Final Report Summary - RARBETA2 IN OSCC (Understanding the functional role of Retinoic Acid Receptor Beta2 in the development of oral cancer) Understanding the functional role of Retinoic Acid Receptor Beta2 in the development of oral cancer (RARbeta2 in OSCC): Final reportThe experimental work described used a unique cohort of cell cultures derived from a variety of oral potentially malignant lesions (OPML). These show a variety of proliferative lifespans with some undergoing replicative senescence a whilst others are immortal. The molecular features which define these have been previously described.Objective 1: Assessment of the contribution of RARβ2 to the onset of senescence.Result 1. Expression of RARB2 in OPML cultures (Figure 1). RARB2 is constitutively expressed by OPML cultures with a limited lifespan (D6 and D30) and increases on approach to senescence. It is not expressed in unstimulated normal oral keratinocyte (iNOK) cultures. By comparison, RARB2 expression has been lost in immortal OPML cell cultures (D19, D20, D34 and D38). A similar pattern of expression is seen in matched tissue samplesResults 2. Treatment with retinoids and demethylating agents: Treatment of the immortal OPML cultures with retinoic acid (RA) and/or the demethylating agent 5-aza-2-deoxy Azacytidine (Aza) results in variable effects on the expression of RARβ2 (Figure 2).• D19 and D34: Combination of RA and Aza results in repression of RARβ2. Aza alone has a lesser effect, with RA alone having little effect. These effects are confirmed by methylation analysis (by MS-PCR and confirmed by sequencing) of the RARβ2 promoter which demonstrates methylation in untreated cells which reduces on Aza treatment. • D20: RA alone causes modest increase in RARβ2 expression and Aza results in a further rise. Promoter analysis demonstrates complete methylation which is unchanged after Aza treatment.• D38: no consistent effect of either drug or in combination. The promoter is unmethylated, thus the loss of expression in D38 is unexplained.Result 3. The effects of treatment with RA and demethylating agents on senescence. Senescence was assessed by expression of β-galactosidase and formation of HIRA nuclear foci. The results varied between the cells (Figure 3).• D19: little effect on senescence of either treatment• D34: marked increase in both assays on RA alone or in combination with Aza• D20 and D38: a small increase in senescence on both treatmentsObjective 2: Assessment of the effect of transfection of RARβ2Three of the immortal precancer cell lines were transfected with RARB2, both as transient transfection and also under antibiotic selection pressure in order to generate stably transfected cells. Whilst the transient transfections were successful (Figure 4), no cell lines stably expressing RARB2 were generated as the cells did not survive the selection pressure. Unfortunately, no transfection experiments were conducted on the mortal cells as their proliferative capacity did not allow this. Objective 3: Investigation of signalling pathways and cell cycle changesResult 1: changes in differentiation markers on treatment (Figure 5). The immortal cell cultures show marked reduction of markers of differentiation (Involucrin and IntegrinB1). Treatment with Aza and Aza+RA resulted in no expression in D19 and D38, whilst Aza induced IVL expression in D20 and D34. The addition of RA reduced the effect in these cells. In monolayer culture, the treatment schedule did not result in any consistent changes in differentiation.Result 2: Changes in the cell cycle and signalling pathways on treatment. Treatment with Aza and AZA+RA resulted in accumulation of cells in the G2 phase of the cell cycle in D19, D20 and D34 (Figure 6). There was little effect on D38. Mirroring this, p16 (cdkn2A) expression also increased in D19 and D34, with little effect on D20 and D38. There was no consistent effect on p21 (cdkn1A) other than in D38, which is the only one of the immortal cultures to retain p53 expression (Figure 7).Result 3: Retinoid signalling pathways. We also identified changes in other components of the RA signalling pathways, in particular the loss of expression of CRBP1 in these cultures (Figure 7A). Treatment with Aza and Aza+RA resulted in increased expression in D34 and D20, with little effect in D19 and D38 (Figure 7B). The mechanism of this is unclear as the promoter did not show CpG island methylation in any case.Objective 4: Assessing effects in tissue engineered oral mucosaAll four of the immortal OPML cultures were used to generate tissue engineered oral mucosa using de-epithelialized donor dermis (DED). The expression of RARβ increased in tissue engineered oral mucosal models on treatment with RA and Aza-C + RA (Figure 8). It was clear that the addition of Aza resulted in some toxicity as the models were much more fragile. It was difficult to assess the extent of terminal differentiation in these models.Objective 5: Assessing RARβ2 expression in a cohort of oral precancer patient tissues.Two cohorts of patient samples (used with ethical approval) were analysed by immunohistochemistry: one whose samples directly matched the cell cultures used in the main part of the project and a separate, unrelated cohort of oral premalignant lesions. Sections were stained for both RARβ and CRBP1. The staining on the sections was analysed in a semi-automated manner using Image J. Results 1. Samples matched to cell cultures (Figure 1B). The proportion of cells expressing RARB was significantly lower in the tissues which gave rise to immortal cell cultures. This indicates that the loss of RARB is not merely an in-vitro phenomenon and that is may be possible to use it as a biomarker for selection of patients for retinoid therapy. Expression of CRBP1 in this cohort was however inconsistent (data not shown)Results 2. Clinical Cohort. Expression of RARB in the extended clinical cohort showed heterogeneity in the expression of RARB2 (Figure 9). The % of cells expressing RARB2 was variable and not directly related to the grade of epithelial dysplasia.Conclusions• RARβ2 is down-regulated in most immortal oral precancer cells and this is due to methylation of CpG islands in the promoter. Mortal oral precancer cells retain RARβ2 expression. This pattern also seen in the original tissues• RARβ2 is expressed following administration of Aza-C to a variable extent: fully in 2/4 cell lines tested, variably in one, and not at all in another. In cells which re-express RARB2, treatment with retinoic Acid and 5’Aza-C in combination induces G2M arrest and senescence• The response to treatment is related to the ability to re-express RARβ2, cRBP1 and p16 in iOPML cells. Assessment of all of these factors would be needed in patient samples to predict the response to therapy with retinoid and demethylation agents.ImpactThe work presented will inform any future trials of retinoids in OMPLs. This will be of relevance to researchers and clinicians in this area, along with their patients. It is clear that the molecular heterogeneity of these lesions will make it difficult to identify clear parameters which may allow for prediction of success of treatment of these lesions with such agents. It is not as simple as identifying lesion which express RARB – the methylation status of both RARB and p16 needs to be taken into account, and also the expression of CRBP1 in order to predict effect on the indication of senescence. Dr Keith Hunter: Reader in Oral and Maxillofacial Pathology k.hunter@sheffield.ac.ukFebruary 2016 Related documents final1-retinoids-eu-report-figures-final.pdf
