Skip to main content
NCBI home page
Endocrine Oncology logoLink to Endocrine Oncology
. 2026 Sep 18;6(1):e250096. doi: 10.1530/EO-25-0096

AGK::BRAF fusion activates MAPK signaling, promotes genomic instability, and reduces NIS expression in thyroid cancer cells

Luiza Sisdelli 1, Welbert Gomes Rocha 1, Débora Mota Dias Thomaz 1, Roxane Hatanaka 1, Renata Elen Costa da Silva 2, Guilherme Henrique 3, Caroline Serrano-Nascimento 3, Matheus Fabião 4, Aline Rangel-Pozzo 4, Sabine Mai 4, Janete Maria Cerutti 1,✉
PMCID: PMC13615842  PMID: 42801159

Abstract

The AGK::BRAF fusion is a recurrent oncogenic alteration in sporadic pediatric papillary thyroid carcinoma (PTC), frequently associated with multifocal disease and pulmonary metastasis. However, the molecular and cellular consequences of AGK::BRAF expression in thyroid follicular cells remain poorly defined. Here, we investigated the biological effects of AGK::BRAF in PCCL3 thyroid cells and compared them with those induced by clinically established pediatric PTC oncogenic drivers, including RET/PTC1, RET/PTC3, ETV6::NTRK3, BRAF V600E, and wild-type BRAF. AGK::BRAF-expressing cells exhibited the most robust MAPK/ERK activation, characterized by markedly enhanced ERK and MEK phosphorylation relative to all other oncogenic drivers analyzed. By contrast, STAT3 and AKT signaling remained comparatively modest and did not show consistent oncogene-specific activation patterns. Functionally, AGK::BRAF expression induced features of thyroid dedifferentiation, including reduced sodium–iodide symporter (NIS) expression and impaired iodide uptake. Concomitantly, AGK::BRAF-expressing cells exhibited increased frequencies of micronuclei and chromatin bridges, together with disruption of three-dimensional telomere architecture, including reduced telomere signal number and intensity, altered spatial organization, and increased nuclear volume. Three-dimensional structured illumination microscopy (3D-SIM) further demonstrated chromatin remodeling, characterized by increased chromatin condensation accompanied by enlarged interchromatin spaces. Collectively, these findings suggest that AGK::BRAF is a dominant oncogenic driver of sustained MAPK/ERK signaling coupled to thyroid dedifferentiation, nuclear architectural remodeling, telomere dysfunction, and genomic instability. These coordinated molecular and structural alterations provide a mechanistic basis for the aggressive clinical course observed in AGK::BRAF-driven pediatric PTC.

Keywords: thyroid cancer, AGK::BRAF fusion, MAPK signaling, genomic instability, sodium–iodide symporter (NIS)

Introduction

Thyroid cancer is the most common malignancy of the endocrine system, with papillary thyroid carcinoma (PTC) representing 80–90% of cases (1, 2, 3, 4). In pediatric PTC, gene fusions are the most frequent genetic alterations, while point mutations are less common. These alterations commonly converge on the activation of the mitogen-activated protein kinase (MAPK) signaling pathway (5, 6, 7, 8, 9, 10).

Independent studies have confirmed that BRAF, RAS, and RET mutations drive distinct MAPK pathway signaling outputs, with the magnitude of signaling varying with the specific oncogenic driver and the experimental and cellular context in which it is assessed (11, 12, 13, 14, 15, 16). Within this context, it has been proposed that the BRAF V600E mutation induces robust MAPK pathway activation, in part because the mutant protein is less susceptible to ERK-mediated negative feedback regulation (16). By contrast, receptor tyrosine kinases such as RET and NTRK fusions, as well as oncogenic RAS mutations, may elicit comparatively more variable or attenuated signaling outputs, reflecting their ongoing sensitivity to feedback control mechanisms and upstream regulatory constraints. These findings have informed the development of the BRAF–RAS score (BRS), a transcriptional classification framework that stratifies tumors according to their MAPK signaling output, distinguishing BRAF-like from RAS-like tumors and categorizing additional tumors along this continuum based on their downstream transcriptional programs (8, 17, 18). In addition, a strong correlation between MAPK pathway activation and the expression of genes involved in thyroid follicular cell differentiation has supported the development of the thyroid differentiation score (TDS), a transcriptional metric that quantifies the expression of genes critical for thyroid function, including those regulating iodide metabolism and uptake (17, 18, 19, 20).

In children and adolescents, BRAF V600E and RAS variants are less frequent than in adults (30% vs 50–60%), while fusions such as RET/PTC and ETV6::NTRK3 are more prevalent (6, 21, 22). Patients harboring RET and NTRK fusions have a higher risk of lymph node and distant metastases and are less likely to achieve remission within a year post-surgery compared with those with BRAF V600E or RAS mutations (21, 23, 24).

Our recent findings highlighted significant differences in the genetic landscape of pediatric PTC in Brazil, compared with other regions (6). Among sporadic cases, the most frequently observed gene fusions were RET/PTC (28%), AGK::BRAF (19%), and ETV6::NTRK3 (18%). Notably, the AGK::BRAF fusion has been associated with more aggressive clinical behavior and increased risk of lung metastasis. This fusion was predominantly identified in younger patients, with a mean age of 13.15 years. Initial evidence from a published case report indicates that AGK::BRAF fusion may contribute to chromatin disruption, potentially leading to increased genetic instability (25).

While the AGK::BRAF fusion appears to play a role in the development and aggressiveness of sporadic pediatric PTC, the exact mechanisms involved remain unknown.

In this study, we aimed to evaluate the effects of ectopic AGK::BRAF expression in differentiated rat thyroid epithelial cells (PCCL3), comparing its impact to those of other recurrent genetic alterations identified in pediatric PTC, as well as to BRAF wild-type (BRAF WT) controls. We assessed activation of the MAPK and PI3K/Akt signaling pathways and evaluated whether AGK::BRAF expression impairs iodide uptake through downregulation of NIS. Furthermore, we investigated disruptions in 3D telomere architecture, including telomere shortening, clustering, and aberrant spatial distribution, as potential drivers of genomic instability and tumor progression.

Materials and methods

Stable PCCL3 expression models of AGK::BRAF and pediatric PTC oncogenic drivers

The PCCL3 cell line, derived from normal rat thyroid epithelial cells, was cultured in Ham’s F12 medium (Life Technologies, USA) supplemented with 5% fetal bovine serum (FBS), thyroid-stimulating hormone (TSH, 1 mIU/mL), hydrocortisone (10 nmol/mL), transferrin (5 μg/mL), and insulin (10 μg/mL) (Sigma-Aldrich, USA). To compare AGK::BRAF with canonical MAPK activation and clinically relevant pediatric PTC drivers, PCCL3 cells were generated in parallel under identical conditions to stably express AGK::BRAF, BRAF V600E, BRAF WT, RET/PTC1, RET/PTC3, and ETV6::NTRK3. BRAF V600E served as a canonical MAPK-activating control, whereas BRAF WT was included as a non-mutated BRAF comparator, enabling discrimination between mutation-specific signaling effects and those attributable to ectopic BRAF expression within the stable transfection system. RET/PTC1, RET/PTC3, and ETV6::NTRK3 represented recurrent pediatric PTC fusion oncogenes, enabling cross-class comparison of MAPK-driven signaling alterations. Expression vectors harboring the cDNA of BRAF V600E (pUHG10-3-mRNA V600E), RET/PTC3 (pUHG10-3-RET/PTC3), AGK::BRAF (pLVX-AGK-BRAF), and ETV6::NTRK3 (pMSCV-ETV6–NTRK3) were generously provided by Dr. James Fagin. The vector harboring RET/PTC1 was kindly provided by Alfredo Fusco. cDNA was excised from the original vectors, purified, subcloned into the pBudCE4.1 vector, and verified by sequencing, as previously described (19). Stable transfections into PCCL3 cells were performed using Lipofectamine 2000 (Invitrogen, Thermo Fisher Scientific, USA) following the manufacturer’s protocol. After transfection, cells were selected with Zeocin (300 μg/mL) to generate stable integrants. Transfected populations were subsequently maintained under continuous Zeocin selection (300 μg/mL) throughout all experiments to ensure stable retention of the expression constructs.

Western blot analysis

PCCL3 cells were serum-starved for 24 h, harvested on ice, washed in cold phosphate-buffered saline (PBS), and lysed in a buffer containing 50 mM Tris-HCl, 100 mM NaCl, 50 mM NaF, 1 mM Na3VO4, and 0.5% Triton X-100. Protein concentration was quantified using the BCA Protein Assay (Pierce, USA). Equal amounts of protein (50 μg) were separated by 8–12% SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked in 5% non-fat milk or 5% BSA in TBST and incubated overnight at 4°C with primary antibodies (Supplementary Table 1 (see section on Supplementary materials given at the end of the article)), followed by incubation with appropriate HRP-conjugated secondary antibodies. Signal detection was performed using Immobilon Western Chemiluminescent HRP Substrate (Merck, Germany), and signals were visualized with the ImageQuant LAS 4000 system (GE Healthcare, USA). Band intensities were quantified using ImageJ (NIH), corrected for background, and normalized. Phospho-ERK (pERK), phospho-MEK (pMEK), phospho-STAT3 (pSTAT3), and phospho-AKT (pAKT) levels were normalized to their respective total proteins (ERK, MEK, STAT3, and AKT), while NIS was normalized to α-tubulin. All experiments were performed using at least three independent biological replicates.

Iodide uptake assay

NIS-mediated iodide (I−) uptake was assessed using a non-radioactive method, as previously described (26). PCCL3 cells were synchronized by serum starvation for 24 h, and then incubated with 10 μM NaI in complete medium for 45 min. Cells were washed three times with cold PBS, and intracellular iodide was quantified using the Sandell–Kolthoff colorimetric reaction. Experiments were performed in triplicate.

Immunohistochemistry (IHC) analysis

IHC was performed on 3-μm-thick FFPE (formalin-fixed, paraffin-embedded) sections from five PTC cases with distinct genetic alterations and from normal thyroid tissue. Sections were deparaffinized in xylene, rehydrated through graded alcohols, and subjected to antigen retrieval in Tris-EDTA buffer (pH 9.0) at 100°C for 15 min using a pressure cooker. After blocking with 5% goat serum and 0.3% Triton X-100 in PBS for 1 h at room temperature, sections were incubated overnight at 4°C with primary anti-pERK antibody (1:100 in 1% BSA and 0.3% Triton X-100; Supplementary Table 1). Endogenous peroxidase activity was blocked using hydrogen peroxide/methanol (v/v) for 20 min. Detection was carried out with the Dako EnVision + Dual Link System-HRP (DAB) (K4061; Agilent Technologies, USA), followed by hematoxylin counterstaining and light microscopy analysis.

Anaphase bridge and micronuclei formation analysis

PCCL3 cells were cultured on glass slides, fixed in 3.7% formaldehyde at room temperature for 20 min, and incubated in a 2× SSC buffer for 20 min. Nuclei were stained with DAPI (1 μg/mL) for 3 min in the dark, rinsed, and mounted using ProLongTM Diamond Antifade Mountant (Invitrogen). Slides were examined using an Olympus fluorescence microscope (Olympus Corp., Japan), and images were acquired using Cell^M software. For each condition, approximately 100 cells were analyzed for micronuclei and anaphase bridges using Fiji (ImageJ).

Three-dimensional (3D) quantitative fluorescence in situ hybridization (Q-FISH)

3D Q-FISH was performed following established protocols (27, 28). Briefly, PCCL3 cells were grown on glass slides, fixed in 3.7% formaldehyde, treated with glycerol, and subjected to three freeze–thaw cycles in liquid nitrogen. Following equilibration in 0.1 M HCl, cells were prehybridized in 70% formamide/2× SSC solution, and then hybridized with telomere-specific probes at 30°C for 2 h (Agilent Technologies, USA). After washing, cells were counterstained with DAPI and mounted using Vectashield mounting medium (Vector Laboratories Inc., USA). Z-stack images, consisting of 80 optical sections at 0.2 μm intervals, were acquired for 40 nuclei per transfection using a Zeiss AxioImager Z2 microscope (Carl Zeiss Canada Ltd., Canada) equipped with a Zeiss AxioCam MRm Rev. 3 digital camera. Image acquisition and processing were performed using 4.8.2 and ZEN 2.3 software (Zeiss, Germany). Image deconvolution was carried out using a constrained iterative algorithm, as described by Schaefer (29). Telomere features were quantified using TeloView® v1.03 software (Telo Genomics Corp., Canada) (30). Parameters analyzed included telomere signal number, total and average telomere signal intensity, telomere aggregates (defined as clusters of telomeres separated by <200 nm, below the optical resolution limit), nuclear volume, the a/c ratio (reflecting the spatial distribution of telomeres within the nucleus in a cell cycle-dependent manner), and overall telomere organization.

Three-dimensional structured illumination microscopy (3D-SIM)

PCCL3 cells were cultured on glass slides and fixed with 3.7% formaldehyde at room temperature for 20 min. Following fixation, cells were washed in 1× PBS, permeabilized with 0.5% Triton X-100 for 10 min, and stained with DAPI (10 μg/mL) overnight at 37°C. Slides were then mounted with Vectashield (Vector Laboratories) and coverslipped. Imaging was performed using a Zeiss Elyra PS1 SIM with a 63×/1.40 NA oil immersion objective and an Andor EM-CCD iXon 885 camera. DAPI was excited using a 405 nm laser with an SR Cube 07 filter set. Z-stack images were acquired at 91 nm intervals and reconstructed using ZEN 2012 Black software (Carl Zeiss). A total of 40 nuclei per transfection were analyzed. Central z-planes were exported as TIFF files and processed in MATLAB. DNA structure and interchromatin spaces were quantified using a granulometry program using a granulometry-based program implementing a morphological sieve approach (31).

Statistical analysis

Normality and variance homogeneity were assessed using the Shapiro–Wilk and Levene’s tests, respectively. Parametric data were analyzed with Student’s t-test or one-way ANOVA, and nonparametric data with the Mann–Whitney or Kruskal–Wallis test (with Dunn’s post-hoc correction). Chi-square tests were applied to compare telomere intensity distribution based on quartiles. Differences in granulometry distributions were evaluated using the two-sided Kolmogorov–Smirnov test. Statistical significance was set at P < 0.05. Analyses were performed using GraphPad Prism (GraphPad Software, USA).

Results

AGK::BRAF activates MAPK/ERK signaling in PCCL3 cells

Western blot analysis demonstrated that AGK::BRAF expression was associated with the marked activation of the MAPK pathway, as evidenced by significantly increased ERK phosphorylation compared with ETV6::NTRK3, RET/PTC1, RET/PTC3, and BRAF V600E-expressing cells (P < 0.001; Fig. 1A and B). Consistent with enhanced downstream MAPK signaling, MEK phosphorylation was also increased in AGK::BRAF-expressing cells relative to the remaining groups (P < 0.01; Fig. 1A and B).

Figure 1.

Figure 1

Open in a new tab

Activation of MAPK and PI3K/AKT signaling pathways in PCCL3 thyroid cells expressing pediatric PTC oncogenes. PCCL3 rat thyroid follicular cells were transfected with pBudCE4.1 expression vectors encoding BRAF wild-type (WT), AGK::BRAF, ETV6::NTRK3, RET/PTC1, RET/PTC3, or BRAF V600E, and culture under growth factor-deprived conditions. (A) Representative western blot analysis showing phosphorylation levels of key components of the MAPK (p-MEK1/2, p-ERK1/2) and PI3K/AKT (p-AKT) pathways, as well as STAT3 activation (p-STAT3 Ser727). Total protein levels are shown as loading control. (B) Quantification of phosphorylated proteins normalized to total protein levels across three independent biological replicates. Data are presented as mean ± s.d. Statistical significance: *P < 0.05; **P < 0.01; ***P < 0.0001.

Phosphorylated STAT3 (Ser727) was detected across all oncogenic conditions; however, no significant differences were observed between AGK::BRAF and the other evaluated drivers (Fig. 1A and B). Similarly, although increased AKT phosphorylation was observed in AGK::BRAF- and BRAF V600E-expressing cells, pAKT levels were not significantly different between AGK::BRAF and the remaining oncogenic groups (Fig. 1A and B).

AGK::BRAF is associated with reduced NIS expression and iodine uptake

Analysis of NIS protein expression demonstrated reduced NIS levels in all oncogene-expressing PCCL3 models relative to parental PCCL3 cells (Fig. 2A and B). Among the evaluated oncogenic drivers, AGK::BRAF-expressing cells exhibited the most pronounced reduction in NIS expression, with significantly lower levels compared with BRAF WT-transfected cells (P < 0.05). Notably, BRAF WT-transfected cells also demonstrated partial reduction of NIS expression relative to parental PCCL3 cells, consistent with modest MAPK pathway activation observed in this experimental setting.

Figure 2.

Figure 2

Open in a new tab

NIS protein expression and iodide uptake in PCCL3 thyroid cells expressing pediatric PTC oncogenes. Normal PCCL3 thyroid cells and PCCL3 cells transfected with pBudCE4.1 constructs encoding BRAF wild-type (WT), AGK::BRAF, ETV6::NTRK3, RET/PTC1, RET/PTC3, or BRAF V600E were cultured under growth factor-deprived conditions. (A) Representative western blot showing sodium–iodide symporter (NIS) protein expression across transfected cell lines. (B) Densitometric quantification of NIS protein levels normalized to loading control across three independent biological replicates. (C) Functional iodide uptake assay (% uptake) in BRAF WT-, AGK::BRAF-, and BRAF V600E-expressing cells. Data are presented as mean ± s.d. from three independent experiments. Statistical significance was determined using one-way ANOVA with Bonferroni post-hoc test: *P < 0.05; **P < 0.01; ***P < 0.0001.

Iodide uptake assays demonstrated significantly reduced iodide uptake in AGK::BRAF- and BRAF V600E-expressing cells compared with BRAF WT-transfected cells (P < 0.05; Fig. 2C). AGK::BRAF-expressing cells exhibited a notable reduction in iodide uptake among the evaluated groups.

Increased pERK by immunohistochemistry in AGK::BRAF-positive PTCs

Immunohistochemical analysis of pediatric PTC tissue sections demonstrated strong nuclear and cytoplasmic pERK1/2 immunoreactivity in tumors harboring AGK::BRAF, whereas tumors driven by other oncogenic alterations exhibited only moderate pERK1/2 immunoreactivity. By contrast, normal thyroid tissue sections, used as an internal negative control, showed no detectable pERK1/2 expression (Fig. 3).

Figure 3.

Figure 3

Open in a new tab

Immunohistochemical detection of phosphorylated ERK (p-ERK) in pediatric PTC tumor samples. Formalin-fixed paraffin-embedded (FFPE) tissue sections from pediatric PTC patients were stained using an anti-phospho-ERK antibody (Phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204), Cell Signaling Technology, #9101). Tumor samples included cases harboring AGK::BRAF, RET/PTC1, RET/PTC3, ETV6::NTRK3, and BRAF V600E, as well as normal thyroid tissue. Representative staining patterns are shown. Scale bar: 10 μm.

Distinct oncogenic drivers differentially induce micronuclei and nuclear bud formation

Micronuclei formation was observed across all oncogene-expressing PCCL3 cells, with variable frequency. The highest levels were detected in ETV6::NTRK3-expressing cells (47%), followed by AGK::BRAF (24%), RET/PTC3 (18%), BRAF V600E (17%), and RET/PTC1 (14%) (Fig. 5). Notably, nuclear buds (NBUDs), small chromatin protrusions associated with chromosomal instability and defective DNA segregation, were predominantly observed in ETV6::NTRK3-, AGK::BRAF-, and RET/PTC1-expressing cells. By contrast, in the BRAF WT-expressing cells, the few cells that exhibited micronuclei, these were predominantly single micronuclei (Figs 4 and 5).

Figure 5.

Figure 5

Open in a new tab

Anaphase bridge formation in PCCL3 cells expressing thyroid cancer oncogenes. Anaphase bridge frequency was assessed in PCCL3 cells transfected with pBudCE4.1 constructs encoding BRAF WT, AGK::BRAF, ETV6::NTRK3, RET/PTC1, RET/PTC3, or BRAF V600E. The table summarizes the percentage of cells exhibiting anaphase bridges under each condition. Representative images show interphase nuclei connected by chromatin bridges (blue arrowheads), visualized in grayscale and as DAPI-stained nuclear counterstaining.

Figure 4.

Figure 4

Open in a new tab

Micronuclei formation in PCCL3 cells expressing thyroid cancer oncogenes. Micronucleus formation was evaluated in normal PCCL3 cells and PCCL3 cells expressing BRAF WT, AGK::BRAF, ETV6::NTRK3, RET/PTC1, RET/PTC3, or BRAF V600E. The table summarizes the proportion of cells exhibiting one, two, or three or more micronuclei in each condition. Representative images show interphase nuclei containing micronuclei (blue arrowheads), visualized in grayscale and as DAPI-stained nuclear counterstaining.

AGK::BRAF and ETV6::NTRK3 promote chromatin bridge formation

Chromatin bridges, a marker of mitotic stress and genomic instability, were observed most frequently in AGK::BRAF-expressing cells (20%), followed by ETV6::NTRK3 (18%), RET/PTC1 (18%), RET/PTC3 (8%), and BRAF V600E (1%) (Fig. 5). These findings indicate oncogene-specific differences in the induction of chromosomal segregation defects during mitosis.

3D Q-FISH reveals telomere instability in AGK::BRAF-expressing cells

PCCL3 cells expressing AGK::BRAF fusion exhibited significant alterations in telomere architecture, including a significantly lower number of telomere signals (P < 0.0001), fewer telomere aggregates (P ≤ 0.0001), and reduced overall telomere signal intensity (P < 0.0001) compared with BRAF WT and other oncogenic-expressing cells (Figs 6 and 7). In addition, AGK::BRAF-expressing cells showed increased nuclear volume and a higher a/c ratio (P < 0.05), indicative of altered three-dimensional nuclear organization.

Figure 6.

Figure 6

Open in a new tab

Three-dimensional Q-FISH analysis of telomere organization in PCCL3 cells expressing thyroid cancer oncogenes. Representative 3D-deconvolved images from quantitative fluorescence in situ hybridization (Q-FISH) in PCCL3 cells expressing BRAF WT, AGK::BRAF, ETV6::NTRK3, RET/PTC1, RET/PTC3, or BRAF V600E. Representative images are shown in the absence and presence of DAPI nuclear counterstaining. Telomere signals are shown in red, illustrating spatial telomere distribution across conditions.

Figure 7.

Figure 7

Open in a new tab

Quantitative analysis of telomere architecture in PCCL3 cells expressing thyroid cancer drivers. Quantitative analysis of 3D Q-FISH data showing (A) total number of telomere signals; (B) total number of telomeric aggregates; (C) average telomere signal intensity; (D) total telomere signal intensity; (E) a/c ratio (indicator of proliferative status), and (F) nuclear volume. (G) P-values from pairwise comparisons between AGK::BRAF and each experimental group/control. Data are expressed as mean ± s.d. Statistical significance: *P < 0.05, **P < 0.01, ***P < 0.0001.

3D-SIM and granulometry reveal oncogene-specific nuclear remodeling

Super-resolution imaging revealed pronounced chromatin reorganization in AGK::BRAF-expressing cells (Fig. 8). Compared with all other experimental groups, AGK::BRAF-expressing cells exhibited a more condensed and spatially reorganized chromatin architecture, characterized by enlarged interchromatin spaces (P < 0.0001), consistent with global nuclear remodeling associated with oncogenic signaling. Notably, RET/PTC1-expressing cells displayed even more pronounced architectural alterations than AGK::BRAF-expressing cells (P < 0.0001), suggesting that distinct fusion oncogenes differentially regulate higher-order chromatin organization. To quantitatively support these observations, granulometry-based analysis was applied as a mathematical approach to assess chromatin texture and spatial organization. This analysis confirmed oncogene-specific differences in chromatin compaction and spatial distribution, providing an objective and reproducible measure of nuclear architectural remodeling across experimental conditions (Fig. 8).

Figure 8.

Figure 8

Open in a new tab

Structured illumination microscopy (SIM) analysis of nuclear architecture in PCCL3 cells expressing thyroid oncogenes. SIM analysis was performed on PCCL3 cells transfected with pBudCE4.1 and constructs encoding BRAF WT, AGK::BRAF, ETV6::NTRK3, RET/PTC1, RET/PTC3, or BRAF V600E. The left panel shows granulometry-based quantification of chromatin organization and interchromatin space distribution across conditions. The right panel presents representative high-resolution SIM images illustrating chromatin condensation and interchromatin compartment organization associated with each genetic alteration.

Discussion

The AGK::BRAF fusion, initially characterized in NIH3T3 fibroblasts as an activator of MAPK signaling, remains incompletely characterized in thyroid follicular epithelial cells, the cell of origin of PTC, and its behavior relative to other oncogenic drivers remains incompletely defined (11). In this study, we addressed this gap by generating stable PCCL3 thyroid cell models expressing AGK::BRAF alongside clinically relevant pediatric PTC oncogenes, enabling systematic and parallel assessment of their effects on MAPK signaling, thyroid differentiation, and higher-order nuclear organization.

Our data demonstrate that AGK::BRAF induces robust and sustained activation of the MAPK/ERK pathway, as evidenced by significantly increased ERK and MEK phosphorylation compared with RET/PTC1, RET/PTC3, ETV6::NTRK3, and BRAF V600E. Importantly, this signaling signature was recapitulated in patient-derived pediatric PTC samples, where AGK::BRAF-positive tumors exhibited stronger pERK immunoreactivity than tumors driven by other MAPK-activating alterations. This concordance between experimental models and human tumors underscores the biological relevance of AGK::BRAF and supports its association with a distinct molecular phenotype in pediatric PTC.

Structurally, AGK::BRAF fusion retains the BRAF kinase domain while lacking the N-terminal RAS-binding regulatory region, a structural feature shared by other BRAF rearrangements (32, 33). Loss of this autoinhibitory domain is thought to favor constitutive MAPK pathway activation and reduce responsiveness to ERK-mediated negative feedback regulation, providing a mechanistic basis for the strong ERK activation observed (11, 32, 34, 35, 36, 37).

Beyond pathway activation, AGK::BRAF induces a coordinated program of thyroid dedifferentiation, genomic instability, and nuclear architectural remodeling. Across all oncogenic drivers, AGK::BRAF consistently produced the highest MAPK output coupled with the most pronounced repression of NIS expression and iodide uptake, reinforcing the role of sustained ERK signaling in suppressing thyroid differentiation programs (18). Notably, partial NIS repression was also observed in BRAF WT-expressing cells, suggesting that MAPK pathway dosage, rather than mutation identity alone, contributes to dedifferentiation in this system. This interpretation is further supported by previous reports demonstrating that sustained ERK activation and increased MAPK signaling intensity correlate with repression of thyroid differentiation-associated genes even beyond the context of canonical BRAF V600E mutation (13, 38, 39). However, differences in construct expression levels cannot be fully excluded.

Clinically, these observations are particularly relevant because impaired NIS expression compromises the efficacy of radioactive iodine (RAI) therapy. Previous studies have associated AGK::BRAF-positive tumors with multifocal disease, distant metastasis, and reduced responsiveness to RAI therapy (40). Although the two-dimensional (2D) PCCL3 model does not fully recapitulate the structural and functional complexity of thyroid follicular epithelium in vivo, particularly with respect to polarity, follicular organization, and iodide transport dynamics, the consistent reduced NIS expression and iodide uptake support a functional role for AGK::BRAF in promoting thyroid dedifferentiation.

Phosphorylation of STAT3 and AKT was also observed in AGK::BRAF-expressing cells. However, these alterations were less pronounced and more variable than those observed for the MAPK/ERK pathway. While STAT3 activation has previously been implicated in PTC pathogenesis and ERK-dependent phosphorylation of STAT3 at Ser727 has been described downstream of receptor tyrosine kinase signaling (41, 42, 43), our results do not support a unique or predominant role for STAT3 or AKT activation in AGK::BRAF-driven cells relative to the other oncogenic drivers evaluated. Instead, our data indicate that MAPK/ERK signaling represents the dominant pathway associated with the biological effects induced by AGK::BRAF in this model.

Beyond dedifferentiation, our findings indicate that AGK::BRAF expression is associated with marked genomic instability. AGK::BRAF- and ETV6::NTRK3-expressing cells displayed increased frequencies of chromatin bridges, micronuclei, and nuclear buds, abnormalities commonly associated with replication stress, telomere dysfunction, and breakage–fusion–bridge cycles (44, 45, 46, 47). Low-level abnormalities observed in BRAF WT-expressing cells likely reflect baseline stress associated with stable ectopic expression rather than mutation-specific effects (48).

To investigate the mechanisms underlying this genomic instability phenotype, we analyzed telomere organization and chromatin architecture using 3D Q-FISH and 3D-SIM imaging (30, 31, 49, 50, 51). AGK::BRAF-expressing cells exhibited marked telomere remodeling, characterized by altered telomere number, reduced signal intensity, and changes in telomere spatial organization within the nucleus. These alterations were accompanied by increased nuclear volume and elevated a/c ratio, features associated with chromatin decompaction and proliferative states.

Super-resolution imaging further revealed a composite chromatin phenotype characterized by increased compaction coexisting with expanded interchromatin spaces, suggesting heterogeneous and spatially segregated chromatin remodeling states. This pattern likely reflects mixed chromatin remodeling, in which condensed chromatin regions coexist with expanded DNA-poor areas interpreted as interchromatin spaces or potentially enlarged nucleolar compartments (52). A limitation of this analysis is the inability to distinguish interchromatin spaces from nucleoli, since nucleolar markers were not included.

Granulometry-based quantitative analysis confirmed these architectural alterations and revealed distinct chromatin reorganization signatures induced by AGK::BRAF and RET/PTC1. Notably, although RET/PTC1 triggered pronounced chromatin remodeling, it was not associated with the highest levels of genomic instability, supporting the concept that chromatin architectural remodeling and mitotic instability represent partially uncoupled outputs of oncogenic signaling rather than directly linked processes.

These findings support a model in which oncogenic signaling drives telomere disorganization and nuclear architectural remodeling that may predispose cells to genomic instability. In particular, disrupted telomere clustering and reduced telomeric signal integrity may contribute to telomere fragility and increased susceptibility to chromosome end-to-end fusions, thereby facilitating breakage–fusion–bridge cycles. Importantly, these changes likely reflect higher-order telomere architectural reprogramming rather than alterations in telomere length per se.

Notably, these observations differ from prior analyses in primary tumors, where AGK::BRAF-positive regions exhibited an increased number of telomere signals, higher total telomere intensity, and increased telomere aggregates (25). This discrepancy likely reflects distinct stages of telomere remodeling. Early tumorigenesis may be characterized by increased detectable telomeric signals and aggregation, whereas sustained oncogenic signaling in vitro may drive progressive telomere reorganization, leading to signal redistribution, clustering, and structural attenuation. Together, these findings support a dynamic, stage-dependent model of telomere architectural evolution culminating in genomic instability.

Importantly, our data support a model in which oncogenic driver identity differentially coordinates nuclear architecture and genomic instability, revealing that chromatin remodeling and mitotic instability represent distinct downstream outputs of oncogenic signaling in thyroid epithelial cells.

From a translational perspective, these results suggest that targeting MAPK-driven nuclear reprogramming may represent a complementary therapeutic avenue in AGK::BRAF-associated disease. Future studies using patient-derived and in vivo models will be essential to validate these mechanisms and define actionable vulnerabilities.

Supplementary materials

Declaration of interest

SM is a co-founder, shareholder, and the chair of the clinical and scientific advisory committee of Telo Genomics Corp. (Canada). The remaining authors declare no conflicts of interest that could compromise the impartiality of the research presented.

Funding

This study was funded by research grants from the São Paulo State Research Foundation (FAPESP, grant numbers 2014/06570-6 and 2022/9713-9) and CNPq (406952/2022-1). LS, DMDT, and WGR received a fellowship grant from FAPESP (grant numbers 2018/09911-0, 2020/06594-3, 2022/16827-0), while JMC was awarded a Research Productivity Scholarship from CNPq (grant number 305690/2022-1). We also extend our gratitude to CAPES for their financial support.

Author contribution statement

LS, AR-P, SM, and JMC contributed to the study conception and design. Data curation and formal analysis were performed by LS. Investigation was performed by LS, WGR, DMDT, RH, RECS, GH, CS-N, MF, and AR-P. Visualization was performed by LS, WGR, and DMDT. Methodology was designed by LS, SM, AR-P, and JMC. Resources, funding acquisition, and supervision were provided by SM and JMC. Project administration was carried out by LS and JMC. The first draft of the manuscript was written by LS, and DMDT and JMC contributed to the review and editing of the manuscript. All authors read and approved the final manuscript.

Acknowledgments

We thank Telo Genomics Corp. for providing access to TeloView® and Mary Cheang for her expert assistance with statistical analyses. We are also grateful to the Genomic Centre for Cancer Research and Diagnosis (GCCRD) for imaging support. The GCCRD is funded by the Canada Foundation for Innovation and is supported by the CancerCare Manitoba Foundation, the University of Manitoba, and the Canada Research Chair Tier 1 (S.M.). The GCCRD is a proud member of the Canadian National Scientific Platforms (CNSP) and Canada BioImaging.

References

  • 1.Hundahl S, Fleming I, Fremgen A, et al. A National Cancer Data Base report on 53,856 cases of thyroid carcinoma treated in the U.S., 1985-1995. Cancer 1998. 83 2638–2648. ( 10.1002/(sici)1097-0142(19981215)83:12<2638::aid-cncr31>3.0.co;2-1) [DOI] [PubMed] [Google Scholar]
  • 2.Busnardo B & Vido DD. The epidemiology and etiology of differentiated thyroid carcinoma. Biomed Pharmacother 2000. 54 322–326. ( 10.1016/s0753-3322(00)80056-6) [DOI] [PubMed] [Google Scholar]
  • 3.Sherma SI. Thyroid carcinoma. Lancet 2003. 361 501–511. ( 10.1016/s0140-6736(03)12488-9) [DOI] [PubMed] [Google Scholar]
  • 4.Leboulleux S, Baudin E, Hartl DW, et al. Follicular-cell derived thyroid cancer in children. Eur J Cancer 2004. 40 1655–1659. ( 10.1016/j.ejca.2004.02.009) [DOI] [PubMed] [Google Scholar]
  • 5.Cordioli MI, Moraes L, Bastos AU, et al. Fusion oncogenes are the main genetic events found in sporadic papillary thyroid carcinomas from children. Thyroid 2017. 27 182–188. ( 10.1089/thy.2016.0387) [DOI] [PubMed] [Google Scholar]
  • 6.De Sousa MSA, Nunes IN, Christiano YP, et al. Genetic alterations landscape in paediatric thyroid tumours and/or differentiated thyroid cancer: systematic review. Rev Endocr Metab Disord 2024. 25 35–51. ( 10.1007/s11154-023-09840-2) [DOI] [PubMed] [Google Scholar]
  • 7.Rangel-Pozzo A, Sisdelli L, Cordioli MIV, et al. Genetic landscape of papillary thyroid carcinoma and nuclear architecture: an overview comparing pediatric and adult populations. Cancers 2020. 12 3146. ( 10.3390/cancers12113146) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ricarte-Filho JC & Franco AT. The evolving genomic landscape of pediatric papillary thyroid cancer. Curr Opin Endocr Metab Res 2023. 33 100483. ( 10.1016/j.coemr.2023.100483) [DOI] [Google Scholar]
  • 9.Yang AT, Lai S, Laetsch TW, et al. Molecular landscape and therapeutic strategies in pediatric differentiated thyroid carcinoma. Endocr Rev 2025. 46 397–417. ( 10.1210/endrev/bnaf003) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Zhao X, Kotch C, Fox E, et al. NTRK fusions identified in pediatric tumors: the frequency, fusion partners, and clinical outcome. JCO Precis Oncol 2021. 5 PO.20.00250. ( 10.1200/PO.20.00250) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Ricarte-Filho JC, Li S, Garcia-Rendueles MER, et al. Identification of kinase fusion oncogenes in post-Chernobyl radiation-induced thyroid cancers. J Clin Investig 2013. 123 4935–4944. ( 10.1172/JCI69766) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Ouyang B, Knauf JA, Smith EP, et al. Inhibitors of Raf Kinase activity block growth of thyroid cancer cells with RET/PTC or BRAF mutations in vitro and in vivo. Clin Cancer Res 2006. 12 1785–1793. ( 10.1158/1078-0432.CCR-05-1729) [DOI] [PubMed] [Google Scholar]
  • 13.Zhang Z, Liu D, Murugan AK, et al. Histone deacetylation of NIS promoter underlies BRAF V600E-promoted NIS silencing in thyroid cancer. Endocr Relat Cancer 2014. 21 161–173. ( 10.1530/ERC-13-0399) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Nagarajah J, Le M, Knauf JA, et al. Sustained ERK inhibition maximizes responses of BrafV600E thyroid cancers to radioiodine. J Clin Investig 2016. 126 4119–4124. ( 10.1172/JCI89067) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Mitsutake N, Knauf JA, Mitsutake S, et al. Conditional BRAFV600E expression induces DNA synthesis, apoptosis, dedifferentiation, and chromosomal instability in thyroid PCCL3 cells. Cancer Res 2005. 65 2465–2473. ( 10.1158/0008-5472.CAN-04-3314) [DOI] [PubMed] [Google Scholar]
  • 16.Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in human cancer. Nature 2002. 417 949–954. ( 10.1038/nature00766) [DOI] [PubMed] [Google Scholar]
  • 17.Agrawal N, Akbani R, Aksoy BA, et al. Integrated genomic characterization of papillary thyroid carcinoma. Cell 2014. 159 676–690. ( 10.1016/j.cell.2014.09.050) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fagin JA, Krishnamoorthy GP & Landa I. Pathogenesis of cancers derived from thyroid follicular cells. Nat Rev Cancer 2023. 23 631–650. ( 10.1038/s41568-023-00598-y) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Cordioli MICV, Moraes L, Alves MTDS, et al. Thyroid-specific genes expression uncovered age-related differences in pediatric thyroid carcinomas. Int J Endocrinol 2016. 2016 1–9. ( 10.1155/2016/1956740) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lee YA, Lee H, Im SW, et al. NTRK and RET fusion–directed therapy in pediatric thyroid cancer yields a tumor response and radioiodine uptake. J Clin Investig 2021. 131 e144847. ( 10.1172/JCI144847) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Franco AT, Ricarte-Filho JC, Isaza A, et al. Fusion oncogenes are associated with increased metastatic capacity and persistent disease in pediatric thyroid cancers. J Clin Oncol 2022. 40 1081–1090. ( 10.1200/JCO.21.01861) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Wasserman JD, Isaza AR, Waguespack SG, et al. The landscape of pediatric differentiated thyroid carcinoma and predictors of invasive disease: a CATC study. J Clin Endocrinol Metab 2026. 22 dgag172. ( 10.1210/clinem/dgag172) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Ju G, Sun Y, Wang H, et al. Fusion oncogenes in patients with locally advanced or distant metastatic differentiated thyroid cancer. J Clin Endocrinol Metab 2024. 109 505–515. ( 10.1210/clinem/dgad500) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Pekova BB, Sykorova V, Mastnikova K, et al. RET fusion genes in pediatric and adult thyroid carcinomas: cohort characteristics and prognosis. Endocr Relat Cancer 2023. 30 e230117. ( 10.1530/ERC-23-0117) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Sisdelli L, Cordioli MIV, Vaisman F, et al. A multifocal pediatric papillary thyroid carcinoma (PTC) Harboring the AGK-BRAF and RET/PTC3 fusion in a mutually exclusive pattern reveals distinct levels of genomic instability and nuclear organization. Biology 2021. 10 125. ( 10.3390/biology10020125) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Waltz F, Pillette L & Ambroise Y. A nonradioactive iodide uptake assay for sodium iodide symporter function. Anal Biochem 2010. 396 91–95. ( 10.1016/j.ab.2009.08.038) [DOI] [PubMed] [Google Scholar]
  • 27.Caria P, Dettori T, Frau DV, et al. Characterizing the three-dimensional organization of telomeres in papillary thyroid carcinoma cells. J Cell Physiol 2019. 234 5175–5185. ( 10.1002/jcp.27321) [DOI] [PubMed] [Google Scholar]
  • 28.Gadji M, Adebayo Awe J, Rodrigues P, et al. Profiling three-dimensional nuclear telomeric architecture of myelodysplastic syndromes and acute myeloid leukemia defines patient subgroups. Clin Cancer Res 2012. 18 3293–3304. ( 10.1158/1078-0432.CCR-12-0087) [DOI] [PubMed] [Google Scholar]
  • 29.Schaefer LH, Schuster D & Herz H. Generalized approach for accelerated maximum likelihood based image restoration applied to three-dimensional fluorescence microscopy. J Microsc 2001. 204 99–107. ( 10.1046/j.1365-2818.2001.00949.x) [DOI] [PubMed] [Google Scholar]
  • 30.Vermolen BJ, Garini Y, Mai S, et al. Characterizing the three-dimensional organization of telomeres. Cytometry A 2005. 67A 144–150. ( 10.1002/cyto.a.20159) [DOI] [PubMed] [Google Scholar]
  • 31.Righolt CH, Guffei A, Knecht H, et al. Differences in nuclear DNA organization between lymphocytes, Hodgkin and Reed-Sternberg cells revealed by structured illumination microscopy. J Cell Biochem 2014. 115 1441–1448. ( 10.1002/jcb.24800) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Ciampi R, Knauf J, Kerler R, et al. Oncogenic AKAP9-BRAF fusion is a novel mechanism of MAPK pathway activation in thyroid cancer. J Clin Investig 2005. 115 94–101. ( 10.1172/jci23237) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Vojnic M, Kubota D, Kurzatkowski C, et al. Acquired BRAF rearrangements induce secondary resistance to EGFR therapy in EGFR-mutated lung cancers. J Thorac Oncol 2019. 14 802–815. ( 10.1016/j.jtho.2018.12.038) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Fagin JA & Wells SA. Biologic and clinical perspectives on thyroid cancer. N Engl J Med 2016. 375 1054–1067. ( 10.1056/NEJMra1501993) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Poulikakos PI, Zhang C, Bollag G, et al. RAF inhibitors transactivate RAF dimers and ERK signaling in cells with wild-type BRAF. Nature 2010. 464 427–430. ( 10.1038/nature08902) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Ross JS, Wang K, Chmielecki J, et al. The distribution of BRAF gene fusions in solid tumors and response to targeted therapy. Int J Cancer 2015. 138 881–890. ( 10.1002/ijc.29825) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Tran NH, Wu X & Frost JA. B-Raf and Raf-1 are regulated by distinct autoregulatory mechanisms. J Biol Chem 2005. 280 16244–16253. ( 10.1074/jbc.M501185200) [DOI] [PubMed] [Google Scholar]
  • 38.Xie G, Lu J, Song A, et al. 2, 2′, 4, 4′-tetrabromodiphenyl ether induces thyroid toxicity by targeting BRAF-mediated MEK-ERK-NIS pathway in human thyroid follicular epithelial cells. Ecotoxicol Environ Saf 2025. 302 118577. ( 10.1016/j.ecoenv.2025.118577) [DOI] [PubMed] [Google Scholar]
  • 39.Kaushik N, Kim MJ, Kaushik NK, et al. Low dose radiation regulates BRAF-induced thyroid cellular dysfunction and transformation. Cell Commun Signal 2019. 17 12. ( 10.1186/s12964-019-0322-x) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Sisdelli L, Cordioli MI, Vaisman F, et al. AGK-BRAF is associated with distant metastasis and younger age in pediatric papillary thyroid carcinoma. Pediatr Blood Cancer 2019. 66 e27707. ( 10.1002/pbc.27707) [DOI] [PubMed] [Google Scholar]
  • 41.Liang YN, Zhang Z, Song J, et al. Role of STAT3 expression in thyroid cancer: a meta-analysis and systematic review based on the Chinese population. Evid Based Complement Alternat Med 2022. 2022 1116535. ( 10.1155/2022/1116535) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Trovato M, Grosso M, Vitarelli E, et al. Distinctive expression of STAT3 in papillary thyroid carcinomas and a subset of follicular adenomas. Histol Histopathol 2003. 18 393–399. ( 10.14670/HH-18.393) [DOI] [PubMed] [Google Scholar]
  • 43.Zhang J, Gill A, Atmore B, et al. Upregulation of the signal transducers and activators of transcription 3 (STAT3) pathway in lymphatic metastases of papillary thyroid cancer. Int J Clin Exp Pathol 2011. 4 356–362 [PMC free article] [PubMed] [Google Scholar]
  • 44.Agustinus AS & Bakhoum S. Biochemical, genomic, and epigenomic profiling of isolated cancer cell lines’ micronuclei. Methods Cell Biol 2022. 172 51–66. ( 10.1016/bs.mcb.2022.07.001) [DOI] [PubMed] [Google Scholar]
  • 45.Krupina K, Goginashvili A & Cleveland DW. Causes and consequences of micronuclei. Curr Opin Cell Biol 2021. 70 91–99. ( 10.1016/j.ceb.2021.01.004) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Shimizu N, Shingaki K, Kaneko-Sasaguri Y, et al. When, where and how the bridge breaks: anaphase bridge breakage plays a crucial role in gene amplification and HSR generation. Exp Cell Res 2005. 302 233–243. ( 10.1016/j.yexcr.2004.09.001) [DOI] [PubMed] [Google Scholar]
  • 47.Umbreit NT, Zhang CZ, Lynch LD, et al. Mechanisms generating cancer genome complexity from a single cell division error. Science 2020. 368 eaba0712. ( 10.1126/SCIENCE.ABA0712) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Li J, Settivari RS & LeBaron MJ. Genetic instability of in vitro cell lines: implications for genetic toxicity testing. Environ Mol Mutagen 2019. 60 559–562. ( 10.1002/em.22280) [DOI] [PubMed] [Google Scholar]
  • 49.Rangel-Pozzo A, de Souza DC, Schmid-Braz AT, et al. 3D telomere structure analysis to DetectGenomic instability and cytogenetic evolutionin myelodysplastic syndromes. Cells 2019. 8 304. ( 10.3390/cells8040304) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Sathitruangsak C, Righolt CH, Klewes L, et al. Quantitative superresolution microscopy reveals differences in nuclear dna organization of multiple myeloma and monoclonal gammopathy of undetermined significance. J Cell Biochem 2015. 116 704–710. ( 10.1002/jcb.25030) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Righolt CH, Knecht H & Mai S. DNA superresolution structure of reed-sternberg cells differs between long-lasting remission versus relapsing hodgkin’s lymphoma patients. J Cell Biochem 2016. 117 1633–1637. ( 10.1002/jcb.25456) [DOI] [PubMed] [Google Scholar]
  • 52.Dolde CE, Mukherjee M, Cho C, et al. HMG-I/Y in human breast cancer cell lines. Breast Cancer Res Treat 2002. 71 181–191. ( 10.1023/A:1014444114804) [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

EO-25-0096_supplementary_table_1.pdf (197.7KB, pdf)

Articles from Endocrine Oncology are provided here courtesy of Bioscientifica Ltd.

RESOURCES