Immunoglobulin Gamma Heavy Chain Gene with Somatic Hypermutation Is Frequently Expressed in Acute Myeloid Leukemia

ORIGINAL ARTICLE Immunoglobulin gamma heavy chain gene with somatic hypermutation is frequently expressed in acute myeloid leukemia

X Qiu1,2, X Sun3,ZHe1, J Huang2,FHu2, L Chen4, P Lin1,MJYou1, LJ Medeiros1 and CC Yin1

Expression of immunoglobulin (Ig), a marker characteristic of B cells, has been reported in epithelial cells and has been suggested to have a role in their survival and growth. We assessed the frequency and level of Ig gamma heavy chain (IgG) expression in acute myeloid leukemia (AML), and found that IgG was expressed at a high frequency and level in AML cell lines and primary myeloblasts, but not in monocytes or neutrophils from patients with non-hematopoietic neoplasms or healthy controls. AML-derived IgG had the same molecular weight as B cell-derived IgG and was secreted. We further detected IgG VHDJH transcripts in AML cell lines and sorted primary myeloblasts, conﬁrming that IgG expression was indeed produced by AML cells. AML-derived IgG gene rearrangements showed evidence of somatic hypermutation of the variable (V) gene segments, and restricted (AML cell lines) or biased (primary myeloblasts) V usage. Anti-human IgG reduced cell viability and induced apoptosis in AML cell lines. Although the function of the AML-derived IgG is unclear, our ﬁndings suggest that AML-derived IgG may be a novel AML-related gene that contributes to leukemogenesis and AML progression. AML-derived IgG may serve as a useful molecular marker for monitoring minimal residual disease or designing target therapy.

Leukemia (2013) 27, 92–99; doi:10.1038/leu.2012.184

Keywords: IgG; acute myeloid leukemia; VHDJH gene rearrangements

INTRODUCTION expression between these cells that might have clinical utility. Our Immunoglobulin (Ig) was previously thought to be produced only results show for the first time that IgG is expressed and secreted in by B lymphocytes and plasma cells. During the past decade, AML cell lines and primary myeloblasts at a high frequency and however, this concept has been challenged by a series of studies. level. We also found that IgG VHDJH was rearranged only in AML Different types of Ig, including IgG, IgM and IgA, have been shown cells, but not in monocytes or neutrophils. Furthermore, the AML- to be expressed in many types of neoplastic epithelial cells, derived IgG Variable (V) region revealed somatic hypermutation including those arising in the breast, colon, lung, liver, neck and and a restricted or biased usage of certain V segments. Preliminary mouth. In these cells, Ig is present in both membranous and functional study showed that anti-human IgG reduced cell viability secreted forms, similar to B cell-derived Ig.1–6 Subsequently, Ig and induced apoptosis in AML cell lines in culture. These results expression has been detected in several types of non- suggest that AML-derived IgG may have a role in AML patho- hematopoietic cells from healthy individuals, including epithelial genesis and progression, and potentially could serve as a novel cells, germ cells, neurons and myocardial cells.7–9 In addition, Ig molecular marker for the monitoring of minimal residual disease expressed in these normal cells shows some unique and development of target therapy. characteristics, including restricted usage of certain sequences,9,10 a unique glycosylation profile11,12 and different regulatory mechanisms of gene expression than are known in B MATERIALS AND METHODS cells.13,14 Moreover, it has been shown recently that non- AML and B-cell lymphoma cell lines hematopoietic cell-derived Ig is involved in cell survival and We used the AML cell lines, HEL, HL-60, KG-1, NB4, THP-1 and OCI-AML3, and proliferation.1,5 These findings suggest that non-hematopoietic the B-cell lymphoma cell line, SP53 (courtesy of Dr M Wang, Department of Myeloma, UT MD Anderson Cancer Center). All cell lines were maintained in cell-derived Ig is a novel molecule with a unique structure and RPMI-1640 medium (Life Technologies Inc., Carlsbad, CA, USA) supplemented function that may have a role in carcinogenesis. with 10% fetal bovine serum (HyClone/Thermal Scientific, Logan, UT, USA) Myeloid cells originate from hematopoietic stem cells in the and 2 mML-glutamine at 37 1C with humidified 5% CO2. bone marrow. It has been believed that myeloid cells cannot produce Ig. In this study, we sought to explore whether acute Patient samples myeloid leukemia (AML) cells (myeloblasts) as well as other cells of The study was conducted according to an institutional review board- myeloid origin, that is, mature normal monocytes and neutrophils, approved laboratory protocol. Peripheral blood specimens from 14 AML could produce Ig. We also sought to determine differences in Ig patients were analyzed for expression of IgG on the cell membrane of

1Department of Hematopathology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA; 2Department of Immunology, School of Basic Medical Sciences, Peking University, Beijing, China; 3Department of Laboratory Medicine, The University of Texas MD Anderson Cancer Center, Houston, TX, USA and 4Department of Pathology, The University of Texas-Medical School, Houston, TX, USA. Correspondence: Dr CC Yin, Department of Hematopathology, UT MD Anderson Cancer Center, 1515 Holcombe Blvd, Unit 72, Houston, TX 77030, USA. E-mail: [email protected] or Dr X Qiu, Department of Immunology, School of Basic Medical Sciences, Peking University, Beijing 100191, China. E-mail: [email protected] Received 27 April 2012; revised 18 June 2012; accepted 2 July 2012; accepted article preview 9 July 2012; advance online publication, 14 August 2012 IgG gene expression in acute myeloid leukemia X Qiu et al 93 þ þ CD33 cells. To analyze IgG VHDJH gene transcription in CD33 AML cells, IgG-HRP (1:2000; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) as the we assessed peripheral blood specimens from another 15 AML patients. secondary antibody, and a chemiluminescent substrate (Mabtech, þ We also analyzed IgG VHDJH gene transcription in CD33 adult monocytes Mariemont, OH, USA). and neutrophils derived from peripheral blood specimens of 15 patients with non-hematopoietic neoplasms as well as 10 healthy individuals. The Enzyme-linked immunosorbent assay clinicopathologic data were obtained by review of medical records. Approximately 2 Â 105 cells from AML cell lines or AML patients were plated per well in six-well plates, and cultured for 0, 24, 48, 72 or 96 h. The Molecular genetic analyses culture supernatant was collected, and secreted IgG was detected using an Karyotypic analysis was performed on metaphase cells prepared from ELISA kit for human IgG (Mabtech) according to manufacturer’s bone marrow aspirate specimens using standard techniques.15 The results instructions. were reported using the 2005 International System for Human Cytogenetic Nomenclature. Detection of VHDJH transcripts by reverse transcription-PCR Genomic DNA extracted from bone marrow aspirate specimens was Total RNA was extracted from AML cell lines using Trizol reagent amplified by PCR and subjected to mutational analysis for codons 12, 13 or (Invitrogen, Carlsbad, CA, USA). For sorted CD33 þ CD19 À CD138 À or 61 of KRAS and NRAS by pyrosequencing using a PSQ HS 96 Pysosequencer CD33 À CD19 þ CD138 À cells, total RNA was extracted using the RNeasy (Biotage, Uppsala, Sweden),16 or KIT (exons 8 and 17) and CEBPA by direct Micro Kit (Qiagen Inc., Chatsworth, CA, USA) according to manufacturer’s Sanger sequencing on an ABI Prism 3100 Genetic Analyzer (Applied instructions. Reverse transcription (RT) was performed using the Sensiscript Biosystems, Foster City, CA, USA).16,17 A fluorescence-based multiplex PCR RT Kit (Qiagen Inc.). IgG V DJ transcripts were detected by a nested PCR. followed by capillary electrophoresis was used to detect internal tandem H H The primers and conditions for the first-round PCR for the amplification of duplication and D835 point mutation of the FLT3 gene and exon 12 V DJ were described previously.10 For the second-round PCR, an mutation of the NPM1 gene using genomic DNA and an ABI Prism 3100 H H upstream primer that anneals to the framework 2 (FR2) region (FR2: Genetic Analyzer.18,19 50-TGG(A/G)TCCG(A/C/G)CAG(G/C)C(T/C)CC(A/C/G/T)GG-30), coupled with a JH primer (JH: 50-AACTGCAGAGGAGACGGTGACC-30), was used.20 The Flourescence-activated cell sorting amplification products were gel-purified using a DNA purification column For sorting of myeloblasts from AML patients, mononuclear cells were (Qiagen Inc.), ligated into the pCR2.1 vector (Invitrogen), and transformed isolated from peripheral blood by Ficoll-Hypaque density gradient into DH5a-competent bacteria (Invitrogen). In all, 5–10 colonies per sample were chosen randomly, and sequenced using the T7 primer (Invitrogen). centrifugation, and washed with 0.01M PBS. The mononuclear cells were stained with monoclonal antibodies specific to CD19 (APC, BD Pharmigen, San Diego, CA, USA), CD33 (FITC, BD Pharmigen) and CD138 (PE, BD Analysis of rearranged IgG VHDJH genes Pharmigen) at 4 1C for 15 min. After washing twice with PBS, the cells were The rearranged V DJ sequences were compared with those in the basic sorted using a FACScan instrument (BD Pharmigen). CD33 þ CD19 À H H À À þ local alignment search tool (BLAST, http://www.ncbi.nlm.nih.gov/BLAST) CD138 cells were selected. CD33 CD19 cells also were collected and and Immunogenetics (http://www.imgt.org/) databases to identify the best used as a positive control. Anti-mouse IgG1-PE was used as an isotype control. matching germline gene segments and junctions. For the analysis of For sorting of monocytes and neutrophils, erythrocytes from patients somatic hypermutation, the part of the IgG V region from FR2 to JH was with non-hematopoietic neoplasms and healthy individuals were lysed by used. Mutation status was designated as unmutated if there were fewer incubating 1 ml of peripheral blood with 8 ml of distilled water for 30 s. than 2% mutations (498% homology to germline sequences), or as Following lysis, 1 ml of 10 Â PBS was added and centrifugation was mutated if there were 2% or more mutations (p98% homology to performed at 1500 r.p.m. for 5 min. The supernatant was discarded, and germline sequences) compared with the germline sequences.21 leukocytes were washed twice with PBS, before 1 Â 106 leukocytes were stained with monoclonal antibodies specific to CD19 (APC), CD33 (FITC) and CD138 (PE), washed and sorted. Monocytes and neutrophils were Effect of anti-human IgG on viability and apoptosis of AML cell gated by using forward versus side scatter dot plots. The CD33 þ CD19 À lines CD138 À cells in both populations were selected. CD33 À CD19 þ cells also We performed a cell viability assay using the Cell Counting Kit-8 (CCK-8, were collected and used as a positive control. Anti-IgG-PE was used as an Dojindo Molecular Technologies, Inc., Kumamoto, Japan). HL-60 and OCI- isotype control. AML3 cell lines were plated in a 96-well plate at 1 Â 105 cells/well, and incubated in DMEM (Sigma) supplemented with 10% fetal bovine serum at 37 1C overnight. The supernatant was then discarded, and goat anti-human Immunocytochemistry IgG (25 mg/ml; Sigma) or goat IgG (isotype control, 25 mg/ml; Sigma) was Cytospin preparations were prepared from AML cell lines. The slides were added. The cells were then incubated as mentioned above, and cell fixed with 100% acetone for 5 min, incubated with 1% hydrogen peroxide viability was assessed using the CCK-8 kit at 24, 48, 72 and 96 h. for 10 min, washed with PBS, and then blocked with PBS containing 10% We evaluated the effect of anti-human IgG on apoptosis of AML cell goat serum for 10 min. Excess blocking buffer was removed. Indirect lines. HL-60 and OCI-AML3 cell lines were plated in a six-well plate at immunohistochemical staining was performed by incubating the slides 5 Â 105 cells/well, and incubated in DMEM (Sigma) supplemented with with monoclonal anti-human IgG (2 ng/ml, Sigma, St Louis, MO, USA) or an 10% fetal bovine serum at 37 1C overnight. The supernatant was then antibody against human cancer-derived IgG, RP215 (2 ng/ml, a gift from discarded, and goat anti-human IgG (25 mg/ml; Sigma) or goat IgG (isotype Dr G Lee, the University of British Columbia, Vancouver, Canada) in a 37 1C control, 25 mg/ml; Sigma) was added. The cells were then incubated as humidified chamber for 45 min. After washing, the slides were incubated mentioned above, and apoptosis was assessed using annexin V/propidium with anti-mouse IgG-horseradish peroxidase (HRP; Dako, Carpinteria, CA, iodide staining at 72 h. USA) at room temperature for 20 min. The slides were washed, and bound antibodies were detected using 3,30-diaminobenzidine tetrahydrochloride (DAB; Dako) according to manufacturer’s instructions. Statistical analyses Statistical analyses were performed using SAS 8.1 (SAS Institute Inc., Cary, NC, USA). Differences between groups were evaluated using Student’s Immunoprecipitation and western blot analysis t-test or the w2-test. A P-value of o0.05 was considered statistically Primary or cultured AML cells were lysed with radioimmunoprecipitation significant. assay lysis buffer (Thermo Fisher Scientific Inc., Rockford, IL, USA) on ice for 15 min, and centrifuged at 15 000 r.p.m. for 15 min. The lysate was incubated with monoclonal anti-human IgG (Sigma), or mouse IgG1 (as RESULTS an isotype control, Sigma), and protein G/A-beads (Abcam, Cambridge, MA, Clinicopathologic features USA) for 30 min. After centrifugation at 12 000 r.p.m. for 5 min, the Fourteen AML patients were evaluated for membranous IgG supernatant was removed and the lysate was washed twice. The þ antigen–antibody complex was eluted followed by sodium dodecyl sulfate expression by CD33 myeloblasts (Table 1). There were 10 men polyacrylamide gel electrophoresis and western blot analysis using goat and 4 women with a median age of 62 years (range, 7–85 years). anti-human IgG (1:2000; Sigma) as the primary antibody, rabbit anti-goat The diagnoses using current World Health Organization

& 2013 Macmillan Publishers Limited Leukemia (2013) 92 – 99 IgG gene expression in acute myeloid leukemia X Qiu et al 94 Table 1. Clinicopathologic features of 14 AML patients assessed for IgG expression

Case Age/Sex (years) Diagnosis Blast (%) Treatment Outcome (F/U) CD33 þ blasts (%) IgG þ CD33 þ blasts (%)

1 60/M AML-MRC 80 Cy, I, FI, My, H, B, Cl DOD (9) 99 93 2 7/M AML-MRC 66 Cy, Mi, ASCT DOD (29) 51 98 3 50/F AML-M4 57 Cy, I, ASCT, S DOD (15) 42 N 4 55/F AML-M5 68 Cy, I, E, S, Fl, G, D DOD (10) 93 75 5 53/M AML-M1 92 Cy, I, Fl, My, ASCT CR (49) 79 58 6 80/M AML-MRC 92 Cy, FI, G DOD (4) 11 N 7 71/F AML-M2 26 Cy, Cl, Fl, G DOD (16) 54 69 8 64/M AML-MRC 30 Cy, I pAML (4) 1.6 N 9 85/M AML-M1 86 Cl, H DOD (32) 52 100 10 69/M AML-MRC 60 Cy, I pAML (69) 59 49 11 75/M AML-MRC 53 Cy, I, Fl DOD (26) 12 100 12 72/M AML-M2 79 Cy, Fl, My, A, S pAML (18) 79 100 13 59/F AML-M2 67 Cy, I, ASCT DOD (3) 15 100 14 54/M AML-M4 58 Cy, I, ASCT CR (18) 51 33 Abbreviations: A, azacitadine; AML, acute myeloid leukemia; ASCT, allogeneic stem cell transplantation; B, bendamustin; CR, complete remission; Cl, clofazabine; Cy, cytarabine; D, dasatinib; DOD, died of disease; E, etoposide; F, female; Fl, ﬂudarabine; F/U, follow-up (months); G, gemtuzumab; H, hydroxyurea; I, idarubicin; M, male; Mi, mitoxantrone; MRC, myelodysplasia-related changes; M1, AML without maturation; M2, AML with maturation; M4, acute myelomonocytic leukemia; M5, acute myeloid leukemia; My, mylotarg; pAML, persistent AML; S, Sorafenib. This group of patients were assessed for the expression of IgG by immunohistochemistry, ﬂow cytometry immunophenotyping, western blot/immunoprecipitation and enzyme-linked immunosorbent assay.

Table 2. Clinicopathologic features of patients with AML, non-hematopoietic neoplasms and healthy controls used in this study

AML Non-hematopoietic neoplasms Healthy controls Reference range

Case number 15 15 10 na Age (years) 58 (5–87) 59 (19–83) 50 (34–63) na Sex 7M, 8F 8M, 7F 3M, 7F na WBC ( Â 109/l)a 18.9 (1.2–122.0) 6.4 (4.3–12.3) 7.0 (3.7–9.7) 4.0–11.0 Hgb (g/dl)a 9.4 (8.2–11.3) 13.0 (8.5–15.2) 13.0 (11.4–15.1) 14.0–18.0 (M) 12.0–16.0 (F) Platelet ( Â 109/l)a 28 (10–87) 263 (112–436) 230 (211–368) 140–440 Blast (%)a 82 (25–99) na na na Abbreviations: AML, acute myeloid leukemia; F, female; Hgb, hemoglobin; M, male; na, not available; WBC, white blood cell. These three groups of patients a were assessed for the expression of IgG VHDJH transcripts after ﬂuorescence-activated cell sorting by ﬂow cytometry. Data shown as median (range).

terminology included: six cases of AML with myelodysplasia- adenocarcinoma, three with squamous cell carcinoma of the tongue related changes (AML-MRC), three cases of AML with maturation (n ¼ 2) or tonsil (n ¼ 1), two with stomach adenocarcinoma and one (M2), two cases of AML without maturation (M1), two cases of each with hepatocellular carcinoma, pancreatic adenocarcinoma, acute myelomonocytic leukemia (M4) and one case of acute high-grade sarcoma, glioblastoma or thymoma (Table 2). monocytic leukemia (M5). The bone marrow blast count ranged from 26 to 92% (median, 67%). All patients received multi-agent Immunoglobulin gamma heavy chain is expressed in AML cell chemotherapy; five also were treated with allogeneic stem cell lines and primary myeloblasts transplantation. With a median follow-up of 17 months (range, To determine if IgG is expressed in AML cells, we first performed 3–69 months), nine patients died of disease, three patients had immunohistochemical studies using a monoclonal antibody persistent disease and two were in complete remission after against human IgG in six AML cell lines: HEL, NB4, HL-60, KG-1, transplantation (Table 1). OCI-AML3 and THP-1. We observed strong IgG expression in all six Conventional cytogenetic analysis showed a complex karyotype AML cell lines assessed (Figure 1a). (X3 abnormalities) in 5 patients, 1–2 cytogenetic abnormalities in We performed two-color flow cytometry immunophenotyping 5 patients and a diploid karyotype in 4 patients. Molecular analysis using anti-human IgG-PE and anti-human CD33-FITC on peripheral revealed internal tandem duplication of the FLT3 gene in 6/14 blood mononuclear cells from AML patients, and demonstrated (43%) cases, and mutations in CEBPA, NPM1 and KRAS in 1/4 (25%), IgG expression in primary CD33 þ myeloblasts in 11 of 14 AML 1/10 (10%) and 1/12 (8%) cases, respectively. None of the 12 cases patients assessed. IgG was expressed on 33–100% of CD33 þ assessed had NRAS or KIT mutations. myeloblasts (median, 93%) from AML patients (Table 1; Figure 1b). Subsequently, another 15 AML patients were assessed for IgG þ VHDJH gene transcription in sorted CD33 AML cells, which included 5 cases of AML-MRC, 3 cases of AML-M4, 2 cases of AML-derived IgG has the same molecular weight as B cell-derived AML-M1, 1 case each of AML with t(15;17)(q22;q21), AML-M2 or IgG and is secreted AML-M5, and 2 cases of relapsed AML, not further classifiable Immunoprecipitation and western blot analysis confirmed the (Table 2). presence of IgG in AML cell lines and primary myeloblasts. We also analyzed CD33 þ monocytes and neutrophils from 15 In addition, the AML-derived IgG had the same molecular weight patients with non-hematopoietic neoplasms and 10 healthy as classic IgG expressed by B lymphocytes (Figure 1c). individuals for IgG VHDJH gene transcription. The group of patients We collected supernatants from cell cultures of AML cell lines with non-hematopoietic neoplasms included five with colorectal and patient myeloblasts at 0, 24, 48, 72 and 96 h, and assessed

Leukemia (2013) 92 – 99 & 2013 Macmillan Publishers Limited IgG gene expression in acute myeloid leukemia X Qiu et al 95

Figure 1. IgG expression and secretion in AML cell lines and primary myeloblasts. (a) Immunohistochemical staining for IgG showed that IgG was expressed in AML cell lines. HL-60, HEL, NB4 and OCI-AML3, AML cell lines; SP53, B-lymphoma cell line, used as a positive control; neutrophils, used as a negative control. (b) Flow cytometry analysis detected IgG expression on the cell surface of primary myeloblasts. AML- MRC, AML with myelodysplasia-related changes; AML-M1, AML without maturation; AML-M2, AML with maturation; AML-M4, acute myelomonocytic leukemia; AML-M5, acute monoblastic leukemia. (c) Immunoprecipitation analysis demonstrated IgG expression in AML cell lines and primary myeloblasts. Two monoclonal antibodies specific to anti-human IgG and anti-human Ig kappa chain, respectively, were used to bind to AML-derived IgG. Mouse IgG1 was used as an isotype control. Goat anti-human IgG was used for western blot analysis. THP-1, NB4 and OCI-AML3, AML cell lines; primary AML cells, AML cells isolated from peripheral blood of a patient with AML. (d) Detection of IgG in cell culture supernatant of AML cell lines and primary CD33 þ myeloblasts by enzyme-linked immunosorbent assay (ELISA). The cell culture supernatant was collected at 0, 24, 48, 72 and 96 h and stored at À 80 1C. The presence and concentration of IgG in these supernatant were then tested using a human IgG detection kit. Upper panel: IgG secretion from two AML cell lines (OCI-AML3 and NB4). The level of IgG expression was significantly higher at 72 and 96 h as compared with the beginning point (0 h control) in both cell lines (Po0.001). Lower panel: IgG secretion from two AML patients (AML-MRC, AML with myelodysplasia-related changes; AML-M5, acute monocytic leukemia). The level of IgG expression was significantly higher at 24, 48 and 72 h as compared with the beginning point (0 h control) in both patients (Po0.001). them for the presence of IgG by enzyme-linked immunosorbent CD33 þ CD19 À CD138 À adult monocytes and neutrophils from assay. We detected the presence of IgG in the supernatant of AML patients with non-hematopoietic neoplasms and healthy cell cultures in both AML cell lines and primary myeloblasts, and individuals. CD19 þ B cells from patients with non-hematopoietic showed that the level of secreted IgG was significantly higher at neoplasms and healthy individuals were used as a positive 72–96 h in culture than at 0 h (Po0.001; Figure 1d). control. We detected IgG VHDJH transcripts in 5 of 6 AML cell lines (all except KG-1; Figure 2C), CD33 þ myeloblasts from 6 of 15 AML patients assessed (Figure 2Da), and CD19 þ B cells from IgG VHDJH transcripts are detected in AML cell lines and primary patients with non-hematopoietic neoplasms and healthy indivi- myeloblasts but not in non-neoplastic monocytes and neutrophils duals (Figure 2Db). IgG VHDJH transcripts were not detected in We performed cell sorting by flow cytometry on peripheral blood monocytes or neutrophils from patients with non-hematopoietic mononuclear cells (Figure 2A, P1) from AML patients using neoplasms or healthy individuals, however (Figure 2Dc). monoclonal antibodies specific for CD19 (APC), CD33 (FITC) and CD138 (PE) to isolate CD33 þ CD19 À CD138 À myeloblasts þ À (Figure 2A, P4). CD19 CD33 cells (Figure 2A, P3) also were IgG VHDJH rearrangements demonstrate restricted or biased usage collected and used as a positive control. We also isolated CD33 þ of V segments monocytes (Figure 2B, P5) and neutrophils (Figure 2B, P7) from AML cell line-derived IgG VHDJH rearrangements displayed peripheral blood mononuclear cells from patients with non- restricted usage of a unique sequence in four of five AML cell hematopoietic neoplasms or healthy individuals. lines assessed (except NB4). All showed VH3-48/D4-7/JH4 rearran- We performed RT-PCR using primers specific to the FR2 and gement (Figure 3a). Moreover, high homology, including the same þ À JH regions of IgG VHDJH on AML cell lines, sorted CD33 CD19 complementary determining region 3 sequence, was identified in CD138 À primary myeloblasts from AML patients, and sorted the VH3-48/D4-7/JH4 rearrangement from different AML cell

& 2013 Macmillan Publishers Limited Leukemia (2013) 92 – 99 IgG gene expression in acute myeloid leukemia X Qiu et al 96

Figure 2. Detection of IgG VHDJH transcripts in AML cell lines, primary myeloblasts, as well as adult monocytes and neutrophils by RT-PCR. (A) CD33 þ CD19 À CD138 À myeloblasts from peripheral blood of AML patients were sorted by flourescence-activated cell sorting (FACS). P1: Mononuclear cells; P2: CD138 þ cells; P3: CD19 þ cells; P4: CD33 þ cells. The cells in P4 were selected and defined as CD33 þ CD19 À CD138 À myeloblasts. The cells in P3 were selected and used as a positive control. (B) Adult CD33 þ monocytes and neutrophils from peripheral blood of patients with non-hematopoietic neoplasms and healthy controls were sorted by FACS. P1: mononuclear cells; P2: granulocytes; P3: CD138 À mononuclear cells; P4: CD138 À granulocytes; P5: CD33 þ CD19 À monocytes; P6: CD19 þ B cells; P7: CD33 þ neutrophils; P8: À or dim þ CD33 granulocytes; cells in P5, P6 and P7 were selected. (C) Analysis of IgG VHDJH rearrangements in AML cell lines. VHDJH transcripts were detected in five of six AML cell lines. HEL, HL-60, KG-1, NB4, OCI-AML3 and THP-1, AML cell lines; SP53, B-lymphoma cell line, used as a positive control. (D) RT-PCR analysis showed that IgG VHDJH transcripts were detected in primary AML cells in 6 of 15 AML patients (a) and CD19 þ B cells from patients with non-hematopoietic neoplasms (b, used as positive control), but not in non-neoplastic adult monocytes and neutrophils (c, the same 12 patients as in b).

Leukemia (2013) 92 – 99 & 2013 Macmillan Publishers Limited IgG gene expression in acute myeloid leukemia X Qiu et al 97

Figure 3. Sequence analysis of AML-derived IgG VHDJH genes in AML cell lines. (a) IgG VHDJH rearrangements in AML cell lines showed high homology and restricted usage of VH3-48/D4-7/JH4. (b) AML-derived IgG showed somatic hypermutation, and the hot spots of mutation were concentrated in the complementary determining region but not in the FR regions.

þ lines (Figure 3a). Similar to CD19 B cell-derived IgG, VHDJH rearrangements from primary myeloblasts displayed diversity in each individual patient, although biased usage of VH3 and VH4 families was noted.

AML-derived IgG VHDJH transcripts exhibit somatic hypermutations In addition, we evaluated the mutational status of the IgG V genes þ from 28 sequences of VHDJH from CD33 myeloblasts of 6 AML patients. Although the segment we ampliﬁed only included 269 bp from FR2 to JH regions, somatic hypermutations were detected in AML-derived IgG VHDJH gene in all 28 sequences assessed (range, 4–7%). Moreover, similar to B cell-derived VHDJH, the hot spots of mutation were concentrated in the complementary determining region and not in the FR regions (Figure 3b). These results indicate that AML-derived IgG usually undergoes somatic hypermutation.

Anti-human IgG reduces cell viability and induces apoptosis of Figure 4. Anti-human IgG reduced cell viability and induced AML cell lines apoptosis in AML cell lines. (a) Cell viability was assessed in HL-60 and OCI-AML3 cell lines using CCK-8 kit after the addition of anti- To test if AML-derived IgG is involved in cell survival, we analyzed human IgG into cell culture medium for 24, 48, 72 or 96 h. cell viability and apoptosis of HL-60 and OCI-AML3 cell lines after (b) Apoptosis was assessed in HL-60 and OCI-AML3 cell lines using incubation with anti-human IgG. The results show that anti-human annexin V/propidium iodide staining after the addition of anti- IgG reduced cell viability at 48, 72 and 96 h (Figure 4a), and induced human IgG into cell culture medium for 72 h.

& 2013 Macmillan Publishers Limited Leukemia (2013) 92 – 99 IgG gene expression in acute myeloid leukemia X Qiu et al 98 apoptosis at 72 h (Figure 4b) in both HL-60 and OCI-AML3 cell lines, rearrangements from AML cell lines displayed restricted usage of a suggesting that AML-derived IgG is involved in cell survival. unique sequence; all showed VH3-48/D4-7/JH4 rearrangement. However, IgG VHDJH rearrangements from primary myeloblasts displayed diversity, similar to B cell-derived IgG, though with a DISCUSSION biased usage of VH3 and VH4 segments. We do not have a In this study, we report for the first time that the IgG gene is satisfactory explanation for the apparent discrepancy in V, D and J transcribed and expressed at a high frequency and level in AML usage between primary myeloblasts and AML cell lines. It is possible cell lines and primary myeloblasts, but not in adult non-neoplastic that the restricted usage of V, D and J segments in AML cell lines is a monocytes or neutrophils. Moreover, IgG VHDJH rearrangements result of secondary rearrangement induced by identical cell culture from AML cells have undergone somatic hypermutation and conditions. display restricted or biased usage of V segments. Anti-human IgG In summary, we have shown that the IgG gene is transcribed, reduces cell viability and induces apoptosis in AML cell lines in expressed and secreted in both AML cell lines and primary culture. These findings suggest that IgG gene expression in AML myeloblasts. As this does not occur in non-neoplastic adult cells is not random, and may therefore have a role in either monocytes or neutrophils, AML-derived IgG may have a role in leukemogenesis or AML progression. AML pathogenesis and progression. The presence of AML-derived For nearly a century, B lymphocytes have been considered to be IgG, in either cell-bound or secreted forms, may be used as a the only source of Ig. It has been assumed that other types of cells, marker for monitoring of minimal residual disease in AML patients including other hematopoietic cells, could not produce Ig. after chemotherapy. Moreover, the fact that anti-human IgG can In keeping with this concept, monoclonal Ig gene rearrangement inhibit growth in AML cell lines suggests that IgG may serve as a has been used as a diagnostic tool for B-cell lymphomas. In recent potential target for AML therapy. years, however, there has been a growing realization that Ig can be expressed in other types of cells, such as epithelial cells, neurons and germ cells, and especially in many types of epithelial CONFLICT OF INTEREST 1–9 cancer cells. Expanding this concept, our results demonstrated The authors declare no conflict of interest. that IgG is often expressed in AML cell lines and primary myeloblasts. More importantly, the AML-derived IgG was secreted into the supernatant of cell cultures. In addition, our ACKNOWLEDGEMENTS results showed that AML-derived IgG has the same molecular weight as that produced in B lymphocytes. This work was supported by a research fund awarded to C Cameron Yin from The University of Texas MD Anderson Cancer Center (#18079170) and a research fund There have been a few reports of Ig gene expression, albeit at awarded to Xiaoyan Qiu from National Natural Science Foundation of China low frequency, in bone marrow or peripheral blood specimens (#30973389). We thank Amelia Scholtz and Kathryn Carnes from Scientific Publication 22 from patients with AML. In these reports, it remained unknown, at The University of Texas MD Anderson Cancer Center for editorial review of the however, whether Ig was produced by myeloid-derived manuscript. hematopoietic cells, for example, myeloblasts, or was secreted by residual B cells that were present in the samples. It also was not clear if Ig was transcribed and translated or if this was due to non- REFERENCES specific binding of Ig produced by residual B cells. Furthermore, Ig 1 Qiu X, Zhu X, Zhang L, Mao Y, Zhang J, Hao P et al. Human epithelial cancers expression in primary myeloid-derived cells in a natural state secrete immunoglobulin G with unidentified specificity to promote growth and in vivo, for example, in non-neoplastic monocytes and neutrophils, survival of tumor cells. Cancer Res 2003; 63: 6488–6495. has not been studied. To address these questions, we investigated 2 Zhu X, Li C, Sun X, Mao Y, Li G, Liu X et al. Immunoglobulin mRNA and protein IgG gene rearrangement and transcription in sorted CD33 þ expression in human oral epithelial tumor cells. Appl Immunohistochem Mol CD19 À CD138 À myeloblasts from patients with AML as well as Morphol 2008; 16: 232–238. adult monocytes and neutrophils from patients with non- 3 Zheng H, Li M, Ren W, Zheng L, Liu HD, Hu D et al. Expression and secretion of hematopoietic neoplasms and healthy controls. After confirming immunoglobulin alpha heavy chain with diverse VDJ recombinations by human that there were no residual B cells or plasma cells in the samples epithelial cancer cells. Mol Immunol 2007; 44: 2221–2227. 4 Chen Z, Gu J. Immunoglobulin G expression in carcinomas and cancer cell lines. by flow cytometric analysis, we performed RT-PCR using primers FASEB J 2007; 21: 2931–2938. that are specific to IgG VHDJH genes. We successfully detected the 5 Babbage G, Ottensmeier CH, Blaydes J, Stevenson FK, Sahota SS. Immunoglobulin þ VHDJH transcripts in CD33 myeloblasts in 6 of 15 AML patients, heavy chain locus events and expression of activation-induced cytidine deami- but not in CD33 þ monocytes or neutrophils from patients with nase in epithelial breast cancer cell lines. Cancer Res 2006; 66: 3996–4000. non-hematopoietic neoplasms or healthy controls. These results 6 Li M, Feng DY, Ren W, Zheng L, Zhang H, Tang M et al. Expression of confirm that the expression of IgG detected by immuno- immunoglobulin kappa light chain constant region in abnormal human cervical histochemistry and flow cytometry in myeloid cells is indeed epithelial cells. Int J Biochem Cell Biol 2004; 36: 2250–2257. produced by myeloblasts, rather than non-specific membranous 7 Huang J, Zhang L, Mao T, Zhang P, Qiu X. Expression of immunoglobulin gene binding of IgG that is secreted by residual B cells in the sample. with classical V-(D)-J rearrangement in mouse spermatogenic cells and epithelial During the past decade, several groups have suggested that IgG cells of epididymis. J Histochem Cytochem 2009; 57: 339–349. 1,5,6 8 Huang J, Sun X, Mao Y, Zhu X, Zhang P, Zhang L et al. Immunoglobulin G is involved in epithelial carcinogenesis. Although the function expression in mouse brain neurons. Int J Biochem Cell Biol 2008; 40: 1604–1615. of the AML-derived IgG is still unclear, our preliminary results 9 Zhang S, Mao Y, Huang J, Ma T, Zhang L, Zhu X et al. Immunoglobulin gene locus showed that anti-human IgG blocked the activity of IgG in AML events in epithelial cells of lactating mouse mammary glands. Cell Mol Life Sci cell lines, HL-60 and OCI-AML3, reduced cell viability and induced 2010; 67: 985–994. apoptosis. These findings suggest that AML-derived IgG has a role 10 Zheng J, Huang J, Mao Y, Liu S, Xun X, Zhu X et al. Immunoglobulin gene in AML pathogenesis. We also have attempted to compare the transcripts have distinct VHDJH recombination characteristics in human epithelial diagnostic classification and clinical outcome between AML cancer cells. J Biol Chem 2009; 284: 13610–13619. 11 Lee G, Ge B. Cancer cell expressions of immunoglobulin heavy chains with unique patients with and without expression of VHDJH transcripts. Although no significant differences were observed, the number carbohydrate-associated biomarker. Cancer Biomark 2009; 5: 177–188. 12 Lee G. Cancer cell-expressed immunoglobulins: CA215 as a pan cancer marker of patients included in this study is too small to draw any and its diagnostic applications. Cancer Biomark 2009; 5: 137–142. definitive conclusions. 13 Zhu X, Wu L, Zhang L, Hao P, Zhang S, Huang J et al. Distinct regulatory Interestingly, sequence analysis of AML-derived IgG showed mechanism of immunoglobulin gene transcription in epithelial cancer cells. Cell evidence of somatic hypermutation. Moreover, IgG VHDJH Mol Immunol 2010; 7: 279–286.

Leukemia (2013) 92 – 99 & 2013 Macmillan Publishers Limited IgG gene expression in acute myeloid leukemia X Qiu et al 99 14 Liu H, Zheng H, Duan Z, Hu D, Li M, Liu S et al. LMP1-augmented kappa intron 18 Chen W, Konoplev S, Medeiros LJ, Koppen H, Leventaki V, Vadhan-Raj S et al. enhancer activity contributes to upregulation expression of Ig kappa light chain Cuplike nuclei (prominent nuclear invaginations) in acute myeloid leukemia are via NF-kappaB and AP-1 pathways in nasopharyngeal carcinoma cells. Mol Cancer highly associated with FLT3 internal tandem duplication and NPM1 mutation. 2009; 27: 92–108. Cancer 2009; 115: 5481–5489. 15 Sun J, Yin CC, Cui W, Chen SS, Medeiros LJ, Lu G et al. Chromosome 20q deletion: 19 Meshinchi S, Woods WG, Stirewalt DL, Sweetser DA, Buckley JD, Tjoa TK et al. a recurrent cytogenetic abnormlaity in patients with chronic myelogenous leu- Prevalence and prognostic significance of flt3 internal tandem duplication in kemia in remission. Am J Clin Pathol 2011; 135: 391–397. pediatric acute myeloid leukemia. Blood 2001; 97: 89–94. 16 Jones D, Yao H, Romans A, Dando C, Pierce S, Borthakur G et al. 20 Yin CC, Medeiros LJ, Cromwell CC, Mehta AP, Lin P, Luthra R et al. Sequence Modeling interactions between leukemia-specific chromosomal changes, analysis proves clonal identity in five patients with typical and blastoid mantle cell somatic mutations, and gene expression patterns during progression lymphoma. Mod Pathol 2007; 20: 1–7. of core-binding factor leukemias. Genes Chromosomes Cancer 2010; 49: 21 Fais F, Ghiotto F, Hashimoto S, Sellars B, Valetto A, Allen SL et al. Chronic 182–191. lymphocytic leukemia B cells express restricted sets of mutated and unmutated 17 Patel KP, Ranvandi F, Ma D, Paladugu A, Barkoh BA, Medeiros LJ et al. Acute antigen receptors. J Clin Invest 1998; 102: 1515–1525. myeloid leukemia with IDH1 and IDH2 mutation: frequency and clinicopathologic 22 Ha K, Minden M, Hozumi N, Gelfand EW. Immunoglobulin gene rearrangment in features. Am J Clin Pathol 2011; 135: 35–45. acute myelogenous leukemia. Cancer Res 1984; 44: 4658–4660.

& 2013 Macmillan Publishers Limited Leukemia (2013) 92 – 99