Retinoic acid promotes differentiation of WiT49‐ but not of CCG99‐11 Wilms tumour cells

Abstract Background Most children with Wilms tumour are successfully treated with multidrug chemotherapy and surgery. These treatments cause severe side effects for the patients, an issue that needs to be addressed by exploring other treatment options with less or no side effects. One option is to complement current therapies with agents that could potentially induce tumour cell differentiation, for example retinoic acid (RA). Aims To facilitate quick assessment of an agent's effect on Wilms tumour differentiation by a rapid in vitro model system. Methods and Results Here WiT49 and CCG99‐11 Wilms tumour cells were treated with 10 μM RA for 72 h or 9 days. Cultured cells were scraped off from Petri dishes, pelleted and embedded in paraffin in the same way as clinical tumour specimens are preserved. Cell morphology and differentiation were evaluated by analyses of haematoxylin eosin (H&E) and immunohistochemical stainings. Based on H&E, WT1 and CKAE1/3 stainings, RA treatment induced further epithelial differentiation of WiT49 cells, whereas there was no sign of induced maturation in CCG99‐11 cells. Ki67 staining showed that RA inhibited cell proliferation in both cell lines. Conclusions Our study shows that in vitro culturing of WiT49 and CCG99‐11 cells, followed by pelleting and paraffin embedding of cell pellets, could aid in a quick evaluation of potential differentiating agents against Wilms tumour. In addition, our results strengthen previous results that retinoic acid could be a potential complement to regular Wilms tumour treatment.


| INTRODUCTION
Wilms tumour is a paediatric solid tumour of the kidney. Most patients with Wilms tumour have a good prognosis thanks to multidrug chemotherapy and surgery, sometimes supplemented by radiotherapy. However, the treatment often comes with lifelong side effects. 1 To increase the quality of life for Wilms tumour survivors, therapies need to be modulated with agents less harmful than chemotherapy. Such agents might not always kill off the cancer cells, but rather inhibit their proliferation and induce a less malignant phenotype, that is, induce the cancer cells to differentiate. The strategy of differentiation therapy in cancer treatment is to induce terminal differentiation of the cancer cells. 2 Retinoic acid (RA), a metabolite of vitamin A, which cause only mild side effects when distributed as a drug, has been used rather successfully as part of the treatment for patients afflicted with leukaemia and neuroblastoma. 3 Neuroblastoma is a solid paediatric tumour derived from the sympathetic nervous system and differentiation of neuroblastoma cells can be analysed in vitro. For instance, differentiation of neuroblastoma derived cell lines can be induced by retinoic acid and is recognised by increased neurite outgrowths (reviewed in). 4 It is less clear how to evaluate differentiation of Wilms tumour cells in vitro.
Cultures with primary Wilms tumours have been established as 3D spheroids or organoids and are perfect for exploration of patient-specific drug sensitivities. [5][6][7] Even so, Wilms tumour is a rare disease, the number of patients and thus the extent of clinical tumour material is limited. There is a need for an alternative way to make pilot testing of drugs before irreplaceable primary cultures are exhausted. Here, established Wilms tumour cell lines become important.
Wilms tumours are known for their commonly triphasic histology consisting of blastema, stroma and epithelium. 8 Those histological components can to some extent be recreated by cell line based orthotopic Wilms tumour xenograft models and differentiation status can be evaluated microscopically. [9][10][11] Drawbacks with xenograft models are the aspects of animal ethics and time. Therefore, research on cell lines would benefit from a more ethical and less time consuming perspective to do pilot testing of potential differentiation inducing agents.
The differentiated epithelial component of Wilms tumour is identified by the presence of epithelial tubules. Accordingly, a procedure to evaluate Wilms tumour differentiation in vitro would require the possibility to analyse the formation of epithelial tubules, along with assessment of a panel of differentiation markers. Thus, the aim of this study is to establish whether Wilms tumour cell lines cultured in vitro can be used to evaluate a drug's capacity to induce differentiation of Wilms tumour cells. RA is explored as differentiation agent.
Two well established cell lines that are uniformly recognised as Wilms tumour derived cells, are WiT49 and CCG99-11. 12,13 Our research group have earlier characterised them as epithelial and blastemal predominant cell lines respectively. 9,11 In order to assess the differentiation status, knowledge about the origin of Wilms tumour is necessary.
Wilms tumour mimics the developing kidney, nephrogenesis, with respect to its three distinct histological components described above.
Nephrogenesis is initiated by the reciprocal induction between the metanephric mesenchyme, that is, the blastema, and the ureteric bud.
Upon induction of nephrogenesis a mesenchymal to epithelial transition takes place and blastema differentiates into mature epithelial tubular structures. SIX2 is expressed in the immature blastema of the developing kidney and is found to be expressed in Wilms tumour blastema. WT1, on the other hand, is important at several stages of nephrogenesis and nuclear expression is frequently recognised in Wilms tumour cells. 14,15 Mature epithelial tumour tubules can be highlighted by expression of specific keratins. 16 Experimentally, RA has been shown to be important in kidney development. 17 In a zebrafish nephrogenesis model systems it has been shown that RA plays a central role in differentiation and segmentation of the nephrons. 18 Previous studies of RA and Wilms tumour show that RA can inhibit proliferation of primary Wilms tumour cultures and create morphological changes. 19,20 On the transcriptional level, it has been demonstrated that RA treatment of cultured Wilms tumour cells enhances gene expression of RA-and TGFβ signalling related genes, which are genes involved in inhibition of cell proliferation. It has also been shown that RA mitigates the expression of genes associated with high risk and relapse. 19,20 The number of patients with Wilms tumour that have been enroled in clinical trials with RA as treatment agent is limited. 21,22 This study shows how the WiT49 and CCG99-11 cell lines can be  At the end of the experiments, cells were gathered by either trypsination or cell scraper and washed with 1Â PBS twice followed by pelleting in order to create paraffin embedded cell blocks. Cell pellets were treated with 4% paraformaldehyde for 20 min, followed by incubation with haematoxylin for 5 min; 70% ethanol (EtOH) o.n.; 95% EtOH for 1 h; 99,5% EtOH for 1 h; xylen for 1 h, followed by paraffin embedding at 65 C for at least 1 h. All steps were performed at room temperature unless indicated, and all cell pelleting steps between treatments were done by centrifugation for 5 min at 300g.     Figure 1G,H). As expected, it was not possible to study the morphology of trypsinized cells as such cells all appear rounded ( Figure 1A,B,E,F). Thus, for assessment of cell morphology, only cells which had been collected using a cell scraper were analysed, since that technique enabled sheets of cells to stay more intact (Figure 1C,D,G,H).

| Retinoic acid induced differentiation of WiT49 cells
To test the effect of RA on Wilms tumour cell differentiation, WiT49 and CCG99-11 cells were treated with 10 μM RA, or DMSO as a control, for 72 h. WiT49 cells treated with RA for 72 h were more elongated than control cells as the RA treated cells lined up in thin pronounced garland formations, mirroring segmentation during nephrogenesis, which is a sign of epithelial differentiation ( Figure 1C,D). RA treatment of CCG99-11 cells did not induce any apparent morphological changes ( Figure 1G,H).
To test if a longer RA treatment could induce signs of differentiation of CCG99-11 cells, the treatment was prolonged from 3 to 9 days. RA kept WiT49 cells in a morphologically more differentiated state also after 9 days of RA treatment, with even more pronounced garland formations than after 72 h RA treatment (Figure 2A  with either 5% or 10% serum for 9 days, both these experimental set ups were evaluated in further aspects of differentiation.

| WT1 protein/nuclear expression is affected by prolonged retinoic acid treatment
WT1 is known to act during the induction of mammalian nephrogenesis, but its expression is also known to fluctuate over time during different stages of kidney development. 14,23 It is known to be expressed in Wilms tumour. 14 After 72 h RA treatment, nuclear WT1 expression became more accentuated and intense in WiT49 cells ( Figure 3A Figure S2. Fisher's exact test, two tailed). In summary, RA treatment significantly inhibited cell proliferation in both WiT49 and CCG99-11 cells.

| DISCUSSION
Although the survival rate of children diagnosed with Wilms tumour is good, treatment with heavy chemotherapy mostly cause lifelong adverse effects. 1 To reduce the adverse effects of present treatment strategies, introduction of drugs with more gentle side effects is to be preferred. Differentiation therapy, which aims to terminally differentiate malignant cancer cells into their original benign cells, is one way to tackle this concern. In addition, if researchers turn to already existing Especially since over the years some of them have been re-evaluated and been designated as other tumour entities based on for instance their genetic profiles. [26][27][28] Tumour genetics could also explain why the two cell lines respond differently. SIX2 is recurrently mutated in Wilms tumour, but mutations in the gene do not seem to affect SIX2 protein expression according to the literature. 29 20 Such results might be discouraging, but RA could potentially be a not so toxic drug that could be used as metronomic cancer treatment. 31 There is an assumed positive effect of RA treatment of patients with Wilms tumour, however, the number of patients that have received RA therapy is too small to give conclusive results. 21,22 Our results support further evaluation of the use of RA in treatment