Keywords

20.1 Introduction

Preclinical studies in the field of adrenal cancer, both in vitro and in vivo using adequate and useful cell and animal models, are mandatory to study the molecular and cellular characteristics of these tumors as well as to investigate the potential of a therapeutic drug or strategy. Because results obtained at preclinical level are important steps before translation to clinical trials, such experiments must be designed, conducted, analyzed and reported to the highest levels of rigor and transparency. The principal experimental models of adrenocortical carcinoma (ACC), pheochromocytoma (PHEO) and paraganglioma (PGL) are briefly described, taking into consideration that this field is rapidly evolving, with preclinical models being developed and validated to reproduce the tumor microenvironment (TME).

20.2 Adrenocortical Carcinoma: Preclinical and Translational Models

The heterogeneity of ACC cell phenotypes requires cellular models capable of reproducing this condition. Preclinical mouse models of ACC have been generated in the form of cell line-derived xenografts and patient-derived xenografts, established using cell suspension cultures or tumor tissues from surgery, respectively. These are administered subcutaneously to immunocompromised mice for localized propagation and in vivo tumor growth, offering the opportunity to observe in vivo the progression of the human tumor, however with some limitations [1]. Transgenic mouse models are also available but their genetic modifications might only partially recapitulate the heterogeneity of human disease [2] (Fig. 20.1).

Fig. 20.1
A flow diagram illustrates a genetically engineered mouse model of human cancer. It depicts various cell lines derived from human tumors and xenografts, which are subsequently implanted into mice. Genetically engineered mice are created by introducing transgenes, or foreign genes, into the mouse genome.

Schematic representation of experimental models of adrenocortical carcinoma. (Created with BioRender.com)

Full size image

20.2.1 Cell Lines

Currently, six human ACC-derived cell lines are used as models for this disease, and several primary cultures of ACC are used in preclinical studies. The principal characteristics of the human ACC cell models are described in depth in two recent reviews [2, 3]. The first human ACC cell line, named NCI-H295 derived from a primary ACC, was reported in 1990. Several sub-strains were adapted from the NCI-H295 cell line using alternative growth conditions, the first of which was NCI-H295R cells. NCI-H295R cells have been reported to harbor a large deletion in the TP53 locus and carry an activating CTNNB1 mutation. Furthermore, NCI-H295 cells and their sub-strains have been shown to produce steroids under basal conditions. Interestingly, the reported steroidogenic capacities are influenced by the culture conditions and the substrate [4].

The first metastasis-derived ACC cell line MUC-1 was established in 2016. MUC-1 cells represent a preclinical model of resistance to treatment with etoposide, doxorubicin, cisplatin and mitotane (EDP-M), as they were obtained from a patient progressing after EDP-M. MUC-1 cells are characterized by a low steroidogenic activity and by a somatic deletion/frameshift mutation in the TP53 gene. Two other metastasis-derived ACC cell lines were subsequently established in 2018, namely CU-ACC1 and CU-ACC2. The CU-ACC1 cell line secretes high levels of cortisol but not aldosterone, and it is endowed with an activating point mutation of the CTNNB1 gene. CU-ACC2 cells secrete very low amounts of cortisol and carry a mutation in TP53. Another recently available ACC cell model is the JIL-2266 cell line, derived from a primary ACC and first reported in 2021. JIL-2266 cells are characterized by intermediate to low expression of SF-1 and their hormone production depends on the composition of the culture medium. Genetically, JIL-2266 cells carry a hemizygous stop-gain mutation in the TP53 gene. Furthermore, a pathogenic germline mutation in the MUTYH gene was observed, leading to the inactivation of the base excision-repair process and, consequently, to a high tumor mutational burden. The most recent ACC cell line, named TVBF-7, was established from lymph node metastasis of a patient progressing after EDP-M treatment. TVBF-7 cells produce high levels of cortisol under basal conditions. Genetic analysis reported an altered Wnt/β-catenin pathway due to the presence of a nonsense APC mutation.

Murine cell lines are also available. Among these, ATC1 and ATC7 cells were generated from transgenic mice and are used as cellular models in basic endocrinological studies [5, 6].

A different approach involves the use of co-cultures. The co-culture of a NCI-H295R cell monolayer above an adipose stem cell monolayer led to reprogramming of both cell types and to a more aggressive disease phenotype [7]. Recent experiments have also looked at interactions between ATC7 cells and human monocytes, showing that activation of intra-adrenal immune cells may play a role in stimulating steroidogenesis or proliferation [2].

20.2.2 3D Cell Models

3D models are particularly promising as an opportunity to better recapitulate the metabolic interplay between TME, cancer cells and tissue zonation. Indeed, 3D models replicate in vivo ACC tumor growth, since they have several important characteristics of solid tumors, including more representative transcriptional profiles, the development of an extracellular matrix and cellular junctions. Similar to in vivo solid tumors, there are various concentrations of oxygen and nutrients as well as different rates of cell proliferation from the outer layer to the center, which can result in central necrosis and regions of hypoxia. 3D models of ACC consisted mostly of spheroids generated from NCI-H295R cells primarily used in drug-screening protocols [8, 9], and they have been standardized in both NCI-H295R and MUC-1 cells as well as in ACC primary cultures [10].

Another research group released promising organoid models of ACC, studying metastasis through matrix metalloproteinase experiments in organoids and microfluidic models and demonstrating that NCI-H295R cells secrete active matrix metalloproteinases [2].

20.2.3 Cell Xenograft in Mouse

Patient-derived xenografts (PDXs) in immunodeficient mice are recognized as the gold standard for human cancer models. However, limited PDXs as well as cell line-derived xenograft (CDX) models of ACC are available. The first PDX model reported in 2013 was generated from a pediatric patient with ACC. No separate cell line of this model has been established [2]. Since then, three new models have been developed, from which derive MUC-1, CU-ACC1 and CU-ACC2 cell lines, which not only retain significant molecular similarity to their primaries, but also recapitulate the differences between those primaries and some of the heterogeneity of the disease [11, 12]. Further work has investigated the behavior of one of these models, CU-ACC2-M2B, in a humanized mouse model to better understand the efficacy of checkpoint inhibitor immunotherapy. A detailed description of these models has been recently published [2].

In vivo experiments with the zebrafish offer, with some limitations, a suitable and expeditious animal model for the screening of potentially effective drugs, identification of dose toxicity, and determination of the most promising compounds for more advanced preclinical phases, especially in rare diseases such as ACC. Different studies reported results obtained with zebrafish embryos xenografted with ACC cells, to investigate the effect of different drugs on tumor growth and metastasis formation [13,14,15].

20.2.4 Genetically Engineered Mouse Models

Efficient mouse genome manipulation has allowed the development of genetically modified ACC models, engineered to contain specific genetic alterations which promote de novo tumor formation within the adrenal cortex; genes and pathways of interest are recognized by human clinical observations or in vitro findings. Once identified, a gene can be deleted, overexpressed, or mutated within the adrenal cortex to experimentally define its role in the pathogenesis of adrenal tumors [16]. Early models mostly focused on the role of IGF2 to elucidate its role in adrenocortical neoplasia. Other recent work has focused more on CTNNB1, APC, WNT, ZNRF3, and TP53 [2, 17]. Val et al. recently showed that inactivation of ZNRF3 in the mouse adrenal cortex, recapitulating the most frequent alteration in ACC patients, is associated with sexually dimorphic tumor progression, promoting or hampering the involvement of phagocytic macrophages in men and women, respectively [18].

20.3 Pheochromocytoma/Paraganglioma Preclinical and Translational Models

The need for cellular models of pheochromocytomas and paragangliomas (together referred to as PPGLs) is becoming urgent as basic research shifts from genetics to the molecular mechanisms driving the tumorigenesis and clinical vulnerabilities of PPGLs [19,20,21]. Research efforts have largely focused on SDHB, since the risk of metastasis is strongly related to genotype, ranging between 30% and 40% with SDHB mutations.

20.3.1 Cell Lines

PPGL cell line models include PC12, MPC cells, the MPC derivative MTT, the immortalized chromaffin cells, the putative human PHEO progenitor line hPheo1, and the recently developed RS0 and RS1/2 cell lines (Table 20.1).

Table 20.1 Characteristics of the different cell lines used in preclinical models of adrenocortical carcinoma and pheochromocytoma/paraganglioma
Full size table

20.3.2 Rat Pheochromocytoma (PC12)

The adrenal rat pheochromocytoma (PC12) cell line was originally isolated from a PHEO developed in an irradiated rat in 1976 [22], and subsequently found harboring MAX gene deletion [23]. This cell line has the characteristic of precursor cells for both sympathetic neurons and chromaffin cells. PC12 can differentiate toward a neuronal phenotype in response to the nerve growth factor, while dexamethasone treatment upregulates catecholamine synthesis and storage. PC12 expresses several of the catecholamine biosynthetic enzymes, including tyrosine hydroxylase along with the enzymes converting l-DOPA to dopamine (aromatic l-amino acid decarboxylase) and dopamine to norepinephrine (DA β-hydroxylase), while phenylethanolamine N-methyltransferase (PNMT), which produces epinephrine from norepinephrine, is not expressed in these cells. PC12 cells are now available from a wide range of cell repositories and have been used in many studies, not only related to adrenal function and catecholamine production, but also in neuronal differentiation and other aspects of neurological development and function.

20.3.3 Mouse Pheochromocytoma Cell Line and Mouse Tumor Tissue Cells

Mouse pheochromocytoma cell (MPC) and the later derived mouse tumor tissue (MTT) cells, developed in the labs of Arthur Tischler and Karel Pacak, respectively, were derived from PHEOs arising in the adrenal medulla of the NF1 knockout mouse [24, 25]. The MPC cell line generally employed for preclinical studies pertaining to human tumors is 4/30/PRR. These cells typically showed extensive spontaneous neuronal differentiation. MPC and MTT are valid tools for studying genes and signaling pathways governing cell growth and differentiation in adrenal medullary neoplasms and are a unique model for studying the regulation of PNMT expression, as they display positive staining for PNMT and produce epinephrine. These cells are also considered a useful model for studying neurotransmitter release and neuroendocrine secretion. Several cell lines were then derived from MPC and MTT to improve and diversify research studies. To assess possible associations between SDHB gene mutations and invasiveness, Richter et al. established an MTT SDHB knockdown by viral transduction with lentiviral particles. Since MPC and MTT spontaneously form clusters in cultures, it is possible to easily generate spheroids [26], extremely useful for switching from monolayer (2D) cultures to 3D spheroids. Spheroids provide an excellent in vitro model to study the influence of hypoxia under conditions close to the in vivo situation, and for anticancer drug screening.

20.3.4 Immortalized Chromaffin Cells

Another mouse cell model, dubbed “immortalized mouse chromaffin cells” (imCCs), was derived from an SDHB knockout mouse [27]. These cells are deficient for the SDHB protein and show loss SDH activity, accompanied by high levels of intracellular and secreted succinate. Letouzé et al. also found other established characteristics of SDHB loss in imCCs, including elevated expression and nuclear translocation of HIF2a and a hypermethylation phenotype. Nevertheless, imCCs exhibit a mesenchymal morphology suggesting that they may not be mature chromaffin cells.

20.3.5 Rat SDH-Deficient RS0 Cells

PHEOs from irradiated rats with a heterozygous germline SDHB mutation were injected subcutaneously into NOD scid gamma (NSG) mice. This approach led to obtain two distinct, serially transplantable, xenograft and cell lines designated RS0 (for rat Sdhb null) and RS1/2 (for rat Sdhb haplo-insufficient). The ultrastructural features of RS0 are reminiscent of human SDH-deficient tumors, with relatively sparse secretory granules and cytoplasmic vacuoles, but the typical mitochondrial swelling and degeneration found in many human tumors are absent. The catecholamine profile of RS0 is also evocative of some SDH-deficient human PGLs, predominantly producing dopamine, with low levels of norepinephrine and undetectable epinephrine [28].

20.3.6 Progenitor Cells Derived from a Human Pheochromocytoma

In 2013, human primary PHEO cells were immortalized by introducing the catalytic subunit of human telomerase reverse transcriptase (hTERT) into the cells [29]. The resulting cell line is a neuroendocrine progenitor cell line called hPheo1. The characterization of these hPheo1 cells showed that the genes associated with catecholamine synthesis were highly expressed in the tumor tissue of origin, but most of them were downregulated in hPheo1 cells. More recently, genomic deletion of SDHB in hPheo1 cells was performed by the CRISPR/AsCPF1 system [30], obtaining a new cell line called hPheo1 SDHB-knockout.

20.3.7 Animal Models

In addition to the rat model harboring a heterozygous germline SDHB mutation from which RS0 and RS1/2 were isolated, there is a new SDH-deficient PGL model derived from mice. Multiple PHEOs arise in mice in which complete loss of SDHB was combined with loss of NF1 [31]. The SDHB/NF1 mouse model has provided insights into early mechanisms of tumorigenesis, but derivative cell lines have not yet been established. Other new investigative tool includes the MENX rat that carries a frameshift mutation in the CDKN1B gene (encoding for p27) and spontaneously develops PHEO. Intriguingly, these tumors recapitulate most characteristics of SDH-related PGL including norepinephrine and dopamine secretion, a HIF-2α-driven pseudohypoxic signature, metabolomics reprogramming associated with an accumulation of the oncometabolite 2-hydroxyglutarate, DNA hypermethylation and massive angiogenesis [32]. A zebrafish model in which loss of SDHB in homozygous larvae recapitulates the metabolic characteristics of human PGL has also been generated, but tumors have not so far been observed in these animals [33].

20.4 Complex 3D Adrenal In Vitro Preclinical Models

Given the physiological relevance of the functional interaction occurring between the steroidogenic cortical and the chromaffin medullary components of the adrenal gland, it is likely that the crosstalk that regulates the gland organogenesis and functions [34] may also affect the pathogenesis and progression of adrenal tumors. The peculiar organization of mammalian adrenals with a portal system connecting cortex and medulla strongly supports the evolutionary importance of the crosstalk inside the same gland. Local glucocorticoids have been demonstrated to stimulate medulla activity, in particular by upregulating the key enzymes in catecholamine biosynthesis [34], while medullary neuroendocrine peptides support cortical secretion and growth [35]. Therefore, a relevant implementation of the 3D in vitro models for these two types of cancers should consider not only the TME but also the other secretory part:

  • Ex vivo fetal adrenal gland explants or adrenal organoids derived in vitro from mixed primary cell populations obtained from fetal gland cell dissociation and reassembly. The complex gland organization and secretory activity are maintained in the explant model [36] but also in the in vitro-induced organoids derived from primary mixed cell populations isolated from fetal adrenals, which display a spatial organization and a steroidogenic and catecholamine secretion resembling the gland of origin [37].

  • Tumor tissue slice cultures have recently been developed as innovative tools for in vitro testing drug efficacy for personalized treatment strategy [38]. Tumor specimens left intact by mechanically slicing after surgical removal are incubated under standardized organo-culture conditions in either static or microfluidic settings up to several days to test different drug treatments.

  • Mixed spheroids obtained from in vitro co-culture of NCI-H295R and MTT cell lines have been recently developed. These 3D structures, called adrenoids, mimic the organization of the gland of origin and the distinct endocrine activity of medullary catecholamine and cortex corticosteroid secretion, demonstrating a growth advantage due to the coexistence of the two endocrine components [39].