Introduction
Thyroid cancer, a cancerous tumour or growth residing within the thyroid gland, is broadly perceived as a prevalent endocrine disorder that can be categorised into papillary thyroid carcinoma (PTC), follicular carcinoma, anaplastic thyroid carcinoma (ATC), and medullary carcinoma [1]. Despite the rarity of ATC, which only makes up around 1–2% of all thyroid carcinoma cases, it is accountable for approximately one-third of all thyroid cancer deaths, positioning it as the deadliest form of thyroid cancer [2]. The survival time of patients cannot be effectively prolonged despite the development of a multimodal intervention pattern for ATC treatment, because it is generally marked by rapid growth and invasion-enabling capacities and is susceptible to therapy resistance [3, 4]. Hence, the revelation of mechanism in respect to the aggressive nature of ATC is imperative to identify biological targets for the advent of alternative therapeutic approaches.
Diaphanous-related formin 3 (DIAPH3) belonging to the diaphanous-related formin subfamily is implicated in actin and microtubule networks, hence participating in fundamental cellular processes such as division, adhesion, motility, intracellular trafficking, and polarity [5, 6]. To date, in terms of human malignancies, the exact role of DIAPH3 seems to be somewhat contradictory, varying among tumour types. For instance, DIAPH3 contributes to the carcinogenic events in the process of pancreatic cancer [7], cervical cancer [8], lung adenocarcinoma [9], and osteosarcoma [10]. On the other hand, previous research has highlighted the tumour-suppressing effects of DIAPH3 in colorectal cancer [11]. Nonetheless, the biological impact of DIAPH3 on ATC has not been well recognised.
The forkhead transcription factor superfamily, sharing a highly conserved (winged helix) deoxyribonucleic acid (DNA)-binding domain, has been extensively reported to mediate oncogenic processes in a plethora of solid cancers [12, 13]. The existing literature has established that the forkhead family member, forkhead box M1 (FOXM1), primarily functions as an oncogene in thyroid carcinoma [12, 14]. Based on FpClass (http://dcv.uhnres.utoronto.ca/FPCLASS/ppis/), FOXM1 was introduced as an interacting partner of DIAPH3.
In this context, this research aimed to reveal the precise functions of DIAPH3 on the formation and advancement of ATC and its link with FOXM1.
Material and methods
Data source
The Gene Expression Omnibus (GEO) database was searched for differential DIAPH3 expression in ATC samples and noncancerous samples.
Cell preparation, culture conditions, and transient transduction
Human thyroid follicular cells (Nthy-ori3-1) from IMMOCELL (Xiamen, China) and human ATC cell lines (THJ-16T and SW1736 cells) from CRISPRBIO (Beijing, China) were amplified in RPMI-1640 (Genom, Hangzhou, China) carrying 10% foetal bovine serum (FBS) (AusGeneX, Australia).
THJ-16T and SW1736 cells were infected with the small interfering ribonucleic acid (siRNA) constructs of DIAPH3 (siRNA-DIAPH3-1/2) and overexpression constructs of FOXM1 (Ov-FOXM1) built by Shanghai GeneChem Co. Ltd., along with their matching control (siRNA-NC and Ov-NC), following the protocol outlined in Genejuice Transfection Reagent (Novagen).
Cell counting kit 8 (CCK-8)
The 96-well plate (3 × 104 cells/well) were implanted with transfected THJ-16T and SW1736 cells, to each well of which 10 μL of cell counting kit 8 (CCK-8) solvent (Proteintech) was supplied and left to stand for 2 h. Upon termination of the reaction, the absorption value was checked under a multimodal reader plate at 450 nm.
5-ethynyl-2’-deoxyuridine (EdU) incorporation
After being covered by 10 mM 5-ethynyl-2’-deoxyuridine (EdU) for 2 h with reference to the protocol of Cell-Light EdU Apollo 488 or 567 in vitro Imaging Kit (RiboBio, Guangzhou, China), the transfected THJ-16T and SW1736 cells were solidified with 4% paraformaldehyde for 15 min and perforated with 0.5% Triton X-100 for 10 min. After being processed with 1 × Apollo reaction cocktail for half an hour and dyed by 5 mg/mL DAPI for 10 min, the EdU-positive cells were identified under a fluorescence microscope.
Scratch assay
The transfected THJ-16T and SW1736 cells were evenly put into 6-well plates (5 x 105 cells/well). A plastic tip (1 mm) was employed to run across the cell layer. The area within the scratch was quantitively recorded under a light microscope at designated time intervals.
Matrigel invasion assay
Cellular invasion ability was appraised in Transwell assay with 8-µm pore chambers with Matrigel coating. The suspension of transfected THJ-16T and SW1736 cells (2 × 104) prepared with serum-deprived medium and 500 μL medium nourished with 10% FBS were severally transferred to the upper and lower cavity. Non-transferred cells were detached with cotton swabs. Transferred cells in the lower cavity were solidified with 100% methanol and dyed by 0.1% crystal violet, prior to being quantitatively calculated under a light microscope.
Terminal-deoxynucleotidyl transferase mediated nick end labelling (TUNEL)
After being evenly distributed into 24-well plates, a total of 2 × 105 transfected THJ-16T and SW1736 cells were exposed to 4% paraformaldehyde for 25 min and 0.2% Triton X-100 for 5 min. Subsequently, the samples labelled by 100 µL 1 × equilibration buffer for 10 min reacted with 50 µL TdT reaction mix (Vazyme, Nanjing, China) for one hour. The positive TUNEL signal was monitored under a fluorescence microscope after being dyed by 2 μg/mL DAPI.
Co-immunoprecipitation (Co-IP) and western blotting
Co-IP assay was conducted adhering to the guidelines of Co-IP kit (Active Motif, Inc., Carlsbad, CA, USA). After the homogenates prepared from THJ-16T and SW1736 cells with RIPA buffer (MedChemExpress, USA) were centrifuged, the cellular lysates were treated with DIAPH3 (cat. no. ab245659; Abcam) and FOXM1 (cat. no. ab245309; Abcam) antibodies or the matching secondary antibodies (cat. no. ab205718; Abcam). The bound immunocomplex on the beads passed through sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) prior to protein immunoblotting.
Following the rupture of THJ-16T and SW1736 cells in RIPA buffer (MedChemExpress, USA), the protein concentration was quantitively evaluated in terms of the Bradford method (MedChemExpress, USA). The equivalent amounts of proteins passed through SDS-PAGE were then loaded onto a nitrocellulose membrane. After being sealed in 5% albumin from bovine serum, the membranes were treated by primary antibodies (Abcam) against DIAPH3 (cat. no. ab245659; 1/2000), FOXM1 (cat. no. ab245309; 1/1000), BCL2 (cat. no. ab194583; 1/2000), Bax (cat. no. ab53154; 1/1000), cleaved caspase3 (cat. no. ab2302; 1/500), b-catenin (cat. no. ab224803; 1/400), and c-Myc (cat. no. ab32072; 1/1000) overnight at 4°C and the matching secondary antibodies linked to HRP (cat. no. ab97051; 1/2000; Abcam) for 1 h. After chemiluminescence reaction in the Odyssey CLx Western blot detection system (Westburg, Leusden, Netherlands), each band was quantitated with ImageJ software (NIH, Bethesda, MD, USA). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was utilised to ensure equivalent protein loading.
Statistics
The data subjected to processing with SPSS 19.0 statistical software are represented as means and standard deviation (SD). The inter-group comparison was conducted by one-way ANOVA with Tukey’s test. The significance was set as a probability level of 0.05.
Results
DIAPH3 possessed aberrantly elevated expression in ATC tissues and cells
To shed light on the role of DIPAH3 in ATC, DIAPH3 expression was initially determined based on GEO database in ATC tissues. As analysed, DIAPH3 was strongly hyper-expressed in ATC tissues relative to noncancerous thyroid tissues (Fig. 1A). Through Western blotting analysis, when compared with human thyroid follicular cells (Nthy-ori3-1), DIAPH3 expression turned out to be greatly raised in human ATC cell lines (THJ-16T and SW1736 cells) (Fig. 1B).
Deletion of DIAPH3 impaired the potentials for THJ-16T and SW1736 cells to proliferate, migrate, and invade
To validate the assumption that dysregulated DIAPH3 is associated with the aggressive behaviours of ATC cells, DIAPH3 expression was robustly eliminated in THJ-16T and SW1736 cells after being infected with siRNA-DIAPH3-1/2 because DIAPH3 upregulation was observed in ATC cells. Moreover, siRNA-DIAPH3-1 resulted in a more potent reduction on DIAPH3 expression, hence being utilised in the follow-up assays (Fig. 2A). The experimental data based on CCK-8 method expounded that SW1736 and THJ-16T cell viability were both regressed at 48 h and 72 h in response to DIAPH3 interference (Fig. 2B and D). It was also observable from EdU incorporation assay that the number of proliferating SW1736 and THJ-16T cells remarkably descended when DIAPH3 was depleted (Fig. 2C and E). Simultaneously, as manifested by scratch assay, the migrated rate of SW1736 and THJ-16T cells prominently slowed down after DIAPH3 was silenced (Fig. 2F and H). Through Matrigel invasion assay, DIAPH3 deletion was seen to lower the invasive ability of SW1736 and THJ-16T cells (Fig. 2G and I).
Deletion of DIAPH3 improved the apoptotic rate of THJ-16T and SW1736 cells
As shown in Figure 3A and B, the results from TUNEL assay showed that that upon knockdown of DIAPH3, the production of apoptotic SW1736 and THJ-16T cells was evidently facilitated. Expectedly, Western blotting analysis also displayed the declined BCL2 expression and the forced Bax, cleaved caspase3 expression in DIAPH3-depleting SW1736 and THJ-16T cells (Fig. 3C and D).
DIAPH3 was positively correlated with its binding partner FOXM1
After being tested with Western blotting, FOXM1 expression was also discovered to be distinctly increased in SW1736 and THJ-16T cells relative to human thyroid follicular cells (Fig. 4A). Furthermore, in both SW1736 and THJ-16T cells infected with siRNA-DIAPH3-1, FOXM1 expression was strongly eliminated (Fig. 4B). Through Co-IP assay, it was noticed that DIAPH3 efficiently co-immunoprecipitated with FOXM1 in both SW1736 and THJ-16T cells (Fig. 4C and D).
DIAPH3 expedited the proliferating, migrating, and invading rate of THJ-16T and SW1736 cells dependent on interaction with FOXM1
Prior to corroborating the implication of FOXM1 in the functional mechanism of DIAPH3 in ATC cells, FOXM1 expression was substantially elevated by transduction of Ov-FOXM1 (Fig. 5A). Significantly, at designated time points, the diminished viability of both SW1736 and THJ-16T cells imposed by downregulation of DIAPH3 were both facilitated again when FOXM1 was upregulated (Fig. 5B and D). The same trend could also be seen in EdU incorporation assay, as manifested by the strengthened proliferative capability on account of DIAPH3 knockdown and FOXM1 overexpression that was declined in DIAPH3-silencing SW1736 and THJ-16T cells (Fig. 5C and E). The results of scratch assay reflected that DIAPH3 loss resulted in a dramatic decrease in the number of migrated SW1736 and THJ-16T cells, which partially increased again after FOXM1 was overexpressed (Fig. 5F and H). Likewise, the competence to invade was impaired by DIAPH3 insufficiency and then was reinforced in both DIAPH3-depleting and FOXM1-overexpressing SW1736 and THJ-16T cells (Fig. 5G and I).
DIAPH3 repressed the apoptotic capability of THJ-16T and SW1736 cells dependent on interaction with FOXM1
Inversely, it turned out from TUNEL assay that the accumulated apoptotic SW1736 and THJ-16T cells attributed to the absence of DIAPH3 was lowered when FOXM1 was highly expressed (Fig. 6A and B). In the meantime, BCL2 expression was increased while Bax and cleaved caspase3 expression were both suppressed in the siRNA-DIAPH3-1+Ov-FOXM1 group in contrast to the siRNA-DIAPH3-1+Ov-NC group (Fig. 6C and D).
DIAPH3 acted as an activator of Wnt/b-catenin signalling dependent on interaction with FOXM1 in ATC cells
Surprisingly, through Western blotting analysis, downregulation of DIAPH3 repressed Wnt/b-catenin signalling-associated b-catenin and c-Myc expression were both raised again by concurrent elevation of FOXM1 (Fig. 7A and B), underlining that DIAPH3 might bind with FOXM1 to drive Wnt/b-catenin signalling in ATC cells.
Discussion
ATC is a particularly deadly malignancy with an aggressive nature. It is now becoming apparent that the molecular change seems to be a common event during the process of ATC, and therefore multiple prospective therapeutic targets for ATC have recently been ascertained [15–17]. Recent pan-cancer analysis has revealed the elevated DIAPH3 expression in a diverse range of cancer tissues [18]. Our work for the first time profiled that both ATC tissues and cells embraced highly expressed DIAPH3 expression, and knocking down DIAPH3 could efficiently repress the proliferative, migratory, and invasive but heighten the apoptotic abilities of ATC cells. Mechanistically, DIAPH3 interacted with FOXM1 and mediated Wnt/b-catenin signalling in ATC cells.
To examine the in vitro biological roles of DIAPH3 in ATC, cell lines with stably DIAPH3-depelting SW1736 and THJ-16T cells were established. In general, unlimited proliferation is an essential biological hallmark of malignant tumour cells, accompanied by the blockade of apoptosis [19]. As a major cytoskeletal regulator, DIAPH3 is involved in biological mechanisms – cell proliferation and apoptosis included [6]. For instance, deficiency of DIAPH3 boosts the proliferation of colorectal cancer cells [11]. In most cases, DIAPH3 silencing inversely restrains the proliferation of cervical cancer, pancreatic cancer, and osteosarcoma cells [7, 8, 10]. During our experiments, it was found that interfering with DIAPH3 strongly impaired SW1736 and THJ-16T cell viability and proliferation. Furthermore, the promoting role of DIAPH3 loss in ATC cell apoptosis was proven, as shown by the elevated apoptotic rate, the lowered BCL2 expression, and the elevated Bax, cleaved caspase3 expression. These findings first highlighted that DIAPH3 might be linked to the formation of ATC.
At the same time, dysregulated DIAPH3 alters the cellular migratory and invasive characteristics of cancer cells, which are mainly responsible for the metastatic property of tumours. Emerging studies have favoured that DIAPH3 contributes to the migration of hepatocellular carcinoma cells and the migration and invasion of osteosarcoma cells [10, 20]. Intriguingly, Huang et al. have underlined that depletion of DIAPH3 produced completely contrary effects on colorectal cancer migration [11]. Jiang et al. have predicted that DIAPH3 upregulation hampers breast cancer cell migration and invasion [21]. Specifically, in SW1736 and THJ-16T cells, downregulation of DIAPH3 significantly attenuated the capacities to migrate and invade, which was in accordant with previous investigations on hepatocellular carcinoma and osteosarcoma.
Surprisingly, FOXM1 was predicated by FpClass as an interacting partner of DIAPH3. More importantly, DIAPH3 has been reported to bind HSP90 to activate the beta-catenin/TCF signalling in hepatocellular carcinoma cells [20]. Herein, DIAPH3 was noted to efficiently co-immunoprecipitate with FOXM1, which could be positively modulated by DIAPH3 in ATC cells. FOXM1 is a representative proliferation-related transcription factor that has been supported to be potently positive in human malignancies and associated with carcinogenesis and tumour metastasis [22, 23]. As regards the role of FOXM1 in ATC, FOXM1 is overexpressed in ATC tissues and cells, and FOXM1 potentiates the ATC-specific malignant signature [24, 25]. Consistently, FOXM1 expression was raised in ATC cells, and further FOXM1 elevation partly weakened the repressive effects of DIAPH3 deficiency on the potentials for ATC cells to proliferate, migrate, and invade, as well as promoting the role on the apoptotic capability.
Former research has unveiled the modulatory role of FOXM1 in the Wnt signalling through accelerating b-catenin nuclear localisation and downstream target gene expression via binding to b-catenin [26–28]. The Wnt/b catenin cascade, an evolutionarily conserved and versatile pathway, is well known to play the oncogenic role in ATC, in the context of which, the transfer of b catenin from the cytoplasm to the nucleus is a crucial event [29–31]. Interestingly, in our work, DIAPH3 insufficiency lowered the expression of b-catenin and c-Myc, a direct target gene of Wnt/b-catenin, which were both augmented again in SW1736 and THJ-16T cells after FOXM1 was overexpressed, hinting that DIAPH3 might interact with FOXM1 to activate the Wnt/b-catenin signalling in ATC.
On the basis of the experimental observations as well as mechanistic findings, our investigation offered the first evidence for the tumour-promoting role of DIAPH3 by driving cell proliferation, migration, as well as invasion, and it led to a novel identification for the DIAPH3-FOXM1 protein interaction in ATC. These insights may contribute to the establishment of the molecular basis of ATC aggressiveness, and might indicate that the therapies targeting DIAPH3 may be effective in preventing the aggressive advancement of ATC.
Ethnical statement
Not applicable.
Data availability statement
Data will be made available on reasonable request.
Author contributions
P.S. and Y.L. conceived the study. P.S., H.F., and K.F. performed the experiments and wrote the original manuscript. P.S. and H.F. analysed the data. P.S., Z.Z., and F.Y. revised the manuscript.
