termed XFP) in order to visualize synapse formation by motoneurons. Neuro-specific elements of the thy-1 gene were chosen as a promoter site, as they were proven to generate the strongest reporter gene expression in the cells of interest. The results of these attempts were surprising, because patterns of transgene expression in 25 independently generated transgenic lines (incorporating identical regulatory elements) were diverse. Meaning that some mouse strains expressed XFP in almost all neural cells, whereas others showed fluorescence in only a small percentage of cells. The next step was to cross the animal lines with one another, which resulted in the new strains expressing two or three XFPs in different subsets of neurons. Some cells showed co-expression of fluorescent proteins causing formation of several new colours [16,18]. The possibility to cross different transgenic mouse lines gave opportunity to create specific progeny expressing particular XFPs in particular neural cell types. This kind of manipulation enables a complete reconstruction of synaptic connections in the chosen brain regions. Importantly, XFP expression in cells turned out to be stable which facilitated tracing of individual cell fate over time. Thus, in the first years of the 21st century Lichtman’s lab tried to express different fluorescent proteins in small subsets of motor axons in order to image axons withdrawal and expansion at postsynaptic sites. These studies revealed that synaptic competition is a random process (without obvious spatial bias), quite changeable in time and that not all postsynaptic sites are re-occupied after axon withdrawal [19,20]. The following years provided further advances in time-lapse imaging of neural cells using XFPs. High resolution confocal laser scanning microscopy enabled the visualization of small structures like retinal ganglion cells (RGCs) in detail, showing not only cell bodies, but also axons and whole dendritic arbors. This has a potential
94
E. Rojczyk-Golebiewska et al.
application in studies of normal retinal development and in eye disease modeling [21].
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Creating connectomic map To make an authentic, detailed map of neuronal circuitry, there was a need to use more than several basic colours for fluorescent labeling of neurons. Initially, scientists tried to achieve it by creating chimeric mice, by crossing mouse lines or initiating crossing over process. Nevertheless, these methods were time consuming, cumbersome and generated only a limited number of fluorescent shades [15]. Jean Livet – a member of Lichtman’s lab since 2002 – invented a new technique to solve this problem [16]. It was based on well-known DNA-engineering system called Cre/lox recombination which is often used to create knockout animals and to carry out deletions, insertions, translocations and inversions at specific sites of DNA. Cre is a recombinase, that performs DNA excision or inversion at the sites flanked by lox sequences. Its activation is usually obtained either by tamoxifen administration (in case of so called tamoxifen-inducible CreER recombinases) or by crossing the mouse possessing lox sites with a mouse showing constitutive expression of the Cre enzyme. Livet’s general idea was to create genetic constructs with several XFP genes flanked by multiple lox sites. In such constructs only genes directly following the promoter can be expressed. Then, activated Cre recombinase excises/inverts DNA between one pair of lox sequences. Importantly, there are multiple incompatibile lox site variants (for example loxP, loxN, lox2272). If they are inserted into one construct, Cre recombinase is somehow forced to choose between two mutually exclusive excision events, so only one possibility of recombination occurs in one cell. As a choice of excision/inversion site is made randomly in each cell, this manipulation will result in neural cells labeled in different shades (Figure 1). The more different XFP genes and lox site variants are used, the more colours can be potentially obtained [15,16]. After examination of the first generation of these transgenic mice, it turned out, that some mouse lines had incorporated more than one construct copy. It enabled independent recombination of each copy (so called combinatorial XFP expression), that resulted in multitude of colours obtained. That is why Prof. J.Lichtman named this interesting and simple technique “Brainbow” [16]. In case of three transgene copies, it is possible to obtain ten distinct colours (Figure 2) and the addition of further copies generates an enormous number of shades – upwards of a hundred. As a result, using fluorescent microscopy, we can generate colourful images of different brain regions (Figure 3). In order
to control the number of “Brainbow” construct tandem repeats, researchers used another recombination system (analogous to Cre/lox) called Flp/FRT. Insertion of FRT sites into appropriate place in the DNA sequence allowed excision of excessive transgene copies after Flp recombinase activation [15].
Examples of “Brainbow” applications After invention of “Brainbow”, many efforts focused on generation of multiple transgenic mice lines in order to test this system for different applications. Sometimes it is convenient to restrict recombination process to specific cell types or brain regions. Another interesting idea is to use “Brainbow” expressing mice as a tool for mapping neural circuits. Researchers succeeded in visualizing granular cells of the inner granular layer (IGL) of the cerebellum together with the mossy fibres innervating these cells. Confocal microscopy accompanied by computer-assisted methods enabled us to distinguish and trace individual axons forming the exact map of connections. This study revealed that cerebellar granular cells are polyneuronally innervated by mossy fibres [15]. In conclusion, “Brainbow” represents an enormously promising tool for the visualization of the nervous system in nanoscale and is considered as a true milestone on the way to create the first full “connectomic” map of the brain.
CLARITY Classical methods of brain imaging enable visualization of small volumes of tissue, typically a few micrometerthin microscopic sections. Detailed reconstruction of three-dimensional brain structure was so far possible only by means of computer data consolidation methods, but it was quite complicated and not applicable to immunohistochemistry because of limited antibody penetration capacity.
Optical clearing methods Three-dimensional fluorescent brain imaging requires the sufficient transparency of the whole organ. The gross anatomical CNS structures are opaque, therefore chemical clearing is a useful technique to dissolve this problem. The classical, commonly applied clearing medium contains benzyl alcohol and benzyl benzoate (BABB). [22]. BABB is a hydrophobic agent and thus can only be used after brain dehydration. Unfortunately exposure to ethanol and BABB may cause a significant decline or even disappearance of GFP fluorescence. It has been recently suggested that tetrahydrofuran (THF) does not International Journal of Neuroscience
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Neurolight
95
Figure 1. The application of the Cre/lox system to create a colourful map of neural circuitry.
produce this effect and seems to be a better dehydrating agent than ethanol [23]. Moreover, combined dibenzyl ether (DBE) and THF provide more beneficial chemical clearing of brain tissue and preserve fluorescence [23, 24]. Importantly, the clearing process with DBE is also faster than with BABB [25]. The THF-BABBbased clearing technique makes the unsectioned brainstem and spinal cord transparent and appropriate to 3D deep fluorescence analysis. For instance, this promising method enables the precise visualization of the trajectories of regenerating axons. It also quantitatively determines the injury-related glial reaction [26].
CLARITY for creating translucent brain In recent months, brand new clearing methods have been described by scientists in the USA. It is called CLARITY, which was initially an acronym to describe Clear Lipid-exchanged Acrylamide-hybridized Rigid Imaging/Immunostaining/In situ hybridizationcompatible Tissue hYdrogel. Put simply, it creates a C
2014 Informa Healthcare USA, Inc.
translucent brain tissue that is permeable for macromolecules using hydrogel-tissue hybridization and subsequent electrophoretic clearing. This method is so mild, that it enables tissue molecular structure preservation, making it possible to visualize the whole intact, unsectioned brain using microscopes with long working distance objectives [27]. The simple procedure takes about eight-nine days and can be divided into three steps (Figure 4). Firstly, formaldehyde (crosslinker), hydrogel monomers and thermally triggered polymerization initiators are administered to the tissue which causes not only tissue crosslinking, but also covalent linking of biological elements (for example proteins and nucleic acids) to monomers. The temperature is then increased up to 37 degrees Celsius and thermally triggered polymerization takes place resulting in hydrogel-tissue hybrid formation. This gelatinous meshwork is still covalently linked to native biological molecules of the tissue apart from lipids, which lack the necessary reactive groups.
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
96
E. Rojczyk-Golebiewska et al.
The image of mouse dentate gyrus obtained with the use of “Brainbow” system (left) –courtesy of Dr Jeff Lichtman, Harvard University, and combinatorial XFP expression resulting from tandem transgene copies integration (right).
Figure 2.
The last step is electrophoresis in ionic detergent (for example in SDS) called electrophoretic tissue clearing (ETC), in which micelles are actively transported into the tissue, whereas membrane lipids (in these micelles) are extracted. During ETC tissue is placed in the chamber supplied with temperature-controlled buffer circu-
lator and electrophoresed by applying 20–60 volts to the electrodes. As a result, the brain sample becomes transparent, but all structures and biomolecules are well preserved (Figure 5). It makes structural and molecular information accessible for visualization and analysis (27,28).
Figure 3. The “Brainbow” imaging of the rat hippocampus neuronal population - courtesy of Dr Jeff Lichtman, Harvard University.
International Journal of Neuroscience
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
Neurolight
Figure 4.
97
Three steps to make the brain translucent–overview of the CLARITY method.
Whole brain imaging and molecular phenotyping with CLARITY
Three-dimensional view of the rat hippocampus showing eYFP (green), parvalbumin (red) and GFAP (blue)expressing neurons with CLARITY method - courtesy of Dr Karl Deisseroth, Stanford University.
Figure 5.
C
2014 Informa Healthcare USA, Inc.
As CLARITY is compatible with most fluorescence microscopy techniques (like confocal, single-photon, multiphoton imaging), it is possible to obtain images of the whole brain expressing fluorescent proteins. Directly before imaging, it is recommended to perform an immersion in refractive-index-specific solutions matching the CLARITY hybrid (for example in 85% glycerol) in order to decrease light scattering occurring due to heterogeneously distributed protein and nucleic acid complexes in the hybrid. It makes the brain uniformly transparent and facilitates high-resolution visualization [27, 28]. CLARITY enables the immunostaining of entire intact tissue on practical timescales of several weeks. It
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
98
E. Rojczyk-Golebiewska et al.
was proven, that electrophoretic clearing of hydrogeltissue hybrids causes only 8% protein loss, a good result in comparison with classical methods of tissue clearing. Moreover, elution of antibodies in multi-round immunostaining (using the same conditions as in ETC) does not cause tissue damage and the diffusion rate of molecular probes deep into intact tissue is quite fast. For example, 2 weeks-long incubation for each antibody is enough to stain 5 mm-thick adult mouse brains. Together with modern microscopic techniques it is possible to obtain multi-colour images presenting specific protein localization in big tissue volumes [27, 28]. In summary, CLARITY provides access to structural and molecular information in large volumes of tissue without need to perform sophisticated computer data consolidation. It may be applied in preclinical comparative studies and disease animal modeling to obtain broad insight into brain structure and molecular features in different conditions.
2.
3.
4.
5. 6. 7.
8.
9.
Conclusions Development of neurobiology is based to a large degree on superseeding imaging methods. Classical techniques are still being improved and used for new purposes, including basic studies performed on small animals and optogenetics, which opens new possibilities in manipulating the bioelectrical activity of single neurons. In recent years, a large emphasis has been placed on recognizing nervous system microstructure, taking synaptic connections architecture under special consideration. That is why traditional brain microstructure imaging methods are gradually displaced by modern methods (like “Brainbow” and CLARITY), which use current achievements of biophysics, biotechnology and molecular biology. Thanks to these advances, it has became possible to accurately visualize nervous system microstructure, which is the first step leading to the true investigation of the relationship between structure and function of particular groups of neurons.
10.
11.
12.
13.
14.
15.
16. 17.
18.
Declaration of Interest
19.
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
20.
21.
References 22. 1. Van der Linden A, Van Camp N, Ramos-Cabrer P, Hoehn M. Current status of functional MRI on small animals: application
to physiology, pathophysiology and cognition. NMR Biomed. 2007;20(5):522–45. Aoki I, Wu YJ, Silva AC, et al. In vivo detection of neuroarchitecture in the rodent brain using manganese-enhanced MRI. NeuroImage 2004;22(3):1046–59. Lizarbe B, Benitez A, Pel´aez Brioso GA, et al. Hypothalamic metabolic compartmentation during appetite regulationas revealed by magnetic resonance imaging and spectroscopy methods. Front Neuroenergetics 2013;13;5:6 doi: 10.3389/fnene.2013.00006. eCollection 2013 Jacobs RE, Cherry SR. Complementary emerging techniques: high-resolution PET and MRI. Curr Opin Neurobiol. 2001;11(5):621–9. Miyawaki A. Fluorescence imaging in the last two decades. Microscopy (Oxf). 2013;62(1):63–8. Tsien RY. The green fluorescent protein. Annu Rev Biochem. 1998;67:509–544 Bernstein JG, Boyden ES. Optogenetic tools for analyzing the neural circuits of behavior. Trends Cogn Sci. 2011;15(12):592–600. Bernstein JG, Garrity PA, Boyden ES. Optogenetics and thermogenetucs: technologies for controlling the activity of targeted cells within intact neural circuits. Curr Opin Neurobiol. 2012;22(1):61–71. Yao JP, Hou WS, Yin ZQ. Optocenetics: a novel optical manipulation tool for medical investigation. Int J Ophthalmol. 2012;5(4):517–22. Bovetti S, Moretti C, Fellin T. Mapping brain circuit function in vivo using two-photon fluorescence microscopy. Microsc Res Tech. 2014 (epub ahead of print). Micheva KD, Smith SJ. Array tomography: a new tool for imaging the molecular architecture and ultrastructure of neural circuits. Neuron. 2007;55(1):25–36 Denk W, Horstmann H. Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. PLoS Biol. 2004;2(11):e329. Ohara S, Inoue K, Witter MP, Iijima T. Untangling neural networks with dual retrograde transsynaptic viral infection. Frontiers Neurosci. 2009;3(3):344–9. Wickersham IR, Lyon DC, Barnard RJ, et al. Monosynaptic Restriction of Transsynaptic Tracing from Single, Genetically Targeted Neurons. Neuron 2007; 53(5):639–47. Livet J, Weissman TA, Kang H, et al. Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature 2007;450(7166):56–62. Lehrer J. Neuroscience: making connections. Nature 2009;457(29):524–7. Lichtman JW, Wilkinson RS, Rich MM. Multiple innervation of tonic endplates revealed by activity-dependent uptake of fluorescent probes. Nature 1985;314(6009):357–9. Feng G, Mellor RH, Bernstein M, et al. Imaging Neuronal Subsets in Transgenic Mice Expressing Multiple Spectral Variants of GFP. Neuron 2000;28(1):41–51. Walsh MK, Lichtman JW. In Vivo Time-Lapse Imaging of Synaptic Takeover Associated with Naturally Occurring Synapse Elimination. Neuron 2003;37(1):67–73. Keller-Peck CR, Walsh MK, Gan WB, et al. Asynchronous Synapse Elimination in Neonatal Motor Units: Studies Using GFP Transgenic Mice. Neuron 2001;31(3):381–94. Walsh MK, Quigley HA. In vivo time-lapse fluorescence imaging of individual retinal ganglion cells in mice. J Neurosci Methods. 2008;169(1):214–21. Becker K, J¨ahrling N, Kramer ER, et al. Ultramicroscopy: 3D reconstruction of large microscopical specimens. J Biophotonics. 2008;1(1):36–42
International Journal of Neuroscience
Neurolight
Int J Neurosci Downloaded from informahealthcare.com by Selcuk Universitesi on 01/31/15 For personal use only.
23. Becker K, Jahrling N, Saghafi S, et al. Chemical clearing and dehydratation of GFP expressing mouse brains. PLoS One. 2012;7(3):e33916. 24. Dodt HU, Leischner U, Schierloh A, et al. Ultramicroscopy: three-dimensional visualization of neuronal networks in the whole mouse brain. Nat Methods. 2007;4(4):331–6 25. Viswanath DS, Ghosh T, Prasad D, et al. Viscosity of liquids: theory, estimation, experiment and data. Dordrecht: Springer; c2007
C
2014 Informa Healthcare USA, Inc.
99
26. Erturk A, Mauch CP, Hellal F, et al. Three-dimensional imaging of the unsectioned adult spinal cord to assess axon regeneration and glial responses after injury. Nat Med 2011;18(1):166–71. 27. Chung K, Wallace J, Kim SY, et al. Structural and molecular interrogation of intact biological systems. Nature 2012;497:332–7. 28. Chung K, Deisseroth K. CLARITY for mapping the nervous system. Nat Methods 2013;10(6):508–13.
