IGLL5 controlled by super-enhancer affects cell survival and MYC expression in mature B-cell lymphoma

Highlights • Immunoglobulin light like polypeptides 5 gene (IGLL5) is expressed in mature-B-cell lymphoma lacking IGλ protein.• IGLL5 expression is controlled by super-enhancer, predicted by in-silico analyses.• Decreased expression of IGLL5 results in apoptosis with down-expression of MYC.


Introduction
Mature B-cell lymphoma is the most common among hematopoietic tumors.Accumulated evidence indicates that acquired genetic alterations are required for the disease development, and novel findings continue to identify molecular pathogenesis of this disease.In the past decade, genome-wide chromatin immunoprecipitated sequencing (ChIP-seq) has revealed that large active enhancers, called superenhancers (SE), are scattered in the genome, and that tumor cellspecific SEs have been identified [1].Most of the genes controlled by SE have been shown to play important roles in tumorous cells as well as in normal cells [1].A high-ranked SE is predicted at the proximity of immunoglobulin lambda like polypeptide 5 gene (IGLL5) in diffuse large B-cell lymphoma or multiple myeloma [2,3].IGLL5 might therefore have some function in the tumorigenesis of B-cell malignancies.
IGLL5 consists of three exons.The first exon is unique for IGLL5 and located at ~6 kb upstream of the immunoglobulin lambda (IGλ) joining region 1 segment (Jλ1) [4,5].The second and third exons are Jλ1 and IGλ constant region 1 segments (Cλ1), respectively.IGLL5 mRNA therefore contains Jλ1 and Cλ1 nucleotide alignment.Authentic IGλ requires somatic rearrangements to express as mRNA encoding IGλ protein; but, IGLL5 does not [4,5].Two transcriptional isoforms of IGLL5 have been identified, the isoforms 1 and 2 encode 214 and 139 amino-acids protein, respectively [6].Both proteins belong to the immunoglobulin superfamily.Although many immunoglobulin superfamily proteins play significant roles in immunity, cell adhesion, and signaling [7], the function of IGLL5 has not been well explored.
IGLL5 has recently been reported as one of the most frequently mutated genes in mature B-cell lymphoma and multiple myeloma [8,9,10].The frequently-encountered mutated genes seen in B-cell tumors, such as BCL2 or EZH2, have some functions in B-cell malignancies.In addition, IGLL5 has also been shown to be a target of chromosome translocation in mature-B-cell lymphoma or multiple myeloma [10,11].Generally, a target gene of chromosome translocation in hematopoietic tumors plays important roles in the tumorigenesis.These observations on IGLL5, including the proximity to SE, the genetic mutation, and the target gene of translocation, suggest that IGLL5 might have some role in B-cell tumorigenesis.However, the biological role of the IGLL5 in B-cell neoplasms remains unclear.
We found that two mature B-cell lymphoma cell lines expressed IGLL5.Using these cell lines, we attempted to analyze the biological role of IGLL5 in mature B-cell lymphoma using small interfering RNA (siRNA) methods.

Cell lines
Four mature B-cell lymphoma cell lines (MD901, FL218, WILL1 and WILL3), one myeloma cell line (RPMI8226) and one myeloid cell line (K562) were used in this study.MD901 and FL218 are positive for cell surface IGκ and IGλ protein, respectively.WILL1 produces IGκ protein, while WILL3 lacks surface IGs and produces no IG protein.RPMI8226 secretes IGG/IGλ protein.The characteristics of the four mature B-cell lines are summarized in the supplemental table.

Reagents and transfection
A bromodomain and extra terminal domain inhibitor, JQ1, was used as an SE inhibitor [2,3].The subjected cells were incubated in culture medium containing JQ1 at final concentration 0.75 μM for 14 h or 24 h.

Northern blot analysis
Total RNA was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA).One μg of total RNA except RPMI8226 was subjected to the analysis.For RPMI8226, 0.5 μg of total RNA was loaded.Northern blot analyses were performed using the DIG Northern Starter Kit according to the manufacturer's protocol.(Roche Diagnostics, Mannheim, Germany).Chemiluminescence signals were detected on X-ray films.A 0.7 kb of IGλ3 constant (Cλ3) and a 2.2 kb of MYC cDNA fragments were obtained from the pC-lambda 3 and pSPT-myc cDNA, respectively (both were purchased from JCRB, Ibaraki, Japan).A 464 bp of DNA fragment encompassing IGLL5 exon 1 was amplified by PCR using IGLL5F1 (5′-GTAGATGCCCCTCTGGGAGA-3′) and IGLL5R1 (5′-ACCTGGGGTCTGCTCTCTGG-3′) primers.These three DNA fragments were subcloned to appropriate plasmids that have Sp6 or T7 promoter at the cloning site for reading by RNA polymerase.Digoxigenin (DIG)- labelled RNA probes were synthesized using the DIG RNA Labeling Kit (SP6/T7) (Roche Diagnostics, Mannheim, Germany).The DIG-labelled RNAs derived from the respective DNA fragments were designated as Cλ, MYC and IGLL5 probes corresponding to the parental plasmids.The Cλ probe hybridizes to all Cλ segments because of its high homology.

Apoptosis assay
The cells were analyzed 24 h after electroporation The subjected cells were stained with annexin-V and propidium iodine using the Annexin-V-FLUOS staining kit according to the manufacturer's protocol (Roche Diagnostics, Mannheim, Germany).The cells were then analyzed by flow cytometry (FACS, Beckton-Dickinson, Franklin Lakes, NJ).FACS data were analyzed by FlowJo software (Beckton-Dickinson, Franklin Lakes, NJ).Dead cells were counted as the sum of early apoptotic, late apoptotic and necrotic cells.Four independent analyses were performed, and statistical significance was examined by Student's t-test method.

IGLL5 mRNA expressed in mature B-cell lines lacking IGλ production
We detected two transcripts containing Cλ segment in three IGλ negative cell lines (Fig. 1A).Owing to lack of IGλ protein synthesis, we thought that the transcripts do not encode IGλ protein.To determine the origin of the transcripts, we performed 5′RACE.The 5′RACE using WILL1 cDNA yielded multiple PCR products.The largest ~660 bp fragment (Fig. 1B) was cloned and sequenced.The nucleotide alignments (Fig. 1B) showed homology to the IGLL5 transcript by the BLAST search program.RT-PCR was then performed using IGLL5F2 and IGLL5R3 primers to validate IGLL5 expression.The RT-PCR detected IGLL5 transcripts in WILL1 as well as WILL3 (Fig. 1C).Direct sequencing of the PCR products revealed that both IGLL5 mRNAs are identical to IGLL5 transcript isoform 2 lacking the second exon (Jλ1) (Figure S1).Although 5′RACE detected the IGLL5 transcript containing exon 2 in WILL1, (Fig. 1B), the main IGLL5 transcript is considered to be isoform 2. The IGLL5 mRNA is predicted to produce 139 amino-acids protein.Northern analysis confirmed that both of the cell lines express IGLL5 mRNA (Fig. 1D).

Down-expression of IGLL5 mRNA by JQ1 and Cλ siRNA
A high-ranked SE was predicted in previous reports near the IGLL5 in diffuse large B-cell lymphoma and myeloma cells [2,3].We also confirmed a large active Histone H3 at lysine 27 acetylation (H3K27Ac) region, suggesting SE, near IGLL5 in mature B-cell lines using a published data-base, ChIP-Atlas (https://chip-atlas.org/) (Figure S2).This in silico data suggests that IGLL5 expression is controlled by SE.To confirm regulation of IGLL5 by SE, we examined expression levels after JQ1 treatment by Northern blot analyses using the Cλ and IGLL5 probes recognizing the third and first exons.The IGLL5 mRNA expression in WILL1 and WILL3 was more sensitive to JQ1 treatment than the Cλ mRNA expression encoding IGλ protein in FL218 or RPMI8226 (Fig. 2A  and B).
Next, we challenged knockdown of the IGLL5 mRNA using a previously reported siRNA targeted Cλ region, si[IGLC CR ], because the Cλ1 is the third exon of IGLL5.The si[Cont] and si[IGLC CR ] were introduced into cells by electroporation.Although the efficacy differed in each cell line, the signals derived from the Cλ probe decreased at 14 h after si [IGLC CR ] transfection (Fig. 2C).RPMI8226 showed the most significant reduction of signals among the subjected cells.Signals derived from the IGLL5 probe were also decreased after transfection in WILL1 and WILL3 cells (Fig. 2D) .This result suggested that si[IGLC CR ] can decrease IGLL5 mRNA.

IGLL5 knockdown induces apoptosis along with decreased expression of MYC
In our pilot study, we observed that knockdown using si[IGLC CR ] resulted in inhibition of proliferation in IGλ negative cell lines.We then asked whether apoptosis occurred after si[IGLC CR ] treatment.Dead cells significantly increased after si[IGLC CR ] treatment compared with control siRNA in WILL3 (Figs. 3A and 3B).Although statistical significance was not confirmed, WILL1 showed increase of dead cells after si[IGLC CR ] treatment in each paired experiment (Figure S3).
Given that both cell lines have chromosome translocations involving the MYC locus and show deregulated expression of MYC, we focused on MYC expression after si[IGLC CR ] treatment.Both cells showed decreased MYC expression 24 h after transfection (Fig. 3C).MYC down-expression appeared late after IGLL5 down-expression (Fig. 3D).

Discussion
In this study, we detected IGLL5 expression in two mature B-cell lines lacking IGλ protein.We then found that the expression is controlled by SE as previously predicted.Finally, IGLL5 down-regulation resulted in cell death and decreased expression of MYC.
Five IGLL genes have been recognized by the Human Gene Nomenclature Committee [13].Among the five genes, IGLL1 and IGLL5 encode immunoglobulin superfamily proteins, and the remaining are pseudogenes.IGLL1 functions as a surrogate IG molecule expressed in an early differential stage of B-cells [14].IGLL5 mRNA expression has been reported in mature B-cells exhibiting IGκ protein [5].We also detected IGLL5 mRNA in two mature B-cell lymphoma cell lines, which lacked IGλ protein.Most of the immunoglobulin superfamily proteins function in immunity, cell adhesion or signal transduction [7], but the biological characteristics of the IGLL5 have not been clarified.
Previous ChIP-Seq analyses predicted that SE exists near IGLL5 in diffuse large B-cell lymphoma and multiple myeloma [5,6].We found that IGLL5 mRNA was down-regulated by JQ1, which indicates that IGLL5 is controlled by SE.The Cλ transcripts encoding IGλ protein seemed to not be affected by JQ1 (see FL216 and RPMI8226 in Fig. 2A).The protein coding IGλ allele undergoes Vλ-Jλ-Cλ recombination which results in the deletion of intervening sequence.The IGLL5 locates between Vλ and Jλ regions; thus, the SE of IGLL5 might be deleted during Vλ-Jλ-Cλ recombination in the IGλ protein producing cells.Although the si[IGLC CR ] was reported to target all Cλ segments [12], the decreased level of Cλ mRNA differed in each cells (Fig. 2C).Different efficacy of Cλ down-regulation was observed in a previous study using myeloma cell lines [12], so the difference might be owing to the complexity of the transcriptional regulation of the IGλ.
Zhou et al. reported that knockdown of Cλ causes apoptosis in IGλ producing myeloma cells resulting from endoplasmic reticulum (ER) stress due to accumulation of uncoupled IG heavy chain protein in the ER [12].In this study, we used the same siRNA targeting Cλ, and observed apoptosis after transfection.The subject cells in our experiments lack IGλ proteins.Therefore, the apoptosis seen in this study should not depend on ER stress.We noted the decreased expression of MYC after transfection of si[IGLC CR ].MYC contributes to tumorigenesis of various malignancies and its expression is regulated by diverse mechanisms.We observed that MYC expression was decreased late after IGLL5 down-expression (Fig. 3D).This time lag might indicate that IGLL5 down-regulation affects MYC expression via intervening processes.Comprehensive gene expression profiles might explore the detailed link between IGLL5 and MYC expression.
We observed apoptosis after si[IGLC CR ] transfection to IGLL5 expressing mature-B cell lines.However, WILL1 did not show a significant increase in dead cells.Moreover, although WILL3 showed a statistically significant increase in cell death after siRNA treatment, the increase ratio was less than 10 %.The siRNAs may therefore have partial effects on IGLL5 mRNA expression.One of the reasons for the partial effects is presumably due to transient efficacy, as suggested in Fig. 3D.To obtain sufficient efficacy, other efficient gene silencing methods, such as lentiviral vector or gene editing methods, should be used in future experiments.
In conclusion, our results suggest that IGLL5, an immunoglobulin superfamily protein encoding gene, is controlled by super-enhancer, and might contribute to cell survival affecting MYC expression.Further study with larger sample sizes should need to explore biological and clinical significance of IGLL5 in B-cell malignancies.

Declaration of competing interest
None.

Financial supports
This work was supported grants (KAKENHI 21K16248 and 19K08821) from the Japan Society for the Promotion of Science and a grant from Takeda Science Foundation to HH.

Fig. 3 .
Fig. 3. Apoptosis and MYC down-expression after si[IGLC CR ] transfection.A. Apoptosis assay using annexin-V (AV) and propidium iodine staining (PI).The subject cells were transfected to si[Cont] or si[IGLC CR ] by electroporation.The subject cells were cultured in RPMI1620/10 %FBS for 24 h after electroporation, then, apoptosis assays were performed.Lower right (LR), upper right (UR) and upper left (UL) quadrant represent early apoptotic cells, late apoptotic, and necrotic cells, respectively.Dead cells were counted as a sum of cells in LR, UR and UL quadrants.B. Increase in dead cells after si[IGLC CR ] transfection.WILL3 showed statistically significant increase of dead cells after si[IGLC CR ] transfection.Lanes 1 and 2: WILL1, lanes 3 and 4: WILL3, lanes 5 and 6: K562.Odd and even numbers represent si [Cont] and si[IGLC CR ] transfection.C. Down-expression of MYC occurred after si[IGLC CR ] transfection.The cells were harvested at 24 h after transfection.Lanes 1 and 2: WILL1, lanes 3 and 4: WILL3, lanes 4 and 5: K562.Odd and even numbers represent si[Cont] and si[IGLC CR ].EtBr staining gel served as a loading control.D. Time course of IGLL5 and MYC expression after si[IGLC CR ] transfection.The WILL3 was cultured for 14 h and 24 h after transfection.The RNA blot was concurrently hybridized with the Cλ and MYC probes.The signals derived from the Cλ probe represent IGLL5 mRNA.Lane 1: 24 h after si[Cont]; lane 2: 14 h after si[IGLC CR ], lane 3: 24 h after si[IGLC CR ].EtBr staining gel served as a loading control.