FXYD Domain Containing Ion Transport Regulator 3 (FXYD3) is Over-Expressed in Germinal Centre Derived Aggressive Lymphomas and Plasma Cell Myeloma

Derived Aggressive Lymphomas and Plasma Cell Myeloma. abstract FXYD3 is a Na/K-ATPase regulator which has been recently associated with different cancers development and progression; consequently, FXYD3 has been proposed in those as a potential therapeutic target. By contrast, no data are available concerning FXYD3 expression in hematological malignancies. In this study we aimed to assess FXYD3 gene expression in a large panel of B-cell derived lymphoid malignancies and to evaluate possible clinic-pathological correlations. Normal B-cell subsets served as control. We found that FXYD3 gene was not significantly modulated in normal B-cells. By contrast, BL, DLBCL, PMBCL, and PCM presented with a significant over-expression of the gene when compared to their cellular counterpart (p<0.0006). Interestingly, tumors characterized by higher FXYD3 expression presented with a significant enrichment in specific cellular functions and pathways, including NFkB pathway, WNT/B-catenin signaling, and MYC network, while FXYD3 levels also turned out to be directly related to PRDM1/BLIMP1. Finally, higher FXYD3 expression was not significantly associated with patients’ survival in PCM and DLBCL, though a trend in favour of patients with lower expression was recorded in DLBCL cases. In conclusion, we unveiled FXYD3 gene over-expression in specific non-Hodgkin lymphoma subtypes and PCM, providing evidences of its involvement in their pathobiolo-gy. Future studies are needed to define its precise role.


Introduction
The FXYD domain containing ion transport regulator 3 (FXYD3) belongs to the FXYD protein family, originally identified by Sweadner and Rael [1] basing on sequence similarity, and accounting in mammal for at least seven members. All these proteins have a signature sequence of six highly conserved aminoacids, comprising the FXY motif in the NH2-terminus and two glicines and one serine in the transmembrane domain. Despite that, the NH2 and COOH-termini remain variable among different members of FXYD proteins [2].
The major role of FXYD proteins consist in their ability to interact with Na/K-ATPase, adapting its activity to changing physiological demands [2]. Despite that, FXYD proteins seem to be involved in a higher number of physiological process, most of which have not been completely elucidated. One of the most interesting proteins of the FXYD family, that has been connected to tumor development and progression, is mammary tumor marker 8 (MAT-8, or FXYD3).

FXYD3 was first identified in murine breast tumors initiated by
NEU or RAS, and functional characterization showed that FXYD3 is capable of inducing ion-specific conductance when overexpressed in Xenopus oocytes.
While in normal tissue FXYD3 is mainly expressed in the colon, stomach, uterus and skin, it was also found to be expressed in breast tumor, in prostate tumor and in colorectal tumor. Noteworthy, it has been shown that siRNA-mediated inhibition of FXYD3 expression causes a reduction of cell proliferation in prostate cancer cell lines, so FXYD3 is probably also directly or indirectly involved in cell proliferation and it does not simply modulate the activity of Na/K-ATPase. Different researchers are now attempting to determine the importance of FXYD3 in numerous type of cancers. In this regard, Loftas and Colleagues demonstrated that strong expression of the membrane protein FXYD3 was associated with infiltrative tumor growth and a reduction in tumor necrosis in rectal cancer, while tumors with weak FXYD3 expression had better prognosis after radiotherapy [3]. Another study proposed that a strong FXYD3 expression might help in protecting the Na/K-ATPase from the high level of oxidative stress characteristic of many tumors and also induced by cancer treatment in breast cancer cell lines [4]. It is also interesting to notice that, regarding breast cancer, the administration of estrogen and tamoxifen has been connected with the up-regulation of FXYD3 [5]. Very recently, FXYD3 has been also proposed as potential target for innovative therapeutic approaches in prostate cancer. In this study, we aimed to determine the expression and potential role of FXYD3 in B-cell derived human lymphomas.

Gene Expression Analyses
Gene expression analysis was carried out as previously reported concerning supervised, unsupervised and gene set enrichment analyses [12,13,[15][16][17][18][19]. Briefly, the expression value of each selected gene was normalized to have a zero mean value and unit standard deviation. The distance between two individual samples was calculated by Pearson correlation with the normalized expression values. Unsupervised clustering was generated using a hierarchical algorithm based on the average-linkage method.
To perform the supervised gene expression analysis, we used Gene Spring GX 12 (Agilent Technologies, Santa Clara, CA, USA).
Differentially expressed genes between different groups were identified using a two-tails Student t-test and adjusted Benjamini-Hochberg correction for false discovery rate, applying the following filtering criteria: p-value <0.05, and fold change>2.

Statistical Analysis
Statistical analyses were performed using IBM SPSS Statistics 20.0 and Prism (GraphPad softwares, USA). ANOVA and unpaired T-tests were used for continuous variables examination. When a sample size was less than 10 cases in at least 1 group a nonparametric (Mann-Whitney) test was used to analyze the GEP data to compare FXYD3 expression in different subgroups. Survival analyses were performed by Kaplan-Meier method. Two-sided tests were used in all calculations. The limit of significance for all analyses was defined as p<0.05. Survival analyses for DLBCL and PCM patients were carried out using the GSE34171 and GSE24080 datasets, respectively.  Table 1.

FXYD3 Expression Depends on PRDM1/BLIMP1
In order to assess whether FXYD3 gene expression was related to any specific transcription factor known to be involved in B-cell tumorigenesis, we studied the expression of BCL6, MYC, IRF4, PRDM1/BLIMP1, and STAT3 and correlated it to FXYD3. We found that PRDM1/BLIMP1 expression was significantly correlated with Again, the correlation was highly significant (p=0.025; Figure 6F).

FXYD3 Over-Expression is not Related to Clinical Aggressiveness And Survival in DLBCL and PCM
Last, we sought to determine whether the over-expression of

Discussion
Our study aimed to determine for the first time the expression of FXYD3gene in non-neoplastic B-cell subsets as well as in different types of B-cell neoplasms. We observed a significant over-expression of the gene in aggressive germinal center derived lymphomas (namely, BL, DLBCL, and PMBCL), as well as in PCM.
This expression pattern was definitely interesting. In fact, on the one hand it did not directly reflect tumor aggressiveness, since MCL showed pretty low values while PCM showed rather high ones. On the other hand, it did not reflect the cellular origin, since neither FL nor HL, two germinal center derived neoplasms, did parallel the expression patterns observed in the other GC-derived malignancies. Therefore, although FXYD3 expression tended to be higher in highly proliferating lymphomas (BL, DLBCL, and PMBCL), its over-expression, documented also in PCM, a tumor generally characterized by a lower proliferating rate, did not appear to be related to a single feature but rather to be referred to specific single entities ( Figure 8). We then sought to understand whether FXYD3 over-expression was associated with the activation of specific cellular functions. Grippingly, FXYD3 expression was found to be directly related to that of PRDM1 [22]. This gene encodes for BLIMP1 protein, a transcription factor necessary for plasma cell differentiation [23]. In fact, while PAX5 is critical for the maintenance of B cell identity [24], the expression of BLIMP1 leads to the repression of both PAX5 and BCL6and it is capable of triggering plasma cell differentiation by ensuring that plasma cell cannot return to an earlier developmental stage [25]; BLIMP1, also, has been correlated with the induction of genes that are involved in the immunoglobulin secretion [26]. Indeed, our correlation between FXYD3 and PRDM1 is consistent with the evidence that FXYD3 expression was not significantly associated with the clinical outcome in terms of overall survival in PCM. In fact, PRDM1 is always expressed in this tumor, reflecting its physiological role in plasma cells. By contrast, PRDM1 expression was correlated with poor prognosis in ABC-type DLBCL [27]; however, we failed to document significant differences among DLBCL cases based on FXYD3 expression.
The main limitation of this study probably relies on the fact that all analyses were conducted by GEP only and were not supported by an additional method, due to the lack of available pathological material. In this regard, a couple of consideration must be taken into account. Firstly, GEP, at least when performed by Affymetrix microarrays, is now considered to be a robust and reliable method to measure and quantify genes expression in a given sample, with substantial correspondence (despite an inferior sensitivity) to quantitative PCR methods [28]. Therefore, a validation at protein level would be the most appropriate.
However, a number of studies previously sought to compare GEP and IHC (the unique proteomic technique feasible in formalin fixed paraffin embedded tissue samples) [29,30] demonstrating a high degree of consistency between the two methods (around 77-87%) that made such validation approaches even unnecessary in many applications [31,32]. Moreover, IHC might even fail in detecting subtle but still biologically significant differences caught at gene expression level [33]. It would be interesting, instead, to evaluate in future independent series whether GEP or IHC will be able to stratify lymphoma and/or myeloma patients with different clinicopathological features based on FXYD3 expression [34].

Conclusion
In conclusion, we unveiled the over-expression of FXYD3 gene in some diseases (BL, DLBCL, PCM, and PMBCL), and connected its expression to the activation of specific cellular pathways and a well-known transcription factor. On the clinical ground, FXYD3 over-expression was not significantly correlated to overall survival in DLBCL and PCM, though different attempts might be warranted.
Finally, based on the molecular profile associated to FXYD3 gene over-expression, the specific association of this molecule with clinical response to epigenetic modifiers might be tested.