The novel lncRNA BlackMamba controls the neoplastic phenotype of ALK− anaplastic large cell lymphoma by regulating the DNA helicase HELLS

The molecular mechanisms leading to the transformation of anaplastic lymphoma kinase negative (ALK−) anaplastic large cell lymphoma (ALCL) have been only in part elucidated. To identify new culprits which promote and drive ALCL, we performed a total transcriptome sequencing and discovered 1208 previously unknown intergenic long noncoding RNAs (lncRNAs), including 18 lncRNAs preferentially expressed in ALCL. We selected an unknown lncRNA, BlackMamba, with an ALK− ALCL preferential expression, for molecular and functional studies. BlackMamba is a chromatin-associated lncRNA regulated by STAT3 via a canonical transcriptional signaling pathway. Knockdown experiments demonstrated that BlackMamba contributes to the pathogenesis of ALCL regulating cell growth and cell morphology. Mechanistically, BlackMamba interacts with the DNA helicase HELLS controlling its recruitment to the promoter regions of cell-architecture-related genes, fostering their expression. Collectively, these findings provide evidence of a previously unknown tumorigenic role of STAT3 via a lncRNA-DNA helicase axis and reveal an undiscovered role for lncRNA in the maintenance of the neoplastic phenotype of ALK−ALCL.

LncRNAs are transcripts, longer than 200 nucleotides, often display an intron-exon organization, and share close similarities to protein-coding genes. They have pleotropic properties, controlling gene expression, protein stability, localization and function, and cell identity [15]. Unbiased genome-wide analyses discovered thousands of lncRNAs, whose number outnumbers those of protein-coding RNAs. More than 8000 lncRNAs are aberrantly expressed in cancer, making these genes ideal tumor-specific biomarkers and putative targets for therapeutic interventions [16].
Here, by performing deep expression profiling in conjunction with de novo transcriptome assembly, we discovered a panel of previously unknown lncRNAs of ALCL. We focused on a chromatin-associated lncRNA, selectively expressed by ALK − ALCL lymphoma, named Black-Mamba. Mechanistically, BlackMamba is regulated via STAT3 and its expression is required to sustain proliferation and clonogenicity of ALK − ALCL through the transcriptional regulation and the functional control of the lymphoid helicase HELLS. These findings provide new evidence on the mechanisms leading to STAT3-mediated ALCL transformation and foster the implementation of STAT3 target therapies for these lymphomas.

Tissue samples
Fresh and viable cryopreserved cells were isolated from diagnostic/relapsed primary lymphoma biopsies. Diagnoses were assigned according to the WHO classification. Tissues used for NGS analyses were selected for their high tumor cell content (>50%). All studies were approved through institutional human ethics review boards, and patients provided written informed consent in accordance with the Declaration of Helsinki.
Cell growth, colony formation assays, and cell division For cell growth assays, cells were washed with phosphatebuffered saline seeded at 2.5 × 10 5 cells/ml and treated with drugs. Viable cells were counted by trypan blue exclusion.
Colony formation assays were performed as previously described [26].
Cell division was evaluated using 5 μM of carboxyfluorescein succinimidyl ester (CFSE) fluorescent dye following the manufacturer's instructions (ab113853, Abcam). The fluorescence was read by FACSCanto II instrument after 6 or 9 days and data were analyzed using BD FACSDiva Software (BD).

Analysis of mRNA stability
Actinomycin D was used to inhibit nascent RNA synthesis. MAC2A, TLBR-1, and TLBR-2 cells (5 × 10 5 cells/ml) Fig. 1 ALCL samples expressed a restricted set of aberrantly activated previously unknown lncRNAs. a Schematic representation of human samples used to perform directional RNA sequencing. b Bioinformatic pipeline for the discovery of previously unknown lncRNAs. c Density plot for transcript length shows shared pattern between previously unknown and known lncRNAs compared with protein-coding genes, which are much longer. d Coding Potential Score obtained from GENEID shows that previously unknown lncRNAs and known lncRNAs have comparable and lower average coding potential than do protein-coding genes. e Comparing previously unknown lncRNAs against all transcripts with at least two or more exons show a greater number of exons for the protein-coding genes. f Unsupervised analysis of previously unknown lncRNAs profile across normal T-cell lymphocytes and ALCL primary samples. g Flowchart for the discovery of BlackMamba in ALK − ALCL samples. h Schematic representation of locus, structure, and aligned reads of BlackMamba. Numbers represent the length (bp) of exons and introns (gray). Sashimi plots of representative ALK − and AL + ALCL samples were generated by Integrative Genomics Viewer software. The genomic coordinates are measured along the horizontal axis and the RPKM (Reads Per Kilobase per Million mapped reads) values up the vertical axis. qRT-PCR analysis of BlackMamba in a validation set of ALCLs samples (i) and in a panel of cell lines (j).
Twenty-four hours after transfection, cells were harvested and luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) in a GloMax Discover Luminometer (Promega) according to the manufacturer's instructions. For each sample, firefly luciferase activity was normalized on Renilla luciferase activity and transactivation of the various reporter constructs was expressed as fold induction on empty vector (pGL3-basic or pGL3-promoter) activity.

RNA immunoprecipitation (RIP)
RIP was performed as described by Abcam RIP protocol. The precleared lysate was incubated for 2 h with 6 μg of antibodies specific for HELLS (sc-46665, Santa Cruz Biotechnology, Inc.) or with IgG as negative control. All experiments were repeated at least three times.

Quantitative PCR (qRT-PCR)
One microgram of total RNA was reverse transcribed using RT (iScript, Biorad). The amplified transcript level of each specific gene was normalized on CHMP2A housekeeping. ΔΔCt quantification method was used for RT-qPCR analyses.
The list of primers used is provided in Supplementary Table 2.

Statistical analyses
Statistical analyses for identification and the analysis of lncRNAs are described in Supplementary methods. Statistical analyses were performed using GraphPad Prism Software (GraphPad). Statistical significance was determined using the Student's t test. Each experiment was replicated multiple time (>3 up to 6).

ALCL express a large pool of previously unknown lncRNAs
To define lncRNAs preferentially associated with ALCL, we performed high coverage and directional RNA sequencing (RNA-Seq) of 21 ALK + ALCL and 16 ALK − ALCL primary samples. We included normal T-lymphocytes, corresponding to different stages of differentiation and 10 ALCL cell lines ( Fig. 1a and Supplementary Table 3). Firstly, to confirm the appropriateness of pathological samples within our discovery set, we used a 3-gene model classifier (TNFRSF8, BATF3, TMOD1), proven to accurately define ALK + and ALK − ALCL (Supplementary Fig. 1A) [6]. A principal component analysis based on canonical coding gene expression further resolved the discovery cohort into distinct clusters corresponding to normal T-cell and ALCLs ( Supplementary Fig. 1B); the latter group was further stratified into two distinct subgroups, largely represented by ALK + and ALK − ALCL samples (Supplementary Fig. 1C).
Next, we executed a de novo transcriptome analysis of the aligned primary tumor samples [29], which identified 106,014 new transcripts. Applying filtering cutoffs based on transcript length, exon count, and coding potential (based on cross-species comparisons), we discovered 1208 previously unknown ALCL-specific lncRNAs (Fig. 1b). These lncRNAs showed transcript length >200 bp (a canonical lncRNAs feature), a stringent number of exons n = 2 for conserved spliced transcripts (Fig. 1c, d), and they displayed low coding potential compared to canonical coding RNAs (Fig. 1e).
Next, we identified a set of lncRNAs significantly overexpressed in ALCL samples (18/352) by unsupervised analysis (Supplementary Table 4). Even if this set was not able to fully discriminate ALK + and ALK − ALCL, these novel lncRNAs were found to be significantly overexpressed in ALCL compared with normal T-lymphocytes ( Fig. 1f and Supplementary Fig. 1D, E). Then we restricted the analysis to lncRNAs expressed in at least 20% of ALCL samples (five or more samples), this analysis led to the discovery of a pool of 14 lncRNAs (FDR < 0.05) (Fig. 1g). Since the deregulated STAT3 signaling is oncogenic in ALCL [3,12], we then searched for canonical STAT3 binding sites within putative promoter regions of these previously unknown lncRNAs (Fig. 1g). We identified the lncRNA XLOC_043524 only, which we named BlackMamba.
BlackMamba is a nonannotated lncRNA located on q26.3 of chromosome 10, predicted to be transcribed from the minus strand, with an estimated transcript length of 70,292 bp ( Fig. 1) and with half-life of~4 h (Supplementary Fig. 2a-d). BlackMamba is composed of three exons, lacking alternative isoforms. Within the discovery cohort, 9/ 16 ALK − ALCL (56%) and 1/21 ALK + ALCL (4.7%) expressed detectable levels of BlackMamba. The preferential expression of ALK − ALCL was further confirmed in an independent set of 15 ALCLs and 9 PTCLs, and no detectable transcripts were seen in a cross-validation cohort of healthy donor resting and activated PBMCs ( Fig. 1i and Supplementary Table 5). When we extended this analysis to a panel of T-cell lines, we confirmed that BlackMamba expression was consistently detectable in ALK − ALCL and breast implanted associated (BIA)-ALCL cell lines, albeit with variable levels of expression, with TLBR-1 and TLBR-2 expressing the highest levels. Very low transcripts were detected in ALK + ALCL (L82 and Karpas299), systemic and cutaneous ALK-ALCL (FEPD, Mac1), and in cutaneous T-cell lymphoma MJ, while T-ALL lines (CUTLL1, KOPT-K1, and Jurkat), and mycosis fungoides HUTL-78 cells were negative (Fig. 1j).
BlackMamba is a promoter-associated lncRNA transcriptionally regulated by STAT3 To explore the mechanism(s) which regulates the expression of BlackMamba, we investigated the genomic elements responsible for its transcription. H3K4me3 profile by ChIPseq in ALK − ALCL cell lines and patient-derived tumor xenograft (PDTX) (Belli) lines showed a high-density profile within a 2000 bp region, spanning the putative transcription start site (TSS) (Fig. 2a). These data are in agreement with a relatively high level of H3K4Me3 (but not of H3K4Me1) within the same region in ENCODE (in several cell lines), suggesting that BlackMamba is likely to be a promoter-associated gene rather than an enhancer-associated RNA. To elucidate the promoter region of BlackMamba, we then cloned multiple DNA fragments upstream of a luciferase reporter cassette corresponding to a region of 2000 bp (P1-P5), spanning from −811 to +1076 bp of the BlackMamba TSS (Fig. 2a). High luciferase signals were observed with the segment spanning from −459 bp to −90 bp (P4) in both MAC2A and TLBR-2 (Fig. 2b). By cloning P4 in an "enhancer-like" position downstream to the reporter gene, we demonstrated that this segment did not act as an enhancer (Fig. 2c). We next show that the H3K27Ac and H3K4Me3 marks were enriched in two ALK − ALCL lines but not in BlackMamba negative CUTLL1 line (Fig. 2d, e). Likewise, the RNA-Pol II was found to be actively recruited on the P4 element only in the ALK − ALCL cells (Fig. 2f).
ALK + and ALK − ALCLs can be addicted to JAK-STAT signaling pathway, and the loss of STAT3 signaling impairs their growth and survival [3,12]. Because we initially predicted that BlackMamba had a canonical binding site for STAT3, we tested whether STAT3 could bind to the BlackMamba promoter. By STAT3 ChIP -Seq, ALK − cell lines and PDTX displayed multiple STAT3 peaks in close proximity to the P1 or bs_II and bs_III sites of BlackMamba (Fig. 3a). These regions corresponded to accessible, active chromatin sites by ATAC-seq. (Fig. 3a). To test the BlackMamba dependence on the transcriptional activity of STAT3, we tested the level of STAT3 phosphorylation in a panel of cell lines (Supplementary Fig. 3A). Next, STAT3+ ALK − ALCL cells were treated with a JAK1/2 (ruxolitinib) or a selective JAK1 inhibitor (INCB039110), demonstrating that after 24 h of treatment the mRNA expression of BlackMamba was downregulated (Fig. 3b). Consistent with ChIP-seq data, STAT3 was significantly enriched on identified regions of BlackMamba (Fig. 3c). Moreover, the JAKi treatment resulted in the specific inhibition of RNA-Pol II binding, reduced transcriptional activity and the concomitant modulation of H3K27Ac marks (Fig. 3d, e). Lastly, since the pharmacological inhibition of JAK/STAT can elicit a plethora of targets and trigger alternative events, we evaluated the expression of BlackMamba upon silencing of STAT3 by specific siRNA. Having first demonstrated that STAT3-siRNA could effectively reduce STAT3 expression (Supplementary Fig. 3B) and its canonical targets (i.e., CD30, Fig. 3f). we confirmed that the loss of STAT3 was associated with the downregulation of Black-Mamba (Fig. 3f).
Remarkably, although the expression of BlackMamba was barely detectable in ALK + ALCL, the selective inhibition of ALK signaling (crizotinib) led to the downregulation of BlackMamba (Supplementary Fig. 3C).
Collectively, these data demonstrate that STAT3 regulates the expression of BlackMamba, independently from its upstream activators.

BlackMamba is required for active proliferation and clonogenicity of ALK − ALCL cells
To test the biological properties of BlackMamba, we silenced its expression using two different approaches: a transient siRNA transfection and a doxycycline inducible shRNA. Because BlackMamba is a large gene (~70 kb), we targeted different regions ( Supplementary Fig. 4A and Supplementary Table 2). We found two independent shRNAs that effectively reduced its expression after doxycycline induction in MAC2A (−50% for each shRNAs) without affecting top-scoring off-targets significantly ( Supplementary Fig. 4B). Both shRNA#2 and #6 could effectively knockdown (KD) the lncRNA in TLBR-2, although with different potency (−30%, shRNA#2 and −60%, shRNA#6) (Fig. 4a). Next we studied the cellular localization of BlackMamba at steady state, demonstrating it was enriched in the nucleus and strongly associated to the chromatin fraction, suggesting a putative role in chromatin organization and gene expression regulation ( Fig. 4b and Supplementary Fig. 4C).
Functionally, the KD of BlackMamba resulted in a dosedependent cell growth inhibition (Fig. 4c and Supplementary Fig. 4D, E), in the absence of an increased rate of apoptosis ( Supplementary Fig. 4F). This phenotype was reproducibly detected in shRNA#6-treated ALCL cells, although significant shRNA#2 mediated changes were observed only in MAC2A cells, which expressed lower levels of mRNA ( Supplementary Fig. 4G). Next we demonstrated a delayed cell division in BlackMamba KD cells ( Fig. 4d and Supplementary Fig. 4H, I), and an impaired ALCL lymphoma colony formation (Fig. 4e). Conversely, the growth and survival in control K562 (chronic myeloid leukemia) and CUTLL1 KD cell lines were not affected (Supplementary Fig. 4J). Interestingly, the cytological inspection of May-Grunwald Giemsa stained ALK − ALCL cells upon BlackMamba KD showed an increased number of polynucleated cells (Fig. 4f and Supplementary Fig. 4K), with a polyploid DNA content by partial FISH-based karyotyping (Fig. 4g).
Lastly, since cytokinesis requires an appropriate cytoskeleton organization, we investigated the cytoskeleton architecture of BlackMamba silenced cells. As actin filaments are major components of the contractile structure that guide cytokinesis in eukaryotes [30], we examined the cellular cytoskeleton using phalloidin staining of actin filaments. We found that after BlackMamba KD, cells lose actin polarization and have a displacement of filaments from their membrane localization, supporting its role in cytoskeleton reorganization and cytokinesis (Fig. 5a). This hypothesis is supported by the transcriptional changes observed after RNA sequencing in Black-Mamba KD (59 downregulated genes and 61 significantly upregulated genes, ≥2-fold and FDR < 0.05, Fig. 5b and Supplementary Table 6) where top-scored genes modulated (i.e., RGS1, CCL22, CCL17, PAK2, RHOU) epitomize by the modulation of actin cytoskeleton, integrinmediated cell adhesion, and focal adhesion genes (Fig. 5c, d and Supplementary Fig. 5A) Overall these data suggest that BlackMamba is involved in the maintenance of appropriate completion of cytokinesis.
BlackMamba regulates the transcription of the lymphoid-specific DNA helicase HELLS Chromatin-enriched lncRNAs are spatially correlated with transcription factors [31], can act as cell type-specific activators of proximal gene transcription [32] and chromatinassociated lncRNAs (such as XIST or KCNQ1ot1) can influence local chromatin organization, leading to in cis transcriptional repression of genes within large genomic regions [33].
To test whether BlackMamba could operate according to this model, we correlated its expression with transcriptional factors and chromatin-remodeling genes (16 ALK − ALCL and 21 ALK + ALCL samples). The lymphoid-specific helicase (LSH) HELLS, the SET domain containing protein PRDM13, and the polycomb repressive complex 2 histonelysine N-methyltransferase EZH2 were found to be positively correlated with BlackMamba. Conversely, the homeobox protein HHEX showed a negative correlation (Fig. 5e). To test whether this association was directly linked to BlackMamba expression, we quantified the mRNA levels of HELLS and HHEX in inducible shRNA BlackMamba ALCL cells. HELLS expression was consistently downregulated upon BlackMamba silencing, while HHEX was upregulated ( Fig. 5f and Supplementary  Fig. 5B). No consistent changes were observed for EZH2 ( Supplementary Fig. 5C) while PRDM13 mRNA was undetectable (data not shown). In line with the gene expression changes, H3K4me3 and H3K27me3 underwent chromatin reorganization, with a significant reduction of H3K4me3 binding on HELLS promoter and a parallel loss of H3K27Me3 on HHEX promoter after silencing (Fig. 5g).
Being also HHEX and HELLS the only TFs located on the same chromosome of BlackMamba, we hypothesed that BlackMamba could regulate gene expression in cis.
To enforce this concept, we quantified the mRNA levels of several neighboring genes spanning 1 Mb from Black-Mamba locus. After BlackMamba KD, 5/6 genes were concordantly downregulated in MAC2A and TLBR-2 cell lines ( Supplementary Fig. 5D).
These data support the model which predicts that BlackMamba regulates in cis the expression of genes located on the same chromosome.
BlackMamba interacts with HELLS to control the BlackMamba-dependent transcriptional program HELLS, also known as LSH or proliferation-associated SNF2-like (PASG), belongs to a large family of SNF2 chromatin-remodeling ATPases. HELLS is critical to the normal development and survival of lymphoid cells [34] and regulates chromatin organization and gene expression [35,36]. HELLS mutations or misregulated expression were seen in several cancers and some cases of ICF syndrome [37].
Conversely, the ectopic expression of HELLS effectively restored the baseline expression of these genes in Black-Mamba KD cells (Fig. 6d, e).
It is known that lncRNAs control gene expression by recruiting chromatin-remodeling complexes to target promoters or enhancers, thereby influencing histone modifications and chromatin accessibility [40]. Since the chromatin-remodeling properties of HELLS can be mediated by the lncRNA HOTAIR [41], we reasoned that BlackMamba might interact with HELLS to mediate its recruitment to target genes. To prove a direct association between HELLS and BlackMamba, we used RNA immunoprecipitation [3] and showed that HELLS binds to two distinct regions of BlackMamba at the 3′-end of the lncRNA (Fig. 6f). No readout was seen in HELLS KD cells. Next, we proved that HELLS was preferentially bound to target gene promoters only in BlackMamba-positive ALK − ALCLs independently from HELLS basal expression level ( Fig. 6g and Supplementary Fig. 6C).
Collectively, these data demonstrate that BlackMamba-HELLS could be a part of the regulatory complex that occupied loci to coordinate ALK − ALCL transcriptional program.
HELLS is required for the maintenance of the neoplastic phenotype of ALK − ALCL To determine the contribution of HELLS and BlackMamba in the maintenance of ALCL phenotype, we first investigated the cell growth capacity of HELLS KD cells. Indeed, the loss of HELLS led to an impaired cell growth ( Fig. 7a and Supplementary Fig. 6D), reduced duplication rate (Fig. 7b), and clonogenicity ( Fig. 7c and Supplementary  Fig. 6E), a phenotype associated with an increased number of polynucleated cells, phenocopying the BlackMamba KD cells (Fig. 7d and Supplementary Fig. 6F).
Next we proved that the overexpression of HELLS could counteract the phenotype associated with the KD of BlackMamba, as the growth impairment of TLBR-2 cells expressing inducible shRNA against BlackMamba was effectively rescued (Fig. 7e) a finding associated with the mitigation of the polynucleated phenotype of the Black-Mamba KD (Fig. 7f).
Overall, these findings demonstrate that HELLS is an essential downstream mediator of BlackMamba and that BlackMamba-HELLS axis represents a vulnerability of ALCL cells.

Discussion
Although a more complete genomic annotation of ALK − ALCL is emerging [42,43], the mechanistic modalities of action remain elusive even for known recurrent defects [4,42].
Here we describe a new chromatin-associated lncRNA, named BlackMamba, preferentially expressed in ALK − ALCL. Mechanistically, STAT3 regulates the expression of BlackMamba and its expression is required for ALK − ALCL neoplastic phenotype. This is achieved mainly through the action of the LSH, HELLS (Fig. 7g) and the transcriptional regulation of genes controlling G-protein and cytoskeletal organization, cell migration, tissue recruitment, and inflammation.
While emerging evidence have shown that lncRNAs are pathogenetic in human B-cell hematological neoplasms [44], little is known in mature T-cell neoplasms [23][24][25]45]. In this study, we identified 1208 previously unknown lncRNAs linked to normal or neoplastic T-cells and among them, 18 lncRNAs were largely restricted to ALCL. Accordingly, a third of these new lncRNAs were co-shared by both ALCL subtypes, supporting the notion that both ALK + and ALK − ALCL can share some communalities [46]. Deregulated activation of STAT3 is a hallmark of many human cancers [47]. ALK + ALCL and a subset of ALK − ALCL have been proven to be addicted to the JAK/STAT signaling pathway [3,48] and BIA-ALCL display a constitutive JAK/STAT deregulation [49]. Here, BIA-ALCL cell lines, as well as PDTX, were found to express high levels of BlackMamba. In ALCL, the deregulation/activation of JAK/STAT pathway is mediated by gene fusions, somatic mutations [50], and loss of negative regulators [7,51]. These lead to a distinct transcriptional program associated with defined pathological entities [6,46]. LncRNAs have been also shown to be linked to STAT signaling by modulating metabolic pathways [52] and conversely, STAT3, controlling the expression of lncRNAs, can regulate cell differentiation [17].
BlackMamba is a target of STAT3. Both siRNAmediated and pharmacological inhibitions of JAK/STAT3 pathway profoundly repressed BlackMamba expression demonstrating its contribution in cell growth and clonal expansion of ALCL. Interestingly, loss of the BlackMamba Fig. 6 BlackMamba controls the recruitment of HELLS on multiple target promoters. a qRT-PCR analysis of HELLS in ALK − ALCL cell lines expressing pLKO-shRNA#2 against HELLS (48 h after doxycycline induction). b Western blot shows the reduction of HELLS in cells expressing pLKO-shRNA#2 HELLS after 3 days of doxycycline induction. c qRT-PCR analysis of a panel of genes after 2 days of doxycycline induction in TLBR-2 expressing pLKO-shRNA#2 HELLS. Each data point represents the mean ± SEM. (n = 3). *p ≤ 0.05; **p ≤ 0.01. d qRT-PCR analysis of TLBR-2 cells coexpressing shRNA#6 BlackMamba and pCDH-HELLS-HA vectors (6 days after doxycycline induction). e Western blot shows the overexpression of ectopic HELLS-HA in TLBR-2 overexpressing pLKO-shRNA#6 BlackMamba. f RIP assay for HELLS in TLBR-2 expressing pLKO-shRNA#2 HELLS (2 days). Fold enrichment is relative to IgG (average of six independent experiments ±SEM; *p ≤ 0.05; **p ≤ 0.01). g ChIP qRT-PCR detection of HELLS antibody on several BlackMamba target gene promoters in a panel of T-cell lymphoma lines. MLL1, PD5S4 were used as positive controls (average of six independent experiments ±SEM; *p ≤ 0.05; **p ≤ 0.01).
expression was linked to a unique phenotype characterized by an increased number of polynucleated cells with a balanced chromosomal enumeration possibly linked to key genes regulating cytoskeleton and cell motility. As STAT3 can regulate cell migration via RAC1 and Rho [53,54], our data provide a new layer of complexity demonstrating a new role for STAT3 in cell shape and cytoplasmic partition via the axis of BlackMamba-HELLS. This model is supported by phenotype seen in HELLS KD cells, arguing that HELLS represents a critical downstream effector of BlackMamba.
Among BlackMamba targets, we found HELLS (16,550,000 bp from the locus of BlackMamba). HELLS controls T-cell growth [34], regulates DNA methylation [55], and modulates the epigenetic states at specific enhancers of key cell cycle regulators [56]. In cancer cells, HELLS sustains glioma stemness and through the interaction with E2F3 controls cell proliferation of prostate cancer cells [38,57]. By interacting with the epigenetic silencer factor G9a, HELLS represses gene transcription [58]. Remarkably, a recent genome-scale CRISPR-Cas9 screen has shown cancer dependencies to DNA helicase and identified ATP-DNA helicases as promising new synthetic lethal targets in tumors [59]. Our work extends these findings providing a new lncRNA-dependent mechanism controlling the recruitment of HELLS on chromatin sites and its expression in lymphomas.
Our data provide novel insights into the transformation of ALCL via the untapped role of a lncRNA. Collectively, the findings further support the design of target therapeutic strategies to pharmacologically ablate/inhibit the expression of STAT3 and encourages novel discovery programs for the selection of compounds which could impair STAT3downstream effector elements like HELLS. Lastly, since HELLS is expressed in many human tumors and plays a relevant role in DNA repair and genomic stability of cancers, its pharmacological inhibition represents a viable therapeutic strategy in many human cancers.

Data availability
BlackMamba sequence has been deposited in GenBank database with the accession number MN902222.