ALBERT

Description: Abstract 1122 The survival of replicating B cells, with DNA damage arising from oxidative stress and/or activation-induced cytosine deaminase (AID), appears in part to be under p53 control. Importantly, a common C〉G single nucleotide polymorphism (SNP) within codon 72 of p53 influences p53 function. Among other differences, p53-72R (CGC=Arginine) is notably more effective than p53-72P (CCC=Proline) at inducing apoptosis. The SNP has been linked to clinical outcome in multiple settings, including malignancy. Most individuals in the US population display heterozygosity. In this study, we have examined whether B cells, whose genomic DNA is heterozygous for the codon 72 SNP, exhibit allelic exclusion at the level of expressed RNA. This was suggested by reported evidence that p53 expression is strongly regulated by gene methylation status; mRNA of peripheral blood cells from p53-72P/R heterozygous individuals is skewed toward the representation of only one SNP, depending upon ethnic status; and a p53 intron 2 SNP, representing a potential methylation site, is in strong linkage disequilibrium with the codon 72 SNP (PLoS ONE 6:e15320, 2011). Evidence for allelic exclusion of this functionally relevant p53 SNP would suggest that not all identically-stimulated B cells have equal likelihoods of survival within p53-72P/R heterozygous individuals. To investigate this issue, we first confirmed that p53 was expressed in non-transformed human B cells replicating in response to surrogate C3d-bound antigen, IL4 and BAFF. These physiologically-relevant stimuli synergize to induce a burst of T cell-independent clonal expansion, followed by apoptosis of many of the divided progeny (J. Immunol. 175:6143, 2005). Expression of p53 was monitored by immunoblotting and flow cytometry. Consistent with a role in regulating clonal burst size, p53 protein and mRNA/protein of p53-regulated pro-apoptotic genes were significantly elevated in blasts, prior to apoptosis. This contrasted with undectable p53 protein in non-stimulated B cells. To assess whether expressed p53 within a single lymphoblast derives from one allele, i.e. demonstrates allelic exclusion, we first identified tonsil donors heterozygous for the codon 72 polymorphism. This involved PCR-restriction fragment length polymorphism (RFLP) analysis of genomic DNA, as described by others (Leukemia Research 30:1113, 2006). Subsequently, purified resting B2 cells from cryopreserved tonsil cell suspensions, determined to be heterozygous for p53-72P/R, were labeled with CFSE and stimulated as above. At d5, single B lymphoblasts were sorted into 96 well PCR plates containing lysis buffer and cDNA prepared using random hexamer primers. A p53 sequence comprising exons 2a, 3, and 4 was subjected to two rounds of PCR amplification with the following primers, Forward: 5-cagccagactgccttccg-3 & Reverse: 5-gcaagtcacagacttggctg-3. Nucleotide sequencing of PCR-amplified p53 was performed commercially (Applied Biosystems Big Dye Terminator v3.1 cycle sequencing) and analyzed by chromasPro software. In two experiments, only 39 cells of a total of 236 assayed were positive for p53. By contrast, β-actin cDNA from two rounds of PCR amplification yielded 88% of the single cell-containing wells positive for actin. These yields indicate that p53 mRNA levels within a single cell are significantly more limiting than those for actin and are consistent with quantitative PCR of cDNA obtained from activated cell pools. Interestingly, analysis of the cDNA p53 chromatogram sequence of the p53-positive single cells (n=39), whose genomic DNA was heterozygous for the p53 codon 72 SNP, showed only a single homogenous sequence: either P (n=21) or R (n=18), but never both. This was in marked contrast to the chromatogram of cDNA derived from a pool of activated B cells within each experiment. In the latter cases, two overlapping peaks indicating co-expression of p72-P and p72-R were noted. Taken together, our findings suggest that p53 mRNA expression from a single non-transformed human B lymphocyte arises from the transcriptional activation of a single allele, i.e. shows allelic exclusion. Although the mechanism for this phenomenon requires further investigation, these results imply that B cells within individuals heterozygous for the functionally important p53-p72 polymorphism might vary considerably in their resistance to apoptosis. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 2415 Activation-induced cytidine deaminase (AID) is normally expressed in germinal center B cells and causes immunoglobulin somatic hypermutation and isotype class switching. High AID expression in B-cell chronic lymphocytic leukemia (CLL) associates with unmutated IGHV and cytogenetic aberrations that correlate with unfavorable prognosis. Mistargeted AID mutational activity may promote genetic changes leading to aggressive disease. To directly examine patient outcome and AID expression, peripheral blood cells from 126 CLL patients were tested for AID mRNA by nested PCR and then these results compared to time to first treatment (TFT) and overall survival (OS). In our cohort, patients have been followed for up to 393 months. AID+ CLL patients had a significantly shorter TFT (75 months; n = 53) compared to AID− CLL patients (243 months; n = 51)(P 〈 0.0025). AID+ patients also had a significantly shorter OS (134 months; n= 55) than AID− CLL patients (245 months; n =54)(P 〈 0.0001). AID+ patients also died at a significantly younger median age than AID− patients (74 versus 83 years, respectively)(P 〈 0.0001), yet both groups were diagnosed at about same median age (58 versus 61 years, P 〈 0.5879). These data imply that AID+ CLL patients have an inherently more aggressive and/or less treatment responsive disease. IGHV mutation status in this cohort was determined and compared to AID expression. Unmutated IGHV (U-CLL) was confirmed to significantly correlate with AID mRNA expression: 31/46 U-CLL versus 21/58 mutated IGHV CLL (M-CLL) cells express detectable AID (P 〈 0.0002). One hypothesis to explain the apparent paradox of finding elevated levels of AID expression in cells lacking IGHV mutations is that the initial leukemogenic event is independent of AID activity, but subsequent AID activation in the leukemic clone results in more aggressive disease. Although IGHV mutation status correlates with AID expression, a fair number of AID− U-CLL (33%) and AID+ M-CLL (36%) exist. AID+ expression in M-CLL significantly correlates with worse OS (183 versus 286 month median survival, P 〈 0.0275), whereas AID+ expression did not significantly discriminate OS in U-CLL patient (102 versus 151 month median survival, P 〈 0.2691). Eight common CLL cytogenetic aberrations at 13q14.3, 17p13.1, 11q22.3, 12-CEN, 13q34, 11q13, 14q32, 6q23.3 were determined by FISH in a subset of these patients. We confirmed that AID+ CLL tended to have more cytogenetic aberrations (n = 21, average aberrations = 1.6) than AID− CLL (n = 21, average aberrations = 1.0). Because 13q14.3 aberrations are not associated with unfavorable clinical courses, such cases were excluded from a subsequent analysis. In this assessment, AID expression correlated with the other genetic aberrations (AID+ = 1.0 average aberrations versus AID− = 0.2 aberrations), suggesting that 13q14.3 deletion may be an early genetic change that occurs independently of AID expression, while subsequent genetic changes are AID dependent. Finally, high CD38 levels correlate with poor CLL patient outcome. CD38 is induced upon activation of mature B cells and CD38+ CLL cells are enriched in cells engaged in the cell cycle. Furthermore, CD38 expression is highest among B cells in germinal centers, where AID is normally expressed. Thus, CD38 levels on CLL B cells were measured by flow cytometry on a subset of patients and compared to AID expression. AID expression (14/19) significantly correlated with high CD38 (≥ 30%) levels as compared to expression (5/20) in those with low CD38 levels (P 〈 0.0038). This suggests that AID expression is found in CLL patients with a higher proportion of activated, cycling cells. Thus, AID expression correlates with unmutated IGHV genotype, more cytogenetic aberrations, high numbers of CD38+ cells, and more aggressive disease as manifested by shorter TFT and OS. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 4606 Curcumin is a natural phenolic compound within the spice, Curcuma longa. It has noted anti-inflammatory effects, in large part due to its potent suppressive effect on the NF-kB signaling pathway. AID is a NF-kB-regulated enzyme, essential for B cell Ig class switch recombination and somatic hypermutation and recently shown to promote oncogenic transformation within both B cell and non-B cell lineages. This study has examined the effect of curcumin on the division-linked upregulation of AID protein and mRNA within several human B cell populations: in vitro-activated normal and CLL B lymphocytes and the AID-positive, pre-germinal center B cell line, CL-01. CFSE-labeled, IgM+ human B2 cells isolated from spleen/tonsil were pre-activated for 4–5 days with stimuli likely encountered in sites of inflammation, i.e. limiting surrogate C3dg-coated antigen (anti-IgM: anti-CD21: dextran) + IL-4 + BAFF. Peripheral blood B-CLL cells were activated with TLR-9 ligand, ODN-2006, + IL-15. Curcumin at doses from 6 to 50 μ M, and parallel DMSO vehicle controls, were pulsed into dividing B cell cultures (day 3, 4, or 5 of activation), and AID mRNA and protein assessed after 1 to 2 days. In experiments with CL-01 B cells, the kinetics of curcumin-induced AID suppression was further analyzed. Messenger RNA was monitored by both quantitative and qualitative RT-PCR; AID protein was assessed by two-color flow cytometry of CFSE-labeled cells and immunoblotting. The above experiments revealed that curcumin can significantly down-regulate AID mRNA and protein, in dose dependent fashion within each of the above B cell populations. Following a 16h pulse of curcumin (25 μ M), AID mRNA within CL-01 cells was inhibited by 60% (p=0.001), and accompanied by ~ 60% decrease in AID protein. Within cultures of replicating normal human B lymphocytes, a similar pulse of curcumin reduced total culture AID mRNA by an average of 70% in 3 experiments. AID protein in blasts representing 3–4 divisions was reduced by 79%, and in those representing 1–2 divisions by 58%, within a representative experiment. AID mRNA was evident within all in vitro-activated B-CLL clones tested (total = 6 clones at the time of submission). This was significantly reduced by a 15–24 hr pulse with curcumin (20-25 μ M): 42% inhibition (p=0.02)). The inhibitory effects of curcumin were evident in both IgHV mutated and unmutated clones. Within stimulated B-CLL assessed for AID protein (4 total clones, of which 2 were positive), a 20 hr pulse of curcumin at 40 μ M and 20 μ M reduced AID expression in one clone by 80% and 40%, respectively. In the other clone, a maximal tested dose of 20 μ M curcumin reduced AID protein by only 12%. Suppression of AID mRNA in the CL-01 cell line was noted as soon as 3 hours following exposure to curcumin and was preceded by inhibition of NF-kB activation, both baseline and BAFF-induced. The latter was determined through monitoring intracellular levels of phospho-p65-Ser(529) – an active phosphorylated form of NF-kB RelA. Taken together, these findings suggest that NF-kB- and AID-suppressing curcumin may be useful in reducing the risk of malignant transformation and B-CLL progression into more malignant subclones, as well as treating B cell autoimmune diseases driven by pathogenic, somatically-mutated IgG autoantibodies. Disclosures: No relevant conflicts of interest to declare.

Description: Exosomes are cell-generated nano-vesicles found in most biological fluids. Major components of their cargo are lipids, proteins, RNA, DNA, and non-coding RNAs. The miRNAs carried within exosomes reveal real-time information regarding disease status in leukemia and other cancers, and therefore exosomes have been studied as novel biomarkers for cancer. We investigated the impact of exosomes on cell proliferation in pediatric acute lymphocytic leukemia (PALL) and its reversal by silencing of exo-miR-181a. We isolated exosomes from the serum of PALL patients (Exo-PALL) and conditioned medium of leukemic cell lines (Exo-CM). We found that Exo-PALL promotes cell proliferation in leukemic B cell lines by gene regulation. This exosome-induced cell proliferation is a precise event with the up-regulation of proliferative (PCNA, Ki-67) and pro-survival genes (MCL-1, and BCL2) and suppression of pro-apoptotic genes (BAD, BAX). Exo-PALL and Exo-CM both show over expression of miR-181a compared to healthy donor control exosomes (Exo-HD). Specific silencing of exosomal miR-181a using a miR-181a inhibitor confirms that miR-181a inhibitor treatment reverses Exo-PALL/Exo-CM-induced leukemic cell proliferation in vitro. Altogether, this study suggests that exosomal miR-181a inhibition can be a novel target for growth suppression in pediatric lymphatic leukemia.

Description: Introduction: Liquid biopsies are minimal invasive and changing our understanding of tumor cell biology through the exploration of exosomes research in adult oncology. The clinical application of liquid biopsies to study exosomes as disease biomarkers with correlation to disease status is especially appealing in pediatrics. However in childhood cancer, the first exosome studies only begin to emerge in this exciting new field. Exosomes are microvesicles (30-100 nm) produced by normal as well as malignant cells in most biological fluids. Exosomes represent the fingerprint of the parental tumor and are loaded with unique biomarkers such as proteins, RNA, DNA and non-coding RNA, which may regulate tumor growth. One of the unique features of exosomes of interest to oncology research is the horizontal cargo transfer into target cells changing their biological properties. Our data present the first study to investigate the role of exosomes in acute lymphoid leukemia (ALL) in pediatric patients. Objectives: 1. Demonstrate transfer of ALL-exosomes into target cells with induction of cell proliferation 2. Confirm that leukemic exosome-induced paracrine and autocrine cell proliferation of leukemic cell lines is regulated by expression of proliferative and pro-survival genes with suppression of pro-apoptotic genes. Methods: Exosomes were isolated from serum of healthy donors (HD) and pediatric ALL (P-ALL) patients by ultracentrifugation. Exosomes were also isolated from conditioned medium (CM) of SUP-B15 (B-ALL), JM1 (B-ALL), and CL-01 (normal B cells) human cell lines. Proliferation assays were designed by exposing all three cell lines to different sources of leukemia-derived exosomes in a paracrine or autocrine fashion. After 24hr incubation, induction of cell proliferation was assessed by cell counting using trypan blue under the microscope and confirmed by colorimetric assay and gene expression analysis. Total RNA was extracted from cultured SUP-B15, JM1, or CL-01 cells and 2-5 µg of total RNA was used for cDNA synthesis. Primers sequences, probe number and gene accession numbers were identified from the universal probe library (UPL) of Roche Applied Science (Indianapolis, IN, USA). Each sample was assayed in duplicates. Proliferation was evaluated at the gene expression level for the following genes: Ki-67, PCNA (proliferation), MCL1, BCL2 (pro-survival) and BAD, BAX (pro-apoptotic). Fold change was calculated by comparing controls (naive) vs. exosome-induced cell lines both in autocrine and paracrine fashion. Data were analyzed with RQ manager version 1.2.1 (Applied Biosystems, Foster City, CA). The data were expressed as relative mRNA expression with reference to control samples and normalized to an endogenous reference gene (GAPDH). Results: Exosomes were characterized by NanoSight Tracking Analysis (NTA-Malvern) and western blot. Exosomal uptake and internalization into the cells was visualized by exosome labeling with a PKH67 fluorescent dye. We elucidated that CM-derived exosomes from SUP-B15 and JM1 cell lines induce cell proliferation in SUP-B15, JM1 (autocrine and paracrine for both cell lines) and CL-01 cells (paracrine) (control vs. exosomes, p

Description: Abstract 3888 In B-CLL, mutations in the IGHV genes encoding the BCR are correlated with disease aggressiveness: patients with unmutated BCR (U-CLL) typically have a worse prognosis than those with mutated BCR (M-CLL). Evidence that many U-CLL express BCR with specificity toward apoptotic cell antigens suggests that dying cells provide pro-survival signals for U-CLL clones. Thus progression of U-CLL may depend both on mechanisms that ensure the clone's survival and those that promote cell death. A single nucleotide polymorphism (SNP) within the p53 gene at codon 72 significantly affects p53 function: p53-72R (CG C=Arg) is substantially better at inducing apoptosis than p53-72P (CC C=Pro). Although p53-72P is linked with worse outcomes in a number of malignancies, p53-72R has been linked to heightened B-CLL incidence (Leuk Res. 2006;30:1113–8) and to diagnosis as U-CLL (Br J Haematol. 2011;153:533–5). Pro-apoptotic p53-72R may be fostering the apoptotic self antigen needed for BCR signaling in U-CLL. MDM2 is a ubiquitinase with a major role in keeping p53 levels low. Following DNA damage, p53 is phosphorylated by ATM and protected from MDM2. A SNP within the promoter affects MDM2 RNA/protein levels: SNP309G 〉 SNP309T. In some B-CLL cohorts, the SNP309G allele was associated with disease progression (Eur J Haematol. 2010;85:251–6; Blood 2010;115:4191–7). We investigated a cohort of 93 Caucasian B-CLL patients from the New York City metropolitan area for expression of the above p53 and MDM2 SNPs. We examined whether the B-CLL-expressed SNPs differ from Caucasian controls in the same geographic area (Lupus 2009;18:61–6). Additionally, we examined whether U-CLL and M-CLL patients in this cohort vary in SNP expression and assessed whether the above SNPs influence disease progression, as determined by time to first treatment (TTFT) and overall survival (OS). Isolated genomic DNA underwent PCR and pyrosequencing for the p53 SNP at codon 72 (rs1042422) and the MDM2 SNP309 (rs2279744) using primers selected by PSQ Assay Design. Primers used for p53: forward = Biotin-5'-GAAGACCCAGGTCCAGATGA and reverse = 5'-CCGGTGTAGGAGCTGCTG, resulting in an 82 bp PCR product; sequencing = 5'-GGTGCAGGGGCCACG. Primers used for MDM2: forward = 5'-ATTTCGGACGGCTCTCGC and reverse = Biotin-5'-CTAGTGACCCGACAGGCACCT, resulting in a 147 bp PCR product; sequencing = 5'-GGGCTGCGGGGCCGCT. Pyrosequencing was done on a PSQ HS96A instrument as per manufacturer's guidelines using PyroMark Gold Q96 Reagents. Our cohort of B-CLL was not significantly different from controls in distribution of the p53 SNP. However, there was a higher proportion of MDM2 SNP309G/G in B-CLL compared to controls (CLL=31%; normal=16%; p=0.0238). A p53-72R/R genotype was evident in a lesser proportion of U-CLL than M-CLL, albeit the difference was not statistically significant (n=80, p=0.1860). On the basis that U-CLL with p53-72R/R genotype might be favored if the clone underwent a deletion of p53-augmenting ATM, we compared the p53-72 SNP genotype of U-CLL ± FISH evidence of ATM deletion (11q22.3) (patients with 17p13 p53 deletion were excluded). Interestingly, the p53-72R/R genotype was significantly greater in U-CLL patients with ATM deletion than those without (n=27; p=0.0002). Comparison of p53-p72 SNP to TTFT showed trends in agreement with Majid, et al. (Br J Haematol. 2011;153:533–5). B-CLL patients with p53-72P/P had a shorter TTFT (P/P=41 months; R/R=97 months; n=88). However, due to limited power, this did not reach significance (p=0.1531). Also consistent with the above-cited study, the single M-CLL with p53-72P/P had the shortest TTFT, while M-CLL with p53-72R/R had the longest (n=48; p

Description: Abstract 3234 Prostaglandin E2 (PGE2) plays a pivotal role in inflammation. It is produced by COX-2-positive cells of the innate immune system, pre-neoplastic and neoplastic tissue, and activated B cells. p53 is a tumor suppressor with broad effects that include regulation of the cell cycle, DNA repair, and induction of apoptosis. A common single nucleotide polymorphism (SNP) at codon 72, within the proline rich domain of p53, can alter p53 protein structure, phosphorylation state, and function. The p53-72R (arginine) form is significantly more effective than p53-72P (proline) at promoting apoptosis in several cell lineages, but its function in B lymphocytes is unknown. COX-2 and p53 are often found co-expressed. A mechanistic link between the two might help explain why a COX-2/PGE2 axis augments the viability of cycling non-transformed human B lymphocytes (J. Immunol. 176:6736, 2006). This study tested the hypothesis that apoptosis-prone p53-72R-positive B lymphoblasts are preferentially rescued by PGE2. This analysis was aided by our discovery that p53 expression in activated B cells exhibits allelic exclusion. Thus, individual B cells whose DNA is heterozygous (HET) for p53-p72P/R express only p72P mRNA or p72R mRNA, but never both (Haque et al., submitted ASH abstract, 2011). In the present study, single cell RT-PCR of activated B cells bearing p72P/R HET DNA was used to examine whether expression of the p72P or p72R allelic forms is influenced by level of PGE2 exposure. Purified HET quiescent tonsillar B2 cells were activated in vitro by a surrogate for C3dg-coated antigen (anti-IgM:anti-CD21:dextran), IL4 and BAFF. To permit later differentiation of undivided and dividing subsets, cells were pre-labeled with CFSE. To supplement the low levels of PGE2 produced in density-limiting cultures, one set of parallel cultures received exogenous PGE2 (50 nM) on d2 and d4. On d5, harvested cells were analysed by FACS and sorted as one cell per well into 96-well PCR plates containing lysis buffer. RNA was isolated and cDNA prepared, using random hexamer primers. Two rounds of amplification of a sequence comprising p53 exons 2a, 3, and 4 involved primers, Forward 5-cagccagactgccttccg-3 & Reverse 5-gcaagtcacagacttggctg-3, and yielded a 400bp product upon semi-quantitative gel electrophoresis. Sequencing of PCR-amplified p53 cDNA from single cells was performed commercially (Applied Biosystems Big Dye Terminator v3.1 cycle sequencing) with analysis by chromasPro software. Within each experiment, an equivalent number of single sorted cells was analyzed for each cell subset (undivided or divided ± exogenous PGE2). The frequency of single sorted cells yielding p53 amplicons from cultures without added PGE2 was 6.1 ± 6.0 % and 12.8 ± 4.9 % for the undivided and divided subsets, respectively. In contrast, the corresponding frequency from PGE2-pulsed cultures was 8.1 ± 4.9 % and 23.8 ± 8.1 %, respectively. The higher yield of p53+ divided cells upon greater PGE2-exposure is in agreement with quantitative RT-PCR of p53 mRNA isolated from cell pools. PGE2-pulsed cultures had a 2.26 ± 0.15 fold increase p53 mRNA, as compared to non-pulsed cultures (p = 0.001). Sequence chromatograms of PCR-amplified p53 cDNA revealed the identity of the expressed p53 SNP. Consistent with conclusions of p53 allelic exclusion in activated human B cells (Haque et al, submitted ASH abstract), individual B cells of p72P/R heterozygous donors manifest either p72P or p72R, but never both. While the reportedly more pro-apoptotic p72R allele was expressed in 0% (0 of 5) p53+ divided blasts from cultures without supplementary PGE2, the allele was manifest in 62% (13 of 21) of the p53+ divided blasts from PGE2-supplemented cultures. PGE2 did not promote p72R allele representation in the undivided subset. Within p53+ undivided cells, p72R was found a frequency of 50% (4 of 8) and 25% (1 of 4) in PGE2 unsupplemented and supplemented cultures respectively. Thus, the p72R allele is underrepresented in replicating B cell blasts exposed to autocrine PGE2 alone, but expressed comparably to the p72P allele when higher viability-promoting concentrations of PGE2 are present. Taken together, the observations suggest that the p72R allele promotes greater activation-induced B cell apoptosis; PGE2 has a role in dampening p53-driven apoptosis within activated B cells; and the recovery of B cell blasts expressing the p53 p72R allele will be heightened in inflammatory settings. Disclosures: No relevant conflicts of interest to declare.