Quantification and visualization of metastatic lung tumors in mice

Abstract

Histopathological examination is important for the diagnosis of various diseases. Conventional histopathology provides a two-dimensional view of the tissues, and requires the tissue to be extracted, fixed, and processed using histotechnology techniques. However, there is an increasing need for three-dimensional (3D) images of structures in biomedical research. The objective of this study was to develop reliable, objective tools for visualizing and quantifying metastatic tumors in mouse lung using micro-computed tomography (micro-CT), optical coherence tomography (OCT), and field emission-scanning electron microscopy (FE-SEM). Melanoma cells were intravenously injected into the tail vein of 8-week-old C57BL/6 mice. The mice were euthanized at 2 or 4 weeks after injection. Lungs were fixed and examined by micro-CT, OCT, FE-SEM, and histopathological observation. Micro-CT clearly distinguished between tumor and normal cells in surface and deep lesions, thereby allowing 3D quantification of the tumor volume. OCT showed a clear difference between the tumor and surrounding normal tissues. FE-SEM clearly showed round tumor cells, mainly located in the alveolar wall and growing inside the alveoli. Therefore, whole-tumor 3D imaging successfully visualized the metastatic tumor and quantified its volume. This promising approach will allow for fast and label-free 3D phenotyping of diverse tissue structures.

Introduction

Hematoxylin and eosin (H&E)-stained tissues on glass slides have been examined by conventional light microscopy for at least a century. H&E staining is essential for the identification of various tissue types and morphologic changes suggestive of cancer [1]. Although H&E staining provides information about changes in cell and tissue morphology, it also has some fundamental limitations [2].

Although pathological specimens are three-dimensional (3D) structures, H&E-stained tissues on a glass slide are two-dimensional (2D) images that lack information about tissue depth (the third dimension). This limits our ability to understand the complexity of real-world objects [3].

The histopathological diagnosis of large specimens is based on assessment of representative areas in the slides [4]. Therefore, concerns exist regarding the possibility of lesions in non-evaluated parts of the specimen [2]. Furthermore, there is a growing need to evaluate 3D structures in biomedical research [5].

Three-dimensional images can be used to visualize structures, determine regions of interest in relation to the surrounding structures, and obtain new measurements (e.g., volumetric data) [6]. Recent developments in bioimaging technology have made it possible to implement 3D and 2D organizational structures [7, 8].

Several imaging modalities can display 3D images [9, 10], including micro-computed tomography (micro-CT), which has been used for the detection of lung tumors and assessment of tumor burden in mouse lung cancer models and preclinical studies [11]. Micro-CT provides excellent air-tissue contrast. Therefore, it is used to longitudinally evaluate disease progression and treatment effects in numerous lung diseases, including cancer, fibrosis, and emphysema, as well as in transplantation models [12]. Optical coherence tomography (OCT) is an emerging noninvasive technique that provides high-resolution, cross-sectional microscopic images of tissues [13, 14]. OCT measures light scattered back from the microstructural features of examined tissues, at a scale of several to tens of microns and with a penetration depth of 1–3 mm [15]. OCT accurately diagnoses in vivo pathology and can be used to assess the fine microarchitectural features of normal and pathologic lungs [16]. Field emission scanning electron microscopy (FE-SEM) provides high-resolution 3D surface images of biological structures [17, 18]. FE-SEM has a spatial resolution of < 1 nm, and has a number of other advantages including improved performance at low accelerating voltages. FE-SEM has allowed life science studies to include high-resolution images that were previously only available through transmission electron microscopy [19].

The objective of this study was to develop a reliable, objective tool for quantifying the burden of metastatic tumor in mice lung. We used micro-CT, OCT, and FE-SEM to evaluate structural changes in lung cancer.

Materials and methods

Animals and treatment

Studies of pulmonary metastasis generally uses mouse models. For this study, 8-week-old C57BL/6 female mice were obtained from Koatech Co. Ltd. (Seoul, Korea) and acclimated for 7 days before study initiation. The mice were kept in a temperature-controlled environment (22 ± 3 °C, 55 ± 5% relative humidity) under a 12-h light/dark cycle. The mice were fed a rodent diet and filtered water ad libitum. Twenty mice were randomly divided into four groups: group 1 (G1) consisted of control mice; group 2 (G2) of mice with metastasis at 2 weeks after tumor injection; group 3 (G3) of age-matched control mice; and group 4 (G4) of mice with metastasis at 4 weeks after tumor injection. Two to four weeks after intravenous injection of mouse melanoma cells (B16-F10 cells; ATCC Inc., Manassas, VA, USA), the mice were sacrificed. This study was approved by the Animal Experiment Committee of Namseoul University, South Korea (NSU-19-05), and was conducted according to the Animal Protection Act.

Macroscopic examination

Lung surfaces were visually inspected to assess tumor spread. Only tumors evident on the lung surface were evaluated; the length (a) and width (b) of tumor were measured and tumor volumes was calculated as (a × b²)/2; those located inside the lung were not evaluated, which may have made the assessment of tumor location and size inaccurate. Therefore, bioimaging technologies, such as micro-CT, OCT, and FE-SEM, were used.

Micro-CT analysis and histopathological examination

Lungs were fixed with formalin followed by 2% phosphotungstic acid hydrate to enhance visualization of soft tissue details [20]. The metastatic tumor was quantitatively analyzed using a micro-CT scanner (Skyscan1173; Bruker, Kontich, Belgium) with an X-ray source of 75 kV/106 μV, pixel size of 9.94 μm, and 1.0-mm aluminum filter. Cross-sectional images of micro-CT scans were reconstructed, using a cut-off value of 0–0.03 to distinguish between bone and air with NRecon software (version 1.6.9.16; Bruker). CT Analyser (version 1.14.4.1; Bruker) was used for 3D analysis. Using the optimal threshold value, tumor cells were segmented and total tumor volume was quantified. The total lung volume and tumor volume fraction (%)were calculated using the built-in software of the micro-CT scanner.

After fixing in 10% neutral phosphate-buffered formalin, lung tissues were embedded in paraffin, and 4-μm sections were stained with H&E for histopathological examination. To assess variation in tumor size by location, lung tumors were sliced at 50-μm intervals from the bottom to produce 20 slices.

OCT analysis

Spectral domain-OCT (SD-OCT) was performed to examine lung tissues, similar to previous studies [21].

Briefly, a super-luminescent diode (Exalos Inc., Schlieren, Switzerland), with a 50-nm full-width half-maximum spectral bandwidth at the center wavelength of 840 nm, was incorporated into the SD-OCT system. OCT images obtained from 1000 axial scans were viewed at a rate of ~ 20 frames/s.

FE-SEM analysis

The tissue samples were dehydrated with increasing concentrations of ethanol (70–100%) and fixed in buffered neutral formalin solutions (10%). The samples underwent a purification process using xylene and paraffin was used to create blocks. The blocks were cut into 3-μm pieces using an automated rotary microtome (RM2255; Leica Inc., Nussloch, Germany), processed with 100% ethanol, and dried at room temperature. After gold coating, FE-SEM (Sigma 500; Carl Zeiss, Jena, Germany) images were taken at a voltage of 1 kV in slow scanning mode (1224 × 768 pixels), with a magnification of 100 × (voltage: 1 kV; distance: 3.8 mm) or 500 × (voltage: 1.5 kV; distance: 3.9 mm).

Statistical analysis

GraphPad Prism 6 (GraphPad Software, Inc., San Diego, CA, USA) was used to perform Student’s t-test. P values < 5% were considered statistically significant.

Results

Macroscopic examination

The total lung tumor volume in the different groups is summarized in Table 1. Total tumor volumes in G2 and G4 were significantly different from those in G1 and G3 (p < 0.01 and p < 0.05, respectively). Similarly, total tumor volume in G4 was significantly different than in G2 (p < 0.05).

Table 1.

Total tumor volume of lung by gross view

Group	Treatment	No. of mice	Tumor volume (mm3)
1	Control at 2 weeks	5	0.00 ± 0.00a
2	Metastatic tumor at 2 weeks	5	0.10 ± 0.04**
3	Age-matched control at 4 weeks	5	0.00 ± 0.00
4	Metastatic tumor at 4 weeks	5	3.10 ± 2.84*, #

^aData represent mean ± SD

*^,**Significantly different from G1 or G3 at p < 0.05 and p < 0.01, respectively

^#Significantly different from G2 at p < 0.05

Representative lung photographs from each group are shown in Fig. 1. No tumors were observed in G1 or G3 lungs (Fig. 1a and c). Many black-colored tumors were observed in G2 and G4 lungs (Fig. 1b and d).

Fig. 1.

Macroscopic findings of lungs from normal mice and mice treated with melanoma cells. Many tumors were observed in the lungs of mice treated with melanoma cells (c, d), whereas no tumors were observed in the lungs of mice treated with saline (a, b). a Representative image of the gross view of normal mouse lung 2 weeks after saline treatment. b Representative image of the gross view of normal mouse lung 4 weeks after saline treatment. c Representative image of the gross view of mouse lung tumors 2 weeks after melanoma treatment. d Representative image of the gross view of mouse lung tumors 4 weeks after melanoma treatment

Micro-CT analysis and histopathological examination

The results of quantitative analyses of the lung tumors detected by micro-CT are summarized in Table 2. Lung volume in G4 was significantly different compared to G3 (p < 0.01) and G2 (p < 0.01). There were significant differences in tumor volumes of G2 and G4 compared to G1 and G3 (p < 0.01, respectively), as well as of G4 compared to G2 (p < 0.01). The percentage of tumor volume was significantly different in G2 and G4 compared to G1 and G3 (p < 0.01 for both), and in G4 compared to G2 (p < 0.05).

Table 2.

Quantitative analysis of the lung metastatic tumor by micro-CT

Group	Treatment	No. of mice	Lung volume (mm3)	Tumor volume (mm3)	Percent tumor volume (%)
1	Control at 2 weeks	5	119.46 ± 31.26a	0.00 ± 0.00	0.00 ± 0.00
2	Metastatic tumor at 2 weeks	5	125.14 ± 21.38	17.02 ± 3.90**	13.61 ± 2.07**
3	Age-matched control at 4 weeks	5	95.09 ± 35.17	0.00 ± 0.00	0.00 ± 0.00
4	Metastatic tumor at 4 weeks	5	238.92 ± 57.62**, ##	74.01 ± 31.44**, ##	31.25 ± 14.28**, #

^aData represent mean ± SD

**Significantly different from G1 or G3 at p < 0.01

^{#, ##}Significantly different from G2 at p < 0.05 and p < 0.01, respectively

Tumor volume was greater in micro-CT analysis compared to macroscopic examination. Micro-CT clearly differentiated between tumor and normal cells, on the surface of and inside the lungs, making 3D quantification of tumor volume possible. Micro-CT 2D and 3D representations showed the inner and surface lung tumor mass (Fig. 2).

Fig. 2.

Micro-CT analysis of mouse lung tumors. Representative micro-CT image of metastatic tumors from mouse lung 4 weeks after melanoma treatment. a Cross-sectional view of metastatic lung tumor. b A 3D view of metastatic lung tumor. Many tumors were observed in the lung of a mouse treated with melanoma cells (arrow). Micro-CT micro-computed tomography, 3D three-dimensional

A comparison between micro-CT and microscopic analysis revealed that micro-CT clearly distinguished lesions, and showed clear margins between tumor and normal tissues (Fig. 3). H&E staining revealed differences in tumor size by location (Fig. 4).

Fig. 3.

Comparison of macroscopic and microscopic views and micro-CT. Macroscopic view showing lung tumors in a mouse treated with melanoma cells (red circle). Many tumors were clearly observed in the lung of a mouse treated with melanoma cells (arrow), but not in unaffected areas (arrow head), as seen on micro-CT and microscopic images. a Macroscopic view; b micro-CT image; c H&E-stained slice. Magnification: 100 × . Micro-CT micro-computed tomography, H&E hematoxylin & eosin

Fig. 4.

Microscopic analysis of the lungs of mice treated with melanoma. Number (1–20) indicates serial sections at 50-μm intervals. The tumor size significantly differed between sections (slide 1–20), reflecting variation in tumor size by location. Serial sections were stained with hematoxylin and eosin. Magnification: 40 × 

OCT analysis

Horizontal and vertical OCT images clearly distinguished normal tissue and tumors. Gross view of normal lung tissue showed no tumor in lung tissue (Fig. 5a) and OCT analysis of normal lung tissue revealed a smooth and uniform color distribution (Fig. 5b and c). Conversely, gross view of lung of mice treated with melanoma showed several metastatic tumors in lung tissue (Fig. 5a) and OCT analysis of the tumor showed a bright gray color, distinct from the surrounding normal tissue (Fig. 5e and f).

Fig. 5.

OCT analysis of lung tumor and normal tissues. OCT analysis clearly distinguished tumor from surrounding normal lung tissue. a Gross view of normal lung tissue after refixation. b Transverse cross-sectional view of normal lung tissue. c Longitudinal cross-sectional view of normal lung tissue. d Gross view of lung tumor. e Transverse section of lung tumor. f Longitudinal cross-section of lung tumor. Arrows indicate lung tumors. OCT optical coherence tomography

FE-SEM analysis

FE-SEM analysis clearly delineated the tumor from surrounding normal tissues, and the results were consistent with histopathological examination (Fig. 6).

Fig. 6.

Comparison of microscopic and FE-SEM images. Many tumors were clearly observed in the lung of a mouse treated with melanoma cells (arrow), but not in the unaffected area (arrow head), as seen on microscopy and FE-SEM. a Microscopic images of H&E-stained sections; b FE-SEM image. Magnification: 200 × . FE-SEM field emission-scanning electron microscopy

Discussion

In this study, we quantified the volume of metastatic lung tumor, and compared the tumor structures between micro-CT and histopathological examinations.

The micro-CT results were highly consistent with those of histopathological classification. The tumor and normal tissues were readily differentiated by micro-CT, which was confirmed by histopathological examination. In this study, micro-CT represents a promising approach for fast and label-free 3D phenotyping of diverse tissue structures. This will allow the 3D structure of metastatic tumors to be examined, and could provide additional information to that obtained from standard microscopic techniques. Even though micro-CT can differentiate tumor from non-tumor lesions such as fibrosis and granuloma, it must be confirmed by histopathological method. The multi-functional potential of bioimaging tools allows assessment of the volumetric parameters of metastatic tumors, and may provide a basis for pharmacological and pathological research.

Based on our results, bioimaging analysis of tumors using OCT and FE-SEM can distinguish tumor and non-tumor tissues in situ. However, OCT had limited ability to quantify the tumor volume in this study. Additionally, although FE-SEM analysis provided high-magnification images, they were 2D.

Because no single imaging technique can comprehensively evaluate tumor volume, reliable quantification of tumor volume in ex vivo lung tumors requires a combination of 3D imaging techniques, such as micro-CT and histology [22]. In a study, high-resolution OCT provided clear 3D images of rat embryos, including bone structures, vertebra, and the feet and skull [23]. OCT can induce tissue-wide shrinkage. However, the shrinkage is relatively homogenous, and the scans can be corrected for shrinkage [24]. In addition, advances in OCT technology have made it possible to image non-transparent tissues, thus enabling OCT to be used in a wide range of medical specialties [25].

Taken together, whole-tumor 3D imaging was successfully applied to quantify the volume of metastatic tumor and to visualize tumor structures. This promising approach will allow for fast and label-free 3D phenotyping of diverse tissue structures and can apply to toxicological evaluation of carcinogen or cancer therapeutics.

Funding

This work was funded by National Research Foundation of Korea (Grant no. 2019R1F1A1058721).

Declarations

Conflict of interest

Authors declare no conflict of interest.