Chapter 19 - Springer Link

Chapter 19 Combined Immunochemistry and Live Imaging of Fluorescent Protein Expressing Neurons in Mouse Brain Ruth M. Empson, Malinda L.S. Tantirigama, Manfred J. Oswald, Stephanie M. Hughes, and Thomas Knöpfel Abstract The use of transgenic mice expressing fluorescent proteins to report a specific protein or to identify specific groups of neurons in the brain is revolutionizing many different aspects of neuroscience. Here we use an example of a GFP-expressing reporter mouse from the GENSAT project that allows identification of a specific group of neurons in the mouse cortex. Live GFP detection facilitates identification of the neurons for whole-cell patch clamp electrophysiological recording to probe their functional properties. Post hoc immunohistochemistry allows specific reconstruction of the shape of the recorded neuron; this together with the detection of other co-expressed proteins helps confirm the functional identity of specific neuron types. Approaches such as these are beginning to progress the major task of untangling the complexity of a variety of brain circuits. Key words GENSAT, Motor cortex, Whole-cell electrophysiology, E-GFP reporter, Layer 5

1

Background and Historical Overview Molecular cloning of the gene for the green fluorescent protein (GFP) from the jellyfish Aequorea victoria and the subsequent development of a full color palette of bright and photo stable fluorescent proteins (FPs) enabled and is revolutionizing the use of fluorescence-based methods in most neurophysiology laboratories. Among the most widely used FP-based approaches are those that use FPs to tag specific proteins or specific classes of neurons in living cells, tissues, or intact animals. Here we focus on the use of transgenic mice with genetically targeted expression of FPs in specific neuronal populations. These mice are immensely valuable for the combination of cellular level anatomy and functional circuit level characterizations in living brain tissue. In some ways the ongoing development and application of these technologies can

Adalberto Merighi and Laura Lossi (eds.), Immunocytochemistry and Related Techniques, Neuromethods, vol. 101, DOI 10.1007/978-1-4939-2313-7_19, © Springer Science+Business Media New York 2015

357

358

Ruth M. Empson et al.

be compared to the revolution started by the discovery of monoclonal antibodies in the 1970s. Now nearly half a century ago it was work that aimed to understand the basis for the generation of antibody diversity that led to the development of monoclonal antibodies and changed biomedical research forever. Now, most of us do not think twice about producing a monoclonal antibody as a key tool to detect and characterize a new protein of interest, and the therapeutic power of monoclonal antibodies is now widely accepted for the treatments of cancer and inflammatory disorders. In this chapter, we discuss how fluorescent proteins enable not only identification of specific neuron classes but also their functional characterization in complex brain circuits, as summarized in Fig. 1. In contrast to classical immunohistological approaches, structural and functional fluorescence microscopy can now routinely be performed in living tissue and even living animals. Looking several decades forward, these new approaches will likely translate into new strategies for the treatment of neurological diseases that are caused by—or associated with—disturbed functions of neuronal circuits. 1.1 Transgenic Mice Expressing Fluorescent Proteins in Specific Cell Classes

A large number of transgenic mouse lines that express a FP in specific classes of neurons have been generated during recent years. In these mice, the cell class specificity of FP tagging is determined by the particular promoter used to drive gene transcription and, hence, protein expression. In its simplest implementation of this principle, the cell class specific promoter directly drives the expression of the FP. Conditional and intersectional strategies (e.g., those that take advantage of the Cre-lox system or multiple regulatory sequences) are reviewed elsewhere, so here we focus on the former simple transgenic approaches. Examples of promoters used are those of the neuron-specific Thy1 gene [1–3], the potassium channel Kv3.1 gene [4], the glutamate decarboxylase (GAD) 67 gene, or the glycine transporter GlyT2 [5], the gene for transcription factor Etv1, or the gene for glycosyltransferase protein Glt25d2 [6]. The advantage of using these transgenic mouse lines compared with in utero electroporated or virus-injected mice is the high consistency of the expressing cell population across animals in a given transgenic line [3]. This translates into robust electrophysiological, morphological and molecular characteristics. In this chapter we use an EGFP reporter mouse line derived from the GENSAT project as an example to describe the various experimental protocols (see Fig. 1) involved in these types of studies.

1.2 The GENSAT Project

“Our understanding of the molecular mechanisms that contribute to the formation and function of the brain must include information about the precise distributions of specific genes and proteins throughout development, and the ability to identify, visualize and genetically manipulate each of the major central nervous system (CNS) cell types.” [7]. The GENSAT project therefore set about

Immunocytochemistry and Live Imaging

c

359

b

Excitation

a

+

Emission

Flourescent Reporter Readout = Identity - Fezf2

Immunohistochemistry Reporter Verification

e

d

Complex Cortex Action Potentials

Post-hoc histochemistry Detailed morphology

Flourescent Reporter Readout = Identity - Fezf2 Readout = Electrophysiology + Biocytin Fill

Fig. 1 Using fluorescence reporters and post hoc immunohistochemistry to help untangle the complexity of the cortex (a). Schematics of how the complex cerebral cortex can be simplified through the combination of genetically encoded fluorescent reporters/protein signal readouts (b), reporter verification using post hoc immuno-histochemistry (c), coupled with targeted electrophysiological readouts that identify and control neuron activity (d) together with post hoc detailed analysis of the neuron identity using post hoc histochemistry and reconstruction methods (e). Complex cortex schematic modified from Fig. 28 after Cajal: Section of the motor gyrus of an adult man, p. 234. In: DeFelipe and Jones (Eds) Cajal on the cerebral cortex: an annotated translation of the complete writings. 1988. OUP

solving the problem of identifying the spatial organization and expression level of about 5,000 CNS-expressed genes. To do this it has generated a library of bacterial artificial chromosome (BAC) clones that contain the genetic material relevant to each of the

360


major cell populations in the mammalian brain. It has also established transgenic mouse lines that contain these BACs along with fluorescent reporters of the BAC, most notably the use of green fluorescent protein, GFP. This means that green fluorescent neurons report the anatomical position and physiological behavior of all cells that specifically express the gene of interest. 1.2.1 Fezf2, a Master Control Gene in the Cortex, and the Fezf2-GFP Mouse

We are interested to identify and understand the role of “master genes” that drive the specification and maintenance of cortical networks. One of these is Zfp312, also called Fezf2 that encodes production of a zinc finger transcription factor protein, FEZF2. This protein is considered to control the expression of genes necessary for the acquisition of neuron-type specific properties during development (see also Fig. 3). Transcription factors of this type activate or repress an array of downstream genes that specify a neuron’s timing of birth, morphology, molecule expression, and axonal connectivity, and are well suited to control broad aspects of neuronal phenotype. In this view, master control genes, like Fezf2 are thought to be expressed by a specific neuron-type and therefore to specify its unique identity, often in combination with other key transcription factors. Differential expression of a variety of these master control genes is thought to establish a molecular code that underpins the enormous diversity of cortical projection neurons [8–13]. Understanding their control and downstream effectors is a major question in neuroscience [14, 15]. Fezf2 is expressed in all subcortical projection neurons from early stages of development [embryonic day (E) 10.5] through to adulthood [detected up to postnatal day (P)120] and is necessary for the formation of corticospinal projection neurons during development [10, 11, 16]. Deletion of Fezf2 in a knockout mouse (Fezf2-KO) results in abnormal morphology and loss of other key specification genes (i.e., Ctip2, ER81, Crim1, Cdh13, S100a10, and Netrin-G1) in layer 5 cortical projection neurons. The Fezf2-KO mice lack axon pathways from cortex to the spinal cord, brainstem, thalamus, and contralateral cortex via the corpus callosum, and the mice are hyperactive [9, 11]. This evidence suggests that Fezf2 drives the specification of projection neurons in the cortex. Despite this, very little is known about the morphology and electrophysiological properties of Fezf2-expressing neurons, and whether the expression of Fezf2 continues to be important for maintenance of neuronal phenotype into adulthood. GENSAT developed a Fezf2 reporter mouse called the (Zfp312-EGFP)CO61Gsat/Mmnc mouse line that we hereafter refer to as the Fezf2-GFP mouse. This reporter mouse offered an ideal opportunity to identify those projection neurons within the mature mouse cortex that expressed this master control gene and also allowed us to target them with electrophysiology.


361

Our hypothesis was that neurons expressing this transcription factor would behave differently to their non-expressing neighbors and by using a fluorescence reporter this aided targeting and reduced the normally random nature of selection using electrophysiology methods. Furthermore, targeting the neurons in this way allowed us to identify their morphology. In this way we hoped to get a step closer to identifying how this master gene contributes to projection neuron shape and function within the complex cortical network (Fig. 1a, b). 1.3 Combination of FP-Based Methods with Electrophysiology and Immunohistochemistry

2

While cell class specific expression of FPs allows targeting specific cell types for patch clamp electrophysiological examination, autofluorescence and light scattering of uncleared brain tissue usually prohibits high-resolution structural analysis of the patch clamped cell. It is therefore good practice to fill the neurons via the patch pipette with a marker chemical, such as biocytin. Detection of this marker post hoc using histochemical methods can then reveal important additional information about the shape of the neuron, its spatial position within the network and its proximity to neighboring neurons or other structures (Fig. 1d, e). Immunohistochemical detection of endogenous marker proteins that characterize the recorded neuron beyond the specificity of the genetic tagging are also often necessary if we are to understand the neuronal population under study and the complex neuronal circuit within which it is embedded (Fig. 1c).

Equipment, Materials, and Setup

2.1 Wide-Field Fluorescence Microscopy

2.2 Detection of GFP in Fixed Brain Tissue

●

Standard fluorescence microscope with a GFP filter cube (excitation, 480 ± 30 nm; emission, 535 ± 20 nm) and a CY3 filter cube (excitation 545–580 nm, long pass emission at 610 nm).

●

20×/0.75 NA objective.

●

Optional: Anti-GFP primary antibody.

●

Optional: Species-specific fluorescent-tagged secondary antibody for GFP detection.

●

Perfusion instruments.

●

Peristaltic pump.

●

Phosphate buffered saline (PBS): 10 mM phosphate buffer (75 mM Na2HPO4 and 25 mM NaH2PO4), 2.7 mM KCl, 137 mM NaCl, dissolved in Milli-Q water, pH 7.4.

●

Fixative: 4 % paraformaldehyde in PBS.

●

30 % sucrose solution in Milli-Q water.

362


2.3 Live-Cell Microscopy for GFP Fluorescence Detection and Targeting for WholeCell Patch Clamp Electrophysiology

●

Acrylic Brain Blocker (Ted Pella Inc., Redding, CA).

●

Freezing microtome or cryostat.

●

Phosphate buffer (PB).

●

Mounting medium, e.g., VECTASHIELD®, Laboratories Inc., Burlingame, CA.

●

High sucrose cutting solution: Sucrose75 mM, NaCl 87 mM, KCl 2.5 mM, NaH2PO4 1.25 mM, MgCl2 6 mM, CaCl2 0.5 mM, NaHCO3 25 mM, glucose 25 mM.

●

Artificial cerebrospinal fluid (ACSF): NaCl 126 mM, KCl 3 mM, NaH2PO4 1 mM, MgSO4 2 mM, CaCl2 2 mM, NaHCO3 25 mM, glucose 15 mM. During recording, ACSF was supplemented with 50 μM 2-Amino-5-phosphonopentanoic acid (AP5 Abcam Biochemicals, Cambridge, UK) and 20 μM 1,2,3,4-Tetrahydro-7-nitro-2,3-dioxoquinoxaline-6carbonitrile (CNQX) disodium (Abcam Biochemicals) to block glutamatergic synaptic transmission and 50 μM picrotoxin (Sigma, Chemicals, St. Louis, MO) to block GABAAergic synaptic input.

●

Internal pipette solution: KCl 10 mM, Na-phosphocreatine 10 mM, K-gluconate 110 mM, HEPES 10 mM, Mg-ATP 4 mM, Na-GTP 0.3 mM, Biocytin 0.2 %; pH 7.3 and osmolarity adjusted to 295 mOsm with sucrose.

●

95 % O2 and 5 % CO2 gas cylinder.

●

Mouse CO2 chamber.

●

Peristaltic pump for animal perfusion.

●

Dissection tools.

●

Petri dish.

●

Razor blades.

●

Cyanoacrylate glue.

●

Vibrating microtome, e.g., VT1000S, Leica Microsystems, Wetzlar, Germany.

●

Peristaltic pump for slice perfusion: MS-Reglo (IsmaTec, Glattbrugg, Switzerland) or 50S (Watson Marlow, Wilmington, MA).

●

Recording chamber.

●

Inline heating system (Warner Instruments, Hamden, CT).

●

Upright microscope, e.g., Eclipse (Nikon, Tokyo, Japan) equipped with a 10× objective (Plan Fluor, Nikon), and a 60× objective (1.00 NA), water immersion (Plan Fluor, Nikon).

●

Infrared DIC optics.

●

CCD camera.

●

Monitor.

2.3.1 Slice Preparation and Electrophysiology

Vector


2.4 Post hoc Histochemistry and Reconstruction of Fezf2-GFP Expressing Neurons

2.5 Combining Fluorescence-Based Immunohistochemistry with the Fezf2-GFP Reporter Mouse Immunohistochemistry of Fezf2 Downstream Effectors

3

363

●

Micromanipulator, e.g., Siskiyou MC 1000e (Siskiyou Corporation, Grants Pass, OR) or Burleigh® PCS6000 (Thorlabs, Newton, NJ).

●

Patch Clamp amplifier/digitizer.

●

Alexa® 568 streptavidin conjugate Technologies, Grand Island, NY).

●

0.3 % Triton® X-100 in PBS.

●

Mounting medium, e.g., VECTASHIELD®.

●

ImageJ Software.

●

Adobe Photoshop (Adobe, San Jose, CA).

●

Rabbit-anti-TBR1 polyclonal antibody (Abcam): used 1:500.

●

Rat-anti-Ctip2 1:1,000.

●

Mouse-anti-SATB2 [SATBA4B10] (Abcam): used 1:1,000.

●

Goat-anti-rabbit Alexa® 555 conjugate (Invitrogen™): used 1:500.

●

Goat-anti-rat biotinylated secondary antibody (Jackson Immuno Research Labs): used 1:600.

●

Streptavidin-Alexa® 568 conjugate (Invitrogen™): used 1:800.

●

Goat-anti-mouse Alexa® 555 conjugate (Invitrogen™): used 1:500.

●

Antibody buffer: PBS with 5 % goat serum and 0.2 % Triton® X-100.

[25B5]

monoclonal

(Invitrogen™,

antibody monoclonal

Life

(Abcam): antibody

Procedures In the following sections we explain how we use fluorescence imaging techniques in both living and fixed tissues to characterize neurons expressing a fluorescent reporter, such as those within the cortex of the Fezf2-GFP mouse. Firstly an initial characterization of faithful FP reporter expression can be easily carried out using fixed tissue (Sect. 3.1). Once the fidelity of the marker FP is established in the brain region of interest, live-cell imaging and electrophysiological targeting of individual neurons can take place using in vitro brain slices (Sect. 3.2) (or, in principle, also in vivo). During this process the neurons can be filled with chemicals that are detected later using fluorescence-based histochemical methods and detailed reconstruction using fluorescence microscopy can follow (Sect. 3.3). In order to further examine the relationship between the FP expressing neurons and their neighbors within the brain circuit of interest, it is also possible to combine immunohisto-

364


chemistry for other related proteins with the FP reporter. We describe how to do this in Sect. 3.4, taking advantage of the expression of some other master control gene products within the cortex, 3.1 Wide-Field Microscopy for Fluorescence Detection in Fixed Brain Tissue

We use standard wide-field fluorescence microscopy to detect GFP and map Fezf2 expression without any need for amplification. If required a primary antibody to GFP could be utilized together with a fluorescent tagged secondary antibody, but in the case of this reporter mouse, the endogenous GFP fluorescence was sufficient. After perfusion with chilled fixative the brain was removed and post-fixed overnight at 4 °C. The day after, the brain was placed in a 30 % sucrose solution until it sunk, for cryoprotection before sectioning. A block of forebrain was cut in the coronal or sagittal plane using an acrylic Brain Blocker. The blocker allowed reproducible cuts in the plane that matches sections depicted in the stereotaxic mouse atlas [17] to be made. Serial coronal or sagittal sections (25–50 μm) from the forebrain block were then cut on a freezing microtome. Sections were washed in PB, mounted on glass slides, briefly air-dried, and coverslipped using VECTASHIELD® mounting medium. Sections were viewed with a microscope equipped for epifluorescence and the appropriate GFP filter set. Sixteen-bit images were captured on selected sections (F-view II Trigger FW) and were imported as TIFF files to ImageJ or Adobe Photoshop CS5.1 software. Final images were produced by optimizing image brightness/contrast, and merging and cropping where necessary. Regions of sections that displayed fluorescence (Fig. 2) were identified according to atlas [17].

3.2 Live-Cell Microscopy for GFP Fluorescence Detection and Targeting for WholeCell Patch Clamp Electrophysiology

Prior to slicing, the cutting solution was cooled (1–3 °C) on ice and saturated with a 95 % O2 and 5 % CO2 gas mixture (carbogen) for at least half an hour. Mice were initially anesthetized in a CO2 chamber and rapidly decapitated. The head was quickly doused with cold cutting solution and a midline incision with a scalpel exposed the dorsal surface of the skull. The brain was uncovered by making a cut across the nasal bone, extending through the skull on either side of the cranium and terminating near base of the neck. The brain was separated from the base of the skull and transferred to a dish with cold oxygenated cutting solution. The frontal cortex, that is used for sectioning was cut from the rest of the brain by a coronal razor blade through the parietal cortex. The angle of the blade was slanted rostrally (~10–15°) to align with the radial axis of cortex [18]. This off-coronal angle yielded slices with apical dendrites of layer 5 neurons approximately parallel with the cut surface. The tissue block was then glued to a platform using cyanoacrylate glue and transferred to the stage of a vibrating microtome

3.2.1 Brain Slice Preparation


365

Fig. 2 Fez2-GFP reporter expression in the mouse brain. (a) Fez2-GFP reporter expression in a fixed (4 % paraformaldehyde) coronal slice from mature mouse frontal cortex. The GFP fluorescence (without amplification) can be seen as distinctive labeling in layer 5 of the cortex, (bii) but not in the more superficial layer 2/3 (bi), or in the deeper layer 6. Note the large pyramidal shape of the GFP positive neurons (inserts bi and bii). (c) Images of GFP positive and GFP negative neurons in a living slice observed with infra red (differential interference contrast) DIC optic, overlay on right hand image. The dotted line indicates the position of a whole-cell patch clamp recording electrode, seen more clearly in the lower DIC image, targeting a GFP negative neuron. aca anterior part of anterior commissure

that was filled with chilled oxygenated cutting solution. Off-coronal 300 μm slices containing M1 in both hemispheres were sectioned and directly transferred into a holding chamber containing heated ACSF (32 °C) that was continuously bubbled with carbogen. The tissue was allowed to recover for one h at 32 °C and then relax to room temperature before recording. 3.2.2 Whole-Cell Recording for Targeting GFP+ve (Fezf2+ve) Cortical Neurons

Brain slices were transferred and anchored to a recording chamber with a nylon grid. A peristaltic pump continually perfused the slices with ACSF (bubbled with carbogen) at a rate of 1–2 mL/min. The recording temperature was controlled at either 26 or 35 °C with an inline heating system. Slices were visualized with an upright microscope equipped with DIC optics and were displayed on a monitor using a CCD camera. Before recording, a few steps are needed to orient the slice and locate the cortical region of interest using a low power 10× objec-

366


tive Under the 60× objective layer 5 neurons were identified on the basis of their pial depth and their larger neuronal somata than neurons in layers 2/3 or 6. On a given experimental day the types of neurons that were targeted for electrophysiological recording were: (1) GFP+ neurons, identified with a filter cube for GFP; (2) GFP- neurons, identified as displaying a fluorescent signal indistinguishable from background at the region of the soma. The identified neurons of interest were marked on the monitor. The somas were then visualized using infrared DIC optics in the same field of view and the patch electrode was guided for whole-cell recording using a micromanipulator. Both fluorescence and DIC images of each recorded neuron were taken as an offline record of their fluorescent label (Fig. 2). Prior to recording, the major axis of the apical dendrite near the soma was checked to run close to parallel to the slice surface. This guaranteed that the distal dendrites of recorded neurons were not cut and were suitable for intracellular biocytin filling. A variety of electrophysiological tests were applied to the recorded cells in order to identify their electrical phenotype, including action potential firing properties and membrane currents and properties but the details are beyond the scope of this chapter. 3.3 Post hoc Histochemistry and Reconstruction of Fezf2-GFP Expressing Neurons

Since the neurons targeted by whole-cell patch clamp recording were also filled with biocytin (0.2 %) we were also able to identify the shape and position of these Fezf2-GFP neurons within the cortical network. We used streptavidin conjugated to a red fluorescent probe (Alexa® 568). Streptavidin reacts directly with biotin (biocytin) within the filled neuron, labeling only the filled neuron in the red fluorescence channel of the wide-field microscope. Typically we obtained recovery percentages of 80–100 % of neurons using the protocol below.

3.3.1 Neuron Capture and Post hoc Fluorescence Histochemistry

After the completion of electrophysiological recording, the electrode was slowly withdrawn while continuously monitoring the seal resistance in voltage clamp using a 5 mV step (20 ms, 100 Hz). In most cases high resistance seal (>500 MΩ) between the electrode tip and the neuron membrane resulted and, at this instance, the electrode was quickly withdrawn. The slice was then stored in a PBS solution containing 4 % paraformaldehyde overnight at 4 °C and then in PBS until processing. In order to avoid morphology overlapping between two or more neurons, we only targeted one neuron per slice. Slices were rinsed in PBS three times and permeabilized in PBS containing 0.3 % Triton®X-100 for 4 h in room temperature. The slices were then incubated for another 4 h in 2–4 μg/mL streptavidin Alexa® 568 in PBS containing 0.3 % Triton® X-100. Slices were thoroughly washed in PBS at least three times and then mounted on glass slides, air-dried, and coverslipped using VECTASHIELD® mounting medium.


367

Fig. 3 Post hoc histochemistry and reconstruction of a FezF2-GFP positive neuron to access morphological parameters. (a) Immunofluorescence using streptavidin-Alexa® 568 binding to biotin applied to a single cortical projection neuron via the whole-cell recording pipette (see also Fig. 2c, lower panel). The image is a composite of several image planes through the thickness of the slice. Note the very bright soma but the clear delineation of the apical dendrite reaching towards the pia. The box inset (dashed lines) shows the beaded appearance of the axon (shown red in the reconstruction in b) and the fine basal dendrites. (b) Traced reconstruction of the neuron extracted from the fluorescence image stacks. (c) Different morphological parameters can be extracted from reconstruction and the inset shows the measurements made to assess soma size and apical dendrite shaft width (red dashed line)

3.3.2 Reconstruction of Fezf2-GFP Positive Neuron Shape

Fluorescence images were acquired by a fluorescence microscope using a 20×/0.75 NA objective and a CY3 filter cube. Sequential image stacks were taken at each z-step of the fine focus wheel to capture the depth of the dendritic tree, and usually 2–4 stacks were required to cover the complete morphology of the recorded neuron. Stacks of 20–60 images were flattened by calculating the maximal intensity of the z-stack (ImageJ) and the resulting tiles were stitched in Adobe Photoshop by manually aligning common landmarks in each tile. The final stitched image with the full morphology of the neuron contained a pixel density of 0.52 μm per pixel. Neuronal morphology was traced manually using the freely available ImageJ plugin NeuronJ [19]. The dendrites were differentiated from the smooth beady appearance of axons and were color coded for identification. The process of reconstruction and analysis of morphology is shown in Fig. 3. A variety of measurements were extracted from the reconstructed neurons for comparative purposes. An example of a reconstructed typical Fezf2-expressing neuron is shown in Fig. 3. Although as we saw in Fig. 2 GFP fluorescence was most easily

368


detectable in the soma, the full reconstruction shows that these neurons display many of the characteristics typical of layer 5 projection neurons within the mouse cortex. 3.4 Combining Fluorescence-Based Immunohistochemistry with the Fezf2-GFP Reporter Mouse: Immunohistochemistry of Fezf2 Downstream Effectors

The master action of FEZF2 is likely to initiate and maintain a cascade of gene expression that leads to the induction of downstream effectors (Fig. 4). One potential effector of FEZF2 function is the transcription factor CTIP2 (COUP-TF interacting protein 2), also expressed in subcortical projection neurons during development [8]. CTIP2 likely controls later aspects of subcortical projection neuron development as it plays a role in regulating extension of axons toward subcortical targets, and ectopic expression can rescue the axon pathways lost in the Fezf2-knock out mouse [8, 20]. Equally important are targets that are repressed by FEZF2. SATB2 (special AT-rich sequence-binding protein 2) is another key transcription factor thought to be critical for projection neuron specification. This transcription factor is thought to repress the expression of Ctip2 as a negative regulator of subcortical axon formation. Downregulation of Satb2 is thus necessary for the prevention of cortico-callosal projection neuron specification and relieves repression of Ctip2 [21, 22]. Similarly, TBR1 (T-box brain protein 1) is responsible for the specification of layer 6 cortico-thalamic projection neurons, and its downregulation supports layer 5 subcortical projection neuron specification [23, 24]. In this notion, it is the differences in the levels of expression and the interplay between these transcription factors that could specify one projection neuron type versus another [12, 13]. Once specification has taken place, these transcription factors presumably also ensure the same projection neuron identity is retained in the adult brain. In order to gain some insight into the relationship between FEZF2 and these other transcription factors, and since there is no reliable antibody to mouse FEZF2 we utilized the Fezf2-GFP mouse in conjunction with immunohistochemistry for these other transcription factors.

3.4.1 Tissue Preparation

We used standard wide-field fluorescence microscopy to detect GFP in fixed Fezf2-GFP mouse cortex and map Fezf2 expression alongside other transcription factors that we identified with immunohistochemistry and using red shifted secondary antibody fluorophores. Tissue was prepared in a similar manner as in Sect. 3.1 although brains from young post natal day 5 Fezf2-GFP mice were more easily immersion-fixed in chilled 4 % paraformaldehyde in PBS overnight at 4 °C, followed by cryoprotection before sectioning using the same protocol as described above.

3.4.2 Immunohistochemistry for SATB2, CTIP2 and TBR1

As described above, the expression of these three transcription factors is predicted to occur alongside (SATB2 and CTIP2) or distinct (TBR1) from Fezf2 expression (Fig. 4). Therefore detection of these proteins alongside Fezf2 expression may help us to resolve


369

Fig. 4 Detecting other markers downstream from FezF2 using immunohistochemistry. (a, b) a simplified model of (part of) the transcription factor code for the specification of cortical projection neuron identity. (b) Generic view of how the different transcription factor proteins (encoded by the master control genes) interact with each other for development of projection neuron types within the mouse cortex (based on [8, 20–24]). (c) Combined FezF2-GFP fluorescence (no amplification— green) with immunohistochemistry for two associated transcription factors (red ) illustrating the diversity of expression within the cortex. Sections are obtained from the frontal cortex of a young postnatal day 5 mice, hence the high position of the GFP-positive layer 5 with respect to the pia, the large width of layer 5, and the band of superficial FezF2-GFP positive cells that may be migrating progenitor neurons. Neurons expressing CTIP2 in C also express Fezf2, as seen by the yellow, colocalized neurons throughout layer 5, consistent with the idea that Fezf2 promotes expression of Ctip2 as one of its main effector genes, note these double labeled cells are in both upper and lower layer 5. There are also many GFP positive neurons that do not express CTIP2 (green only) and a minority of CTIP2 positive cells that do not express Fezf2 (red only). (d) TBR1 does not colocalize with Fezf2-GFP, and there are many GFP positive (green only) neurons and many TBR1-positive neurons (red only) that are particularly prominent in the GFP negative layer 6. This pattern is consistent with the idea that FEZF2 downregulates TBR1 expression (that can also repress Fezf2 expression) and prevents acquisition of a layer 6 phenotype

370


how these interact within the complexity of the cortical network. As shown in Fig. 4c, d we observed significant colocalization of GFP, green with CTIP2, red. This provides some confidence, but by no means proof, that FEZF2 and CTIP2 transcription factors may be working together in at least some of the neurons, but not all. For example, only green and only red soma can also be observed. In contrast TBR1 expression was almost exclusively detected in layer 6 neurons and little colocalization with GFP was evident in these cells, indicating a selectivity and specificity for Fezf2-GFP reporter expression. Standard immunohistochemistry was carried out on sections prepared as described (Sect. 3.1). They were rinsed in PBS (2 × 10 min) and then placed in antibody buffer for 30 min at 18–20 °C. Sections were then incubated in the same solution containing the primary antibody (250 μL/well in a 24-well plate, three sections/well) for 70 h at 4 °C (control sections incubated in antibody buffer only). Sections were then washed three times for 10 min each in PBS at room temperature before incubation with secondary antibody diluted in the antibody buffer (including controls) for 3 h at 18–20 °C. Sections were washed twice for 10 min in PBS (three times for 10 min for CTIP2 detection). The sections used for CTIP2 detection were then incubated with streptavidinAlexa® 568 for 2 h at 18–20 °C diluted in antibody buffer, followed by two rinses in PBS. All sections were mounted on glass slides and coverslips applied with VECTASHIELD® antifade solution.

4

Notes and Troubleshooting 1. The Fezf2-GFP mouse is a good example of a reporter mouse that is helping reduce the complexity of the cortex by facilitating live-cell targeting of neurons. Use of immunohistochemistry indicates that, at least to some extent, the GFP reporter does report Fezf2 expression, but a key limitation is the absence of a specific antibody to FEZF2. Other alternatives for verification are therefore required. Of these, in situ hybridization is a sensitive and reliable method that allows visualization of neurons expressing Fezf2 mRNA and with more recent advances in this technique dual color fluorescence in situ hybridisation (FISH) is becoming mainstream, e.g., http://www.creativebioarray.com. 2. Perhaps the most sensitive approach, particularly if whole-cell patch clamp methodologies are available, is to use single cell PCR. Targeting relies upon the live-cell expression and detection of the GFP and successful “patching” of the neuron. Once the whole-cell configuration has been obtained elements of the cytoplasm can be extracted and amplified using appropriate primers to detect expression of the gene of interest. Important


371

controls remain critical, and vigilance is needed to combat contamination from low levels of the gene of interest. Recently, we have successfully used single cell PCR to verify high fidelity expression of Fezf2 in GFP positive cortical pyramidal neurons. In a recent publication we have used both FISH and single cell PCR to validate Fezf2 expression in this GENSAT mouse [25]. Both FISH and single cell PCR remain limited to the detection of mRNA, and if verification of protein expression is needed in a complex tissue, then immunohistochemistry still remains the best method. However, the antibody must be specific and it must be possible to colocalize the detection via primary and secondary fluorescence antibody with the in situ detection of the fluorescent reporter. 3. The methodology described here not only can be applied to transgenic mice expression of FPs in specific cell types, but also extended to mouse lines that express fluorescent-protein based indicator proteins for live monitoring of calcium concentrations [26, 27] or membrane voltage [28–30]. Indeed, refined genetic approaches can be used to generate “indicator mice” where the expression and monitoring of fluorescence not only identifies specific cell types but also reports their “online” physiological activity within the circuit.

5

Conclusion Resolving or untangling the structural and functional complexity of the brain and particularly the cortex is one of the main goals in neuroscience. Several modern approaches towards this goal are facilitated by genetically encoded fluorescent proteins, including fluorescent indicator proteins and reporter proteins. However, these new genetic methods still require the use of more classical methods like immunohistochemistry and single-cell level electrophysiology to validate the optical response outputs and the fidelity of expression. Within the remit and approaches discussed in this chapter, currently the best way, given suitable antibody specificity, is to directly colocalize the FP with the protein or message of interest, in the same neuron and in the intact cortical network.

Acknowledgements Supported by the Marsden Fund Council from Government funding, administered by the Royal Society of New Zealand. M.T. is the recipient of a Department of Physiology, University of Otago PhD scholarship.

372


References 1. Miller MN, Okaty BW, Nelson SB (2008) Region-specific spike-frequency acceleration in layer 5 pyramidal neurons mediated by Kv1 subunits. J Neurosci 28:13716–13726 2. Yu J, Anderson CT, Kiritani T et al (2008) Local-circuit phenotypes of layer 5 neurons in motor-frontal cortex of YFP-H mice. Front Neural Circuits 2:6 3. Porrero C, Rubio-Garrido P, Avendaño C et al (2010) Mapping of fluorescent proteinexpressing neurons and axon pathways in adult and developing Thy-eYFP-H transgenic mice. Brain Res 1345:59–72 4. Akemann W, Zhong Y-M, Ichinohe N et al (2004) Transgenic mice expressing a fluorescent in vivo label in a distinct subpopulation of neocortical layer 5 pyramidal cells. J Comp Neurol 480:72–88 5. Uusisaari M, Knöpfel T (2010) GlyT2+ neurons in the lateral cerebellar nucleus. Cerebellum 9:42–55 6. Groh A, Meyer HS, Schmidt EF et al (2009) Cell-type specific properties of pyramidal neurons in neocortex underlying a layout that is modifiable depending on the cortical area. Cereb Cortex 20:826–836 7. Heintz N (2004) Gene expression nervous system atlas (GENSAT). Nat Neurosci 7:483 8. Arlotta P, Molyneaux BJ, Chen J et al (2005) Neuronal subtype-specific genes that control corticospinal motor neuron development in vivo. Neuron 45:207–221 9. Chen B, Schaevitz LR, McConnell SK (2005) Fezl regulates the differentiation and axon targeting of layer 5 subcortical projection neurons in cerebral cortex. Proc Natl Acad Sci U S A 102:17184–17189 10. Chen J-G, Rašin M-R, Kwan KY et al (2005) Zfp312 is required for subcortical axonal projections and dendritic morphology of deeplayer pyramidal neurons of the cerebral cortex. Proc Natl Acad Sci U S A 102:17792–17797 11. Molyneaux BJ, Arlotta P, Hirata T et al (2005) Fezl is required for the birth and specification of corticospinal motor neurons. Neuron 47:817–831 12. Molyneaux BJ, Arlotta P, Menezes JRL et al (2007) Neuronal subtype specification in the cerebral cortex. Nat Rev Neurosci 8:427–437 13. Leone DP, Srinivasan K, Chen B et al (2008) The determination of projection neuron identity

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

in the developing cerebral cortex. Curr Opin Neurobiol 18:28–35 Nelson SB, Sugino K, Hempel CM (2006) The problem of neuronal cell types: a physiological genomics approach. Trends Neurosci 29: 339–345 Rouaux C, Bhai S, Arlotta P (2012) Programming and reprogramming neuronal subtypes in the central nervous system. Dev Neurobiol 72:1085–1098 Özdinler PH, Benn S, Yamamoto TH et al (2011) Corticospinal motor neurons and related subcerebral projection neurons undergo early and specific neurodegeneration in hSOD1G93A transgenic ALS mice. J Neurosci 31:4166–4177 Franklin K, Paxinos G (2001) The mouse brain in stereotaxic coordinates, 2nd edn. London Academic Press, San Diego Anderson CT, Sheets PL, Kiritani T et al (2010) Sublayer-specific microcircuits of corticospinal and corticostriatal neurons in motor cortex. Nat Neurosci 13:739–744 Meijering E, Jacob M, Sarria JCF et al (2004) Design and validation of a tool for neurite tracing and analysis in fluorescence microscopy images. Cytometry A 58A:167–176 Chen B, Wang SS, Hattox AM et al (2008) The Fezf2-Ctip2 genetic pathway regulates the fate choice of subcortical projection neurons in the developing cerebral cortex. Proc Natl Acad Sci U S A 105:11382–11387 Alcamo EA, Chirivella L, Dautzenberg M et al (2008) Satb2 regulates callosal projection neuron identity in the developing cerebral cortex. Neuron 57:364–377 Britanova O, de Juan Romero C, Cheung A et al (2008) Satb2 is a postmitotic determinant for upper-layer neuron specification in the neocortex. Neuron 57:378–392 Han W, Kwan KY, Shim S (2011) TBR1 directly represses Fezf2 to control the laminar origin and development of the corticospinal tract. Proc Natl Acad Sci U S A 108: 3041–3046 McKenna WL, Betancourt J, Larkin KA et al (2011) Tbr1 and Fezf2 regulate alternate corticofugal neuronal identities during neocortical development. J Neurosci 31:549–564 Tantirigama MLS, Oswald MJ, Duynstee C et al (2014) Expression of the developmental

Immunocytochemistry and Live Imaging transcription factor Fezf2 identifies a distinct subpopulation of layer 5 intratelencephalicprojection neurons in mature mouse motor cortex. J Neurosci 34(12):4303–4308 26. Dombeck DA, Harvey CD, Tian L et al (2010) Functional imaging of hippocampal place cells at cellular resolution during virtual navigation. Nat Neurosci 13: 1433–1440 27. Muto A, Ohkura M, Kotani T et al (2011) Genetic visualization with an improved GCaMP calcium indicator reveals spatiotemporal activation of the spinal motor neurons in

373

zebrafish. Proc Natl Acad Sci U S A 108: 5425–5430 28. Akemann W, Mutoh H, Perron A (2010) Imaging brain electric signals with genetically targeted voltage-sensitive fluorescent proteins. Nat Methods 7:643–649 29. Akemann W, Mutoh H, Perron A (2012) Imaging neural circuit dynamics with a voltagesensitive fluorescent protein. J Neurophysiol 108:2323–2337 30. Knöpfel T (2012) Genetically encoded optical indicators for the analysis of neuronal circuits. Nat Rev Neurosci 13:687–700