t(11;19) translocation and CRTC1-MAML2 fusion oncogene in mucoepidermoid carcinoma

Introduction 

Mucoepidermoid carcinoma (MEC) is a relatively uncommon malignancy arising from exocrine glands in the upper aero-digestive tract and tracheo-bronchial tree. Most frequently, they arise in the major and salivary glands where they account for approximately 35% of all malignant salivary gland tumours.1, 2 Of non-salivary sites, MEC is most frequently reported arising within the lung,3, 4, 5 although other tracheal/laryngeal origins have all been documented. Rarely MEC arises at others sites including the thyroid gland, breast, lacrimal gland, and conjunctiva.6, 7, 8, 9, 10

In the salivary glands, absolute data for the incidence and prevalence of MEC, both overall and by location (i.e. major versus minor glands, specific site, etc.) are uncertain, as is the relative contribution of MEC to the absolute number of all salivary gland malignancy. True estimation is confounded by differing methodology in tumour registration across registries. In the US, reported data from local11 and National registries12 suggest an estimated incidence of all salivary gland malignancies of approximately 1–1.2 per 100,000 with similar incidence reported from Sweden.13 Although retrospective reviews have reported data suggesting that MEC is the most common salivary gland malignancy, accounting for as much as 40–52% of all major and minor salivary gland malignancies,14, 15 other institution’s experience suggests that adenoid cystic carcinoma is more frequent.16 Nevertheless, it is apparent that MEC represents one of the most common and important forms of salivary gland cancers.

The pathogenesis of MEC is unclear, although data suggest that radiation exposure is a risk factor.17, 18, 19 Recent attention to the presence of a non-random t(11;19) reciprocal translocation as a frequent occurrence in MEC20 and the identification of the novel fusion oncogene generated, CRTC1-MAML2,21 suggests that these events are of biological significance in MEC. Recurrent, non-random chromosomal translocations which generate novel chimeric gene fusions with oncogenic activity are well recognised and causally implicated in a range of mesenchymal and haematological malignancies.22, 23 Such ‘fusion oncogenes’ are typically both derived from and consequently encode for transcription factors, transcriptional regulators and receptor tyrosine kinases that act either alone or in combination with other genetic events to alter gene expression, contributing to tumourigenesis. Increasingly, evidence suggests that such events are also implicated in epithelial malignancy, including papillary and follicular thyroid carcinomas, and prostatic adenocarcinoma.24

The presence of such recurrent translocations across different types of salivary gland tumours has also been reported (reviewed in Stenman).20 Pleomorphic adenomas are associated with two non-random translocations and their variants. Rearrangements involving chromosome 8q12, most frequently t(3;8)(p21;q12), that involves the transcription factor gene, PLAG1, are present in 39% of all pleomorphic adenomas with cytogenetic abnormalities.20 These rearrangements usually lead to ectopic overexpression of PLAG1, with similar events also implicated in lipoblastoma.25 A second translocation, involving the 12q14–15 locus, most commonly t(9;12)(p21q13–15), leads to deregulation of the HMGA2 gene, and is found in 8% of pleomorphic adenomas with cytogenetic abnormalities.20 HMGA2 rearrangements are also present in a number of different benign tumour types, including lipomas and uterine leiomyomas.

In MEC involving the salivary glands and the lung, the presence of a non-random t(11;19) reciprocal translocation has long been recognised, with the same translocation also possibly a feature of certain Warthin’s tumours (WTs).20 Recently, using molecular methods developed following the cloning of its associated CRTC1-MAML2 fusion oncogene,21, 26 a number of key studies have advanced our understanding of this event in both MEC and WT.27, 28, 29, 30 In particular, the significance of CRTC1-MAML2 and certain clinical and pathological features of MEC have been clarified. Moreover, these features provide a basis for novel prognostic and therapeutic strategies in the clinical management of MEC. The aim of this article is to present the available data on the t(11;19) and its associated CRTC1-MAML2 fusion product in MEC with some discussion of the clinical relevance of such features.

Recent key studies of the t(11;19) translocation and related CRTC1-MAML2 fusion oncogene 

Four pivotal studies have evaluated CRTC1-MAML2 expression in either various types of salivary gland tumours or MEC involving different sites of origin.27, 28, 29, 30 The majority of data were derived from three independent studies from different patient cohorts.28, 29, 30 All studies were retrospective in nature using archival tumour tissue, although in two of these, analysis was facilitated by the availability of fresh-frozen tumour material. An outline of the key findings from these studies is presented in Table 1.

Table 1. Key studies investigating CRTC1-MAML2 expression in salivary gland tumours

	Study	Tumours analysed	Key findings
	Martins et al.27	10 MEC	•70% MEC fusion +ve •0% WT fusion +ve
		7 WT
	Behboudi et al.28	29 MEC	•55% MEC fusion +ve •Fusion expressed in all MEC-cell types •Only low- and intermediate-grade tumours were fusion +ve •Significant correlation between fusion +ve tumours and reduced risk of local recurrence, metastases, or tumour related death (p=0.0012) •One WT fusion +ve
		3 WT
	Okabe et al.29	71 MEC	•38% MEC fusion +ve •Only low- and intermediate-grade tumours were fusion +ve •Significant correlation between fusion +ve tumours and improved overall survival (p=0.002) •0% WT fusion +ve •0% other salivary gland tumours fusion +ve
		26 WT
		25 other salivary gland tumours
	Tirado et al.30	22 MEC	•81% MEC fusion +ve •No association between histological grade/clinical stage and fusion status •Significant correlation between fusion −ve tumours and distant metastases (p=0.05) •36% WT fusion +ve •0% other salivary gland carcinomas fusion +ve
		11 WT
		22 other salivary gland carcinomas

The nature of the t(11;19) translocation and related CRTC1-MAML2 fusion oncogene and the role in the neoplastic process 

Positional cloning of the t(11;19) translocation identified both in primary tumour samples (MEC from salivary gland and bronchopulmonary tissue) and MEC-derived cell lines (H3118: pulmonary origin and H292: parotid origin) has characterised the nature of this translocation.21, 26 This translocation results in the fusion of exon 1 of the CTRC1 gene (also known as MECT1, TORC1, or WAMTP1) on chromosome 19p13 with exons 2–5 of the MAML2 gene on chromosome 11q21, to generate a novel fusion oncogene, CRTC1-MAML2 which encodes an associated novel fusion transcript protein (Fig. 1).

View Large Image Download to PowerPoint

Figure 1. t(11;19) and CRTC1-MAML2 fusion oncogene. Schematic illustration of the wild-type CRTC1 and MAML2 genes and the resulting CRTC1-MAML2 fusion oncogene associated with t(11;19). The Notch-binding domain of MAML2 is replaced by the CREB binding domain of CRTC1 which is fused to exons 2–5 of the MAML2 gene. This oncogene remains under the control of the CRTC1 promoter.

The products of both constituent genes in this translocation each have roles in cell-cycle control in their normal state. CRTC1 is a cAMP response element binding protein (CREB) co-activator. CREB regulates gene expression involved in cell proliferation and differentiation in response to cytokines and growth factors31 with abnormal CREB activity implicated in carcinogenesis.32 CRTC1 acts by binding CREB and enhancing its transcription, independent of phosphorylation. MAML2 is a member of the Mastermind-like gene family, widely expressed in tissues, which function as co-activators of Notch receptors. Notch receptor signalling has a complex role in cell proliferation and differentiation, which is context-dependant, acting either as an oncogenic stimulus or as a tumour suppressor, depending on stimulus and cell-type.33 In the novel CRTC1-MAML2 fusion transcript, the CREB binding domain from CRTC1 replaces the free intracellular Notch-binding domain from MAML2 to produce a protein with novel transformation properties.

Tonon and co-workers demonstrated the in vitro transforming abilities of the CRTC1-MAML2 fusion product by the induction of foci within an immortalised rat RK3E cell-line.21 A subsequent study found that injection of CRTC1-MAML2 transfected RK3E cells into nude mice led to in vivo tumour formation with molecular studies showing that sustained ectopic expression of the transcript is required for tumourigenesis.34 This study also found that RNA interference of the transcript abolishes tumour cell growth in MEC lines expressing this transcript, lending further support to the role of CRTC1-MAML2 in MEC formation as well as offering a potential target for therapeutic intervention.

The molecular mechanisms underlying such transformation activity by CRTC1-MAML2 are complex and are the subject of ongoing investigation. However, certain pathways have been identified. Notch-target genes implicated in cell fate decisions, including HES1 and HES5, are upregulated in MEC-cell lines transfected with CRTC1-MAML2.21, 26 In this context Notch-target activation is independent of direct Notch ligand stimulation. Studies have shown that CRTC1-MAML2 transfected RK3E cell transformation is dependant upon the activation of the CREB pathway by the transcript, and that growth suppression of fusion-positive cell lines is seen with interference in CREB activity. The CRTC1-MAML2 transcript also induces expression of multiple CREB target genes in transfected cells suggesting an alternative mechanism to cAMP-mediated expression.32, 35, 36 Such findings suggest that CRTC1-MAML2 may act by the disruption of both Notch and CREB-regulated cell-cycle and differentiation pathways.

The prevalence of t(11;19) in mucoepidermoid carcinoma 

A number of cytogenetic studies on MEC arising from both minor and major salivary glands have revealed the presence of a recurrent t(11;19) reciprocal translocation, either as part of a more complex abnormal karyotype20, 37, 38, 39 or as the sole chromosomal abnormality.20, 38, 40, 41 Similar observations have been reported in MEC arising in the lung.42, 43, 44, 45, 46 The limitations of conventional cytogenetic techniques that require the use of fresh tumour samples to determine the frequency of the t(11;19) in MEC have been overcome by the use of molecular techniques to detect the associated CRTC1-MAML2 fusion transcript. Using reverse transcriptase-polymerase chain reaction (RT-PCR) and/or fluorescence in situ hybridization (FISH) analyses, Martins et al. found the CRTC1-MAML2 transcript in seven out of 10 MEC samples.27 Behboudi and colleagues, using a nested RT-PCR technique, demonstrated the transcript in 16 out of 29 MEC cases.28 Of additional value from this study was the finding that the fusion protein was expressed in all MEC-specific cell types (mucous, epidermoid, and intermediate cell types) (Fig. 2). The authors also reported cytogenetic data that suggested that not all fusion-positive tumours carried the t(11;19), with the implication that other cryptic translocations may contribute to the process in such cases. A similar study, with a larger sample size, by Okabe et al. found this transcript in 27 out of 71 MEC samples.29 More recently, using a similar RT-PCR methodology (Fig. 3), Tirado and colleagues demonstrated that 18 out of 22 salivary gland MEC harboured the CRTC1-MAML2 fusion protein.30 Such studies indicate the presence of this transcript in 38–81% of MEC. In the latter study the authors commented that the differences in the detection rates across these studies may reflect the nature of the tumour material studied, with lower rates of detection being reported when archival paraffin-embedded tumour was analysed in comparison to studies using fresh-frozen tissue.30

View Large Image Download to PowerPoint

Figure 2. MEC with immunostaining for CRTC1-MAML2 transcript expression. Immunostaining of the CRTC1-MAML2 fusion protein in a primary t(11;19)-positive MEC tumour. Note the predominant nuclear staining of the tumour cells. Inset: Intermediate cells as well as mucous cells stained positive, whereas stromal cells were negative. Photomicrograph is from Behboudi et al. Molecular classification of mucoepidermoid carcinomas-prognostic significance of the MECT1-MAML2 fusion oncogene.28 (Reprinted with permission of Wiley-Liss Inc., a subsidiary of John Wiley & Sons Inc.)

View Large Image Download to PowerPoint

Figure 3. RT-PCR demonstrating CRTC1-MAML2 transcript expression in selected MEC. Nested RT-PCR amplification product of the CRTC1-MAML2 fusion transcript in salivary mucoepidermoid carcinoma (MEC). The transcript is present in lane 2 (positive control), lane 5 (MEC parotid), and lane 6 (MEC of salivary gland NOS). Lane 1: M, molecular weight marker (100bp); lane 2: +ve, positive control H292 cell line positive for the CRTC1-MAML2 transcript (117bp); lane 3: N, negative control (normal parotid); lane 4: N, negative control (squamous carcinoma of larynx); lane 5: MEC, MEC parotid and lane 6: MEC, MEC of salivary gland NOS. Courtesy of Dr. El-Naggar, University of Texas, M.D. Anderson Cancer Center.

t(11;19) and CRTC1-MAML2 and tumour specificity 

To date, the great majority of tumour cases characterised cytogenetically by the presence of the t(11;19) translocation, and molecularly by the demonstration of the associated CRTC1-MAML2 fusion transcript have been MECs. Such an association has been shown for MEC arising in different tissue types including salivary gland,20, 27, 28, 29, 30 lung,42, 43, 44, 45, 46 and thyroid gland.30 However, data from case reports and related molecular studies suggest that this phenomenon is not restricted to MEC as a tumour type.

Two cases of Warthin’s tumour (WT) have been reported containing a t(11;19) translocation, either as the sole karyotypic anomaly47 or as part of a more complex abnormal karyotype.48 This latter case has also been molecularly characterised as carrying the CRTC1-MAML2 fusion transcript.26 This initial suggestion that WT could also harbour this transcript was of some debate across different investigators,49, 50 with large molecular studies failing to demonstrate this transcript in seven and 26 cases of WT, respectively.27, 29 However, Tirado and colleagues analysed a range of tumours (22 salivary gland MECs, three MECs of thyroid origin, 22 other salivary gland carcinomas of various histologies, and 11 WTs) for CRTC1-MAML2 transcript expression using nested RT-PCR on both fresh-frozen and paraffin-embedded tumour tissue. This group reported that 36% (four out of 11 cases) of WT were fusion positive.30 A recent independent study reported that 4% (two out of 48 cases) of WT analysed expressed the CRTC1-MAML2 transcript, and that both cases were metaplastic variants of WT.51 Such data suggest that this transcript may be a factor in the pathogenesis of WT, albeit at a lower prevalence than for MEC. Another recent study provides a further twist to this relationship between the CRTC1-MAML2 oncogene, MEC, and WT.52 The authors analysed five cases of WT associated with malignancy (three cases of WT with co-existent MEC; one WT with co-existent metastatic melanoma; and one case of primary malignant WT) for CRTC1-MAML2 expression. All five cases studied were fusion positive. Specifically, all benign elements of WT, the co-existent MEC and the malignant WT were fusion positive whilst the melanoma was negative for the transcript. Of the three cases of WT with co-existent MEC, metaplastic changes were noted in the oncocytic epithelium juxtaposed between the WT and MEC.

Of some interest also is another recent study published at the time of writing as an abstract, which reported on the presence of the CRTC1-MAML2 transcript in six out of nine cases of histologically confirmed MEC arising within a pre-existing pleomorphic adenoma.53 Although more information is needed before any conclusions can be drawn from such data, it does add to the evidence that CRTC1-MAML2 plays an important and perhaps initiating role in MEC development, both de novo and from an existing benign phenotype.

The CRTC1-MAML2 transcript has also been demonstrated in clear cell hidradenoma, a benign sweat gland tumour of eccrine duct origin.54 Molecular analysis of a case previously characterised by a t(11;19) translocation55 revealed the presence of the CRTC1-MAML2 fusion product.56 Subsequent analysis of 20 such tumours from a single institute found that 50% contained the CRTC1-MAML2 gene fusion and expressed the resulting transcript.57 Of interest from this study was the association of this fusion with only clear cell variants of the tumour, with all fusion-negative tumours lacking clear cells. This transcript has also recently been demonstrated in a hidradenoma arising within the breast parenchyma, suggesting, as in MEC, that the translocation may have significance across tissue types in this tumour also.58

Clinical significance of the CRTC1-MAML2 fusion protein 

The available data suggest that the t(11;19) translocation and the associated CRTC1-MAML2 fusion transcript are recurrent events present in 38–81% of MEC.27, 28, 29, 30 Although this transcript may also be present in a small number of WTs, it has not been demonstrated in any other salivary gland tumour. Clearly this phenomenon may allow additional avenues for the clinical management of MEC.

Discussion and conclusions 

The existence of a recurring t(11;19) translocation and subsequent gene fusion between CRTC1 and MAML2 to create a novel chimeric oncogene may have important implications for understanding MEC biology and developing novel diagnostics and targeted therapy. Such findings complement and increase our knowledge of the role of translocations and fusion oncogenes in epithelial malignancy.20, 22, 72

Molecular studies suggest that this fusion oncogene acts as a transcription activation factor in CREB and Notch regulatory pathways, and that such events may be early events in tumour formation.21, 32, 34, 35, 36 The finding of this transcript in a high proportion (38–81%) of MEC,27, 28, 29, 30 and that this gene is expressed in all cell types,28 suggests that in such tumours CRTC1-MAML2 acts at an early stage in tumour initiation. The recently reported example of MEC characterised by a t(11;15) translocation and associated variant fusion oncogene, CRTC3-MAML2, in which a CRTC1-related CREB binding protein, CRTC3 present on chromosome 15, acts as an alternative fusion partner with MAML2 and provides further support for such a mechanism.73

CRTC1-MAML2 is associated with a number of tumour types, most importantly MEC involving salivary gland and non-salivary tissues, but also WT and hidradenoma. The common expression of CRTC1-MAML2 across such different phenotypes is consistent with findings for other fusion oncogenes, which suggest that fusion oncogenes are not necessarily tumour type specific. The ETV6-NTRK3 fusion oncogene is associated with a number of different tumours including congenital fibrosarcoma, secretory breast carcinoma, and the cellular form of congenital mesoblastic nephroma.74 The ASPL-TFE3 oncogene, like CRTC1-MAML2, also a novel transcription activator, is associated with both alveolar soft part sarcoma and forms of renal cell carcinoma.75 Conceivably such oncogenic products, in particular chimeric transcription factors, influence differentiation pathways in precursor tumour cells, with the phenotypic outcome dependant upon environmental constraints, and that this also applies to CRTC1-MAML2 fusion-positive tumours across different tumour types.76

Clinico-pathological studies have demonstrated that, although all grades of MEC may express CRTC1-MAML2, there is an association between transcript expression and tumour stage, with fusion-positive tumours behaving in a less aggressive fashion. That the presence of the CRTC1-MAML2 fusion oncogene is of putative prognostic value conferring an improved clinical outcome is of some interest. In contrast to haematological and lymphoid malignancy, absolute data for the prognostic significance of the presence or absence of a particular fusion oncogene in solid tumours are limited. In many sarcomas whilst the presence of a particular fusion is of prognostic value, this is often relative to the prognosis of the same tumour type associated with a variant fusion, rather than the absence of such events per se. Recent studies have highlighted the role of the TMPRESS2-ETS fusion oncogene in prostate cancer as a frequent early event, with TMPRESS2-ETS expression associated with an aggressive clinical behaviour, higher tumour stage and increased Gleason scores.77 This association between a fusion oncogene and poorer prognosis contrasts in part with the more limited data available for CRTC1-MAML2 and MEC. This in itself is of interest, and requires further consideration, whilst also serving to underline our narrower understanding of such events in rarer epithelial tumours. The studies discussed in detail are from a limited number of patients (approximately 120 cases of MEC) highlighting the inherent difficulties in gathering valid information in a relatively uncommon tumour group. In this endeavour the authors are to be admired. As such, it should not be too surprising that the conclusions in part are inconsistent between studies, although there now appears to be a certain consensus as discussed above. It should be emphasised, however, that although CRTC1-MAML2 may be associated with favourable prognostic features in MEC, further data are necessary before any conclusions can be made that will influence clinical practice.

In summary, CRTC1-MAML2 gene fusion, associated with a recurring t(11;19) translocation is a frequent event in mucoepidermoid carcinoma occurring in numerous tissues and may act as an adjunctive diagnostic and prognostic marker for this tumour. This transcript is common to MEC, Warthin’s tumour and clear cell hidradenoma suggesting a fundamental role in tumourigenesis in different tumour types. Although the functional role of this oncogene is complex, it offers a target for the development of novel therapeutics towards MEC in patients where conventional treatment has failed.

Conflict of Interest Statement 

None declared.