Spontaneous arising of a lymphoblastoid B‐cell line harbouring a pre‐leukemic DNMT3A mutation in acute myeloid leukaemia cell culture

Preleukemic mutations (e.g., DNMT3A), first discovered in lymphocytes of patients with acute myeloid leukaemia (AML), occur very early during leukemogenesis at the stem cell level.1 There is paucity of data about the effect of these mutations on lymphocyte functions in AML, mainly due to the lack of an appropriate model to study this phenomenon. We herein report on spontaneous generation of lymphoblastoid cell lines (SPLCLs) that originated from peripheral blood mononuclear cells (PBMCs) of EBVseropositive AML patients. One of the SPLCLs carried DNMT3A mutation. EpsteinBarr virus (EBV), a DNA herpesvirus, latently residing in lymphocytes, can immortalize normal B cells in vitro and lead to lymphoblastoid cell line (LCL) development. Such immortalization of resting B cells gives rise to an actively proliferating population, usually resulting in polyclonal activation of Blymphocytes.2 Approximately 95% of adults are exposed to EBV, mainly during childhood, and EBVinduced Bcell malignancies (e.g., Burkitt lymphoma, Hodgkin lymphoma) are welldocumented.3 To stimulate Bcell growth from PBMCs of either EBVpositive patients or healthy EBVseropositive donors to establish an LCL in vitro, the use of cyclosporine A, as a selective immunosuppressive agent, is required.4 However, spontaneous generation of LCLs (SPLCLs) from PBMCs, without EBV infection or exposure to transforming agents, as observed in our study, is rare. Previously reported spontaneous appearance of immortalized Bcell lines originated from PBMCs of patients with active autoimmune diseases, including Sjogren's syndrome, systemic lupus erythematosus, rheumatoid arthritis5 or multiple sclerosis.6 To the best of our knowledge, this is the first report of spontaneous outgrowth of LCLs in culture of myeloid blasts obtained from AML patients. To culture AML blasts, eight blood samples were drawn from AML patients at diagnosis (Table S1), processed and cultured. The viability and ability of the cell cultures to grow were routinely examined with microscopy, without splitting the cells or changing the growth media. Within two weeks, most cells died, while a minor cell population adhered to the culture flask and was recognized as stroma cells. Remarkably, in three of the eight samples, 15% of the cells remained alive and were later identified as lymphocytes. During the third week, live single suspended cells and several small cell clusters repopulated the culture, constituting 30% of the total cell population. A week later, the cells duplicated and the clusters enlarged, while exhibiting rosettelike shape morphology. In immunophenotyping, 80% of these cells displayed the Blymphocyte phenotype. At weeks 5– 6, Giemsa staining revealed numerous vacuoles in the cytoplasm. The process of blast culture transition to a continuously proliferating LCL is demonstrated in Figure 1A. BLCLs (LCLAML1, LCLAML2, LCLAML3), originating from three EBVseropositive AML patients, spontaneously grew in the culture. These SPLCL clusters appeared to be very dense, tightly packed and formed large clumps (Figure 1B). Surprisingly, none of the cell surface antigens defining AML blasts (CD34, CD117, or CD33) was expressed. We found associations of high blood cell counts and a high blast percentage with spontaneous generation of LCLs, without correlation to the mutation profile (Table S1). More patientderived SPLCLs are required to confirm these data. EBVtransformed human BLCLs can be established by the traditional method using EBV infection of PBMCs isolated from healthy donor blood (LCLH).4 We created such cells to serve as a reference for the characterization of the SPLCLs produced in our study. SPLCLs were found to be larger in size than LCLH, which was particularly evident in LCLAML1 (p < 0.01) (Figure 1C). Additionally, similar to LCLH, all SPLCLs expressed the pan Bcell markers CD19 and CD20 (Figure 1D). The quantitative EBV RTPCR assay revealed a range of 106– 107 DNA copies in SPLCLs, while this number was as high as 109 in LCLH (Figure 1E). Likewise, significantly higher levels of EBVDNA were released to the culture media by LCLH compared with SPLCLs (>2 logs) (Figure 1F). The low EBV copy number was needed to obtain a rapidly proliferating SPLCL culture compared with the designed LCLH, which may be explained by reactivation of latent EBV under stress conditions in the culture. The ability to produce and secrete immunoglobulins is known to depend on the maturation stage of B cells. We identified surface marker phenotypes of LCLH, LCLAML2 and LCLAML1 as CD19+CD20+CD10−CD5−CD38+CD27+IgD−, CD19+CD20+CD10−CD5−CD38+CD27+/−IgD− and CD19+CD20+CD10−CD5−CD38−CD27+IgD−, respectively. The two


Spontaneous arising of a lymphoblastoid B-cell line harbouring a pre-leukemic DNMT3A mutation in acute myeloid leukaemia cell culture
Pre-leukemic mutations (e.g., DNMT3A), first discovered in lymphocytes of patients with acute myeloid leukaemia (AML), occur very early during leukemogenesis at the stem cell level. 1 There is paucity of data about the effect of these mutations on lymphocyte functions in AML, mainly due to the lack of an appropriate model to study this phenomenon. We herein report on spontaneous generation of lym- Approximately 95% of adults are exposed to EBV, mainly during childhood, and EBV-induced B-cell malignancies (e.g., Burkitt lymphoma, Hodgkin lymphoma) are well-documented. 3 To stimulate B-cell growth from PBMCs of either EBV-positive patients or healthy EBV-seropositive donors to establish an LCL in vitro, the use of cyclosporine A, as a selective immuno-suppressive agent, is required. 4 However, spontaneous generation of LCLs (SP-LCLs) from PBMCs, without EBV infection or exposure to transforming agents, as observed in our study, is rare. Previously reported spontaneous appearance of immortalized B-cell lines originated from PBMCs of patients with active autoimmune diseases, including Sjogren's syndrome, systemic lupus erythematosus, rheumatoid arthritis 5 or multiple sclerosis. 6 To the best of our knowledge, this is the first report of spontaneous outgrowth of LCLs in culture of myeloid blasts obtained from AML patients.
To culture AML blasts, eight blood samples were drawn from AML patients at diagnosis (Table S1), processed and cultured. The viability and ability of the cell cultures to grow were routinely examined with microscopy, without splitting the cells or changing the growth media. Within two weeks, most cells died, while a minor cell population adhered to the culture flask and was recognized as stroma cells. Remarkably, in three of the eight samples, 15% of the cells remained alive and were later identified as lymphocytes.
During the third week, live single suspended cells and several small cell clusters repopulated the culture, constituting 30% of the total cell population. A week later, the cells duplicated and the clusters enlarged, while exhibiting rosette-like shape morphology. In immunophenotyping, 80% of these cells displayed the B-lymphocyte phenotype. At weeks 5-6, Giemsa staining revealed numerous vacuoles in the cytoplasm. The process of blast culture transition to a continuously proliferating LCL is demonstrated in Figure 1A.
These SP-LCL clusters appeared to be very dense, tightly packed and formed large clumps ( Figure 1B). Surprisingly, none of the cell surface antigens defining AML blasts (CD34, CD117, or CD33) was expressed. We found associations of high blood cell counts and a high blast percentage with spontaneous generation of LCLs, without correlation to the mutation profile (Table S1). More patient-derived SP-LCLs are required to confirm these data.
EBV-transformed human B-LCLs can be established by the traditional method using EBV infection of PBMCs isolated from healthy donor blood (LCL-H). 4 We created such cells to serve as a reference for the characterization of the SP-LCLs produced in our study. SP-LCLs were found to be larger in size than LCL-H, which was particularly evident in LCL-AML1 (p < 0.01) ( Figure 1C). Additionally, similar to LCL-H, all SP-LCLs expressed the pan B-cell markers CD19 and CD20 ( Figure 1D).
The quantitative EBV RT-PCR assay revealed a range of 10 6 -10 7 DNA copies in SP-LCLs, while this number was as high as 10 9 in LCL-H ( Figure 1E). Likewise, significantly higher levels of EBV-DNA were released to the culture media by LCL-H compared with SP-LCLs (>2 logs) ( Figure 1F). The low EBV copy number was needed to obtain a rapidly proliferating SP-LCL culture compared with the designed LCL-H, which may be explained by reactivation of latent EBV under stress conditions in the culture.
The ability to produce and secrete immunoglobulins is known to depend on the maturation stage of B cells. We iden-  Figure 2B). Yet, a possibility of LCL-AML1 malignant nature was ruled out by FISH analysis for t(8:14) (data not shown). The monoclonality pattern might be related to selection of a mutated pre-leukemic clone. As for the cell surface protein expression, IgG immunoglobulin was detected in 51% of LCL-AML1, 80% of LCL-AML3 and 47% of LCL-H, which also expressed IgM (40%) ( Figure 2C).

F I G U R E 1
Identification of SP-LCLs originated from PBMCs of AML patients using different methods. (A) Morphological changes occurring during SP-LCL development in culture. The transition from a single-cell suspension of AML blasts in culture through rosette colonies to big clumps is demonstrated. The following method was used to obtain these SP-LCLs: PBMCs were drawn from AML patients at diagnosis and the mononuclear cells were separated by centrifugation over a layer of LymphoprepTM (Axis-Shield PoC AS, Oslo, Norway). After washing with PBS, 1.5 × 10 6 PBMCs/ml were seeded in a T75 flask containing RPMI1640 medium supplemented with 15% FBS, 2 mM L-glutamine and 1% penicillin-streptomycin and incubated at 37°C in 5%CO 2 . Cells were followed using inverted microscopy (×20 magnification) until the rosette-like formation was observed. Cytospin preparation of SP-LCLs obtained from an AML patient was stained using Wright-Giemsa stain. The image shows the cells with a high cytoplasmic/nuclear ratio and cytoplasmic vacuoles (light microscopy, ×100 magnification). The DNMT3A gene, encoding DNA methyltransferase, is frequently mutated in various haematological malignancies. Between 17 and 34% of AML patients with a cytogenetically normal karyotype harbour point mutations in DNMT3A. 9 The R882H mutation in DNMT3A is most prevalent and leads to a roughly 40%-reduction in overall DNA methylation activity. 10 The mechanism of EBVmediated transformation of PBMCs to LCLs is reported to involve promoter hypomethylation only. 11 Remarkably, in the present study, DNMT3A gene sequencing revealed a single homozygous mutation in R882H in LCL-AML1 (validated using molecular inversion probe technique, 100% VAF). In this patient, the mutation was detected in blasts derived at diagnosis, but not in the germline saliva sample ( Figure 2F). Therefore, we assume that a somatic mutation in DNMT3A emerges early at the hematopoietic stem cell level, potentially altering the ability of lymphocytes carrying this mutation to secret IL-6. Of note, DNMT3A overexpression is suggested to suppress IL-6 through alterations in the status of DNA methylation in the IL-6 promoter. 12 To date, only two human AML cell lines with the DNMT3A R882 mutation have been described, NPM1-mutant OCI-AML3 13 (R882C) and SET-2 (R882H, heterozygous). 14

AVA I L A B I LIT Y O F B I O LO G I C A L M ATER I A L S
We declare that we allow to Cancer Research Technology Ximbio (ximbio.com) to provide LCL-AML1 (=MHYO-LCL1) and LCL-AML3 (=MHYO-LCL2) upon request.

K E Y WO R DS
acute myeloid leukaemia, B-cell functions, DNMT3A, pre-leukemic mutations, spontaneously arising lymphoblastoid cell line

CO N FLI C T O F I NTE R E S T
The authors have no conflicts to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.