A new versatile clearing method for brain imaging

A new versatile clearing method for brain imaging Irene Costantinia, Antonino Paolo Di Giovannaa, Anna Letizia Allegra Mascaroa, Ludovico Silvestria, Marie Caroline Müllenbroicha, Leonardo Sacconib,a, Francesco S. Pavonea,b,c* a

European Laboratory for Non-linear Spectroscopy, University of Florence, via nello carrara 1, 50019, Sesto Fiorentino, Florence, Italy b National Institute of Optics, National Research Council, Largo Fermi 6, 50125 Florence, Italy c Department of Physics and Astronomy, University of Florence, Via Sansone 1, 50019 Sesto Fiorentino, Italy ABSTRACT

Light scattering inside biological tissue is a limitation for large volumes imaging with microscopic resolution. Based on refractive index matching, different approaches have been developed to reduce scattering in fixed tissue. High refractive index organic solvents and water-based optical clearing agents, such as Sca/e, SeeDB and CUBIC have been used for optical clearing of entire mouse brain. Although these methods guarantee high transparency and preservation of the fluorescence, though present other non-negligible limitations. Tissue transformation by CLARITY allows high transparency, whole brain immunolabelling and structural and molecular preservation. This method however requires a highly expensive refractive index matching solution limiting practical applicability to large volumes. In this work we investigate the effectiveness of a water-soluble clearing agent, the 2,2'-thiodiethanol (TDE) to clear mouse and human brain. TDE does not quench the fluorescence signal, is compatible with immunostaining and does not introduce any deformation at sub-cellular level. The not viscous nature of the TDE make it a suitable agent to perform brain slicing during serial two-photon (STP) tomography. In fact, by improving penetration depth it reduces tissue slicing, decreasing the acquisition time and cutting artefacts. TDE can also be used as a refractive index medium for CLARITY. The potential of this method has been explored by imaging blocks of dysplastic human brain transformed with CLARITY, immunostained and cleared with the TDE. This clearing approach significantly expands the application of single and two-photon imaging, providing a new useful method for quantitative morphological analysis of structure in mouse and human brain. Keywords: clearing, two-photon, imaging, CLARITY, TDE, human brain, light sheet

1. INTRODUCTION Biological tissues are opaque because of light scattering, which is present because the refractive index of native proteins is different from that of the water surrounding them. To reduce scattering and make the tissue almost transparent, the water needs to be substituted with a medium of the same refractive index as the proteins. Based on refractive index matching, different approaches have been developed to reduce scattering in fixed tissue. At first, high refractive index organic solvents have been used for optical clearing of entire mouse brains[1, 2]. Although these methods guarantee high transparency, enabling imaging even with single-photon light sheet microscopy, fluorescence quenching and tissue shrinkage limit their applicability. Later, water-based optical clearing agents, such as Sca/e[3], and SeeDB[4], were developed allowing better preservation of fluorescence though presenting other non-negligible limitations. Indeed they require long incubation times (from several weeks up to months), produce structural alteration due to large tissue expansion and introduce technical limitations because of their high viscosity.

*[email protected]; phone +39 055 457 2480; www.lens.unifi.it/bio

Optical Techniques in Neurosurgery, Neurophotonics, and Optogenetics II, edited by Henry Hirschberg, Steen J. Madsen, E. Duco Jansen, Qingming Luo, Samarendra K. Mohanty, Nitish V. Thakor, Proc. of SPIE Vol. 9305, 930514 · © 2015 SPIE · CCC code: 1605-7422/15/$18 · doi: 10.1117/12.2076908 Proc. of SPIE Vol. 9305 930514-1 Downloaded From: http://spiedigitallibrary.org/ on 08/20/2015 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

The challenge of producing large, transparent and fluorescently labeled volumes has been recently addressed by a new approach called CLARITY[5, 6]. It is a method for the transformation of intact tissue into a nanoporous, hydrogelhybridized, lipid-free form that is fully assembled but optically transparent and macromolecule-permeable. Packed lipid bilayers are implicated in rendering tissue poorly accessible to chemical penetration as well as disadvantageous lightscattering properties at the lipid–aqueous interface. By removing these lipid bilayers CLARITY allows high transparency, whole brain immunolabeling and structural and molecular preservation. This method, however, requires an expensive refractive index matching solution (FocusClear TM) limiting its practical applicability to large volumes. In this work we investigated the effectiveness of a new water-soluble clearing agent, 2,2'-thiodiethanol (TDE)[7], to clear mouse and human brain samples. The chemical characterization of this solvent shows that it does not quench the fluorescence signal, is compatible with immunostaining and does not introduce any deformation at sub-cellular level. The non-viscous nature of the TDE and the possibility of finely adjusting its refractive index make it a suitable agent to perform brain slicing during serial two-photon (STP) tomography[8]. In fact, by improving the imaging penetration depth this approach reduces tissue slicing and decreases both acquisition time and cutting artefacts. TDE can also be used as a refractive index medium for CLARITY allowing imaging with light sheet microscopy (LSM). The potential of this method applied to two-photon fluorescence microscopy (TPFM) has been explored by imaging a whole mouse hippocampus with STP at high resolution and a whole transgenic mouse brain treated with CLARITY was analyzed with LSM. Finally a block of dysplastic human brain transformed with CLARITY, immunostained with different antibodies was analyzed.

2. MATERIALS AND METHODS 2.1 Specimen collection The experimental protocols involving animals were designed in accordance with the laws of the Italian Ministry of Health.The human brain sample was removed from a child with symptomatic drug-resistant epilepsy due to hemimegalencephaly, a severe malformation of cortical development involving one entire hemisphere. The sample was obtained after informed consent, according to the guidelines of the Human Research Ethics Committee of the A. Meyer Children’s Hospital. 2.2 Tissue preparation Adult mice (p56) were deeply anesthetized with an intraperitoneal injection of ketamine (90 mg/kg) and xilazine (9mg/kg). They were then transcardially perfused with 100 ml of ice-cold 0.01 M phosphate buffered saline (PBS) solution (pH 7.6), followed by 100 ml of freshly prepared ice-cold paraformaldehyde (PFA) 4% in 0.01 M PBS (pH 7.6). The brain was extracted from the skull and fixed overnight in 20 ml of PFA 4% at 4 °C. Samples were then rinsed three times (30 minutes each) in 20 ml of 0.01M PBS at 4°C. The brains were stored in 20 ml of 0.01M PBS at 4°C. Mouse brain CLARITY samples were prepared according to the Chung protocol[5]. Human brain samples were prepared with passive CLARITY (PC) protocol and subsequently immunostained with Chung protocol [5]. 2.3 Optical clearing with TDE For TPFM imaging, murine PFA-fixed samples and human brain samples were cleared with serial incubations in 20 ml of 20% and 47% (vol/vol) 2,2'-thiodiethanol in 0.01M PBS (TDE/PBS), each for either 1 hour at 37°C or for 12 hours at room temperature (RT) while gently shaking. For LSM imaging, CLARITY-processed murine brain samples were cleared with serial incubations in 50 ml of 30% and 63% (vol/vol) 2,2'-thiodiethanol in 0.01M PBS (TDE/PBS), each for 1 day at 37°C while gently shaking. 2.4 Two-photon fluorescence microscopy A mode locked Ti:Sapphire laser (Chameleon, 120 fs pulse width, 90 MHz repetition rate, Coherent, CA) was coupled into a custom-made scanning system based on a pair of galvanometric mirrors (VM500+, Cambridge Technologies, MA). The laser was focused onto the specimen by a water immersion 20x objective lens (XLUM 20, NA 0.95, WD 2mm, Olympus, Japan) for uncleared (PBS) sample imaging or a tunable 20x objective lens (Scale LD SC PlanApochromat, NA 1, WD 5.6mm, Zeiss, Germany) for cleared (47% TDE/PBS) sample imaging. The system was equipped with a motorized xy stage (MPC-200, Shutter Instrumente, CA) for axial displacement of the sample and with a closed-loop piezoelectric stage (ND72Z2LAQ PIFOC objective scanning system, 2mm travel range, Physik Instrumente, Germany) for the displacement of the objective along the z axis. The fluorescence signals were collected by

Proc. of SPIE Vol. 9305 930514-2 Downloaded From: http://spiedigitallibrary.org/ on 08/20/2015 Terms of Use: http://spiedigitallibrary.org/ss/TermsOfUse.aspx

two photomultiplier modules (H7422, Hamamatsu Photonics, NJ). The instrument was controlled by custom software, written in LabView (National Instruments, TX). To perform serial two-photon tomography stacks of each layer of sample were acquired using a program allowing for automatic acquisition of adjacent regions drawing a spiral square. Each stack had a depth of 1000 µm with a z displacement of 4 µm between images. Each frame had a field of view of 300 x 300 µm2, adjacent stacks had an overlap of 30 µm. To ensure an efficient 3D reconstruction along the z axis, slicing was performed every 800 µm such that the subsequent layer had an overlapping region of 200 µm with the previous one. 2.5 Light-sheet microscopy Specimens were imaged using a custom-made confocal light sheet microscope (CLSM) described in Silvestri[9]. The light sheet was generated by scanning the excitation beam with a galvanometric mirror (6220H, Cambridge Technology, MA) and confocality was achieved by synchronizing the galvo scanner with the line read-out of the sCMOS camera (Orca Flash4.0, Hamamatsu Photonics, Japan). Five different cw wavelengths were available (MLDs and DPSSs, Cobolt, Sweden) for fluorescence excitation and an acousto-optic tunable filter (AOTFnC-400.650-TN, AA Opto-Electronic, France) was used to regulate laser power. The excitation light was focused with a long working distance, low magnification objective (10x 0.3NA WD 17.5mm, Nikon, Japan) and fluorescence was collected on a perpendicular axis with a specialized objective for high refractive index immersion and a correction collar for refractive indices ranging from 1.41 to 1.52 (XLSLPLN25XGMP, 25x 1.0NA, WD 8mm Olympus, Japan). The samples were mounted on a motorized x-, y-, z-, θ-stage (M-122.2DD and M-116.DG, Physik Instrumente, Germany) which allowed free 3D motion plus rotation in a custom-made chamber filled with 63% TDE/PBS. The microscope was controlled via custom written LabVIEW code (National Instruments) which coordinated the galvo scanners, the rolling shutter and the stack acquisition. 2.6 Data analysis Stitching of different stacks were perform with Terastitcher software[10]. Graphs and data analysis were done with OriginPro 9.0 (OriginLab Corporation). Stacks were analyzed using both Fiji (http://fiji.sc/Fiji) and Amira 5.3 (Visage Imaging) software. 3D renderings of stitched images were produced using the Amira Voltex function.

3. RESULTS 3.1 Two-photon serial sectioning with two-photon fluorescence microscopy In this work we proposed a novel and versatile method to clear brain tissue for two-photon fluorescence imaging, we combined the TDE clearing method with STP to benefit from increased imaging depth which consequently lead to reduced tissue slicing and a decrease in acquisition time and cutting artefacts, allowing for the complete reconstruction of a big volume at high resolution with minimal loss of information. To illustrate this, we reconstructed a mouse hippocampus dissected from a PFA fixed Thy1-GFP-M mouse brain (Fig.1a). With a complete tomography it was easily possible to recognize the anatomical characteristics areas of the hippocampus (Fig.1b), moreover being constituted of high resolution stacks (Fig.1c) anatomical details of the sample such as spines and varicosity were well defined. 3.2 Whole mouse brain imaging with light sheet microscopy To image a whole mouse brain we combined the TDE clearing method with the CLARITY technique[5] as alternative of the optical clearing medium suggested (FocusClearTM), extremely expensive solution whose recipe is unknown due to copyright. TDE offers a cheap alternative to make CLARITY-treated tissue transparent. We selected a solution of 63% of TDE/PBS because this percentage results in the same refractive index as FocusClearTM (RI=1.45). We applied this clearing protocol in combination with light sheet microscopy to reconstruct a whole Thy1-GFP-M mouse brain (Fig.2a). It was possible to recognize micrometer sized feature such as dendrites and axons (Fig. 2b,c). 3.3 Human brain imaging TDE clearing is compatible with immunostaining, the solution did not modify the protein-antibody interaction. We combine the clearing with passive CLARITY protocol and antibody staining on human brain. A 2 mm thick block of


cortex from a hemimegalencephaly patient, was treated with the passive CLARITY protocol, stained with different antibodies and cleared with TDE solution. We were able to stain the tissue with a single antibody against parvalbumin (PV) or glial fibrillary acidic protein (GFAP) and DAPI (Fig.3a,b). We used high concentration of the antibody to penetrate deep into the tissue and we were able to image a 1 mm3 of tissue with the TPFM (Fig.3c). High resolution images point up the cyto-architecture alteration of the dysmorphic neurons that are characteristic of this pathology (Fig.3d,e).

Figure 1. Hippocampus tomography. Reconstruction of entire Thy1-GFP-M mouse hippocampus fixed with PFA and cleared with TDE 47% (20X Scale objective, two-photon excitation). (a) 3D rendering of 6 layers of 1 mm depth each, sampled every 4 µm. Serial sectioning at 800 µm depth. (b) Horizontal view of a section of the hippocampus at 1200 µm depth after 3D reconstruction, scale bar = 300 µm. (c) High magnification inset corresponding to the red box in b, scale bar = 50 µm.


Figure 2. Mouse brain tomography. (a) Left half of the brain from an adult Thy1-GFP-M mouse, characterized by a sparse unspecific neuronal EGFP labeling, scale bar = 2 mm. (b,c) Intermediate and high resolution view of portions of the red square shown in a and b respectively, scale bar = 500 µm and 50 µm.


Figure 3. Human brain tomograpy. A 2 mm thick block of a formalin-fixed tissue of a patient with hemimegalencephaly (HME), treated with passive CLARITY protocol, immunostained with different antibodies and cleared with TDE 47% (20X Sca\e objective, two-photon excitation). Maximum intensity projections (MIP) of collected stacks (100 µm depth, z-step 2 µm) stitched together , scale bar = 300 µm: (a) Parvalbumin (PV) staining in red and nuclei (DAPI) in green; (b) Glial fibrillary acidic protein (GFAP) in red and nuclei (DAPI) in green. (c) 3D view of 1 x 1 x 1 mm3 of tissue label for PV in red and DAPI in green. (d,e) MIP of a stack (100 µm depth, z-step 2 µm) of PV and GFAP in red and DAPI in green, scale bar = 50 µm.

CONCLUSIONS In the last few years the characterization of brain architecture has been address with the development of different microscopy techniques coupled with various methods for sample preparation. In particular several clearing protocols have been published to address the challenge of brain structure reconstruction. Each of them has distinctive


characteristics that make it suitable for a particular application while limiting their use for another. In this work we presented a simple, quick and inexpensive clearing method based on TDE as refractive index matching agent. Using the TDE protocol, we improved the STP technique described by Ragan et al [8]. We were able to reduce acquisition time and tissue slicing and consequently also decrease distortions due to cutting artefacts. The imaging overlap between adjacent layers allowed for the complete, interpolation-free reconstruction of an entire PFA fixed Thy1GFP-M mouse hippocampus at high resolution with minimal loss of information. The complete tomography of this region allowed for anatomical study. This method can be applied to study specific neuroanatomical structures but also for whole mouse brain reconstruction. The limiting factor was the long acquisition time due to the point-scanning nature of the STP and the big data analysis generated with this technique. In combination with the CLARITY method TDE allows whole mouse brain imaging. It is a valid alternative to FocusClearTM as refractive index matching solution, considerably lowering the cost of every experiment and making affordable large volume analysis with light sheet microscopy. Using the same mounting medium for different technique as TPFM and LSM, it is possible to perform correlative approaches, this combination allow to overcome the inherent limitation of single methods, allowing a better understanding of the brain[11]. We tested the compatibility of TDE clearing with immunostained in human brain samples permeabilized with the passive CLARITY method and cleared with TDE. We were able to homogenously stain a volume of 1 mm 3 therefore improving the penetration depth achievable in human brain tissue previously described[5] by a factor of two. With the volumetric continuity of this technique, it was possible to have a 3D view of the brain connectivity in human tissue and follow the neuron fibers through the volume to characterize their organization. Compared with other technique, our TDE protocol covers a wide range of applications, is suitable with immunostaining and did not lead to bleaching or fluorescence quenching. It is therefore suitable for transgenic animals and human brain samples. Due to the possibility of refractive index adjusting, this procedure can be used for sample preparation for both serial two-photon tomography and light sheet microscopy. We believe that TDE protocol can contribute to enlarge the understanding of anatomic structure and connectomics of the brain. The usefulness of TDE may not be limited to brain neuroanatomy investigation but can span different area of research.

ACKNOWLEDGEMENTS The research leading to these results has received funding from the European Union Seventh Framework Program (FP7/2007-2013) under grant agreements no. 604102 (Human Brain Project) and n° 284464 (LASERLAB-EUROPE). The research has also been supported by the Italian Ministry for Education, University and Research in the framework of the Flagship Project NANOMAX, by ‘‘Ente Cassa di Risparmio di Firenze’’ (private foundation) and by Regione Toscana (grant number: POR-CreO 2007–2013).

REFERENCES

[1] [2]

[3]

K. Becker, N. Jahrling, S. Saghafi et al., “Chemical clearing and dehydration of GFP expressing mouse brains,” PLoS One, 7(3), e33916 (2012). H. U. Dodt, U. Leischner, A. Schierloh et al., “Ultramicroscopy: three-dimensional visualization of neuronal networks in the whole mouse brain,” Nat Methods, 4(4), 331-6 (2007). H. Hama, H. Kurokawa, H. Kawano et al., “Scale: a chemical approach for fluorescence imaging and reconstruction of transparent mouse brain,” Nat Neurosci, 14(11), 1481-8 (2011).


[4] [5] [6] [7] [8] [9] [10] [11]

M. T. Ke, S. Fujimoto, and T. Imai, “SeeDB: a simple and morphology-preserving optical clearing agent for neuronal circuit reconstruction,” Nat Neurosci, 16(8), 1154-61 (2013). K. Chung, J. Wallace, S. Y. Kim et al., “Structural and molecular interrogation of intact biological systems,” Nature, 497(7449), 332-7 (2013). R. Tomer, L. Ye, B. Hsueh et al., “Advanced CLARITY for rapid and high-resolution imaging of intact tissues,” Nat Protoc, 9(7), 1682-97 (2014). T. Staudt, M. C. Lang, R. Medda et al., “2,2'-thiodiethanol: a new water soluble mounting medium for high resolution optical microscopy,” Microsc Res Tech, 70(1), 1-9 (2007). T. Ragan, L. R. Kadiri, K. U. Venkataraju et al., “Serial two-photon tomography for automated ex vivo mouse brain imaging,” Nat Methods, 9(3), 255-8 (2012). L. Silvestri, A. Bria, L. Sacconi et al., “Confocal light sheet microscopy: micron-scale neuroanatomy of the entire mouse brain,” Opt Express, 20(18), 20582-98 (2012). A. Bria, and G. Iannello, “TeraStitcher - a tool for fast automatic 3D-stitching of teravoxelsized microscopy images,” BMC Bioinformatics, 13, 316 (2012). L. Silvestri, A. L. Allegra Mascaro, I. Costantini et al., “Correlative two-photon and light sheet microscopy,” Methods, 66(2), 268-72 (2014).
