**1. Introduction**

Long non-coding RNAs (lncRNAs) are transcribed RNA molecules longer than 200 nucleotides that do not code for translated proteins. The human genome is estimated to code for about 58,000 lncRNAs [1], that are being more and more recognized as central players in a plethora of biological processes. They can act as flexible scaffolds providing binding platforms for different proteins, they can interfere with other endogenous RNAs acting as microRNA "sponges" and they can modify chromatin state [2], thus regulating the expression of various proteins. LncRNAs have also been shown to play a role in several layers of epigenetic regulation: they are involved in DNA methylation and demethylation, they can modify chromatin conformation through binding to remodelers [3] and many of them interact with histone modifier enzyme complexes such as PRC2, coREST or SMCX [4].

The physiological processes where lncRNA regulation have been suggested involve cell cycle regulation, epithelial mesenchymal transition (EMT) [5], cancer progression [6] and maintenance of cancer stem cells [5], hypoxia [7] and leukemia [8].

Various lncRNAs are shown to have altered expression levels in different leukemias, resulting in a crucial influence on cellular transformation [9], chromosomal translocation [10], apoptosis [11] and on drug resistance [12]. Accumulating evidence regarding the involvement of lncRNAs in leukemic processes prompted the suggestion to use them as prognostic and classification factors. It was found that lncRNA expression has prognostic value in AML patients [13] and multiple pathways were involved in lncRNA expression, including chromosome organization and trans-membrane receptor protein tyrosine kinase signalling pathway.

As lncRNAs are also considered valuable drug targets, it is essential that the molecular details of their functions are uncovered.

Polycomb repressive complex (PRC2) is the most studied histone modifier that relies on lncRNA binding in its function, being able to bind several lncRNAs including HOTAIR, Xist, RepA, Braveheart, MALAT1 and MEG3 [14]. In vitro experiments revealed that not only EZH2, but other PRC2 subunits are also capable of lncRNA binding [15], thus providing a pattern of binding regions distributed along the surface of the complex. Even though there remain open questions regarding the specificity of the RNA binding by PRC2 [16], it is widely accepted that lncRNA binding plays a defining role in PRC2 targeting and the ensuing gene silencing [14]. It is interesting to note that despite the numerous experimental results that show EZH2 to be an RNA binding protein, it cannot be found in databases that list RNA binding proteins, furthermore no RNA binding site is predicted to be located in the region that is shown to be responsible for the RNA-protein interaction [17].

Apart from PRC2, other histone lysine methyltransferases (HKMTs) or HKMT complex components also appear to bind lncRNAs with a relevant physiological outcome.

LncRNA EZR-AS1 enhances EZR expression through recruiting SMYD2 to the upstream region of its promoter region and elevating the activating H3K4 methylation [18].

G9a interacts with lncRNA PARTICLE to regulate MATA2 expression upon mild irradiation [19]. The interaction was shown using ChIP assay and apart from G9a, the PRC2 subunit Suz12 was also pulled down. In a later experiment, it was found that PARTICLE can also interact with DNA methylase DNMT1 and that it increases H3K27 methylation as well as EZH2 expression. It was suggested that PARTICLE may serve as a functional platform that enables the specific targeting of chromatin modifiers, such as PRC2 [20].

WDR5, a component of the MLL1-4 and SET1a/1b complexes was proven to interact with lncRNAs NeST and HOTTIP with an effect on microbial susceptibility through the enhancement of interferon-γ expression [21]. Further investigation of the WDR5-HOTTIP interaction led to the recognition that lncRNA binding by WDR5 is essential in maintaining embryonic stem cell pluripotency [22]. However, not this work nor any previous studies investigated the possibility that the enzymatic component of the methyltransferase complex may also be capable of lncRNA binding.

The family of mammalian MLL (Mixed Lineage Leukemia) proteins consist of Set1a, Set1b and four MLL proteins, MLL1, MLL2, MLL3 and MLL4. They work in COMPASS-like complexes and catalyze H3K4 mono-, di- or tri-methylation, each complex having different specificity and methylase activity [23]. MLL3 and MLL4 are responsible for the monomethylation of H3K4 at enhancer regions [24] and has been linked to a high number of different cancers. Properly functioning MLL3 and MLL4 act as tumor suppressors [23], therefore mutations affecting their activity or stability can result in cancer development. Despite their central role in several types of cancers, many open questions regarding the regulation of the activity and the targeting of the MLL complexes remain unanswered. The exact molecular details of how MLL3 and MLL4 are targeting enhancer regions [23] as well as the specific molecular effects of the interactions of their different regulatory domains [25] are largely unknown. It is also worth noting that the known structured domains represent only 15–21% of the sequences of MLL proteins, leaving the vast majority of these proteins uncharacterized both structurally and functionally.

In a previous work [26] we suggested that the disordered regions of HKMTs may harbor so far unrecognized interaction sites, adding more layers of the regulation of their activity. Based on the observation that many lncRNAs are involved in processes governed by HKMTs, we hypothesized that lncRNA binding might be one of the functions of these regions.

Since multiple evidence point in the direction that leukemic processes are fundamentally affected by lncRNAs and MLL complexes are involved in this regulation, we concentrated on MLL proteins. Taken the analogy of the PRC2 complex, where more than one complex subunits are capable of lncRNA binding, we aimed at testing the ability of MLL4 to bind different RNA molecules.
