Dissecting the signaling pathways associated with the oncogenic activity of MLK3 P252H mutation

Background MLK3 gene mutations were described to occur in about 20% of microsatellite unstable gastrointestinal cancers and to harbor oncogenic activity. In particular, mutation P252H, located in the kinase domain, was found to have a strong transforming potential, and to promote the growth of highly invasive tumors when subcutaneously injected in nude mice. Nevertheless, the molecular mechanism underlying the oncogenic activity of P252H mutant remained elusive. Methods In this work, we performed Illumina Whole Genome arrays on three biological replicas of human HEK293 cells stably transfected with the wild-type MLK3, the P252H mutation and with the empty vector (Mock) in order to identify the putative signaling pathways associated with P252H mutation. Results Our microarray results showed that mutant MLK3 deregulates several important colorectal cancer- associated signaling pathways such as WNT, MAPK, NOTCH, TGF-beta and p53, helping to narrow down the number of potential MLK3 targets responsible for its oncogenic effects. A more detailed analysis of the alterations affecting the WNT signaling pathway revealed a down-regulation of molecules involved in the canonical pathway, such as DVL2, LEF1, CCND1 and c-Myc, and an up-regulation of DKK, a well-known negative regulator of canonical WNT signaling, in MLK3 mutant cells. Additionally, FZD6 and FZD10 genes, known to act as negative regulators of the canonical WNT signaling cascade and as positive regulators of the planar cell polarity (PCP) pathway, a non-canonic WNT pathway, were found to be up-regulated in P252H cells. Conclusion The results provide an overall view of the expression profile associated with mutant MLK3, and they support the functional role of mutant MLK3 by showing a deregulation of several signaling pathways known to play important roles in the development and progression of colorectal cancer. The results also suggest that mutant MLK3 may be a novel modulator of WNT signaling, and pinpoint the activation of PCP pathway as a possible mechanism underlying the invasive potential of MLK3 mutant cells.

MLK3 is a serine/threonine protein kinase that regulates the mitogen-activated protein kinase (MAPK) pathway, activating ERK, p38 and JNK in response to extracellular signals [1,5]. Additionally, functional studies have demonstrated that overexpression of wildtype MLK3 leads to morphological transformation of NIH 3T3 fibroblasts and growth in soft agar, a capacity that is MEK/ERK dependent [6]. Further, MLK3 has been demonstrated to function as a scaffolding protein, involved in the formation of a multiprotein complex containing MLK3/BRAF/RAF1 [1,5,7,8]. The formation of this complex was shown to be important for the activation of wild-type BRAF and, consequently, to the activation of ERK signaling [1,5,7,8]. Furthermore, MLK3 was reported to be important for the proliferation of tumor cells, bearing either oncogenic KRAS or neurofibromatosis-1 (NF1) or NF2 inactivating mutations [8]. In addition, the MLK3 signaling activation was associated with an increase in the migratory and invasive capacity of tumor cells in gastric [9], breast [10][11][12], lung [13] and ovarian [14] cancers. All together, these observations implicate MLK3 as a cancerrelated gene although, until recently, nothing was known about MLK3 gene deregulation in primary cancer tissues.
Our group reported the occurrence of MLK3 mutations in microsatellite unstable (MSI) gastrointestinal tumors (both sporadic and hereditary forms) in a frequency of about 20%. Using in vitro transforming assays, we demonstrated that several MLK3 mutations affecting different domains of the protein had transforming potential when compared to cells expressing the wild-type and the kinasedead forms of the protein [15]. These results were further supported by in vivo studies in which one of the two most transforming mutations (P252Hlocated in the kinase domain) was found to be tumorigenic and to give rise to highly invasive tumors when subcutaneously injected in nude mice. Thus, our previous work pointed mutant MLK3 as a new oncogene in MSI gastrointestinal cancers, however, the signaling pathways associated to its oncogenic activity remained to be explored.
In this work, we aimed at identifying the signaling pathways associated to mutant MLK3, in particular the P252H mutation. The reasons underlying the choice for this mutation were: (a) it was one of the most transforming mutations previously analyzed; (b) showed high tumorigenic capacity with strong invasive potential in in vivo studies; and (c) it was located in the kinase domain which is an important domain for the regulation of downstream signaling pathways.
The results showed that P252H mutation interferes with important colorectal cancer-associated signaling pathways such as WNT, MAPK, NOTCH, TGF-β and P53.

cDNA constructs and mutagenesis
Wild-type MLK3 and mutant sequences were cloned into pLENTID6/V5 directional TOPO (Invitrogen). Mutant MLK3 P252H sequence was generated by site-directed mutagenesis using the MLK3 wild-type sequence cloned in pLENTID6/V5 as template. pLENTID6/V5 empty vector (Mock) was obtained by the insertion of a small fragment of cDNA in order to circularize the plasmid.

Transfections
For HEK293 stable transfections, ViraPower Lentiviral Expression kit (Invitrogen) was used for the transduction of the MLK3 wild-type and mutant P252H sequences as well as the empty vector. Lentiviral transduction was performed following the manufacturer's instructions. Transduced cells were selected by antibiotic resistance (blasticidin, 12 μg/ml) (Gibco, Invitrogen). The expression levels of MLK3 in the different clones selected were measured by western-blot.

RNA extraction and cDNA synthesis
Total RNA was isolated from cell lines using TriPure Isolation Reagent (Roche Applied Science), following manufacturer's instructions. Complementary DNA was synthesized from 1 μg of total RNA using SuperScriptII Reverse Transcriptase (Invitrogen) and Random Primers (Invitrogen).

Labeling and hybridization
Five hundred ng aliquots of RNA from the samples were quality checked using the Agilent 2100 Byoanalyzer and only samples above integrity quality number (RIN) 8 [16] were used and amplified according to the specifications of the Illumina® TotalPrep™ RNA Amplification Kit (Ambion, Austin, TX, USA). The cRNA samples were applied to the arrays of Sentrix® Human-v6 Expression BeadChip (Illumina, San Diego, CA, USA) and hybridized according to manufacturer's specifications. The Sentrix BeadChips were scanned with the Illumina's Beadarray system 500G Scanner (Illumina®).

Microarray data analysis
The signal intensity was extracted from the hybridization images, background subtracted and normalized using Illumina Inc. BeadStudio software version 3.3.7. The data produced was checked against the Illumina internal quality controls and loaded into the Bioconductor software [17,18]. To identify differentially expressed genes based on a moderate t-test, the bioconductor Limma package [19] was used. Genes were selected based on a p-value cut-off (after adjustment) of p < 0.01 to control the false discovery rate (FDR) [20]. To test the association of selected differentially expressed genes with KEGG pathways [21], information provided in the Illumina annotation files

Statistical analyses
For statistical analyzes of in vitro transformation assays a t-student test was used and p < 0.05 was taken as statistically significant. Specific statistical tests used for microarray interpretation are embedded in the corresponding materials and methods section.

MLK3 P252H mutation affects fundamental colorectal cancer-associated pathways
In order to assess the effect of a tumorigenic MLK3 mutant on a genome-wide level, we performed microarray-based (See figure on previous page.) Figure 1 Expression profiling of biological triplicates of HEK293 cells transfected with mutant and wild-type MLK3. (a) Heatmap of the 445 genes which clearly classify the mutant vs the wild-type, selected on the basis of a 2 log-fold change in expression between cells transfected with P252H mutant and wild-type forms of MLK3, and eliminating genes which were differentially expressed between the Mock and wild-type transfection (red=down-regulated in P252H mutant MLK3, green=up-regulated in P252H mutant MLK3). (b) Illustrates a subset of the heatmap, focusing on genes which are part of the colorectal cancer pathway which is significantly affected (p-value = 0.03). (c) Graphical representation of the colorectal pathway (larger rounded rectangle) and relevant pathways contained within it (smaller rounded rectangles), such as MAPK, WNT, TGF-β, NOTCH and p53, indicating differentially expressed genes (ovals) present in Figure 1, their relationships to each other (solid line indicates a direct protein-protein interaction, T shaped ending for inhibition interactions, arrow ending for activation interactions, and no ending for other known interactions), as well as the direction of the expression change (red = down-regulated in P252H mutant MLK3 transfection, green = up-regulated in mutant MLK3). White Diamond shaped boxes indicate entire gene families, which are significantly affected, genes within the families are shown as colored ovals. Long dashed lines indicate genes, which are present in multiple pathways. (d) Graphical representation of Real-time PCR quantification of mutant vs wild-type MLK3 targets (LEF1, CCND1, BMP6 and FZD10) selected from expression microarrays. . The expression profiles were obtained by comparing all biological replicas from each transfection experiment. Colorectal cancer cell lines were not used in this experiment, as proteins from the MAPK pathway are frequent mutation targets in this type of cancer, and would most likely interfere with the interpretation of the results. The expression profiles were compared to identify genes that were differentially expressed at least 2 log-fold (FDR < 0.01) between wild-type and Mock, as well as between P252H and Mock. A final set of 445 genes was identified which showed significant differential expression only between P252H and wild-type and not between wild-type and Mock (Figure 1a). The most statistically significantly differentially expressed genes are displayed in Figure 1b. The genes identified were significantly enriched (p < 0.05) in several KEGG pathways (Table 1) involved in overall biosynthesis processes, as well as in response in disease relevant processes. Interestingly, the colorectal pathway, which encompasses several relevant pathways such as WNT, MAPK, NOTCH, TGF-β and P53, was significantly over-represented (Figure 1c). These signaling pathways are crucial to maintain intestinal epithelium homeostasis by balancing the rate of proliferation, apoptosis, and differentiation along the crypt-villus axis, and their de-regulation is commonly associated to colorectal cancer initiation and progression [23]. Corroborating our results, a recent study using both in vivo and in vitro approaches showed that MLK3 signaling is important in intestinal mucosal healing and epithelial cell motility [24], therefore implicating MLK3 signaling in the maintenance of intestinal epithelial homeostasis. In order to further validate the microarray data, a set of differentially expressed genes (LEF1, CCND1, FZD10, and BMP6) were selected for validation by real-time PCR in HEK293 cells stably expressing MLK3 wild-type or MLK3 P252H. The results obtained with the microarrays were validated for all genes tested (Figure 1d).
Of particular interest are the alterations that mutant MLK3 induces in the WNT pathway. Activation of canonical WNT signaling through WNT/β-catenin cascade has traditionally been regarded as a critical player in colorectal tumorigenesis [25]. More recently, accumulating evidence supports a role for the non-canonical WNT planar cell polarity (PCP) pathway, a signaling cascade involved in the polarization of cells during tissue remodeling, and cell adhesion and motility, in cancer progression, invasion, metastasis, and angiogenesis [26][27][28]. A more detailed analyzes of our microarray data showed that the expression of several molecular components of the canonical pathway, such as DVL2, LEF1, CCND1 and c-MYC were down-regulated in MLK3 mutant cells, and the expression of DKK, a wellknown negative regulator of canonical WNT signaling [29], was up-regulated (Figure 1c and d). On the other hand, genes encoding two WNT receptors known to act as negative regulators of the canonical WNT/β-catenin signaling cascade and as positive regulators of the PCP pathway, FZD6 and FZD10, were found to be up-regulated in P252H cells (Figure 1c and d). Taken together, the down-regulation of DVL2, LEF1, CCND1 and c-MYC, and the up-regulation of DKK and FZD receptors suggest a role of mutant MLK3 as a molecular switch between canonic and non-canonic WNT signaling. In accordance, it was recently reported that MLK3 reduces the expression of β-catenin/TCF downstream targets by promoting the interaction between βcatenin and KLF4, a known repressor of β-catenin/TCF transcriptional activity [30]. Furthermore, in accordance with a role of PCP in colorectal cancer, FZD10 was recently demonstrated to be up-regulated in colorectal cancers and matched liver metastases, and its overexpression was associated with the activation of noncanonical WNT pathway [31,32].

Conclusion
In conclusion, our results provide an overall view of the expression profile associated with mutant MLK3, and they support the functional role of mutant MLK3 by showing a deregulation of several signaling pathways known to play important roles in the development and progression of colorectal cancer. The results also suggest that mutant MLK3 may be a novel modulator of WNT signaling, and pinpoint the activation of the PCP pathway as a possible mechanism underlying the invasive potential of MLK3 mutant cells. Nevertheless, further studies are required in order to validate this hypothesis in a panel of gastrointestinal cell lines and human primary tumors, to determine if the altered signaling pathways are common to other MLK3 mutations, and to investigate the role of mutant MLK3 in the context of mutant KRAS and BRAF genes.