A novel method of correlative light and electron microscopy in cryosectioning of bovine anterior pituitary tissue using NanoSuit CLEM

Abstract

Correlative light-electron microscopy (CLEM) combines fluorescence microscopy and scanning electron microscopy (SEM) to achieve nanoscale resolution while highlighting regions of interest identified by fluorescence microscopy. CLEM is becoming increasingly important in life sciences but traditionally requires highly dried samples to withstand the high vacuum of SEM. The NanoSuit method, which mimics native extracellular substances, was developed to address this limitation by encasing samples in a thin, vacuum-proof membrane, allowing SEM observation of live or wet multicellular organisms. While previous NanoSuit CLEM studies focused on formalin-fixed paraffin-embedded sections and cultured cells, cryosections had not yet been explored. In this study, NanoSuit CLEM with diluted NanoSuit solution was applied to cryosections of bovine anterior pituitary tissue. Secretory granules in gonadotrophs, which constitute less than 12% of anterior pituitary cells, were successfully visualized. However, other organelles remained unobserved due to fixation conditions. Therefore, NanoSuit CLEM enabled visualization of the ultrastructure of important cells in cryosections, even from large animals.

Immunohistochemical (IHC) staining is an essential technique for examining the cellular localization of key proteins and other molecules. However, the resolution of light microscopy used in IHC is limited to a few hundred nanometers. In contrast, electron microscopy (EM) offers nanoscale, high-resolution imaging of cellular ultrastructure. Despite its advantages, using EM to investigate the ultrastructure of rare cells within large specimen areas remains challenging.

The ideal approach for observing a region of interest involves screening and extracting colored images using light microscopy, followed by high-resolution observation of the same region using EM. To address this, correlative light and electron microscopy (CLEM) was developed. This technique combines fluorescence microscopy with scanning electron microscopy (SEM) to achieve nanoscale resolution while highlighting regions of interest through fluorescence imaging [1]. As a result, CLEM is becoming an important method in the life sciences. However, traditional CLEM requires highly dehydrated samples to withstand the high vacuum of EM, which often distorts the original structure [2].

Recently, an advanced method known as the “NanoSuit method” was developed for observing the ultrastructure of wet organisms using SEM. This method involves encasing multicellular organisms in a thin, vacuum-proof membrane, allowing them to remain alive or hydrated under high vacuum conditions [3,4,5]. It utilizes the native extracellular substance (ECS) or ECS-mimicking materials, polymerized into a membrane via plasma or electron beam irradiation. The resulting NanoSuit is flexible and dense, preventing the evaporation of bodily fluids and gases—a property referred to as the “surface shield enhancer (SSE)” effect—while providing sufficient electron conductivity for SEM observation. Consequently, the NanoSuit method enables high-resolution imaging of both living and fixed wet specimens using SEM [2].

Although previous NanoSuit CLEM studies focused on formalin-fixed paraffin-embedded (FFPE) sections [3] and cultured cells [4,5,6], cryosections remained unexplored. Preparing FFPE sections involves treatment with organic solvents and heating, which often compromises the detection of target proteins by IHC [7]. Therefore, it is important to evaluate whether NanoSuit CLEM can be applied to cryosections, which offer better antigen preservation and lipid retention but lower physical stability compared to paraffin sections [8].

Gonadotrophs, the cells responsible for secreting luteinizing hormone (LH) and follicle-stimulating hormone to regulate reproduction, comprise less than 12% of anterior pituitary cells [9]. In this study, we applied NanoSuit CLEM to cryosections of bovine anterior pituitary tissue. Using the morphology of blood vessels as markers, we located regions of interest containing LH-positive cells and obtained images via light microscopy and SEM. An example of this method is shown in Fig. 1. LH-positive cells were round or oval in shape. Higher-magnification SEM images revealed many small, electron-dense, and regularly spherical granules. This type of granule was consistently observed in LH-positive cells across all examined sections.

Fig. 1.

The NanoSuit method applied to cryosections for observation using scanning electron microscopy (SEM). The shape of blood vessels was used as a landmark near the region of interest and imaged with both a light microscope (A) and SEM (B). The red rectangles in images (A) and (B) indicate the regions shown at higher magnification in the corresponding panels. Immunohistochemistry of the tissue revealed LH, DAPI (as a nuclear counterstain), and their merged image (C). The blue rectangle in image (C) highlights the region shown at higher magnification in panel (D), where LH immunofluorescence, SEM, and a merged image are presented. Yellow and green arrows in (D) to (H) indicate LH-positive and LH-negative cells, respectively. The LH-positive and LH-negative cells were further observed by high-magnification SEM in images (I) to (N). Scale bars: 1 mm (low magnification in A, B); 400 μm (high magnification in A, B; and in C); 100 μm (D–F); 50 μm (G); 20 μm (H); 10 μm (I, L); 5 μm (J, M); and 2 μm (K, N).

Although little is known about the ultrastructure of bovine gonadotrophs, the size and shape of these granules resembled those of secretory granules in human mammosomatotroph adenomas [10], unidentified bovine anterior pituitary cells [11], and ovine gonadotrophs [12]. Therefore, our findings successfully visualized secretory granules in gonadotrophs, which constitute less than 12% of anterior pituitary cells [9]. Interestingly, spherical granules in LH-negative cells (n = 120 granules from four individuals) were significantly larger in diameter (P < 0.01) than those in LH-positive cells (also 120 granules from four individuals), as shown in Fig. 2A.

Fig. 2.

(A) Significant differences in the diameter of secretory granules between LH-positive and LH-negative cells, as observed using the NanoSuit method and SEM. (B) Three-dimensional fluorescence immunohistochemistry images (220 μm in the X-axis, 180 μm in the Y-axis, and 10 μm in the Z-axis) showing LH (red), voltage-dependent anion channel 1 (VDAC1; green), and DNA counterstained with DAPI (blue). Note the absence of yellow fluorescence, indicating a lack of colocalization. (C, D) Yellow arrows and orange dotted lines indicate mitochondria occasionally detected by SEM in macerated sections of the same tissues coated with NanoSuit solution. The blue rectangle in image (C) indicates the region shown at higher magnification in image (D). In these mitochondria, only the outer membrane was visible, as shown in the upper right orange panel. The inner membrane and other organelles, such as the Golgi apparatus and endoplasmic reticulum, were not observed. (E, F) Yellow arrows and orange dotted lines indicate mitochondria occasionally detected by SEM in non-macerated sections of the same tissues coated with NanoSuit solution. The blue rectangle in image (E) indicates the region shown at higher magnification in image (F). As in the macerated sections, only the outer membrane was visible, as shown in the upper right orange panel (the pink arrow indicates the signal of a gold-labeled antibody), while the inner membrane and other organelles—such as the Golgi apparatus and endoplasmic reticulum—were not observed. Scale bars: 20 μm (C), 10 μm (E), 5 μm (D), and 2 μm (F).

As shown in Fig. 2B and Supplementary Movies 1 and 2, three-dimensional fluorescence immunohistochemistry revealed LH in red and the mitochondrial marker voltage-dependent anion channel 1 (VDAC1) in green. The absence of yellow fluorescence indicates difficulty in visualizing mitochondria in LH-positive cells.

Mitochondria were observed in both macerated (Figs. 2C, 2D) and non-macerated sections (Figs. 2E, 2F), although they were relatively sparse. Only the outer mitochondrial membrane was visible, with no detectable inner membrane structures. Other organelles, such as the Golgi apparatus and endoplasmic reticulum, were not detected. Therefore, all spherical granules observed were most likely secretory granules.

We hypothesize that the occasional appearance of disrupted mitochondria and the absence of other organelles may be due to insufficient ultrastructural preservation, as the sections were fixed solely with 4% paraformaldehyde, without membrane-stabilizing agents. These findings underscore the inherent trade-off between ultrastructural preservation and antigenicity in electron microscopy sample preparation. Fixatives such as glutaraldehyde followed by osmium tetroxide are effective in stabilizing membrane-bound organelles, including the inner mitochondrial membrane. However, they substantially reduce antigenicity, thereby hindering immunolabeling [13, 14]. In contrast, fixation with paraformaldehyde alone better preserves antigenic sites but often results in poor preservation of fine structural details.

Secretory granules were clearly visible in our SEM images. This may be attributed to their high internal protein density, which enhances electron scattering and image contrast [15], their uniform spherical morphology [16], and their structural resilience during dehydration and drying [17, 18]. In contrast, mitochondria were less distinctly visualized, likely due to their more complex and less compact internal architecture—including cristae and double membranes—which results in weaker electron scattering. Additionally, these components are prone to collapse or distortion during sample dehydration, reducing their visibility [19]. Ultimately, fixation protocols should be tailored to the specific goals of the study, with optimal conditions determined empirically through experimentation.

We observed numerous secretory granules in LH-positive cells. It has been reported that the majority of the transcriptome in bovine anterior pituitary glands consists of hormone RNAs. Specifically, luteinizing hormone beta (LHB) mRNA constitutes approximately 11% of the total Reads Per Kilobase of exon per Million mapped reads (RPKM) in heifers [20], and about 0.3% of Transcripts Per Million (TPM) in heifers [21]. Although LH is secreted in a pulsatile or surge manner in mammals, it is also continuously secreted by gonadotrophs, even in the absence of GnRH stimuli [22]. Consequently, gonadotrophs require abundant secretory granules, as demonstrated in Fig. 1.

Unlike previous studies [10,11,12], the NanoSuit CLEM method did not require gold-labeled antibodies to detect LH. Immunoelectron microscopy typically employs gold- or metal-labeled antibodies for a single target molecule, limiting its application. In contrast, the NanoSuit CLEM method offers an advantage for observing additional targets within LH-positive cells without such constraints.

For light microscopy, the use of a coverslip is essential, but it must be removed before SEM analysis. We used glycerol as a mounting medium to facilitate easy coverslip removal. We recommend non-hardening mounting media, such as glycerol, to preserve ultrastructural integrity. Hardening mounting media can also be removed, but this often requires extended shaking in PBS for several days. Using a non-hardening medium simplifies the process.

In this study, the SSE solution was diluted 100-fold with water. Since conductivity is linked to moisture retention, excessive dilution may weaken its conductive properties. To effectively protect tissues from electron beam damage, maintaining a certain concentration is essential. If further observation of the sample after SEM analysis is required, a slightly higher concentration, such as a 20-fold dilution, may be preferable. Conversely, if the SSE solution is too highly concentrated, fine structures may become obscured, making them unobservable under SEM. Additionally, because the electron beam causes thinning of the film, areas observed under SEM cannot be removed even after washing with PBS. Ultimately, the concentration should be adjusted based on the intended purpose, with the optimal level determined through experimentation.

In conclusion, our findings suggest that NanoSuit CLEM is a promising tool for visualizing the ultrastructure of key cells in cryosections, including those from large animals. This expands the applications of NanoSuit CLEM in biological research.

Methods

Animals

All experiments were performed in accordance with the Guiding Principles for the Care and Use of Experimental Animals in the Field of Physiological Sciences (Physiological Society of Japan) and were approved by the Committee on Animal Experiments of Yamaguchi University (approval number 301).

Pituitary glands were obtained from post-pubertal Japanese Black heifers (25.4 ± 0.5 months old, n = 5) at a local abattoir. The glands were isolated within 15 min of slaughter and immediately stored in 4% paraformaldehyde at 4°C for 16 h. The fixed samples were then placed in 30% sucrose in PBS at 4°C until the tissue blocks sank. The heifers were in the mid-luteal phase, i.e., 8 to 12 days after ovulation, as determined by macroscopic examination of the ovaries and uterus [23]. Anterior pituitary glands exhibit the highest LH concentrations during this phase [24].

Sagittal sections (10 µm thick) were prepared using a cryostat (CM1900, Leica Microsystems Pty Ltd., Wetzlar, Germany) and mounted on glass slides (MAS coat Superfrost, Matsunami Glass Ind. Ltd., Osaka, Japan). Sections were treated with 0.3% Triton X-100 in PBS for 15 min and then incubated in 0.5 ml PBS containing 10% normal goat serum (FUJIFILM Wako Pure Chemicals, Osaka, Japan) for 1 h to block nonspecific binding. They were then incubated with a 1:1,000 dilution of mouse monoclonal anti-LH antibody (clone 518-B7) [25] for 12 h at 4°C. This antibody does not cross-react with other pituitary hormones [26].

Subsequently, sections were incubated for 2 h at room temperature with Alexa Fluor 546-conjugated goat anti-mouse IgG (1 µg/ml; A-11030, Thermo Fisher Scientific, Waltham, MA, USA) and 1 µg/ml of 4′,6-diamidino-2-phenylindole (DAPI; FUJIFILM Wako Pure Chemicals). After secondary antibody incubation, sections were washed once with PBS containing 0.5% Tween 20 and twice with PBS, then covered with coverslips (55 × 24 mm, Matsunami Glass Ind. Ltd.) using glycerol as the mounting medium.

Stained sections were observed under a light microscope (BX51, Olympus Co., Tokyo, Japan). The region of interest was marked using a water/organic solvent-resistant pen (Frost Marker, Matsunami Glass Ind. Ltd.) on both the front of the coverslip and the back of the slide. Marked sections were imaged using a NanoZoomer-RS digital camera (Hamamatsu Photonics Co., Hamamatsu, Japan).

Next, the coverslip was removed, and the slides were rinsed twice with PBS and once with water. The back-side mark served as a reference point for applying SSE solution (NanoSuit solution Type III, Nisshin EM, Tokyo, Japan). The SSE solution, diluted 100-fold in water, was applied to cover the entire tissue section and left for 1 min. Excess solution was removed by spinning the slide (2,000 rpm, 15 sec; Slidespiner, Labnet International Inc., Woodbridge, NJ, USA) or by vigorous manual shaking.

The slide covered with SSE solution was placed in a field emission scanning electron microscope (FE-SEM; S-4800, Hitachi High-Technologies, Tokyo, Japan). SEM imaging was performed at an acceleration voltage of 1.0 kV. The vacuum level in the observation chamber ranged from ~10⁻³ to 10⁻⁶ Pa. Secondary electrons were detected using a lower-positioned detector. Additional SEM settings were as follows: working distance, 8 mm; aperture size, ϕ 100 μm; scan speed, 10–15 frames/sec. ImageJ software (version 1.54p; National Institutes of Health, Bethesda, MD, USA) was used for all length measurements. The image scale was calibrated using a known reference distance, and measurements were performed with the straight-line tool.

After fluorescence immunostaining, some cryosections (macerated sections) underwent a maceration process using 0.1% osmium tetroxide solution [27] to selectively remove cytoplasmic content and visualize organelles. Other sections (non-macerated) were not subjected to this process. Instead, they were blocked with 10% normal goat serum and 10% normal donkey serum (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA), followed by incubation with a rabbit polyclonal antibody against VDAC1 (N1C2, GeneTex, Inc., Irvine, CA, USA) along with the anti-LH antibody.

For secondary labeling, in addition to Alexa Fluor 546 goat anti-mouse IgG, non-macerated sections were incubated with a mixture containing:

• Alexa Fluor 488-conjugated goat anti-rabbit IgG (A-11034, Thermo Fisher Scientific), diluted to 1 µg/ml

• 18 nm colloidal gold-conjugated donkey anti-rabbit IgG (711-215-152, Jackson ImmunoResearch Laboratories, Inc.), diluted 1:20

• 40 nm colloidal gold-conjugated donkey anti-rabbit IgG (711-405-152, Jackson ImmunoResearch Laboratories, Inc.), also diluted 1:20

After staining, the sections were covered with coverslip using glycerol and examined using the Mica Micro Hub Imaging System (Leica Microsystems, Wetzlar, Germany) with a 63×/1.20 NA water-immersion objective in confocal mode. Sections were imaged using 405 nm, 488 nm, and 561 nm laser lines. Z-stack images were acquired at 0.3 μm intervals to capture 3D structures. All imaging parameters (laser power, exposure time, detector gain) were kept constant across samples. Image acquisition and processing were performed using LAS X software (Leica Microsystems), with default projection mode used unless otherwise specified. Finally, coverslips were removed from macerated or non-macerated sections. The slides were rinsed twice with PBS and once with distilled water, followed by application of SSE solution and SEM observation.

Conflict of interests

The authors have nothing to declare.