Spontaneous Murine Model of Anaplastic Thyroid Cancer

Summary

Here, we present a standard pipeline to obtain murine ATC tumors by spontaneous genetically engineered mouse models. Further, we present tumor dynamics and pathological information about the primary and metastasized lesions. This model will help researchers to understand tumorigenesis and facilitate drug discoveries.

Abstract

Anaplastic thyroid cancer (ATC) is a rare but lethal malignancy with a dismal prognosis. There is an urgent need for more in-depth research on the carcinogenesis and development of ATC, as well as therapeutic methods, since standard treatments are essentially depleted in ATC patients. However, low prevalence has hampered thorough clinical studies and the collection of tissue samples, so little progress has been achieved in creating effective treatments. We used genetic engineering to create a conditionally inducible ATC murine model (mATC) in a C57BL/6 background. The ATC murine model was genotyped by TPO-cre/ERT2; Braf^CA/wt; Trp53^{ex2-10/ex2-10} and induced by intraperitoneal injection with tamoxifen. With the murine model, we investigated the tumor dynamics (tumor size ranged from 12.4 mm² to 32.5 mm² after 4 months of induction), survival (the median survival period was 130 days), and metastasis (lung metastases occurred in 91.6% of mice) curves and pathological features (characterized by Cd8, Foxp3, F4/80, Cd206, Ki67, and Caspase-3 immunohistochemical staining). The results indicated that spontaneous mATC possesses highly similar tumor dynamics and immunological microenvironment to human ATC tumors. In conclusion, with high similarity in pathophysiological features and unified genotypes, the mATC model resolved the shortage of clinical ATC tissue and sample heterogeneity to some extent. Therefore, it would facilitate the mechanism and translational studies of ATC and provide an approach to investigate the treatment potential of small molecular drugs and immunotherapy agents for ATC.

Introduction

Thyroid cancer is one of the most common endocrine malignancies¹, originating from the thyroid epithelium. In recent years, the incidence of thyroid cancer has increased rapidly worldwide². Thyroid cancer can be divided into distinct types according to the degree of tumor cell differentiation. On the basis of clinical behavior and histology, thyroid carcinomas are divided into well-differentiated carcinomas, including papillary thyroid carcinoma (PTC) and follicular thyroid carcinoma (FTC), poorly differentiated carcinoma (PDTC), and undifferentiated or anaplastic carcinoma of the thyroid (ATC)³. In contrast to PTC, which is a common type with mild behavior and better prognosis⁴, ATC is a rare and highly aggressive malignancy that accounts for 2% to 3% of all thyroid tumors⁵. Although ATC is rare, it is responsible for approximately 50% of thyroid cancer-related deaths, with dismal survival (6-8 months)^6,7. Over 50% of ATC cases are diagnosed as lung metastasis⁸. In addition to the aggressive nature of ATC, limited effective treatment has been developed in the clinic. Therefore, ATC patients have a bleak prognosis^9,10,11. This suggests that further in-depth studies are urgently needed on the molecular mechanisms underlying the development of ATC and treatment.

The tumorigenesis of ATC is a dynamic dedifferentiated process. The difficulty in collecting human tumor samples at each stage in clinical studies has hindered the understanding of the mechanism of development from well-differentiated to undifferentiated carcinomas. In contrast, the use of murine ATC models (mATC) favors the collection of mATC samples in the whole tumorigenesis course. Therefore, we can better understand the mechanisms of tumor formation by analyzing the dynamic dedifferentiated process. In addition, the heterogeneity of clinical ATC samples has also contributed to the difficulty in understanding the molecular mechanism. Nevertheless, mice shared the same genetic backgrounds and were maintained in similar living environments, ensuring each tumor's consistency. This facilitates exploring the generalized role of ATC development^12,13,14. Additionally, mATC is an in situ tumor model that can restore the influence of the anatomic location and tissue-specific microenvironment. As such, compared with commonly used immunodeficient mice, mATC is a spontaneous murine model with an intact immune system and immune microenvironment.

Therefore, we constructed conditionally induced mATC with the C57BL/6 strain, which is a murine model capable of reproducing the pathological features of dedifferentiated thyroid carcinoma. Based on this model, we gave a brief overview of the molecular basis, construction ideas, pathological features, and applications of mATC. In addition, we observed and reported tumor growth, survival time, metastasis, and pathological features of mATC. We believe this will be an informatic overview to assist other researchers in using this model easier.

We constructed a conditional inducible mATC model, as first reported by McFadden¹⁵; initially, we constructed mice: TPO-cre/ERT2, Braf^flox/wt, and Trp53^flox/wt. Specifically, TPO-cre/ERT2 mice included the human thyroid peroxidase (TPO) promoter (a thyroid-specific promoter), driving the expression of a cre-ERT2 fusion gene (a cre recombinase fused to a human estrogen receptor ligand binding domain). Cre-ERT2 is usually confined to the cytoplasm and enters the nucleus only when exposed to tamoxifen, which induces cre to exert recombinant enzyme activity. When the mice are crossed with mice carrying loxP-flanked sequences, after tamoxifen-induction, cre-mediated recombination deletes the floxed sequences in the thyroid cells to achieve the purpose of knocking out or knocking in specific genes.

In addition, Braf^flox/wt mice are a knock-in allele of human Braf based on the cre-loxP system. Braf^flox/wt murine transcript is encoded by endogenous exons 1-14 and loxP-flanked human exons 15-18. After cre-mediated excision of the floxed regions, the mutant exon 15 (modified with a V600E amino acid substitution linked with constitutively active Braf^V600E in human cancers) and the endogenous exons 16-18 are used to generate the transcripts. Furthermore, Trp53^flox/wt mice are knockout alleles of human Trp53 and have loxP sites flanking exons 2-10 of Trp53. When crossed with mice with a cre recombinase, cre-mediated recombination deletes the floxed sequence to knock out Trp53. Then, TPO-cre/ERT2, Braf^flox/w, and Trp53^flox/wt mice were crossed to obtain TB (TPO-cre/ERT2; Braf^flox/wt) mice and TBP (TPO-cre/ERT2; Braf^flox/wt; Trp53^flox/wt) mice, which could be used to generate PTC and ATC. After approximately 8 weeks, the mice were induced by an intraperitoneal (i.p.) administration of 150 mg/kg tamoxifen dissolved in corn oil for two administrations. Tumor growth could be monitored by high-frequency ultrasonography (the first time point of ultrasonography was recorded as Day 0). Initial ultrasonography was performed 40 days after tamoxifen introduction.

Protocol

The animal procedures described here were performed with the approval of the Animal Ethics Committee of West China Hospital, Sichuan University, Chengdu, Sichuan, China.

1. Induction of TBP mice

1. Identify mice genotype
   1. At around 3 weeks, separate the female mice from the male mice. At the same time, use ear tag clamp to fix an ear tag. Place the ear tags in the lower half and on the middle third of the ear, making sure to avoid the area with the highest concentration of capillaries.
   2. Gently but securely restrain the mice. Grasp firmly at the base of the mouse tail. Place the mice on a surface that will allow them to grip.
   3. Gently but firmly place the free hand over the shoulders, then quickly grasp the scruff of the neck between the thumb and forefinger. Hold the tail with the little finger.
   4. Swab the tail with alcohol. Use sterile scissors to snip the skin sample <5 mm and put it into a clean sample container with a label. It is not necessary to remove the fur of the mice. Compress the tail with sterile sponges to attain hemostasis.
   5. Next, lyse the murine tail and perform polymerase chain reaction (PCR) to identify the genotype.
      NOTE: After cutting one murine tail, the surface of the scissors must be wiped with alcohol cotton to avoid mutual contamination of genes between murine tails. After placing them in their cage, the mice must be watched for 5 min to look for any signs of bleeding at the wound site. Refer to Table 1 for the list of primers and the PCR settings used in this study.
2. Tamoxifen-induced TBP mice
   1. Weigh 0.3 g of tamoxifen and dissolve it in 15 mL of corn oil by ultrasound (power = 20%, duration = 20 min, temperature = 4 °C), at a concentration of 20 mg/mL. Store at 4 °C.
      NOTE: Tamoxifen is sensitive to light and needs to be placed in a brown container.
   2. At around 8 weeks, weigh the mice with electronic scales and give them an i.p. injection of tamoxifen at a dose of 150 mg/kg, administered twice with a 1 week interval.

2. Dissection and imaging of mouse thyroid tumors and metastatic tumors

1. Preparation
   1. Prepare the dissection tools: sterile scissors and forceps, sterile blade, 75% alcohol spray bottle, 20 cm straightedge, vernier calipers, paper towels, and alcohol cotton.
   2. Mouse tissue fixing solution: prepare 4% paraformaldehyde tissue fixing solution.
   3. Clean a 10 cm dish filled with about 10 mL of sterile phosphate buffered saline (PBS) for temporary tissue storage, and wash the blood on the surface of the tissue.
2. Thyroid extraction
   1. Euthanize the mice with carbon dioxide at a flow rate of 2.0 L/min for 5 min. Remove the mice from the cage and perform cervical dislocation.
   2. Place the mouse on the dissecting board with the ventral side facing up and the head away from the experimenter, and fix the limbs on the dissecting board with sterile syringe needles. For a more convenient removal of thyroid tissue, use another sterile syringe needle to fix the head.
   3. Disinfect the neck with three alternating rounds of a chlorhexidine or povidone-iodine scrub followed by 75% alcohol. Then, make a small incision above the center of the clavicle with sterile scissors and forceps.
   4. Continue the incision midline up to the mouth. Find the submandibular gland and remove it to expose the anatomical location of the thyroid, which is placed near the thyroid cartilage and trachea.
   5. Find the thyroid, carefully dissect the thyroid from the rest of the neck region with sterile scissors, then put the removed thyroid in a 10 cm dish filled with 10 mL of sterile PBS.
      NOTE: During removal of the thyroid gland, be gentle and slow and avoid cutting the neck blood vessels. If the neck vessels are cut, the neck will be immediately filled with blood, and the blood must be promptly removed with alcohol sponges to expose the anatomical location of the thyroid. Before removing the thyroid tissue, carefully observe the characteristics of the left and right lobes of the thyroid (size, shape, etc.) to avoid being unable to distinguish between the left and right lobes of the thyroid after removal.
   6. In sterile PBS, remove the blood from the surface of the thyroid tissue with sterile scissors and carefully cut off the trachea. Then, put the thyroid tissue on a sterile cloth, and measure the size of the left and right lobes of the thyroid gland with vernier calipers.
   7. Divide the left and right lobes of the thyroid gland into two parts with a sterile blade. Put one part into 2 mL of 4% paraformaldehyde solution for fixation, and put the other part into liquid nitrogen for preservation.
3. Extraction of lung and liver
   1. Disinfect the abdomen with three alternating rounds of a chlorhexidine or povidone-iodine scrub and 75% alcohol. Then, pinch just above the mouse's penis and make a small incision, cut along the midline of the abdomen to the subclavian bone, and expose the abdominal cavity.
   2. Find the liver in the upper part of the abdomen and carefully remove it. Place the liver into sterile PBS.
   3. Carefully cut the diaphragm along the ribs to expose the thoracic cavity. Next, grab hold of the sternum and pull up to widen the space even more. Find the lung and remove it. Put the removed lung into sterile PBS.
   4. Grossly observe whether there are metastases in the lung and liver. Count the number of metastases and take a digital record. After that, use a sterile blade to divide the lung and liver tissues into two parts. Put one part into 3 mL of 4% paraformaldehyde solution for fixation and another part into liquid nitrogen for preservation.
   5. Tissue dehydration
      1. Put the tissues into the dehydration box. Put the dehydration box into the basket at gradient alcohol dehydration as follow: 70% alcohol for 1 h; 80% alcohol for 1 h; 95% alcohol for 30 min; 95% alcohol for 30 min; anhydrous ethanol I for 30 min; anhydrous ethanol II for 30 min; xylene I for 20 min; xylene II for 20 min; paraffin wax I for 30 min; paraffin wax II for 1 h; paraffin wax III for 30 min.
   6. Embed the wax-impregnated tissues in the embedding machine.
      1. Put the melted wax into the embedding frame first. Before the wax solidifies, remove the tissue from the dehydration box and put it into the embedding frame, then attach the label.
      2. Allow the wax to cool at -20 °C on the freezing table. After the wax solidifies, remove the wax block from the embedding frame and trim it.
   7. Place the trimmed wax block on a paraffin slicer, set at a thickness of 5 µm and a size of 2 cm x 2 cm. After sectioning, allow the sections to float on a spreader with warm water at 45 °C to flatten the tissue. Pick up the tissue with a slide, and bake the slices in an oven at 65 °C for 2 h.
   8. After the water dries and the wax melts, remove the slide with the tissue and use it for hematoxylin and eosin (HE) and immunohistochemical (IHC) staining¹⁶. Make sure that the sections are adhered to the slides for staining.

3. HE staining of the primary tumor and lung

1. Dewaxing
   1. Put the slides with the sections in xylene for 10 min.
   2. Move into anhydrous alcohol (100%) (two bottles) for about 3 min each.
   3. Move into 95% alcohol (two bottles) for about 3 min each.
   4. Move into 80% alcohol for about 3 min.
   5. Move into 50% alcohol for about 3 min.
   6. Move into the tap water, and wash away the alcohol for about 1 min.
2. Staining
   1. Move the slides into hematoxylin for 8 min.
   2. Move the slides into the water for 1 min to wash away the hematoxylin; the tissue changes from blue to red.
   3. Move the slides into 1% hydrochloric acid alcohol for about 30 s.
   4. Move the slides into the water, washing for 5 min.
   5. Move the slides into eosin for 90 s, then wash them with water for 5 min.
   6. Move the slides into 50% alcohol, 80% alcohol, 95% alcohol (two bottles), and anhydrous alcohol (100%) (two bottles), for 1 min each.
   7. Move the slides into xylene for 5 min.
   8. After the slides are dry, seal them with neutral resin.

Representative Results

We induced mATC to investigate tumor growth, mouse survival time, and pathological characteristics. After induction, the mice were immediately sacrificed, and samples (thyroid, lung, and liver) were collected once one of the following conditions were found: 1) respiratory distress caused by tumor compression; 2) decreased appetite and abnormal vocalization; 3) unusually lethargy; and 4) body weight loss of over 20%. During the sampling process, we found that all mice (12/12) successfully formed tumors after induction. We recorded the mouse survival time, tumor features/size, and metastasized lesions.

Grossly, we observed the following: 1) the tumors were tender, and the size of the tumors on the left and right sides were inconsistent; 2) most mice (11/12) had lung metastases, but none had liver metastases. Specifically, the tumor size was monitored by animal high-frequency ultrasound and photoacoustic imaging systems (Vevo®3100) throughout the whole process¹⁷. Based on the ultrasound data, the tumor growth curve (Figure 1A) was plotted to observe the dynamic alteration of the tumor size. Furthermore, the tumors grew slowly at the early stage (average tumor size varied from 9.47 mm² to 11.75 mm² from Day 0 to Day 60), and became dramatically faster at the late stage (average tumor size varied from 11.75 mm² to 23.95 mm² (from Day 60 to Day 100). Most mice were sacrificed in the late stage. In short, mATC with a certain tumor latency period needs to be closely monitored after 60 days to prevent asphyxiation-related death.

On the other hand, the survival time of the mice was recorded, and a survival curve (Figure 1B) was plotted. The median survival of mATC was 130 days, ranging from 56-166 days. In addition, lung metastasis was found in most mATC (approximately 92%) (Figure 1C). We observed that only one mouse in this cohort did not show lung metastases on gross examination, and six mice had more than one lung metastatic lesion. No liver metastasis was found. In brief, these results were consistent with the biological behavior of ATC, which are prone to lung metastasis in clinics.

Furthermore, to better observe the dynamic process of mATC, we sacrificed the mice at two time points (1 month and 2 months after induction). We performed HE staining on the primary tumors and metastatic lung tissues of mATC (Figure 1D). In 1 month inducible tissue, we observed incomplete solidified features and the coexistence of follicular structures and malignant cells. The thyroid follicular structures disappeared, and the tumor solidified completely after a 2 month induction. HE staining of the primary tumor revealed that the tumor cells were morphologically diverse, with pleomorphic giant cells (indicated by a red arrow) and spindle-shaped cells (indicated by a yellow arrow). It also showed a wide variety in nuclear size, and many cells contained multiple nuclei. Clear metastatic foci (indicated by a circle) were seen in the lung. HE staining of metastatic lung tissues showed that the normal lung tissue was a reticular structure with clear alveolar structures and airspaces. Nevertheless, lung metastases showed a loss of normal reticular structure, air cavity thickening, and lung parenchyma.

Meanwhile, IHC staining was performed to further characterize (Cd8, Foxp3, F4/80, Cd206, Ki67, and Caspase-3) mATC tumors to quantify cell proliferation and apoptosis, and investigate the infiltration of lymphocytes, T-regular (Treg) cells, and myeloid cells (Figure 2A-D). Anti-Ki67 staining was highly positive, ranging from 86.9% to 95.07%, demonstrating a high degree of cell proliferation. Anti-activated caspase-3 antibody was used to test the rate of apoptosis, ranging from 5.2% to 51.9%. Specifically, mATC presented obvious CD8⁺ T cell infiltration, the ratio of which ranged from 0.47% to 10.55% (mean: 5.93%). This indicated that mATC was not an immune-desert tumor, which was consistent with ATC samples. Besides, Foxp3 staining defined Treg cells, which varied from 0.45% to 25.8%. In addition to lymphocytes, F4/80 and Cd206 were used to define macrophages and M2 macrophages, respectively. We found that myeloid cells extensively infiltrated tumors (F4/80 positive cell rate from 86.6% to 94.6%; Cd206 positive cell rate from 40.4% to 67.7%), which was consistent with the previous literature¹⁶. In brief, we found highly proliferative tumor cells, lymphocyte infiltration, and extensive infiltration of myeloid cells in mATC, which was consistent with clinical samples.

In conclusion, mATC samples showed homogeneity in tumor dynamics, metastasis, and pathological features, consistent with the clinical samples. Considering the tumor formation rate, mATC was a reliable murine model.

Figure 1: Tumor dynamics and pathological characteristics. (A) Tumor growth curve of mice (n = 5). Each line represents a mouse: the early stage ranges from Day 0 to Day 60, and the late stage ranges from Day 60 to Day 100. (B) Survival curves of mice (n = 12). The median survival of mATC was 130 days, ranging from 56 days to 166 days. (C) Lung metastasis curves of mice (n = 12). Representative dissection images of thyroid and lung metastasis. The red arrow indicates the thyroid tumor, and the white circle indicates metastasis in the lung. (D) HE staining of the primary tumor (1 month and 2 months after induction) and lung metastasis. The red arrow indicates the pleomorphic giant cell, the yellow arrow indicates the spindle-shaped cell, and the circled area indicates a metastasis in the lung. Please click here to view a larger version of this figure.

Figure 2: Brief description of immune cell infiltration in mATC. (A) Immunohistochemical staining of ATC murine primary tumors with antibodies (Cd8, Foxp3). The red arrow indicates a CD8 positive cell, the yellow arrow indicates a Foxp3 positive cell. (B) Immunohistochemical staining of ATC murine primary tumors with antibodies (Ki67, Caspase-3). (C) Immunohistochemical staining of ATC murine primary tumors with antibodies (F4/80, Cd206). (D) The quantification of IHC. Please click here to view a larger version of this figure.

Table 1: The list of primers and the PCR settings used in this study. Please click here to download this Table.

Discussion

Critical steps within the protocol for thyroid tumor dissection
During dissection, the anatomical location of the thyroid gland needs to be correctly understood. The thyroid gland is a butterfly-shaped gland located on the dorsal side of the submandibular gland near the thyroid cartilage and the trachea. During the procedure, severing the blood arteries on both sides of the neck was carefully avoided.

Modification and troubleshooting of the mATC breed
ATC is a rare and highly aggressive malignancy. The clinical characteristics of mATC and ATC patients have certain similarities. Specifically, the major cause of death in mATC is asphyxiation, which is consistent with ATC patients. Therefore, during the experiment, the following should be noted: 1) gently grasp the mice and pay attention to their respiration during the operation; 2) closely monitor during the late stage to prevent sudden death and failure to obtain samples in time; 3) the thyroid and lung tissues are easily affected by the blood on the surface of the tissues for observation and photography, so the blood volume of mice can be reduced by eyeball enucleation.

Limitations of using mATC
The occurrence and development of human ATC is complex and changeable, but the genetic background of mATC is relatively simple, which can only simulate a part of human ATC. For example, some ATC tissues harbored TERT and BRAF mutations, the NOTCH2NL copy number variant¹⁸, instead of P53 mutations, which may have different tumorigenesis mechanisms and clinical pathological characteristics^7,19,20. Additionally, ATC usually has a long time-course before diagnosis, but mATC obtains a diagnosis after an induction of 2 months. Therefore, there are inherent distinctions between mATC and human thyroid cancer, animal models cannot fully imitate all characteristics of human thyroid cancer, and many therapies that are restricted to humans cannot be evaluated on mATC. Furthermore, although mATC can be used to research the therapeutic effects of various small molecule medications, several biologics (e.g., antibody or antibody-related drugs) should be designed for mice specifically. In addition, the entire process of constructing conditional knockout mice and then obtaining the target genotype of mice by crossbreeding takes a long time and requires certain mouse culture conditions and costs^21,22.

Significance of using mATC in thyroid cancer research
The heterogeneity of the human population and the rarity of human ATC hinder the exploration of potential mechanisms and therapeutic options for ATC. As a reliable mouse model, mATC facilitates the collection of ATC tissues and enriches the ATC tissue sample pool. mATC can also be used to study the effects of specific gene functions on ATC, study the cellular and molecular mechanisms of ATC development, understand the mechanisms of thyroid tumor progression and drug resistance, and ultimately improve the prognosis of ATC patients¹⁶.

Future applications of mATC
mATC can be used for research on various aspects of radiotherapy, chemotherapy, gene therapy, immunotherapy, and targeted therapy. mATC can also be used to explore the therapeutic effects and adverse effects of regimens, such as combinations between different drugs or with radiotherapy. We subsequently utilized mATC to investigate the therapeutic effects of radiotherapy combined with immunotherapy and small molecule inhibitor drugs on ATC. In addition, mATC can be used to investigate the effects of various delivery modalities or routes on the antitumor effects of medications to develop drug delivery systems with enhanced antitumor activity. In future clinical applications, we hope to minimize the recurrence rate and mortality rate of thyroid cancer patients and thus improve the survival rate of patients.

We expect that more ATC murine models will be developed in the future, as well as more in-depth analyses of ATC in different genetic backgrounds, such as PTEN and P13K²³. These will more comprehensively reveal the molecular mechanism of ATC tumorigenesis and progression, predict the outcome of ATC treatment, and could provide patients with potential therapeutic targets^{24,25,26,27,28}. We hope that through more experimental exploration, we can improve the modeling method of the mouse model and shorten the modeling time to make more contributions to basic research.

Materials

Name	Company	Catalog Number	Comments
100x Citrate antigen retrieval solution (PH 6.0)	MXB	Cat# MVS-0101
50x EDTA antigen retrieval solution(pH 9.5)	ZSGB-GIO	Cat# ZLI-9071
Brafflox/wt mice	Collaboration with Institute of Life Science, eBond Pharmaceutical Technology Ltd, Chengdu, China
Caspase-3	Beyotime	Cat# AC033
CD8	Cell Signaling Technology	Cat# 98941; RRID:AB_2756376
CD206	Cell Signaling Technology	Cat# 24595; RRID:AB_2892682
Chamber for anesthesia induction	RWDlifescience
Enhanced DAB chromogenic kit	MXB	Cat# DAB-2031
Eosin staining solution	ZSGB-GIO	Cat# ZLI-9613
F4/80	Abcam	Cat# 100790; RRID:AB_10675322
Foxp3	Cell Signaling Technology	Cat# 12653; RRID:AB_2797979
Fully enclosed tissue dehydrator	Leica Biosystems	ASP300S
Hematoxylin staining solution	ZSGB-GIO	Cat# ZLI-9610
HistoCore Arcadia fully automatic tissue embedding machine	Leica Biosystems
Ki67	Beyotime	Cat# AF1738
Rotating Slicer	RWDlifescience	Minux S700
SPlink detection kits (Biotin-Streptavidin HRP Detection Systems)	ZSGB-GIO	Cat# SP-9001
TPO-cre/ERT2 mice	Collaboration with Institute of Life Science, eBond Pharmaceutical Technology Ltd, Chengdu, China
Trp53flox/wt mice	Collaboration with Institute of Life Science, eBond Pharmaceutical Technology Ltd, Chengdu, China
Ultrasonic cell crusher	Ningbo Xinyi Ultrasound Equipment Co., Ltd	JY92-IIN
Ultrasound gel	Keppler	KL-250
Ultrasound system	VisualSonics	Vevo 3100