Limits of Vibrating Microtomy

Abstract

The Vibratome, or vibrating microtome allows sectioning of the mouse brain with a reliably established thickness from 50 µm, but they have a strong limitation on the size of the sample suitable for cutting. Herein is described the construction with publicly available element base (from parts for a 3D printer) a so-called 3D-vibrating microtome capable of cutting larger size brain into thin sections, and use it to investigate the limit-attainable values of the minimum slice thickness. Both of these goals have been successfully achieved. Both small mouse slices with a minimum thickness of 30 μm and large whole calf brain slices with a minimum thickness of 150 μm were obtained. Critical features of the effect of blade vibration frequency and sample feed rate on the minimum slice thickness were revealed, In the same way, a clear correlation was established between the minimum achievable thickness and its area:

Introduction

The vibratome, or vibrating microtome, is an indispensable instrument in certain types of histological examinations. Vibrational microtomes are used to cut fresh and fixed tissue specimens that have not been subject to embedding. A vibrating microtome is capable of cutting high-quality sections of greater thickness than can in some instances be obtained with frozen of paraffin-embedded specimens. One utility of the vibrating microtomy is that only means to cut sections with whole, preserved membranes. Classical paraffin-embedded microtomy and cryomicrotomy are not capable of this, as the integrity of the membranes is irreversibly compromised. The preservation of membranes is of exceptional, critical importance in the case of the study of nerve connections using carbocyanine dyes such as DiI, DiO, and DiA.1

Currently, researchers have access to a range of commercially available vibrating microtomes. The following question naturally follows from this variety of models in the context of CNS research: defining the physical limits of slicing brain tissue with a particular slicer. Defining the minimum and maximal achievable slice thickness and what are the maximum dimensions for a whole slice obtained with a particular instrument. Equally important, what are the optimal parameters of such a slice at the operating limits of the vibrating microtome.

Unfortunately, a functional guide to critical elements in the use of vibrating microtomes is lacking. There are no systematic studies on minimal achievable slide thickness and little if any guidance for parameters of cutting. Since the majority of vibrating microtome are designed to work with small objects such as mouse or rat brain samples, guidance on their application is limited, manuscripts such as Chen et al.,2 where serial sections of 50 µm thick mouse brains were obtained on a Leica VT1200S vibrating microtome. Sectioning conditions and parameters were not specified in this paper. In another paper,3 30-µm thick sections of rat brain were obtained on a Leica VT1000S vibrating microtome, which by itself is already a non-trivial task (at the limit of the device operation), but nothing was said about the cutting parameters and conditions, and no overview photos of the sections were given.

Li et al.4 uses an experimental device of non-standard design. The authors of the paper emphasize that they created a special vibration-resistant mechanism (DP-DP flexure), and using a voice-coil motor, managing to achieve ultra-high frequency of blade vibration (180 Hz), allowing to slice a mouse brain sample into thin sections of 10 µm thickness at a speed of more than 50 mm/min.

Few models of vibrating microtomes are available that are suitable for slicing macro objects (70 × 70 mm or larger). Commercially available models include the Leica 1000S (with knife holder L, can cut a 70 × 40 mm sample) and the DSK DTK-3000Wwhich can cut a 70 × 70 mm sample). Both of these models have a very limited upper limit on the sample size that can be sliced. The sectioning of large brain samples is limited to the Compresstome VF-800 vibrating microtome. There are exceptionally few papers describing the use of Compresstome-VF800, only one.5 In this work, brain images from 10-year-old rhesus macaques were used and cut into 250 slices of 300-µm thick. No data on cutting conditions and modes were given in the paper.

The main purpose of this article is to define the value of the minimum achievable thickness of a slice and its maximum achievable (in terms of area) dimensions, as well as to clearly describe some universal quality criteria to judge the suitability of a slice for microscopic analysis.

A number of challenges were encoutntered. Testing of the existing line of commercial vibrating microtomes is extremely difficult, because of the objective impossibility to simultaneously test all available models of vibrating microtome. The concept of a vibrating microtome is neither exceptionally complicated non-unique from a mechanical point of view. In fact, the basis of any vibrating microtome is an oscillating blade that impinges on the sample (or the sample impinges on the blade), and the blade (or sample) must also be able to move along the z-axis by micron-sized increments.

To overcome these issues, we created a kind of universal test bench, easy to build and capable of slicing brain (or any other organ) samples from any organisms and without limitations. Based on the general concept of the vibrating microtome. This lead to the observation that construction of a vibrating microtome can be accomplished with an aluminum metal structure and optomechanical platforms (all these are commercially available mechanisms) and at the same time determine the extreme values of parameters at which, in principle. The final observation is that because the basic concept of all models is generally common the performance characteristics are generally the same, only limited by the design features of the instrument.

In the studies here, Mouse C57Bl6/CBA brain samples at day 11 of postnatal development (P11) were used to determine the value of the minimum achievable slice thickness of small objects. A portion of mouse brain samples and individual vibrational sections were stained using Nissl staining and the fluorescent carbocyanine dye DiI, these stains were necessary to determine microscopic assessment of slice quality. Vibratome sections from the 3D-vibrating microtome were also compared with equal thickness sections obtained on a Leica SM2000R cryo-sliding microtome and on a commercially available Campden Instruments HA752 vibrating microtome, which in turn were also stained with Nissl and DiI stains. Slices for the study were obtained by slicing a calf brain sample to evaluated the external characteristics of slices (slice integrity and surface profile), peculiarities of cutting parameters during their obtaining, and search for the value of the minimum achievable thickness for slices.

Materials and Methods

In this study, the Bioethics commission of the Faculty of Biology of Moscow State University has approved all procedures on animals. Such procedures on animals as perfusion and euthanasia were carried in the Institute of the Mitengineering of the Moscow State University in accordance with the Directive 2010/63 / EU of the European Parliament and the Council of the European Union for the protection of animals used for scientific purposes. All the work was carried out on mice—hybrids F1—C57Bl6/CBA. The total number of animals was 8. The birthday of the mice was considered a one postnatal day P1, all procedures were performed on mice on the eleventh postnatal day P11. A list of all mice used and manipulations performed on them is given in Table 1. In addition, one sample of calf brain were used (one sagittal half of the whole brain cut along the midline). The sample was purchased at the commercial market in Moscow (Danilovsky Rynok, JSC, TIN 7725582006) in accordance with all regulatory procedures for biological material transfer.

Table 1.

Mouse Brain Samples Used in This Study Depending on the Research Aim of the Study.

Brain Specimen	Microtome Type	Slice Thickness	Study Objective
Case 1 - P11	3D-vibrating microtome	30 μm	Study of minimum slice thickness limits
Case 2 - P11	3D-vibrating microtome	50 μm	Study of minimum slice thickness limits + Examination of membrane preservation on the slice with DiI + Nissl staining
Case 3 - P11	Leica SM2000R	50 μm	Study of membrane preservation on a slice using DiI +Nissl staining
Case 4 - P11	3D-vibrating microtome	150 μm	Comparison of uncoloured slices of the same thickness between different vibrating microtomes
Case 5 - P11	Campden Instruments HA752	100 μm and 150 μm	Comparison of uncoloured slices of the same thickness between different vibrating microtomes
Case 6 - P11	3D-vibrating microtome-Renovator	20–80 μm	Study of slicing a brain specimen at a blade oscillation frequency of 180Hz
Case 1—P11-DiI	3D-vibrating microtome	150 μm	Comparison of the quality of DiI stained slices between different vibrating microtomes using confocal microscopy
Case 2—P11-DiI	Campden Instruments HA752	150 μm	Comparison of the quality of DiI stained slices between different vibrating microtomes using confocal microscopy

All mice (P11) were given intraperitoneal solutions of zoletil 100 (Vibrac) in saline (0.9% NaCl) at the rate of 10 mg per 1 kg of body weight before perfusion. Then mice were subsequently perfused with saline solution (0.9% NaCl) through the left ventricle of the heart to remove blood from the blood vessels. Thereafter, the mice were decapitated and the heads were placed in 4% paraformaldehyde (PAF, PanReac AppliChem, СAS 30525-89-4) on 0.1 M PB (pH 7.4) for up to 1 month term at room temperature for fixation. After fixation, the brain was removed from the skull and stored in 4% PAF. Following transfer from a commercial source, calf brain samples were immediately placed in 4%PAF for a minimum of 1 week waiting for further slicing on the 3D-vibrating microtome. The linear dimensions of the mosaic specimen at the most extreme points after soaking in PAF (one sagittal half) are ~ 150 mm along the rostro-caudal axis, ~ 80 mm along the dorso-ventral axis, ~ 60 mm along the medio-lateral axis.

The mice brain was placed in a warm (45°C) solution of 5% fusible agarose (Serva CAS [9012-36-6]) which after cooling formed a block of the desired shape around the brain. This agarose block was placed into a block holder. The block was fixed to the holder with using superglue, and emersed in saline (0.9% NaCl) and then the brain in the agarose block was sequentially cut into sections. All mouse brain samples were sliced using a steel trapezoidal construction knife blade—Dexter SK5 (Fig. 1A, trapeze blade, cutting surface length 60 mm). The blade angle was 25° to 30°.

Figure 1.

A - Dexter’s SK5 segmented and trapezoidal blades used to slice brain samples using a 3D-vibrating microtome. B - is a schematic drawing of the medial surface of the calf brain (sagittal cut); the vertical lines in the drawing symbolize the places of cuts in the frontal plane. The area between the two vertical lines corresponds to a brain fragment 20-25 mm thick. This thickness of a large brain fragment is the most convenient for cutting on a 3D- vibrating microtome (the drawing is taken from Yoshikawa atlas, 1968). C - photo of an excised frontal section of the forebrain of a calf; this brain fragment is almost completely cleared from the external vasculature. The surface of the incision is not perfectly flat, and there are numerous curves on it. These bends inevitably occur during fixation of the brain fragment in 4% paraformaldehyde, even if the incision is initially made well. D - a calf brain fragment poured in 5% agarose, top view. The lower surface of the brain fragment is located directly at the bottom of the plastic tub in which the fragment is placed. Above the upper surface of the brain fragment there should be at least a 10 mm layer of 5% agar-agar. E - a calf brain fragment poured into 5% agarose and placed in the 3D vibrating microtome tub. For proper orientation, the original agarose-filled brain fragment was rotated horizontally by 180 degrees, so that its lower surface, originally bordering the plastic bottom of the tray, becomes upper. This rotation is necessary because there must be a layer of at least 10 mm thick of solidified 5% agarose between the bottom of the 3D- vibrating microtome bath and the bottom (looking at the bottom of the bath) surface of the brain fragment. This layer provides freedom of passage of the blade along the z-axis and gives room to "maneuver" (so that the blade would not rest against the bottom of the tub vibrating microtome too early).

A fresh brain sample (one sagittally cut half) was cut with an ordinary kitchen knife into fragments approximately 8 × 6 × 2 cm (96 cm³ volume); these fragments were then immersed in 4% PAF for at least one week. Fragments of different brain sections were used to practice the cutting technique; the most presentable fragment was cut serially. Prior to cutting, the cut brain fragment was dissected, using tweezers, from the dura mater, primarily from the cerebral vasculature. The presence of the cerebral vasculature or its fragments significantly hindered the cutting. The vasculature does not cut well with the 3D-vibrating microtome blade, and usually pulls behind the blade, cutting the tissue in the most inappropriate way, introducing defects into it. Therefore, removal of the brain sheath is absolutely necessary, otherwise the sections have defects. After removal of the vascular membrane, excised brain fragments 96 cm3 volume were poured into a fusible agarose block with a final size of 9 × 7 × 4 cm (252 cm3 volume). For comparison, the transverse size of the adult human brain is 13.5–14.5 cm, that is, the diameter of a ½ calf brain sample cast into agarose is approximately one third less than the diameter of the human brain in frontal projection. After the block was attached to the bottom of the bath of the 3D-vibrating microtome with superglue, the frontal surface was oriented to the bottom, the dorso-ventral axis was perpendicular to the cutting surface of the blade (Fig. 1B-E). A segmented construction knife blade, Dexter SK5 (Fig. 1A; blade width 25 mm, cutting surface length 120 mm), was used to slice the selected calf brain slice into slices. The angle of inclination of the blade was 25°.

Preparation of 50 μm-thick frontal frozen sections of mouse brain P11 was performed using a Leica SM2000R microtome with an OMT-28 freezing stage. For this procedure, the P11 mouse brain sample required pretreatment ⁶ : The brain sample is first placed in a 10% sucrose solution (SS) in 4% PAF (CF, Lenreaktiv. GOST 5833-75), followed by the next day the brain sample was transferred to a 20% SS solution and subesquently on the third day, the brain sample was placed in a 30% SS solution for one day to one week. The volume of the SS solution was about 20 times the volume of the mouse brain. Once the brain was impregnated with 30% SS (within a day), it could be used for cutting. Before cutting, a gelatin substrate was prepared on the cryo-sliding microtome table, and the brain sample was placed on it. Freezing was done slowly and evenly, the cutting temperature was −15°C.

Nissl staining is excellent for studying cytoarchitectonics and myeloarchitectonics of nerve tissue. Medium thickness sections (50 μm) are used for Nissl staining. Both vibratomic and cryotomic sections of the same thickness (50 μm) were stained using the same protocol ⁷ to compare and detect differences in tissue consistency. In all cases, sections from P11 mice were stained. The staining procedure is as follows: Soak the section in 1% neutral red solution (Lenreaktiv, Technic Specifications 6-09-07-1634-87) in 0.1 M phosphate buffer (pH 4.8): 3 minutes, wash the sections in distilled water: 10 seconds followed by dehydration of the section in 50%, 70%, 90% alcohol: 3 minutes in each alcohol followed by a additional washing in absolute alcohol: 1.5 minutes each. Sections are clarified in O-xylene and incubation under a coverslip.

For the injection of DiI (1,1′-Dioctadecyl3,3,3′,3′-Tetramethylindocarbocyanine Perchlorat, Sigma, Cas number 41085-99-8) into the selected nucleus of the brain, the brain must be properly prepared. Preparation consists in finding the level in the brain where the selected nucleus is located. After finding the level, a marker was injected into the nucleus. Finding the right brain level is only necessary for cases with DiI application. In this work, DiI applications were done unilaterally on Medial Habenula (MHb; one sample), and bilaterally on Medial Habenula and on Lateral Habenula (LHb, one sample). To search for the desired brain level (in case of applying the marker to the frontal surface of the brain), the same preparatory and technical manipulations were carried out as described as above. All samples of the mouse brain were cut coronary and only to the required level corresponding to the caudal part of MHb. Tracking the desired level was controlled by evaluating the morphology of the obtained sections under a stereomicroscope along with cutting. After finding the desired level, brain samples were removed from the vibrating microtome bath and placed in labeled bottles (20 ml) with 4% PAF until DiI was applied.

In total, DiI was applied to 2 brain samples at P11 time point. DiI applications were made to the frontal brain section at the medial Hb level. In the case 1-P11-DiI brain sample DiI applications were made unilaterally in the MHb, in the case 2-P11-DiI brain sample applications were made in the MHb and LHb nuclei of the neighboring hemispheres. Injections were made with undissolved DiI crystals from a 70% alcohol solution. Undissolved crystals here refers to the film of dye that settles on the surface of the injection needle from a saturated solution of 70% ethanol as it evaporates (which occurs very rapidly). The method of injection with settled crystals from solution allows for very precise injections, which has been described in detail earlier in the article. ⁸

When applying DiI to cut sections: These applications were necessary to detect differences in DiI staining quality between cryosections and vibrating sections. For this purpose, dry dye crystals (mostly in TS) were placed on individual brain sections with a glass needle. A total of two series of DiI-labeled sections were made: a series of 50 μm thick mouse forebrain sections made on a Leica SM2000R cryo-sliding microtome; a series of 50 μm thick mouse forebrain sections made on a 3D-vibrating microtome.

After the DiI injection, the agarose block with the brain (and individual sections with the application of DiI crystals) was neatly placed in 4% PAF for 0.1M PB for a period of 2 months at a temperature of 25°C.

All mouse brain sections, both without marker application and with DiI application, were placed under a coverslip in Mowiol-4.88 mounting medium (Kremer; CAS number 25213-24-5). The 2.5% glycerol-based Mowiol-4.88 solution is a standard (http://cshprotocols.cshlp.org/content/2006/1/pdb.rec10255). Transparent mounting Nissl-stained mouse brain sections were incubated under a coverslip in O-xylene. Medium Calf brain sections were not encased but were evaluated directly in aqueous solution.

To assesess section thickness, individual mouse brain sections of different thicknesses (30 μm, 50 μm, 150 μm) or a calf brain slice (150 μm thickness) were encased in a block of liquid warm 5% agarose solution (45°C). After solidification, the block was sliced along the midline of the slice perpendicular to the slice surface. An ocular micrometric grid of 100 squares (the length of one edge of a square was 1000 µm) was further superimposed on the dissected slice surface. The block together with the superimposed grid was evaluated in a direct microscope (Leitz Laborlux S) with upper focus at a magnification of 10×. Based on the length of the square face on the grid, it was possible to derive the exact value of the slice thickness by direct visual observation.

A series of Nissl stained mouse brain sections were assessed under a Leitz Laborlux S microscope in transmitted light. Evaluation in all cases was performed through an Leitz PL objective (2.5×); Leitz EF objectives (4×) and Zeiss Neofluar objective (10×, 20×, 50×). Digital images of microscope slides of a mouse brain were obtained using a MG3CMOS06300KpA-2018 camera (ISOLAB 6.3MP, Sony IMX178 (C) sensor) and ImageView software. Digital images of large total calf brain slices were obtained using a Ipad A2197 camera. Total slices were taken in a bath with saline placed. Computer program Photoshop CS3 (Adobe, USA) was used for processing and editing of digital images and for illustration of the results. Confocal images of DiI stained slices were obtained on a Zeiss LSM880 confocal microscope (20x/0.8 objective, 543 nm laser extinction). Series of some mouse brain sections with DiI were evaluated under fluorescent lighting on an Leitz Laborlux S microscope equipped with a fluorescence module with an Leitz TRITZ cube (Еxitation filter 535/50 Emission filter 610/75).

Results

The Logic of Building a 3D-Vibrating Microtome

The fundamental concept of any vibrating microtome is based on a blade that makes sawtooth, reciprocating movements with a fixed amplitude along the y-axis (some Compresstome models have a slightly different arrangement of axes, but the principle of operation is identical). The whole construction of the vibrating microtome can be divided into 4 modules:

1. The supporting metal structural frame is the “first module.”

2. The mechanism that drives the blade in the y-axis is the “second module,” this is for example a SEMX60-AC linear translator (or AL, or AR, any of them will do) or any Renovator.

   A linear translator is a part consisting of a fixed base and a movable platform, being moved away from the base by means of an adjusting screw. There are usually bearings between the base and the platform, allowing the platforms to slide smoothly against each other. A blade is slid onto the linear translator. In the case of a 3D-vibrating microtome, the adjusting screw and internal springs are removed from the SEMX-60AC linear translator (any version of the translator) to facilitate the blade movement.

   Renovator is an oscillating tool, with a spindle that oscillates tangentially within a center angle of about 3° to 4°. The spindle is fitted with a blade. The Renovator head (key element) is supplied as a single monolithic mechanism, capable of being used for the 3D-vibrating microtome in a ready-to-use form without interfering with the internal structure of the 3D-vibrating microtome.

3. Movable platform for moving the specimen along the x-axis. Any linear guides with a carriage on linear bearings will do here. Actually, the moving platform is the same linear translator, only without internal springs and thus greatly increased in size. Linear guides are nothing more than rails on which bearings are mounted, and the widest variations in the size and choice of guides and bearings are allowed. That is their advantage.

4. Finally, the mechanism that enables the z-axis movement of the blade or sample is the “fourth module” (e.g. linear translator SEMX60-AC)

By combining these modules, a vibrating microtome of any complexity can be assembled. A total of three versions of vibrating microtomes were assembled, their detailed descriptions are given in Figs. 2, 3, 4 and Tables 2, 3, 4.

Figure 2.

A - It is a front view of the 3D- vibrating microtome. Explanations of abbreviations for the aluminum profile and other parts are given in Table 2. B - An overview diagram of the second module of the 3D- vibrating microtome. This design is based on an electric motor (BRS-775SH), on the shaft of which two flanges are mounted, oriented with their ends to each other, and then an eccentric goes. The first flange is fixed on the motor shaft using two M3x5 screws, two M3x20 screws and two self-locking M3 nuts connect the flanges to each other, and two more M3x5 screws fix the eccentric on the second flange. To improve the ergonomic design, simple M3 nuts (not shown) can be installed between the first and second flange. Two flange bearings(MF106ZZ) are mounted on the eccentric center shaft and a connecting rod (Makita 4500) is fitted on them. The connecting rod and flange bearings are fixed on the eccentric with a self-locking M3 nut. Thus, we get a structure, where the central axis of the eccentric is displaced relative to the central axis of the electric motor by a given value (0.5mm). C - Technical drawing of the eccentric with indication of all dimensions (in centimeters). The eccentric consists of three shafts (counting from left to right). A self-locking nut M3 is screwed onto the first shaft (diameter M3, threaded), flange bearings (MF106ZZ) and a connecting rod (Makita 4500) are mounted on the second shaft (diameter Ф6, smooth surface). The third shaft (diameter F4, smoothsurface) is inserted into the central hole of the flange, where it is fixed with two M3x5 screws. The eccentric is the only non-standard (not available in retail) part of the 3D - vibrating microtome. D - Diagram of the 3D- vibrating microtome with the dimensions of the parts and the distances between them (in centimeters).

Figure 3.

A - It is a front view of the 3D-vibrating microtome-Renovator. Explanations of abbreviations for the aluminum profile (first module) and other parts are given in Table 3. B - image of the NEWTON NMT500 renovator head mounted on a frame made of aluminium metal construction (Module 1). C - image of the BRS-775SH motor (18 V, shaft diameter - 5), connected to the anchor of the original NEWTON NMT500 renovator motor through a rigid bushing. Together with the reenovator head, this structure provides the blade movement along the y-axis and is an analogue of module 2 as in a 3D- vibrating microtome. D - image of the tissue sample carrying platform (module 4), consisting of a SEMX60-AC linear translator (provides z-axis movement) and a petri dish into which the sample is placed. Module 4 is mounted on rails (providing movement in the x-axis). The rails, together with the supporting bearings, form the basis of module 3. The 3D- vibrating microtome-Renovator is capable of slicing fabric samples with an oscillation frequency of 180Hz and higher. No commercial vibrating microtome (except for rare experimental models) is able to develop blade oscillations of this frequency.

Figure 4.

A - It is a front view of the 3D-vibrating microtome-Microscope. Explanations of abbreviations for the aluminum profile and other parts are given in Table 4. B - image of module 2, providing movement of the blade along the y-axis. The design of this module is essentially no different from the same module installed on the 3D- vibrating microtome. (Fig. 2B). Module 2 in turn is mounted on a paired platform of two SEMX60-AX linear translators that provide blade motion in the x-axis. These coupled translators are module 3. C - image of the platform (module 4) carrying the tissue sample, consisting of the SEMX60-AC module (provides movement along the z-axis). The platform is mounted on the linear translator SEMX60-AC, which provides correction of module 4 in the x-axis and is in fact a duplicate module 3 (the main module 3 is installed under module 2 with a blade). This redundant system is necessary for x-axis correction of the sample. All functional modules of the 3D slice microscope are mounted via a supporting aluminium profile (module 1) on the inverted slide of the Leitz Dialux microscope. This design allows the tissue sample to be sliced directly under the microscope objective, which can be useful in the case of synchronous scanning of the sample during the slicing process, without mounting the slices on the slide. In the case of sample scanning, the direct microscope can be easily replaced by a confocal microscope.

Table 2.

List of Necessary Parts of the Corresponding Abbreviations Required to Build a 3D-Vibrating Microtome.

Part Type	Abbreviation	Dimensions	Quantity	References With Vendor Code
First module
Aluminum profile	AP1	20×80L, 30 ×cm length	2	https://www.soberizavod.ru 228LX
Aluminum profile	AP2	20×20L 40 cm length	10	https://www.soberizavod.ru 222LX
Aluminum profile	AP3	20×20L 30 cm length	4	https://www.soberizavod.ru 222LX
Aluminum profile	AP4	20×20L 20 cm length	4	https://www.soberizavod.ru 222LX
Corner connector	M51	20×20L groove 6, ×M51	Minimum 40	https://www.soberizavod.ru M51
Second module (y-axis)
Universal bracket	Nema 23		1	https://zona-3d.ru 899513519
Motor	BRS-775SH	DS18V shaft diameter 4 mm	1	https://www.bormash.com 110799
Flange per axle		inner diameter 4 mm	2	https://dvrobot.ru 2232-м33
Flange bearing	MF106ZZ		2	https://www.ultrarobox.ru MF106ZZ
Connecting rod	Makita 4500		1	https://bizip.ru 00771
Eccentric				Nonstandard part
Linear Translator	SEMX60-AC		1	https://russian.alibaba.com 62278162378
Flange bearing	MF106ZZ		2	https://www.ultrarobox.ru MF106ZZ
Screw M4	M4	Length 3 cm. bring the diameter to 5 cm with electrical tape	1	Any suitable
Corner connector	M51	20×20L groove 6, ×M51	Minimum 4	https://www.soberizavod.ru M51
Connector	M45	20x40 M45	1	https://www.soberizavod.ru M45
Third module (x-axis)
Linear module (HLTNC SBR16+SFU1605 400mm sliding table)	HLTNC		1	https://aliexpress.ru/item/ 32356381459
Bath		290x160x40 mm	1	Any suitable
Fourth module (z-axis)
Linear Translator	SEMX60-AS-L (or R) Or SEMX60-AC		1	https://russian.alibaba.com 62278162378
Aluminum profile	AP4	20×20L 20 cm length	3	https://www.soberizavod.ru 222LX
Corner connector	M51	20×20L groove 6, ×M51	Minimum 4	https://www.soberizavod.ru M51
Corner connector	M54	40×40L groove 6, ×M54	2	https://www.soberizavod.ru M54
Plate	S08	30x30L	2	https://www.soberizavod.ru S08

The list of parts is divided into modules that form the vibrating microtome. The only non-standard part is an eccentric, a detailed drawing of it is given in Fig. 2C.

Table 3.

List of Necessary Parts of the Corresponding Abbreviations Required to Build a 3D-Vibrating Microtome-Renovator.

Part Type	Abbreviation	Dimensions	Quantity	References With Vendor Code
First module
Aluminum profile	AP2	20×20L 40 cm length	3	https://www.soberizavod.ru 222LX
Aluminum profile	AP3	20×20L 30 cm length	4	https://www.soberizavod.ru 222LX
Aluminum profile	AP5	20×80L, 50 ×cm length	2	https://www.soberizavod.ru 228LX
Aluminum profile	AP6	20×20L 50 cm length	2	https://www.soberizavod.ru 222LX
Corner connector	M51	20×20L groove 6, ×M51	Minimum 16	https://www.soberizavod.ru M51
2020-T Steel Connecting Plate	T-plate		2	https://zona-3d.ru 0352
Second module (y-axis)
Head of Renovator	Newton NMT500			https://www.ozon.ru 1215025434
Motor	BRS-775SH	DS18V shaft diameter 5 mm	1	https://www.bormash.com 110799
Universal bracket	Nema 23		1	https://zona-3d.ru 899513519
Coupling	GX D20 L25	5x8	1	https://zona-3d.ru GN
Corner Connector	M54	40×40L groove 6, ×M54	2	https://www.soberizavod.ru M5
Third module (x-axis)
Rails	MGN9	50 cm lengt	2	https://zona-3d.ru 3940
Rail carriage (extended)	MGN9H		2	https://zona-3d.ru ц1.8.13
Fourth module (z-axis)
Linear Translator	SEMX60-AS-L (or R)		1	https://russian.alibaba.com 62278162378
Universal bracket with corner	Nema 23		2	https://zona-3d.ru 3471
Petri Dish as bath		90 mm	1	Any suitable
2020-T Steel Connecting Plate for tissue sample	T-plate		2	https://zona-3d.ru 0352

The list of parts is divided into modules that form the vibrating microtome.

Table 4.

List of Necessary Parts of the Corresponding Abbreviations Required to Build a 3D-Vibrating Microtome-Microscope.

Part Type	Abbreviation	Dimensions	Quantity	References With Vendor Code
First module
Aluminum profile	AP2	20×20L 40 cm length	2	https://www.soberizavod.ru 222LX
Aluminum profile	AP4	20×20L 20 cm length	2	https://www.soberizavod.ru 222LX
Second module (y-axis)
Universal bracket	Nema 23		2	https://zona-3d.ru 899513519
Universal bracket with corner	Nema 23		1	https://zona-3d.ru 3471
Motor	BRS-775SH	DS18V shaft diameter 4 mm	1	https://www.bormash.com 110799
Flange per axle		inner diameter 4 mm	2	https://dvrobot.ru 2232-м33
Flange bearing	MF106ZZ		2	https://www.ultrarobox.ru MF106ZZ
Connecting rod	Makita 4500		1	https://bizip.ru 00771
Eccentric				Nonstandard part
Linear Translator	SEMX60-AC		1	https://russian.alibaba.com 62278162378
Flange bearing	MF106ZZ		2	https://www.ultrarobox.ru MF106ZZ
Screw M4	M4	Length 3 cm. bring the diameter to 5 cm with electrical tape	1	Any suitable
Corner connector	M51	20×20L groove 6, М51	1	https://www.soberizavod.ru M51
Corner connector	M54	40×40L groove 6, ×54	1	https://www.soberizavod.ru M54
Connector	M45	20x40 M45	2	https://www.soberizavod.ru M45
Third module (x-axis)
Linear Translator	SEMX60-AC		2	https://russian.alibaba.com 62278162378
Aluminum profile	AP4	20×20L 20 cm length	1	https://www.soberizavod.ru 222LX
Corner connector	M51	20×20L groove 6, М51	4	https://www.soberizavod.ru M51
Fourth module (z-axis)
Linear Translator	SEMX60-AS-L (or R)		1	https://russian.alibaba.com 62278162378
Universal bracket with corner	Nema 23		2	https://zona-3d.ru 3471
Glass Bath		200x150x40 mm	1	Any suitable
Linear Translator	SEMX60-AC		1	https://russian.alibaba.com 62278162378
Universal bracket as basis for tissue sample	Nema 17		1	https://zona-3d.ru 0498

The list of parts is divided into modules that form the vibrating microtome.

1. 3D-vibrating microtome for slicing large brain samples, with a linear translator with a blade driven by a crank mechanism. The crank mechanism is designed to convert the rotary motion of the motor into a reciprocating motion of the carriage with the blade.

2. 3D-vibrating microtome-Renovator for practicing slicing at high blade oscillation frequencies (up to 360 Hz). The blade is driven by the unique mechanism of the Renovator head.

3. 3D-vibrating microtome-Microscope mounted on the slide of a Leitz Dialux direct mycoscope. It was created to demonstrate the feasibility of a metal-based vibrating microtome concept. The blade is driven by a crank mechanism.

Sectioning of Mouse Brain Into Vibrational Sections of Different Thicknesses

A total of 6 mouse brain samples at the P11 developmental stage were used. All these are enumerated in the list. Two brain samples were sliced on a 3D-vibrating microtome in a rostro-caudal direction into 30 μm and 50 μm thick frontal sections (Fig. 5A, B, C, D, E, F). The frequency of blade oscillation during slicing was 1 to 3Hz and the sample feed rate was approximately 1 mm/sec. Section loss during slicing due to the phenomenon of “slippage” was 50% for 30 µm thick sections and 0% for 50 µm thick sections. A lost section is defined here as an incomplete slice with an area that is 80% or less of the area of a complete section. “Blade slip” is due to the fact that the vibrating microtome blade at some point becomes unable to completely sliced through the brain sample. This is likely due to micro-deformations of the shrinkage of the brain sample along the z-axis. Section thickness was measured directly using the technique described in the methods section (2.9). The expected section thickness values were in good agreement with the actual values (Fig. 5B and E). A characteristic “stripe pattern” was present on all vibrational sections, the distance between the strips was 26.47 μm (Fig. 5C) and 11 μm (Fig. 5F).

Figure 5.

A - image of an unstained 30 μm thick vibrational section of mouse brain. The section was made at the Habenula (Hb) level. The actual grid edge length is 1000µm. The length of the grid edge in the photo is 17mm (large white stripe with white arrows on the side). This is a conditional value depending solely on the size of the original photo The value of the section thickness (black stripe) in the photo is 0.5mm (short white stripe with side white arrows). The ratio of the section thickness in the picture to the length of the edge is 0.5/17=0.029, which is very close to the theoretical value of the ratio of the true section thickness in microns to the length of the grid edge - 0.03 (30 microns - section thickness/1000 microns - length of the grid edge). B - is an image of an agarose-stained and cross-sectioned slice from the same batch with an ocular grid superimposed on top, this method allows the slice thickness to be measured directly. The real value of the slice thickness coincides with the expected one. C - image of a slice with an ocular grid superimposed on it. The grid allows you to measure the distance between the strips on the slice. In this case, the distance between the strips is 0,45mm. The distance between the strips is 0.45×1000/17=26.47 μm. The stripes are a visible manifestation of the so-called "Stripe Pattern" phenomenon. D - image of an unstained 50 μm thick vibrational section of mouse brain, sliced at the level of the triangular septal nucleus (TS). E - is an image of a peer-sliced section in agarose from the same batch of sections. The actual value of the section thickness is the same as expected (0.85×1000/17=50 µm). F - image of a random section from a series of 50 µm thick sections with an ocular grid. The distance between the strips is 0.2×1000/17=11 µm. G - image of a 50 µm thick unstained cryo-sliding microtome section of mouse brain at the level of stria medullaris (sm). H - image of a sliced-apart cryo-sliding microtome section in agarose with an ocular grid superimposed on top. The real value of the section thickness coincides with the expected one (0.82×1000/17=48.235 µm). I - an image of a 150 µm thick unstained vibrational slice of mouse brain, sliced at the level of stria medullaris (sm). This slice was taken on a 3D-vibrating microtome as a sample for comparison with a similar thickness slice (L, slice taken at the TS level) taken on a commercial Campden Instruments HA752 vibrating microtome. Both slices were of equal thickness (J, M), which is proof of the quality of the 3D- vibrating microtome at the factory instrument level. K,N - image of random slices from different series with a thickness of 150 µm with an ocular grid superimposed on top. The spacing between the strips is 0.55×1000/17=32.535 µm for the 3D-vibrating microtome and и 1,9×1000/17=111,765 µm for the HA752 vibrating microtome. The difference is due to different sample feed rates. All slices were taken with a Sony IMX178 camera. The top right indicates the magnification of the microscope lens. The bar is 100 µm.

A third brain sample was sliced on a Leica SM2000R cryo-sliding microtome in a rostro-caudal direction into 50 µm thick frontal sections (Fig. 5G). Slice loss during slicing was 0%. The thickness of the cryosectins was measured directly in the optical microscope, and the value obtained was in good agreement with the expected and equal value for identical thickness of the vibrating sections obtained on the 3D-vibrating microtome (Fig. 5H). At the same time, the “Stripe pattern” was absent on the cryo-sections.

Two additional samples (numbers 4 and 5) were evaluated: One brain sample (Case 4) was sliced on a 3D-vibrating microtome in a rostro-caudal direction into 150 µm thick frontal slices (Fig. 5I, J, K). The blade oscillation frequency during slicing was 5 to 10 Hz and the sample feed rate was 3 to 5 mm/sec. The slice loss during slicing due to blade slippage was 0%. Another sample (Case 5; Fig. 5L, M, N) was sliced on a Campden Instruments HA752 vibrating microtome in the same rostro-caudal direction into 150 µm thick frontal slices (Slice loss 0%, blade oscillation frequency 5 to 10 Hz, sample feed rate 5 to 7 mm/sec). This sample was used as a comparison control between the two instruments. The thickness of the slices in both cases was measured directly (Fig. 5J and M), and the thickness values obtained for Case 4 and Case 5 slices did not differ from each other). A characteristic “stripe pattern” was present on all vibrational sections, the distance between the strips was 32,535 μm (Fig. 5K) and 111,765 μm (Fig. 5M).

A sixth brain sample was sliced on a 3D-vibrating microtome-Renovator in a rostro-caudal direction into frontal slices with thicknesses ranging from 80 to 150 μm. The frequency of blade oscillation during slicing ranged from 5 to 180Hz. The loss of slices (80 µm and thicker) during slicing due to blade slippage was 0% at low-frequency blade oscillation, and 100% at 180Hz blade oscillation frequency, regardless of the sample feed rate.

Sectioning of Calf Brain Into 150 µm Thick Vibrational Slices

A total of three different fragments were used for slicing, cut from a saggitally sliced half calf brain. Each brain fragment was approximately 20 mm thick and represented approximately half of the whole brain (sliced along the medial line of the brain). Two fragments were used to determine the value of the marginal minimum thickness for the whole slice, which was 150 µm (Fig. 6A). The third most presentable fragment at the level of the thalamus was cast in agarose (final block size 90 × 70 × 40 mm) and then serially sliced on a 3D-vibrating microtome into 8 slices of 150 µm thickness (Fig. 6B-I). The blade oscillation frequency was 1 to 5 Hz and the sample feed rate was ~5 mm/sec. The phenomenon of “slippage” during slicing was not observed, the loss of slices during slicing was 0%.

Figure 6.

A - image of a calf brain slice cast in agarose and sliced across with an ocular grid superimposed on top. This method allows the slice thickness to be measured directly. The actual value of slice thickness is the same as expected and is 150 μm. B - a medial slice of a calf brain, cleaned from the vasculature and embedded in 5% agarose. The arrow indicates the part of the brain damaged during cleaning from the membranes C-J. Series of eight sequentially sliced rostro-caudalunstained macrosections of 150 μm thick calf brain; The approximately 25-mm-thick brain fragment from which the macrosections were made was chosen at the level of the Thalamus (Th); Th is well visualized and is a convenient reference point for identifying other, neighboring brain structures. Nucleus caudatus (Nc), lying dorso-medial to Th; and a detached Corpus amygdaloideum (Ca), lying lateral to Th, are also clearly visible on the macrosections. The locations of the above brain structures were determined from the calf brain atlas (Yoshikawa, 1968; level C8) ⁹ ; the nomenclature of brain structures is also cited from this atlas. The first two macrocuts (C,D) are only 70-75% intact; they contain damage at the Th level. These lesions are not related to the work of the 3D slice, but are determined by the uneven cut surface of the original brain fragment (see the discussion for more details on the causes of these defects). The next six macrosections (E-J) are 90-95% intact. The only major lesions are recorded at the Ca level, but Ca was damaged initially in the original brain fragment (I, marked by the black arrow). The cause of the Ca damage was the clearing of the original brain fragment from the vasculature, during which the outer periphery of the brain is often subjected to mechanical destruction (e.g., with forceps during the stripping of the vasculature). On the last four macrosections (G-J) the damage in the ventral part of Th is noticeable. The surface of macrosculpturein this area has a pronounced "stripe pattern", sometimes accompanied by straight tears of tissue (marked by white arrow). The presence of "striped pattern" and tissue tears is associated with an excessive cutting speed of more than 15 mm/min. These defects could have been avoided if the cutting speed had been kept low all the time (15 mm/min or less). As soon as the cutting speed was reduced to the proper values, the stripe pattern became almost invisible, which can be clearly seen by the consistency of the tissue between the Rcc and ventral Th on the last three macrocuts. All macrosections were taken with an Ipad 10.2 camera.

The results obtained on the 3D-vibrating microtome revealed a number of difficulties preventing (at least partially) the acquisition of 100% solid slices. The following slicing features were found: During fixation, the excised brain section is slightly deformed. This is an inevitable process for such large tissue fragments, accompanied in this case by curvature of the surface of the original brain section. As a result of this deformation, the boundaries of the section (the original slice sites) are no longer flat. Therefore, there is a stage in the slicing process when only a part of the brain is present on the slice because the different parts of the brain at the site of the slice are already at different levels rather than at the same level as they should be. This effect is particularly noticeable in the first two slices (Fig. 6B and C), but then disappears when the plane of the thick brain fragment is inexorably leveled during the slicing process (Fig. 6D-I)

As the slice speed increases, the brain tissue may be crumpled, and if not slowed down or the slice is not straightened in time, this section loses its integrity by slicing the slice with the blade at the point of bending (Fig. 6I).

The remains of the vasculature are barely sliced through by the blade at low vibration frequencies (1–5Hz). Therefore, if the vasculature is not completely removed from the sample, it will pull on the blade and tear the brain tissue apart. In the process of removing the vasculature with forceps, brain tissue is also often torn apart. For this reason, in some places of the original thick section of the brain, from which the vasculature has been removed, whole fragments of the brain may be missing, which will be absent on the slice (Fig. 6J). It is practically impossible to remove the vasculature entirely without damaging the brain sample. In the process of removing the vasculature, the forceps also tear the brain tissue. Therefore, during slicing, these torn sections may already be permanently detached from the main body of tissue, even if they are originally present in the sample mass.

The time and quality of fixation of the sample in 4% PAF is also important. Older, more than a year-fixed samples are difficult to slice and tend to wrinkle constantly during the slicing process, which leads to over slicing by the blade even at low sample feed rates.

Microscopic Analysis of Mouse Vibrational DiI-Loaded Slices (Cases With DiI Applied to Whole Brain Sample)

A total of 2 mouse brain samples at the P11 developmental stage were examined for microscopic evaluation of slice quality. According to literature data ¹⁰ it is known that in rodents MHb is connected by efferent connections with Triangilar Septal nucleus (TS) and with Bed Nucleus of stria termonalis (BST), and LHb is connected by efferent connections with Lateral Preoptic Area (LPA). Consequently, in the case of DiI point injection into the MHb, neuronal bodies stained with the marker should become visible in the TS; in the case of point injection into the LHb, stained neuronal bodies should be detected in the LPA. It is the Hb, TS, BST, and LPA that were analyzed first for microscopic assessment of the quality of vibrational slices.

Case 1-P11-DiI

Unilateral application of DiI to the frontal surface of the brain in MHb. The sample was sliced frontally on a 3D-vibrating microtome in the caudo-rostral direction into 150 μm thick slices. The blade oscillation frequency was 5Hz and the sample feed rate was 3 to 5 mm/sec. Slice loss during slicing was 0%. The marker application site was evaluated along the entire length of the Hb nucleus in the rostro-caudal direction with special emphasis on the first slice (closest to the marker). On the first slice, a very bright marker glow was observed at the application site (Fig. 7A). The injection site itself was precise, as well as a slight diffusion of the marker from the tagged MHb nucleus into the neighboring LHb nucleus was observed. This is an inevitable process associated with marker diffusion along the slice surface. The BST of the labeled hemisphere showed a cluster of DiI-labeled neuronal bodies (Fig. 7B), which is an expected result. Among the labeled neuronal bodies, DiI-labeled fibers were scattered throughout the BST. In LPA, numerous labeled DiI fibers were also detected (Fig. 7C), which is due to parasitic lateral diffusion between MHb and LHb nuclei at the application site. In TS, no labeled neuronal bodies were detected, probably due to the fact that the marker was localized in the most caudal part of the MHb.

Figure 7.

Distribution analysis of DiI-labelled structures on vibrational slices of two mouse brain samples cut into slices of equal thickness (150 μm). Case-1 - P11 (A-C) cut on a 3D- vibrating microtome; Case-2 - P11 (D-F) cut on a Campden Instruments HA752 commercial vibrating microtome. Section thickness at scanning is 10 µm. A - Confocal image of a frontal brain slice at the level of Medial Habenula (MHb), where unilateral injection of DiI marker was made. The injection site itself was precise, and there was also a slight diffusion of the marker from the MHb nucleus with the marker into the neighbouring Lateral Habenula (LHb). This is an inevitable process associated with marker diffusion along the slice surface. 20x magnification. B - A cluster of DiI labelled neuronal bodies was detected in the Bed nucleus of Stria Terminalis (BST) of the tagged hemisphere, which is an expected result. Among the labelled neuronal bodies, DiI-labelled fibres were also scattered throughout the BST volume. C-Lateral Preoptic Area (LPA) also revealed numerous DiI-labelled fibres, which is an abnormal result due to parasitic lateral diffusion between MHb and LHb nuclei at the application site. D - Confocal image of bilateral DiI application to the frontal (coronal) surface of the brain in the LHb (left) and MHb (right). The injection site itself was accurate. E - A strong bundle of DiI stained fibres - stria medullaris - was clearly visible between Hb and Triangular Septal nucleus (TS). F - In the TS of the labelled hemisphere numerous DiI stained structures were detected, namely bundles of labelled axons, with clusters of neuronal bodies scattered around their periphery in the form of a specific halo. The evaluation of confocal images showed no difference in the quality of the slices obtained on different vibrating microtomes. The bar is 100 µm.

Case 2-P11-DiI

Bilateral application of DiI to the frontal (coronal) surface of the brain in the LHb (left) and MHb (right). The sample was sliced frontally on a Campden Instruments HA7520 vibrating microtome in the caudo-rostral direction into 150 μm thick slices. The blade oscillation frequency was 5 Hz and the sample feed rate was 3 to 5 mm/sec. Slice loss during slicing was 0%. On the first slice, a very bright glow of the marker was observed at the application site (Fig. 7D). The injection site itself was accurate. A strong bundle of DiI stained fibers, the stria medullaris, was visible between Hb and TS (Fig. 7E). In the TS of the labeled hemisphere (MNb), numerous DiI stained structure bodies were identified. In the ventro-medial region of the TS nucleus, powerful bundles of labeled axons were clearly visible, with clusters of neuronal bodies scattered around their periphery in the form of a peculiar halo (Fig. 7F).

Microscopic Analysis of Mouse Cryo- and Vibrational Sections With DiI (Application of DiI Directly to Brain Sections)

To test how membrane structures are patterned on a brain section, one can try to apply a lipophilic marker directly onto the section surface. DiI applications were made on separate vibrational (three samples) and cryosections (three samples) of mouse brain P11, the thickness of sections in both types of cases was 50 μm.

In mouse vibrational sections, when the marker was applied to the section in the TS region, numerous fibers were stained at the site of application (Fig. 8A and B). Single stained neuronal bodies of excellent quality were also found throughout the section area (Fig. 8C). This does not mean that these stained neurons are in any way related to TS, they were detected solely due to the peculiarities of this method of marker application—on the section and not in a specific nucleus of the whole sample. For this reason, small DiI crystals were sometimes found in the most unexpected places, where they naturally colored neurons around themselves or even at some distance from them.

Figure 8.

A,B,C - images of a 50 μm thick frontal vibrational sections of mouse brain P11 with DiI applied to the surface in the Triangular Septal nucleus (TS). A - overview photo of the section at low magnification (2.5×). B - axon bundles forming the TS were perfectly stained with DiI, which is clearly visible even at medium magnification (magnification 10×). C - single, randomly stained neuronal bodies, whose thin membrane structure was not damaged in any way during cutting, are scattered all over the surface of the section (20x magnification). D,E - is an image of a 50-μm-thick frontal cryo-sliding microtome sections of mouse brain P11, the surface of which was stained with DiI in the TS region D - an overview photo of the section (2.5× magnification). E - DiI-stained axon bundles in TS are clearly visualized (10x magnification), but no stained neuron body was found on the surface of the section. All slices were taken with a Sony IMX178 camera. The top right indicates the magnification of the microscope lens. The bar is 100 µm.

In the case of mouse cryosections, when DiI was similarly applied to the TS region, the distribution pattern of labeled fibers was generally very similar to that in the vibrational section (Fig. 8D). The fibers in the TS were well stained and looked quite distinct (Fig. 8E). But not a single stained neuron body with all its branches was detected on the cryosections.

Comparison of Cryo- and Vibrosections Stained by Nissl

To check the quality of cytoarchitectonics, 50-µm-thick vibrational sections were stained by Nissl (a variant using Neutral Red; Fig. 9A, D, C). Similar 50-μm-thick sections obtained on a Leica SM2000R cryo-sliding microtome were used as a control (Fig. 9D, E, F). The staining technique was completely identical between the different types of slices and is described in detail in the Methods. No differences in the quality of Nissel-stained neuronal bodies between the slices obtained on the 3D-vibraing microtome and cryo-sliding microtome were detected. The obtained images were completely identical, indicating the same quality of the slices obtained by different types of microtomes, but equally stained by Nissl.

Figure 9.

A - overview image (4x) of a Nissl stained vibrational 3D-section of mouse brain P11 at the level of Caudate Putamen (Cpu). B - Mouse brain section at high (50x) magnification at the level of the medial part of the Infralimbic cortex (IL). C - Mouse brain slice section at high (50x) magnification at the Lateral Ventricle (LV). Thickness of the slice is 50 µm. D - overview image (4x) of a Nissl-stained cryo-sliding microtome of mouse brain P11 at the level of Lateral Septal nucleus Intermed (LSI). E - Mouse brain section at high (50x) magnification in the Lateral Ventricle (LV) region. F - mouse brain section at high (50x) magnification at the level of the corpus callosum (cc). Thickness of the slice is 50 µm. No differences in the quality of Nissl stained neuronal bodies were found between slices obtained on the 3D- vibrating microtome and cryo-sliding microtome. All slices were taken on a camera with a Sony IMX178 sensor. The bar is 100 µm.

Discussion

There is a frustrating and disconcerting lack of data on optimal performance characterics for the performance of vibrational mictromy. Values for the minimum thickness of mouse brain sections start from 10 μm,4 and only rarely any data are given on the conditions of device operation when obtaining a slice of a particular thickness. No systematic studies were conducted have been previously been published. Based on the results of this work, it is clear that the minimum value of the thickness of a mouse section depends on several factors, namely the frequency of blade vibration and the speed of sample feeding. The lower the vibration frequency and the lower the sample feed rate, the thinner the section can be obtained.

The data obtained on the 3D-vibrating microtome show that mouse sections up to 50 µm thick can be obtained with zero loss, while for thinner sections the phenomenon of “slippage” begins to appear. The essence of this phenomenon is that starting from a certain value along the z-axis, the blade of the vibrating microtome starts to slip through the sample, resulting in either incomplete sections or sections with double thickness (“slip” of the blade through the sample). As soon as the “slip” phenomenon makes itself felt, shear loss begins. And if at the blade oscillation frequency of 1–3 Hz and sample feed speed of ~1 mm/sec it is possible to obtain mouse vibration sections of 50 μm with zero losses, then in the case of 30 μm thick sections the losses due to the phenomenon of slippage are already 50%, which calls into question the expediency of the slicing itself.

It should be noted that the above-described sectioning conditions on the 3D-vibrating microtome are in contradiction with the data obtained on the experimental vibrating microtome described in Li et al.,4 where mouse vibration sections of 10-µm thickness were obtained, and the authors did not describe findings consistent with what we have termed “slippage.” This leads one to assume that the efficiency of their vibrating microtome at all thickness values was 100% including for 10-µm-thick sections. The authors emphasize in the article that they created a special vibration-resistant mechanism (DP-DP flexure), and due to this voice-coil motor, they were able to achieve an ultra-high blade vibration frequency (180 Hz), allowing them to section a mouse brain sample in thin 10-µm-thick sections at speeds of more than 50 mm/min. Their manuscripts specify 120 to 140 mm/min. Thus, two key slicing parameters—sample feed rate and blade vibration frequency—being set to their maximum values, were essentially the key to the success of this work and circumvented the problem of “slippage.”

It is not clear what the physical background of such a qualitative transition in slicing at the maximum parameters of the blade oscillation frequency of 180 Hz. A “3D-Renovator-vibrating microtome” was built specifically to test the performance of this concept. Despite small differences in the design (absence of Voice-coil motor and DP-DP adapter), it was not difficult to achieve a blade oscillation frequency of 180 Hz on this instrument, however, with such section settings, we were not able to get a single solid mouse slice even 80-µm-thick, or thinner sections, down to 10 µm. Perhaps, it is due to differences in the design of the devices, but in any case, the data presented by Li et al.4 are of great interest. For now, we will only limit ourselves to the conclusion that the minimum achievable thickness of vibration sections under standard slicing conditions (low blade oscillation frequency and low sample feed rate) is 30 µm.

We note that the figure of 30 µm for the minimum thickness of mouse brain sections is not a priori universal, as well as for any commercial vibrational microtome. For example, the HA752 vibrating microtome has a minimum slice thickness with 0% loss of 100 µm, possibly due to the fact that it uses a metric screw (rather than a micrometric screw like the 3D-vibrating microtome). Some vibrating microtome models, for example, have a fixed blade oscillation frequency (usually 60 Hz), so it is hardly possible to get thin sections of 30 µm (requiring a blade oscillation amplitude of 1–3Hz on a 3D-vibrating microtome). Some commercial models of vibrating microtomes have only automatic blade feeding, which sharply limits the minimum speed value and, as a consequence, goes beyond the optimal conditions. The conclusion is clear, the more fixed options are present in the model of vibrational microtome (it does not matter whether commercial or experimental), the narrower the range of parameters of tissue sections obtained on it.

Is There a Relationship Between the Minimum Achievable Slice Thickness and Its Area. Analogies With Cryo-Sliding Microtome

Based on the data described above, we can see that for mouse sections, the minimum achievable thickness is 30 µm with 50% loss and 50 µm with 0% loss, whereas for whole calf slices it is at least 150 µm with 0% loss, that is, three times larger. The area of a mouse section cast into an agarose block is usually 20 × 10 mm (in frontal orientation), for half a calf brain in frontal orientation it is already 90 × 70 mm, that is, about 30 times larger. It is reasonable to conclude that there is a certain regularity between the area of the slice and its minimum thickness. In this context, it is interesting to compare the data obtained on other types of microtomes suitable for slicing sections.

There are not so many manuscripts in which large total slices obtained with a microtome or cryostat were studied, but a number of interesting observations can be made from them. The standard thickness of a small paraffin section for a rotary or sliding microtome is 4 to 7 μm (minimum section thickness is 2 μm), for a freezing microtome (cryo-sliding microtome) is 5 to 50 μm (data taken from Carter and Shieh ¹¹ ). In the work of Amunts et al., ¹² a Leica SM2500 microtome (The Leica SM2500 large-scale, heavy-duty sectioning system is the sliding microtome for sectioning hard and/or large-surface specimens) was used to make total slices of the human brain; the thickness of the sections was 20 μm, that is, four times the average thickness of a small, paraffin-embedded slice (5 μm). When compiling the atlas of the human brain, ¹³ a Vogt microtome was used, on which frozen sections with a thickness of 100 μm were obtained. In another recent work, Song-Ling Ding et al., ¹⁴ frozen coronal slices of the human brain with a thickness of 50 μm were obtained on a Lipshaw model 90A microtome (Pittsburgh).

As can be seen from this limited literature sample, in all cases, regardless of the type of microtome used, when slicing large brain samples, the thickness of the total sections was approximately 4 to 5 times higher than the average value (average for small tissue samples). No explanations were given in the above texts regarding the reasons for choosing this particular slice thickness. Larger values of thickness for large total slices in comparison with standard medium-sized slices are most likely proposed to be accepted as a certain empirical fact. The fact that concentional 5 µm sections were not used suggest that sections of this scale are impractical.

Currently, there is no clear physical or mathematical theory explaining the above-described empirical data obtained on completely different types of microtomes. For now, let us accept as a purely experimental fact that the larger the surface area of a slice, the larger the maximum achievable minimum thickness value for this slice. It is likely related to the compression of the section during cutting. The larger the area of the section, the higher the likelihood of fractures and tears at low cutting thicknesses (an empirically established fact). The heterogeneous structure of large brain samples (due to blood vessels and meningeal layers) also negatively affects the integrity of the sections and the ease of cutting.

Cryo-Sliding Microtome Versus Vibrating Microtome?

Based on the capabilities of the equipment, any researcher has a choice of using vibrational sections or cryosections, including cryosections for large objects. A legitimate question arises—why one cannot use a cryo-sliding microtome instead of a vibrating microtome in all types of studies? What makes vibration sections so special?

Some advantages and disadvantages of cryotomy in comparison with vibratome microtomy are given in the Table 5. Cryo-sliding microtome has a slight advantage in slicing speed (approximately twice as fast when obtaining sections of 50-µm thickness on the same object). However, this speed is partially offset by certain drawbacks such as the necessity to use a cryogen (dry ice in the case of cryo-sliding microtomes) and/or specialized, rather expensive freezing systems based on Peltier elements (this option is also encountered in cryo-sliding microtomes).

Table 5.

Comparative Evaluation of Cryo-Sliding Microtome and 3D-Vibrating Microtome, Including Both Average Technical Characteristics and Peculiarities of Sample Preparation for Cutting.

Parameters	Cryo-Sliding Microtome	3D-Vibrating Microtome
Technical part
Blade	1. Specialized reusable microtome knife. 2. Specialized Disposable Microtome Knives	1. Unspecialized trapezoidal blades for construction knife 2. Segmented blades for construction or office knife
Fabric freeze function	Mandatory. Cutting in air. Requires high flow rate of coolant (e.g. flowing water for cryotable)	Optional installation of a cryostat. Standard cutting at room temperature in aqueous solution
Cutting speed	2–5 mm/min	5–10 mm/min
Minimum cutting thickness limit	5 μm	30 μm 50 μm optimal
Maximum cutting thickness limit	50 μm	Limited by the length of the z-axis (25 mm for the 3D-vibrating microtome)
Histological part
Sample pretreatment	Mandatory, requires soaking in sugars of increasing concentration, at least 3 days for mouse brain	Optionally, you can cut as a fresh native sample, or a sample fixed for 1 day in 4% formaldehyde
Cut quality	Incorrect blade setting can result in a wrinkled cut or unevenness	“striped pattern” of varying degrees of severity (depending on the thickness of the slice or model of the vibrating microtome)
Cell membrane integrity	The fine structure of membranes is irreversibly lost during freezing	The fine structure of the membranes is fully preserved
Fabric artifacts caused by freezing	May be present if the sample is not properly seasoned in the cryoprotectant	None since the sample is usually cut at room temperature

On the cryo-sliding microtome it is relatively easy to obtain thin sections of 10 to 20 μm thickness without loss, but at the 50-μm level the cryo-sliding microtome no longer has this advantage over the vibraring microtome. Moreover, 50-μm slice thickness is the upper limit for the cryo-sliding microtome, which is not always an optimal solution, especially when it comes to confocal microscopy where thick tissue slices of 100–150 μm thickness are quoted. As far as external characteristics are concerned, the only difference between cryosections and vibrational slices is the complete absence of the “stripe pattern.”

Additional, where membrane structures are not involved, cryo-sliding microtome and vibrating microtome hardly differ from each other. A good example is the results obtained with Nissl staining, the staining quality is absolutely identical on cryo and vibrating microtome of the same thickness (50-µm in this case). The staining of membrane structures is slightly different. The results obtained in this work show that when a cryo-sliding microtome is used, no whole neuronal membrane remains on a DiI-stained cryosection, although large tracts (e.g. labeled fibers in TS) retain their integrity. Herein lies the major disadvantage of cryo-sliding microtomes when working with lipophilic dyes. Any information about the distribution of neuronal bodies on the slice is irretrievably lost. On vibrational slices, everything is exactly the opposite, neuron bodies are viewed extremely clearly, which indicates the complete preservation of membrane structures on the slice.

The Benefit of an Additional Vibrating Microtome

The metal constructor allows to create a microtome slicing large tissue samples (3D-vibrating microtome), or a vibration microtome with an exceptionally high blade oscillation frequency (3D-vibrating microtome-Renovator with a blade oscillation frequency of 180 Hz or more), or to mount a vibrating microtome directly on the microscope, which can be useful in the case of direct reading directly from the sample, bypassing the step of mounting the sections on glass (as done by Economo et al. ¹⁵ and Amato et al. ¹⁶ ). In this article, an experimental apparatus based on a vibration unit from a commercial Leica 1200S vibrating microtome was used. The blade vibration frequency during slicing was 85 Hz, with a blade vibration amplitude of 1 mm and a sample feed rate of 0.1 mm/sec. This article does not specify the thickness of the sections into which the sample was sliced. This versatility of the metal constructor in building vibrating microtomes for different tasks allows us to define the limits that can be reached when microtomising native tissue samples without having to test the entire range of commercially available models.

Therefore, if a vibrating microtome is built from scratch, one must be sure that it fully meets the quality standards of similar commercial models. Two parameters can be used as such standards—slice thickness and microscopic characteristics of the slice mapped with a fluorescent marker (e.g. DiI). Both of these parameters were evaluated for the 3D-vibrating microtome in comparison to a commercial vibrating microtome with a similar design, the Campden Instruments HA752. No differences in the quality of DiI staining were detected, which was confirmed by the corresponding tissue images obtained on the confocal microscope. It means that the 3D-vibrating microtome based on the metal constructor can be confidently used for potentially any research in the field of vibration microtomy, and it is not inferior to commercial models.

Another important parameter is the optimal slice thickness for a particular study. This is a relevant question, as obtaining 50-µm-thick mouse slices is three times as time-consuming as obtaining 150-µm-thick slices. This depends on the purpose of the study and the tissue staining methods used. For example, for fluorescence or confocal microscopy 150-µm-thick slices are sufficient, while Nissl staining requires slices no thicker than 50 µm. In some cases, it is also sufficient to have slices with only the start and end points of marker diffusion, and the entire tissue volume between these points can be sliced out, and the size of this tissue volume can be measured in millimeters per slice. Most importantly, the 3D-vibrating microtome gives a wide range of options for slice thickness.

In conclusion, we would like to focus on some aspects of the device operation that were not investigated in this work. First of all, we are talking about slicing of unfixed biological tissue. The instructions for the Campden Instruments HA752 model (this model is the basis of the 3D-vibrating microtome) state that the minimum value of the slice thickness for fixed tissue is 20 µm and for fresh tissue 50 µm. The actual slice thickness values for mouse brain tissue obtained on the HA752 model were 100 µm (in the case of 0% loss due to slippage phenomenon) versus 50 µm on the 3D-vibrating microtome. What is the value of the minimum slice thickness for fresh tissue on the 3D-vibrating microtome? This can only be found out experimentally, based on the specific goals of a particular task. The author of this article leaves this question open to other researchers.

And finally, the last question—what types of tissue can be sliced on a 3D-vibrating microtome from a metal construction kit? The exact answer to this question also requires only experimental verification. Each type of tissue, whether normal or pathological (e.g. tumor tissue) requires different slicing conditions. This was well demonstrated by comparing mouse and calf brain slices obtained on a 3D-vibrating microtome. In the case of the calf brain, the minimum slice thickness is three times higher than in the mouse brain (if we are talking about slices with 0% loss due to “slippage”). The calf brain requires careful pre-treatment (all vascular membranes are removed with tweezers beforehand) unlike the mouse brain. Every tissue of any organ of any species needs or does not need such processing, and it can be understood only in practice. So the author leaves this question open as well, as only researchers who deal with this or that object in their daily work can answer it.

Conclusion

Thus, let us summarize some general conclusions related to the slicing parameters on the vibrating microtome and allowing to optimize this process. Before slicing, any tissue sample is recommended to be placed in a 5% agarose block. Agarose fixes well (mechanically) almost any tissue under examination and minimizes the phenomenon of blade slippage. It is an empirically observed fact. In order to obtain the finest possible vibration slicing, a gradual Minimization of the sample feed rate to the blade is recommended. The optimum feed rate of the sample for a particular thickness of slice should be selected individually, depending on the object of study. In addition, when obtaining thin vibrating slices (50 microns or less), minimizing the blade oscillation frequency is recommended. Optimal values of vibration frequency are 1–3 Hz, depending on the object of study. The more delicate is the tissue under study, the lower should be the frequency of blade oscillation. The amplitude of blade vibrations (within the range of values from 1 mm and above) has no effect on the thickness and quality of the resulting vibrating slice

Based on the above recommendations and the list of parameters affecting tissue slicing, we can roughly outline the current and future trends in vibrating microtomy. Currently, we see two key strategies for tissue slicing: Classic vibrating microtomy, which includes the vast majority of commercial models of vibrating microtomes and 3D-vibrating microtomes with a crank-rotary mechanism of the vibrating module. With this strategy, slicing is performed at low-to-medium sample feed speeds (5–10 mm/min or higher) and low-to-medium blade vibration frequencies (1–60 Hz, rarely 100 Hz). Owing to this strategy we can obtain, for example, vibrating slices of mouse brain up to 30-µm thick with 50% loss due to blade slippage phenomenon. Alternatively, high-frequency vibrating microtomy, which includes single experimental models (as well as the 3D-Renovator-vibrating microtome) that slice at high blade oscillation frequencies (180 Hz due to a voice-coil motor) and very high sample feed rates (50–140 mm/min). Due to this strategy, we allegedly can obtain vibrating slices of mouse brain up to 10-µm thick with 0% loss due to blade slippage. However, the physical laws that can be used to explain the production of slices of such thickness at such specified slicing parameters remain unclear.

Thus, in the future we may logically expect the continuation of work in vibration microtomy, dealing with various parameters of blade vibration frequency and sample feed rate (in any direction of the values of these parameters). The appearance of a commercial serial device will probably dot the i’s in the case of high-frequency vibrating microtomy, allowing us to independently test its claimed unique characteristics. The idea of using mechanisms that allow the blade to vibrate with a frequency of up to 40 kHz (ultrasonic knife) is also extremely interesting; this concept is still waiting to be tested in the future.

As for the 3D-vibrating microtome based on the metal constructor, a logical continuation of this topic would be to bring the device to automatic operation requiring minimal operator involvement. The use of stepper motors and digital control (the same that is used on modern metalworking machines or electronic micromanipulators) is a promising vector for further development of devices based on metal construction. But that will be the subject of a special, next article.