Fractions Enriched in Recently Born/Divided and Older/Quiescent Cells

Intraclonal Complexity in Chronic Lymphocytic Leukemia: Fractions Enriched in Recently Born/Divided and Older/Quiescent Cells

Carlo Calissano,1 Rajendra N Damle,1,2 Sonia Marsilio,1 Xiao-Jie Yan,1 Sophia Yancopoulos,1 Gregory Hayes,3 Claire Emson,3 Elizabeth J Murphy,3,4 Marc K Hellerstein,3 Cristina Sison,5 Matthew S Kaufman,6 Jonathan E Kolitz,1,2 Steven L Allen,1,2 Kanti R Rai,1,6 Ivana Ivanovic,7 Igor M Dozmorov,7 Sergio Roa,8 Matthew D Scharff,8 Wentian Li,1 and Nicholas Chiorazzi1,2,9

1The Feinstein Institute for Medical Research, North Shore–LIJ Health System, Manhasset, New York, United States of America; 2Department of Medicine, North Shore University Hospital, North Shore–LIJ Health System, Manhasset, and Hofstra North Shore–LIJ School of Medicine, Hempstead, New York, United States of America; 3KineMed, Inc., Emeryville, California, United States of America; 4Department of Medicine, University of California, San Francisco, California, United States of America; 5Biostatistics Unit, The Feinstein Institute for Medical Research, North Shore–LIJ Health System, Manhasset, New York, United States of America; 6Department of Medicine, Long Island Jewish Medical Center, North Shore–LIJ Health System, New Hyde Park, and Hofstra North Shore–LIJ School of Medicine, Hempstead, New York, United States of America; 7Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma, United States of America; 8Department of Cell Biology, Albert Einstein College of Medicine, Bronx, New York, United States of America; and 9Department of Molecular Medicine, Hofstra North Shore–LIJ School of Medicine, Hempstead, New York, United States of America

The failure of chemotherapeutic regimens to eradicate cancers often results from the outgrowth of minor subclones with more dangerous genomic abnormalities or with self-renewing capacity. To explore such intratumor complexities in B-cell chronic lymphocytic leukemia (CLL), we measured B-cell kinetics in vivo by quantifying deuterium (2H)-labeled cells as an indicator of a cell that had divided. Separating CLL clones on the basis of reciprocal densities of chemokine (C-X-C motif) receptor 4 (CXCR4) and cluster designation 5 (CD5) revealed that the CXCR4dimCD5bright (proliferative) fraction contained more 2H-labeled DNA and hence divided cells than the CXCR4brightCD5dim (resting) fraction. This enrichment was confirmed by the relative expression of two cell cycle–associated molecules in the same fractions, Ki-67 and minichromosome maintenance protein 6 (MCM6). Comparisons of global gene expression between the CXCR4dimCD5bright and CXCR4brightCD5dim fractions indicated higher levels of pro-proliferation and antiapoptotic genes and genes involved in oxidative injury in the proliferative fraction. An extended immunophenotype was also defined, providing a wider range of surface molecules characteristic of each fraction. These intraclonal analyses suggest a model of CLL cell biology in which the leukemic clone contains a spectrum of cells from the proliferative fraction, enriched in recently divided robust cells that are lymphoid tissue emigrants, to the resting fraction enriched in older, less vital cells that need to immigrate to lymphoid tissue or die. The model also suggests several targets preferentially expressed in the two populations amenable for therapeutic attack. Finally, the study lays the groundwork for future analyses that might provide a more robust understanding of the development and clonal evolution of this currently incurable disease. © 2011 The Feinstein Institute for Medical Research, www.feinsteininstitute.org Online address: http://www.molmed.org doi: 10.2119/molmed.2011.00360

INTRODUCTION tients live for long periods with the dis- with genomic changes (3), suggesting Chronic lymphocytic leukemia (CLL) ease, many undergo progressive decline, that clonal evolution is a key factor in the is a relatively frequent, incurable B-cell leading to demise. Progression to a more disease. malignancy (1,2). Even though some pa- aggressive disease is often associated We previously found that CLL clones are composed of subpopulations of cells that proliferate at different rates (4), as Address correspondence and reprint requests to Nicholas Chiorazzi, The Feinstein Institute measured by in vivo deuterium (2H)- for Medical Research, 350 Community Drive, Manhasset, NY 11030. Phone: 516-562-1090; incorporation into newly synthesized Fax: 516-562-1011; E-mail: [email protected]. DNA of dividing cells (5,6). The most Submitted September 22, 2011; Accepted for publication September 22, 2011; Epub proliferative fraction of a cancer clone is (www.molmed.org) ahead of print September 23, 2011. of major interest for several reasons.

1374 | CALISSANO ET AL. | MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 RESEARCH ARTICLE

First, the “proliferative compartment” 2H Measurements by Gas USA). Purified a-RNA was quantified, may contain cells that developed new Chromatography/Mass Spectrometry and the fragment size was ascertained on structural DNA abnormalities leading to and Calculation of the Fraction of a Bioanalyzer (Agilent Technologies) more lethal disease. Furthermore, the Labeled Cells using an Agilent RNA 6000 Nano Kit. most recently born fraction may be prog- Peripheral blood mononuclear cells Labeled a-RNA was hybridized to a eny of putative leukemic stem cells. Fi- were separated from heparinized venous Human WG-6 v3.0 Expression BeadChip nally, such cells would be potential tar- blood and leukocyte-enriched fractions containing 48,804 probes and stained gets for therapies to abort clonal by density gradient centrifugation using using streptavidin-Cy3 as per the manu- evolution. Ficoll-Paque (Pharmacia LKB Biotechnol- facturer’s instructions. The signal gener- We therefore studied the kinetic com- ogy, Piscataway, NJ, USA) and cryo - ated was detected by high-performance plexity of individual CLL clones to deci- preserved until use. Calculation of the laser optics of the BeadArray Reader (Il- pher fundamental insights about the fraction of newly divided cells was per- lumina, San Diego, CA, USA). Data were pathophysiology of the disease. In partic- formed after determination of 2H enrich- normalized using quantile normalization ular, we focused on further characteriz- ment in plasma or saliva and of 2H en- by BeadStudio software (Illumina). To ing the proliferative and resting compart- richment in deoxyadenosine of genomic determine differentially expressed genes, ments using differences in the densities DNA as described (4). the CXCR4dimCD5bright versus CXCR4bright of a surface membrane molecule upregu- CD5dim expression value ratio was com- lated after normal B-cell activation (clus- Isolation of Cell Fractions on the Basis puted for each patient and log-trans- ter designation 5 [CD5]) and another in- of Expression of CXCR4 and CD5 formed, and a Student t test was per- volved in maintaining B-cell contact with Peripheral blood mononuclear cells formed using R (www.Rproject.org). stromal elements of solid lymphoid tis- were incubated with murine anti–human Significant gene differences had 1.3 or sues (chemokine [C-X-C motif] receptor 4 monoclonal antibodies (mAbs): CD5- greater fold change and P < 0.01; for fold [CXCR4]). Using samples from patients FITC (fluorescein isothiocyanate), change, we used the nth root of the prod- for which CLL cells had been labeled in CXCR4-PE (phycoethrin), CD3–peridinin- uct of all individual ratios, or the geo- vivo with 2H, we divided clones into sub- chlorophyll-protein (CD3-PerCP) and metric mean of the ratio, for n samples. fractions enriched in the most prolifera- CD19-allophycocyanin (CD19-APC) (all Genes were assigned to specific func- tive and most quiescent compartments. from BD Biosciences, San Jose, CA, tional categories using Ingenuity Path- These fractions were then further charac- USA). After gating on CD19+CD3–CD5+ way Analysis (IPA, www.ingenuity.com), terized by comparing expression of events, cells were sorted with a BD Panther Classification System (Panther, genes encoding molecules usually upreg- FACSAria™ (Becton Dickinson Immuno- www.pantherdb.org), DAVID Bioinfor- ulated in dividing or resting populations. cytometry Systems, San Jose, CA, USA) matics Resources 6.7, National Institute of Finally, to provide a robust membrane on the basis of intensity of CXCR4 and Allergy and Infectious Diseases (NIAID)/ map of these compartments that might CD5 (Figures 1A, B), and cell pellets National Institutes of Health (NIH) be used for further characterization and were stored at –80°C until performing (DAVID, http://david.abcc.ncifcrf.gov/ therapeutic targeting in patients, an ex- gas chromato graphy/mass spectrometry home.jsp) and GeneCards Batch Queries tended surface phenotype was defined or RNA extraction. (GeneALaCart, www.genecards.org/ with a larger patient cohort. BatchQueries/index.php). Significantly Gene Expression Profiling and Gene different gene expression values were MATERIALS AND METHODS Expression Data Analyses clustered hierarchically on the basis of av- The quality of RNA extracted from erage linkage on a correlation-based dis- Patients sorted cells using the RNeasy minikit tance (defined at http://arxiv.org/abs/ The Institutional Review Board of the (Qiagen, Valencia, CA, USA) was as- cs/ 0402061). Heatmaps of gene expres- North Shore–LIJ Health System approved sessed using an Agilent RNA 6000 Pico sion values for identified relevant func- these studies. After obtaining informed Kit (Agilent Technologies, Colorado tional categories were generated using consent in accordance with the Declara- Springs, CO, USA). A total of 50 ng total heatmap.2 in the gplots package of R. tion of Helsinki, venous blood was col- RNA was subjected to first- and second- lected from randomly chosen CLL pa- strand reverse transcription followed by Measurements of Surface and tients diagnosed by established criteria. A a single in vitro transcription amplifica- Intracellular Antigens by Flow total of 15 subjects participating in the tion that incorporated biotin-labeled nu- Cytometry 2 H2O protocols were studied. Patients cleotide to yield biotinylated antisense Peripheral blood mononuclear cells 2 drank H2O for 6–12 weeks, depending RNA (a-RNA) using a TargetAmp were incubated with different combina- on the protocol, and cells were studied at Nano-g Biotin-aRNA labeling kit (Epi- tions of murine anti–human mAbs di- two time points during this period. centre Biotechnologies, Madison, WI, rectly conjugated with the indicated fluo-

MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 | CALISSANO ET AL. | 1375 INTRACLONAL COMPLEXITY IN CHRONIC LYMPHOCYTIC LEUKEMIA

try systems) and analyzed by FlowJo v7.2.4 version. Within each CLL clone, relative expression in terms of percent positive cells and mean fluorescent intensity of each surface or intracellular marker was determined and compared.

All supplementary materials are available online at www.molmed.org.

RESULTS CLL clones can be divided into distinct fractions on the basis of inverse surface expression of CXCR4 and CD5. By screening a panel of chemokine receptors expressed on clones from a series of CLL patients (4), we found an inverse rela- tionship between CXCR4 and CD5 densities, with differing shapes among patients (Figures 1A–D). The abundance of cells representing the extremes of these dim bright Figure 1. Intraclonal kinetic differences in CLL fractions defined by CXCR4 and CD5. two populations (CXCR4 CD5 ver- (A–D) Contour plots of CLL cells on the basis of CXCR4 and CD5 levels. An inverse rela- sus CXCR4brightCD5dim) was relatively tionship giving a crescent shape is clear in ~35% of analyses (A), evident in ~50% (B), sug- small and varied between patients 2 gested in ~10% (C) and absent in a few patients (D). (E) Representative example of H2O (1–8%, not shown), with the majority of 2 enrichment in plasma (CLL875) of patients drinking H2O. CLL cells were flow-sorted on the cells (>90%) falling into an intermediate basis of CXCR4 and CD5 densities at time points indicated by arrows. The vertical dashed category (CXCR4intCD5int; Figures 1A–D). line indicates end of 2H O consumption. (F) Average fraction of 2H-labeled cells in the 2 We reasoned that high CD5 density three fractions defined by CXCR4/CD5 densities. *CXCR4dimCD5bright fraction contains sig- would reflect cellular activation as in nor- nificantly higher levels of 2H-labeled DNA compared with CXCR4intCD5int and CXCR4brightCD5dim fractions (P < 0.01 for both). (G) Plot indicates the ratio of 2H-labeled mal human B cells (7), and low CXCR4 cells in CXCR4dimCD5bright over CXCR4brightCD5dim fractions, indicating that CXCR4dim levels would identify cells that internal- CD5bright cells proliferated ~11 times more than CXCR4brightCD5dim cells (*P < 0.0001). ized the receptor because of an activation (H, I) CXCR4dimCD5bright fractions are significantly enriched in cells expressing Ki-67 (H) and event and thereby passaged from a lym- MCM6 (I) compared with CXCR4intCD5int and CXCR4brightCD5dim fractions (*P < 0.01 for both). phoid tissue to the periphery (8). As detailed elsewhere (5), incorporation of 2H 2 rochrome: anti–CD5-FITC or -APC; (PBS + 10% fetal bovine serum + 1% into cellular DNA in patients given H2O CXCR4-PE, -APC, or -PECy7; CD3- sodium azide) were incubated with pri- is a direct measure of newly synthesized PerCP; CD19-PerCP, -APC, -Pacific blue mary antibody for 30 min at 4°C and DNA and hence cell division, thereby al- or -AmCyan; CD23, CD27, CD52, CD38, then exposed to secondary antibody lowing study of birth rates of CLL cells in CCR7, CXCR3, BAFF-R, CTLA4, CD11a, for 25 min at 4°C, followed by fixation vivo (6,9,10). Therefore, we sorted CLL γ 2 CD11c, DR5, CD49d, CD62L and Fc RIIb, with 0.1% formaldehyde in PBS. For cells from nine patients consuming H2O all in PE; and CD20-FITC (all from BD intracellular detection of Ki-67 and (Figure 1E) into three fractions and quan- Biosciences). Unconjugated rabbit IgG minichromosome maintenance protein 6 tified 2H-labeled deoxyadenosine in each antibody to receptor tyrosine kinase-like (MCM6), after membrane staining with fraction by gas chromatography/mass orphan receptor 1 (ROR-1) was provided anti–CXCR4-PE, –CD3-PerCP, –CD5-APC spectrometry. This step indicated that at by Christoph Rader (National Cancer and –CD19- Pacific Blue (BD Biosciences), 21 d, the CXCR4dimCD5bright fraction was Institute/NIH, Bethesda, MD, USA). Sec- cells were fixed and permeabilized markedly enriched in divided cells com- ondary staining was performed with PE- (Cytofix/ Cytoperm, BD Biosciences) and pared with the remainder of the clone, conjugated goat anti–rabbit IgG antibod- incubated with murine anti–Ki-67-FITC and this result lasted through 42 d (Fig- ies (Southern Biotech, Birmingham, AL, or anti–MCM6-FITC mAbs (BD Bio- ure 1F); this disparity was reflected by USA). sciences). Data were acquired with a BD significantly different CXCR4dimCD5bright For surface membrane immunofluo- LSRII flow cytometer or a FACS calibur to CXCR4brightCD5dim enrichment ratios at rescence, cells (2 × 105) in FACS buffer (both Becton Dickinson Immunocytome- the two time points (12.6 at 21 d and 10.9

1376 | CALISSANO ET AL. | MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 RESEARCH ARTICLE

at 42 d, Figure 1G; P < 0.0001). See Sup- plementary Figure 1 for data describing the kinetics of each fraction for each patient. These findings were verified by ana- lyzing ex vivo, in the CXCR4/CD5 fractions, two cell cycle–related molecules, Ki-67 and MCM6, both expressed from the G1 to M phase. Comparisons of CXCR4dimCD5bright versus CXCR4intCD5int versus CXCR4brightCD5dim fractions in an expanded cohort of CLL patients were significantly different (P < 0.01) for both Ki-67 and MCM6 (Figures 1H–I). Collectively, these in vivo and ex vivo findings indicate that the CXCR4dim CD5bright and CXCR4brightCD5dim fractions of CLL clones are highly enriched compared with the total circulating leukemic load, in recently divided/young and more resting/older members, respectively. For simplicity, we refer to these as the “proliferative” and “resting” compartments going forward, even though the dominant intermediate population contributes cells with similar phenotypes to a significant extent.

Proliferative and Resting Fractions Exhibit Gene Expression Differences Consistent with Their Putative Contrasting Proliferative Histories Global gene expression profiling (GEP) was performed on isolated fractions from the same patients. Using Illumina Hu- manWG-6 v3.0 expression arrays to measure the relative expression of 25,440 genes and selecting genes for ≥1.3-fold difference in expression and P ≤ 0.01, we defined 1,299 genes differentially expressed between the two compartments; 715 genes were more highly expressed in the proliferative and 584 in the resting fraction (Supplementary Table 1). These genes were then segregated into cate- Figure 2. Heatmaps illustrating genes typically expressed differentially between the prolifer- gories that would support their difference ative versus resting fractions of CLL clones. (A) Pro- and antiproliferation genes up- or in time from cellular activation and divi- downregulated in the proliferative versus resting fractions of the nine CLL clones analyzed. sion and in migratory capacities using a (B) Antiapoptotic genes. (C) Genes related to oxidative injury. (D) Trafficking-related genes. (A–D) Genes presented are significantly differentially expressed (Supplementary Table 1) number of bioinformatic programs and and were assigned to categories using Ingenuity Pathway Analysis, Panther Classification databases (see Materials and Methods). System, DAVID Bioinformatics Resources 6.7 and GeneCards Batch Queries. Significantly dif- Genes involved in cell proliferation. ferentially expressed gene expression values for each set were clustered using hierarchical We determined the extent that the two clustering on the basis of the average linkage on a correlation-based distance. fractions differed in expression of genes

MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 | CALISSANO ET AL. | 1377 INTRACLONAL COMPLEXITY IN CHRONIC LYMPHOCYTIC LEUKEMIA

involved in normal or abnormal cell proliferation. Consistent with the CXCR4dimCD5 bright fraction being enriched in divided cells, 20 pro-proliferation genes were found in this compartment (Figure 2A, top) and 10 in the CXCR4brightCD5dim fraction. In contrast, a larger number of genes with an antiproliferation function were found in the CXCR4brightCD5dim fraction than the CXCR4dimCD5bright fraction (12 genes versus 7; Figure 2A, bottom). Among the pro-proliferation genes upregulated in the CXCR4dimCD5bright fraction were two canonical molecules involved in cell cycle progression: Cyclin E1 (CCNE1) and cyclin D2 (CCND2). Figure 3. Extended immunophenotype of proliferative and resting CLL fractions. Surface dim bright bright dim CCND2 plus Ki-67 and MCM6, which markers were analyzed by flow cytometry in CXCR4 CD5 (X) and CXCR4 CD5 were enriched in this fraction on the (G) fractions. All surface molecules are depicted but CCR7 and CTLA4 were greater in CXCR4dimCD5bright fractions, in terms of percent positive cells (A) or mean fluorescence in- basis of flow cytometry (Figure 3), have tensity (B) or both. CCR7 and CTLA4 are higher in the CXCR4brightCD5dim fraction. *P < 0.01 been found higher in CLL lymph nodes and §P < 0.05 (paired Student t test). than peripheral blood (11). Notably, of the pro-proliferation genes that were higher in resting cells (Figure 2A, top), B-cell activating factor that mediates sur- Products of catalase (CAT), glucose-6- myocyte-specific enhancer factor 2C vival in normal and non-Hodgkin lym- phosphate dehydrogenase (G6PD), glu- (MEF2C) and midkine (MDK; neurite phoma B lymphocytes (18,19). TNFSF15 tathione peroxidase 1 (GPX1), heme oxyge- growth-promoting factor 2) promote codes a ligand for TNFRSF25 and nase (decycling) 1 (HMOX1) and growth of cell types that are often in a TNFRSF6B that is not known to be ex- thioredoxin (TXN) have a role in scaveng- resting state: hematopoietic progenitors pressed in B cells and that inhibits ing free radicals. (12) and embryonic stem cells (13), growth in other cells (20). Finally, inter- Genes involved in cell trafficking. respectively. leukin (IL)-4R, which promotes normal Lastly, 33 trafficking-related molecules Of the antiproliferative genes higher in B-cell and CLL cell survival (21) and pro- were expressed at higher levels in the the CXCR4brightCD5dim fraction (Figure 2A, liferation (22) and is expressed by CLL proliferative and 8 in the resting frac- bottom), BCL6, BACH2 and RGS2 are of in- cells (23), is upregulated in this fraction. tions (Figure 2D). Consistent with cells of terest because they control checkpoints in Prodeath genes are slightly more abun- the proliferative compartment having re- B-cell activation and maturation (14–16). dant in the resting fraction, suggesting cently exited lymphoid tissues is the Genes involved in cell survival. Re- that this compartment contains older, heightened expression of integrin, beta 7 cently born/divided cells would also less robust cells (Figure 2B, top). For ex- (ITGB7) and CXCR3. ITGB7 remains on likely receive prosurvival/antiapoptotic ample, harakiri (HRK), forkhead box O3 the surface of cells after leaving solid tis- signals, whereas cells of the resting com- (FOXO3) and leptin (LEP) (24–26) stimu- sues and is involved in passage of cells partment would be less likely to receive late apoptosis in various cell types, and into mucosa-associated lymphoid tissues these, being temporally further from an transmembrane 123 (TMEM123) encodes a (29,30); CXCR3 is involved in CLL cell activation signal and trophic support of mucinlike gene that leads to oncotic cell migration along interferon (IFN)- tissue microenvironments. A total of 25 death upon cross-linking (27). inducible protein 10 and IFN-γ–induced antiapoptotic genes are higher in the pro- Genes involved in oxidative injury. monokine gradients (31). However, a liferative versus 8 in the resting fraction Cells of the proliferative fraction upregu- number of other genes do not readily fit (Figure 2B, bottom). Of note, MEF2C ap- lated more genes involved in the genera- the profile of recent emigrants of lym- pears necessary for B-cell proliferation tion or repair of oxidative injury than in phoid tissues. In particular, overexpres- and survival after in vitro B-cell receptor resting cells (27 genes versus 6; Fig- sion of adhesion molecules platelet en- (BCR) stimulation (17). Several genes up- ure 2C), likely because of increased meta- dothelial cell adhesion molecule 1 (PECAM1), regulated in the resting fraction are of in- bolic activity in the dividing/recently di- cluster designation 11A (lymphocyte terest. TNF receptor superfamily 13C vided cells. In addition, oxidative injury function-associated antigen 1)/ integrin, (TNFRSF13C) is the receptor for the is a known feature of CLL cells (28). alpha L (CD11a/ITGAL) and cluster

1378 | CALISSANO ET AL. | MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 RESEARCH ARTICLE

designation 11c/ integrin, alpha X (comple- ment component 3 receptor 4 subunit) (CD11c/ITGAX), the latter two of which were confirmed by quantitative reverse- transcriptase polymerase chain reaction (qRT-PCR) (Supplementary Figure 2) and flow cytometry (Figure 3), seems para- doxical, since these are usually expressed on cells docking on the endothelium and exiting the circulation. Cells of the resting compartment expressed more pleckstrin homology domain containing, family A member 2 (PLEKHA2) mRNA (Figure 2D) and CCR7 and CTLA4 protein (Figure 3). CCR7 promotes homing of normal B lymphocytes Figure 4. Hypothetical model of the lifecycle of a CLL B cell. Part 1: CLL cells rest on the and CLL cells to lymph nodes and medi- stroma because of at least CXCR4-CXCL12 interactions. When stimulated, cells are acti- vated and divide, upregulating CD5, internalizing CXCR4 and detaching from stroma. The ates neoplastic B-cell homing in patients process could be ligand-induced (for example, BCR or toll-like [TLR] or other pathways) or with widespread nodal dissemination spontaneous. Low CXCR4 levels increase the chances of recently divided CLL cells (32). Also, CCR7 and CXCR4 promote (CXCR4dimCD5bright phenotype) exiting solid tissue and reaching peripheral blood. Part 2: metastasis of solid tumors (33). Recently born/divided CLL cells reach peripheral blood as members of the PLEKHA2 has been found to be highly CXCR4dimCD5bright fraction. Over time, possibly because of a lack of trophic input from the expressed in poor outcome CLL cells solid tissue microenvironment, cells begin to reexpress CXCR4 to trek back to nutrient-rich (unmutated IGHV or high ZAP-70 levels) niches. This leads to expression of a CXCR4intCD5int and then CXCR4brightCD5dim mem- and to have a role in adhesion to fi- brane phenotype. The model considers the three fractions to be linked as a continuum. bright dim bronectin and laminin (34). Part 3: CXCR4 CD5 CLL cells have the greatest chance of detecting and following a CXCL12/SDF1 gradient, thereby reentering lymphoid solid tissue and receiving prosur- qRT-PCR Analyses of Selected Genes vival stimuli. Those that do not reenter die by exhaustion. Confirm Their Overexpression in Proliferative or Resting Fractions migration or are targets of mAbs in use thesized DNA of dividing cells as a mea- We performed qRT-PCR for genes cod- in humans. The proliferative compart- sure of proliferation, we previously ing molecules expressed on or at B-cell ment contained significantly more cells demonstrated intraclonal kinetic com- surfaces to provide a guide for develop- that expressed, at higher densities, CD38, plexity in CLL, defining a small popula- ing an extended surface membrane phe- CD20, CD23, CD27, CD52, JML-1, DR5, tion (~0.1–1% of the leukemic clone) notype and to corroborate the GEP data. CXCR3, CD49d, CD62L, CD11a and that divided daily (6). Because the 2H- A total of 67.8% (38/56) of genes were CD11c (Figures 3A, B). Only ROR-1, incorporation approach does not permit confirmed as significantly different be- FcγRIIb and BAFF-R exhibited differ- visualization of individual cells that tween the two fractions (Supplementary ences solely on the basis of density. Two have divided, we subsequently enriched Figure 2). Expression levels of CD5 and molecules were higher in the resting for cells of the proliferative fraction CXCR4 served as internal controls. compartment: CCR7 (both in the percent using surface expression of a candidate of positive cells and intensity) and molecule, CD38, that is expressed on ac- Surface Molecules Further Distinguish CTLA4 (intensity). It is noteworthy that tivated human B cells (47) and is associ- the Proliferative and Resting several of these molecules are targets of ated with poor clinical outcome in CLL Compartments mAbs or compounds already in clinical (48). These studies indicated that within Finally, to permit an even more precise use (35–41) or are being tested in human each CLL clone, the CD38+ fraction con- definition of the two fractions in CLL subjects (23,42–46). tained ~2.5 times more recently divided and to define molecules that might be cells than the CD38– fraction (4). In the valuable therapeutic targets for these DISCUSSION present study, using differential surface fractions, the CXCR4/CD5 subsets were The aim of this study was to further membrane densities of CXCR4 and CD5, subjected to a detailed multiparameter delineate and isolate cells with different we isolated fractions differing by phenotypic analysis (Figures 3A, B). Mol- kinetics in CLL clones, focusing on cells ~11-fold in the percentage of recently di- ecules were studied because they are that had recently divided or are resting. vided cells (Figure 1G). In line with our markers of B-cell activation, survival and Using 2H-incorporation into newly syn- previous study, CD38 expression was

MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 | CALISSANO ET AL. | 1379 INTRACLONAL COMPLEXITY IN CHRONIC LYMPHOCYTIC LEUKEMIA

significantly higher in the CXCR4dim model, the intermediate fraction, which the latter possibility is the case, a careful CD5bright proliferative subset (Figure 3A). is the bulk of the clone, links the ex- analysis of the resting compartment Finding increased numbers of cells extremes and is the fraction from which might yield a subset of cells with stem- pressing Ki-67 and MCM6 in the same most of our current knowledge on circu- like/initiating capacities. fraction confirmed the accuracy of this lating CLL cells is derived. This relation- Delineation of intratumor kinetic het- approach (Figures 1H, I). Finally, global ship is supported by the fact that BCR erogeneity may also have therapeutic GEP on the same CXCR4/CD5 subsets stimulation leads to downregulation of implications. Although most anticancer further validated that our approach CXCR4 expression (49), as depicted in therapies are predicated on eliminating identified subsets of CLL clones differ- Figure 4, part 1. Although our data are the entire neoplastic clone, preferen- ing in proliferative histories, since the consistent with this model, the process tially targeting the proliferative and CXCR4dimCD5 bright fraction overex- remains hypothetical because we have resting compartments might be effica- pressed more genes that support cell only studied cells in the peripheral cir- cious. Thus, preferentially eliminating division (Figure 2A), block apoptosis culation. Notably, it was recently shown the most recently born/divided cells (Figure 2B) and induce/repair oxidative by comparative GEP of bone marrow would limit the fount of clonal evolu- damage (Fig ure 2C) than the resting and lymph node cells that lymph nodes tion; preventing cells from reentering fraction, which contained more genes may host most of the CLL proliferative solid tissue would thwart the less ro- that inhibit cell division (Figure 2A) and events (11). In our study, cells with low bust, apoptosis-prone resting cells from survival (Figure 2B). Nevertheless, these CXCR4 and high CD5 levels could in- receiving survival signals. Collectively, fractions remain heterogeneous, likely clude emigrants from any site, and only these approaches would lock a clone containing cells from the intermediate one of the differentially expressed mole- into a steady state of genetic abnormali- fraction with characteristics distinct cules involved in trafficking that we de- ties and eventually lead to clonal from those we are describing. Thus, the fined is site specific: ITGB7 suggests spe- shrinkage based on spontaneous cell situation is complex, and a more expan- cific passage to mucosa-associated death (6,55) and death due to survival sive collection and analysis of such data lymphoid tissues. Considering that at signal deprivation (56). will better elucidate unique features of least some CLL cells may be derived Although speculative, this approach the two compartments. Still, the detec- from a human B-1 cell equivalent (50,51), could be tested, since relevant com- tion of gene (Figure 2) and protein the potential of migration to such tis- pounds are already in clinical use or (Figures 1 and 3) expression differences sues, which is favored for murine B-1 evaluation. For example, the proliferative consistent with recently divided and cells (52), is provocative. compartment expresses higher levels of resting cells, combined with the direct This hypothetical lifecycle also sug- CD20, CD23, CD38, CD49d, CD52, in vivo demonstration of significant en- gests “stem-like” or “initiating” capabili- CD11a and DR5 (Figure 3), all of which richment in cycled cells in the prolifera- ties of CLL subclones (53). For instance, can be targeted by specific, available tive fraction and a markedly reduced the CXCR4brightCD5dim cells could be mAbs (23,35–46). Other targets of poten- abundance in the resting fraction (Fig - nestled on stromal elements (Figure 4, tial interest in the proliferative fraction ure 1), confirm that using CXCR4/CD5 part 1) before being released and giving are CXCR3 (GEP, qRT-PCR, flow), CD27 densities on CD19+ cells is a major im- rise to the dividing CXCR4dimCD5bright (flow), ITGB7 (GEP and qRT-PCR), Ly96 provement in the delineation of these cells of the proliferative and bulk frac- (GEP), CD62L (flow) and CD11c/ITGAX two compartments. tions (Figure 4, part 2); therefore, these (flow and GEP). For the resting compart- The data conjure a lifecycle for indi- CXCR4brightCD5dim resting cell fractions ment, the most relevant molecules ap- vidual CLL cells representing a contin- could be a distinct self-renewing subset pear to be CXCR4 and CCR7 (flow and uum between the CXCR4dimCD5bright, from which all clonal members emanate. GEP). Plerixaflor, a CXCL12 (CXCR4 lig- CXCR4intCD5int and CXCR4 brightCD5dim In a related scenario, these same cells and) inhibitor used for stem cell mobi- fractions (Figure 4). At one extreme is could be members of the CXCR4bright lization, is being tested in CLL treatment the proliferative fraction, highly en- CD5dim fraction of the resting compart- (39). Although no compound is presently riched in young, vital cells that recently ment of the blood on their way back to a available to target CCR7 in vivo, an anti- left a solid lymphoid tissue where acti- sustaining niche in a lymphoid tissue CCR7 mAb kills CLL cells in vitro with- vation and proliferation occurred. At the (Figure 4, part 3), where they might di- out affecting T cells that also express other end is the resting compartment, vide again. The former possibility would CCR7 (57). Thus, an anti-CXCR4 and/or containing older, less robust cells that be consistent with a B-1–like cell, because anti-CCR7 approach could limit the abil- may have been circulating in the periph- in mice these cells have self-renewal ca- ity of CXCR4bright and CCR7bright cells to ery longer and are attempting through pacity (54), implying that the ability to reenter solid tissues, fostering resting high CXCR4 levels to migrate into a migrate effectively would be the key fac- compartment apoptosis. Finally, target- solid tissue niche to avoid death. In this tor determining survival of the clone. If ing IL-4R and CTLA4 is feasible because

1380 | CALISSANO ET AL. | MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 RESEARCH ARTICLE

a recombinant IL-4–Pseudomonas exo- 9. van Gent R, et al. (2008) In vivo dynamics of sta- regulation of human female reproduction. Adv. toxin fusion protein that binds IL-4R (23) ble chronic lymphocytic leukemia inversely cor- Exp. Med. Biol. 665:227–41. and overcomes apoptosis resistance of relate with somatic hypermutation levels and 26. Samuel-Mendelsohn S, et al. (2011) Leptin signal- suggest no major leukemic turnover in bone ing and apoptotic effects in human prostate can- CLL cells (58) is in clinical use, and marrow. Cancer Res. 68:10137–44. cer cell lines. Prostate. 71:929–45. mAbs reactive with CTLA4 (40,41) are 10. Defoiche J, et al. (2008) Reduction of B cell 27. Ma F, Zhang C, Prasad KV, Freeman GJ, Schloss- currently being used to potentiate T-cell turnover in chronic lymphocytic leukaemia. Br. J. man SF. (2001) Molecular cloning of Porimin, a responses against tumors (59). Haematol. 143:240–7. novel cell surface receptor mediating oncotic cell 11. Herishanu Y, et al. (2010) The lymph node mi- death. Proc. Natl. Acad. Sci. U. S. A. 98:9778–83. ACKNOWLEDGMENTS croenvironment promotes B-cell receptor signal- 28. Zhou Y, Hileman EO, Plunkett W, Keating MJ, ing, NF-κB activation, and tumor proliferation in Huang P. (2003) Free radical stress in chronic This work was supported in part by chronic lymphocytic leukemia. Blood. 117:563–74. lymphocytic leukemia cells and its role in cellu- the National Cancer Institute/NIH 12. Stehling-Sun S, Dade J, Nutt SL, DeKoter RP, lar sensitivity to ROS-generating anticancer (CA81554) and by philanthropic contri- Camargo FD. (2009) Regulation of lymphoid ver- agents. Blood. 101:4098–104. butions from The Karches Foundation, sus myeloid fate ‘choice’ by the transcription fac- 29. Parker CM, Cepek KL, Russell GJ, et al. (1992) A tor Mef2c. Nat. Immunol. 10:289–96. Prince Family Foundation, Marks Foun- family of beta 7 integrins on human mucosal lym- 13. Yao X, et al. (2010) Promotion of self-renewal of phocytes. Proc. Natl. Acad. Sci. U. S. A. 89:1924–8. dation, Jerome Levy Foundation, Leon embryonic stem cells by midkine. Acta. Pharma- 30. Postigo AA, Sanchez-Mateos P, Lazarovits AI, Levy Foundation, Andrew and Mona col. Sin. 31:629–37. Sanchez-Madrid F, de Landazuri MO. (1993) Albert Fund Inc., and Joseph Eletto 14. Ochiai K, Muto A, Tanaka H, Takahashi S, Alpha 4 beta 7 integrin mediates B cell binding Leukemia Research Fund. We thank Igarashi K. (2008) Regulation of the plasma cell to fibronectin and vascular cell adhesion Aarti Damle and The Feinstein Institute’s transcription factor Blimp-1 gene by Bach2 and molecule-1: expression and function of alpha 4 Bcl6. Int. Immunol. 20:453–60. integrins on human B lymphocytes. J. Immunol. microarray core facility for gene expres- 15. Basso K, Dalla-Favera R. BCL6: master regulator 151:2471–83. sion analyses. of the germinal center reaction and key oncogene 31. Trentin L, et al. (1999) The chemokine receptor in B cell lymphomagenesis. Adv. Immunol. CXCR3 is expressed on malignant B cells and DISCLOSURE 105:193–210. mediates chemotaxis. J. Clin. Invest. 104:115–21. The authors declare that they have no 16. Reif K, Cyster JG. (2000) RGS molecule expres- 32. Lopez-Giral S, et al. (2004) Chemokine receptors sion in murine B lymphocytes and ability to that mediate B cell homing to secondary lym- Molecu- competing interests as defined by down-regulate chemotaxis to lymphoid phoid tissues are highly expressed in B cell lar Medicine, or other interests that might chemokines. J. Immunol. 164:4720–9. chronic lymphocytic leukemia and non-Hodgkin be perceived to influence the results and 17. Wilker PR, et al. (2008) Transcription factor Mef2c lymphomas with widespread nodular dissemina- discussion reported in this paper. is required for B cell proliferation and survival tion. J. Leukoc. Biol. 76:462–71. after antigen receptor stimulation. Nat. Immunol. 33. Zlotnik A, Burkhardt AM, Homey B. (2011) REFERENCES 9:603–12. Homeostatic chemokine receptors and organ- 1. Chiorazzi N, Rai KR, Ferrarini M. (2005) Chronic 18. Novak AJ, et al. (2004) Expression of BLyS and its specific metastasis. Nat. Rev. Immunol. 11:597–606. lymphocytic leukemia. N. Engl. J. Med. 352:804– 15. receptors in B-cell non-Hodgkin lymphoma: cor- 34. Costantini JL, et al. (2009) TAPP2 links phospho- 2. Zenz T, Mertens D, Kuppers R, Dohner H, Stil- relation with disease activity and patient out- inositide 3-kinase signaling to B-cell adhesion genbauer S. (2010) From pathogenesis to treat- come. Blood. 104:2247–53. through interaction with the cytoskeletal protein ment of chronic lymphocytic leukaemia. Nat. Rev. 19. Fu L, et al. (2009) BAFF-R promotes cell prolifera- utrophin: expression of a novel cell adhesion- β Cancer. 10:37–50. tion and survival through interaction with IKK promoting complex in B-cell leukemia. Blood. κ 3. Shanafelt TD, et al. (2008) Karyotype evolution and NF- B/c-Rel in the nucleus of normal and 114:4703–12. on fluorescent in situ hybridization analysis is neoplastic B-lymphoid cells. Blood. 113:4627–36. 35. Kappos L, et al. (2011) Natalizumab treatment for associated with short survival in patients with 20. Zhai Y, et al. (1999) VEGI, a novel cytokine of the multiple sclerosis: updated recommendations for chronic lymphocytic leukemia and is related to tumor necrosis factor family, is an angiogenesis patient selection and monitoring. Lancet Neurol. CD49d expression. J. Clin. Oncol. 26:e5–6. inhibitor that suppresses the growth of colon car- 10:745–58. 4. Calissano C, et al. (2009) In vivo intraclonal and cinomas in vivo. FASEB J. 13:181–9. 36. Wiernik PH, Adiga GU. (2011) Single-agent ritux- interclonal kinetic heterogeneity in B-cell chronic 21. Coscia M, et al. (2011) IGHV unmutated CLL imab in treatment-refractory or poor prognosis lymphocytic leukemia. Blood. 114:4832–42. B cells are more prone to spontaneous apoptosis patients with chronic lymphocytic leukemia. 5. Neese RA, et al. (2002) Measurement in vivo of and subject to environmental prosurvival signals Curr. Med. Res. Opin. 27:1987–93. proliferation rates of slow turnover cells by than mutated CLL B cells. Leukemia. 25:828–37. 37. Nightingale G. (2011) Ofatumumab: a novel anti- 2H2O labeling of the deoxyribose moiety of 22. Porakishvili N, et al. (2011) CD180 functions in ac- CD20 monoclonal antibody for treatment of re- DNA. Proc. Natl. Acad. Sci. U. S. A. 99:15345–50. tivation, survival and cycling of B chronic lympho- fractory chronic lymphocytic leukemia. Ann. 6. Messmer BT, et al. (2005) In vivo measurements cytic leukaemia cells. Br. J. Haematol. 153:486–98. Pharmacother. 45:1248–55. document the dynamic cellular kinetics of 23. Kay NE, et al. (2005) A recombinant IL-4- 38. Rai KR, et al. (2002) Alemtuzumab in previously chronic lymphocytic leukemia B cells. J. Clin. Pseudomonas exotoxin inhibits protein synthesis treated chronic lymphocytic leukemia patients Invest. 115:755–64. and overcomes apoptosis resistance in human who also had received fludarabine. J. Clin. Oncol. 7. Zupo S, et al. (1994) Expression of CD5 and CD38 CLL B cells. Leuk. Res. 29:1009–18. 20:3891–7. by human CD5- B cells: requirement for special 24. Nakamura M, Shimada K, Konishi N. (2008) The 39. Calandra G, Bridger G, Fricker S. (2010) CXCR4 stimuli. Eur. J. Immunol. 24:1426–33. role of HRK gene in human cancer. Oncogene. 27 in clinical hematology. Curr. Top. Microbiol. 8. Stein JV, Nombela-Arrieta C. (2005) Chemokine Suppl 1:S105–13. Immunol. 341:173–91. control of lymphocyte trafficking: a general over- 25. Brosens JJ, Wilson MS, Lam EW. (2009) FOXO 40. Cameron F, Whiteside G, Perry C. (2011) Ipili- view. Immunology. 116:1–12. transcription factors: from cell fate decisions to mumab: first global approval. Drugs. 71:1093–104.

MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011 | CALISSANO ET AL. | 1381 INTRACLONAL COMPLEXITY IN CHRONIC LYMPHOCYTIC LEUKEMIA

41. Callahan MK, Wolchok JD, Allison JP. (2010) 58. Panayiotidis P, Ganeshaguru K, Jabbar SA, Hoff- Anti-CTLA-4 antibody therapy: immune moni- brand AV. (1993) Interleukin-4 inhibits apoptotic toring during clinical development of a novel im- cell death and loss of the bcl-2 protein in munotherapy. Semin. Oncol. 37:473–84. B-chronic lymphocytic leukaemia cells in vitro. 42. Bajaj M, Heath EI. (2011) Conatumumab: a novel Br. J. Haematol. 85:439–45. monoclonal antibody against death receptor 5 for 59. Yang JC, et al. (2007) Ipilimumab (anti-CTLA4 the treatment of advanced malignancies in antibody) causes regression of metastatic renal adults. Expert Opin. Biol. Ther. 11:1519–24. cell cancer associated with enteritis and hypo- 43. Byrd JC, et al. (2007) Phase 1 study of lumilix- physitis. J. Immunother. 30:825–30. imab with detailed pharmacokinetic and phar- macodynamic measurements in patients with re- lapsed or refractory chronic lymphocytic leukemia. Clin. Cancer Res. 13:4448–55. 44. Burger M, et al. (2005) Small peptide inhibitors of the CXCR4 chemokine receptor (CD184) antago- nize the activation, migration and antiapoptotic responses of CXCL12 in chronic lymphocytic leukemia B cells. Blood. 106:1824–30. 45. Puri S, et al. (2009) A review of studies on targeting interleukin 4 receptor for central nervous system malignancy. Curr. Mol. Med. 9:732–9. 46. de Weers M, et al. (2011) Daratumumab, a novel therapeutic human CD38 monoclonal antibody, induces killing of multiple myeloma and other hematological tumors. J. Immunol. 186:1840–8. 47. Pascual V, et al. (1994) Analysis of somatic mutation in five B cell subsets of human tonsil. J. Exp. Med. 180:329–39. 48. Damle RN, et al. (1999) Ig V gene mutation status and CD38 expression as novel prognostic indica- tors in chronic lymphocytic leukemia. Blood. 94: 1840–7. 49. Quiroga MP, Burger JA. (2010) BCR-mediated de- crease of CXCR4 and CD62L in CLL. Cancer Res. 70:5194; author reply 5195. 50. Chiorazzi N, Ferrarini M. (2011) Cellular origin(s) of chronic lymphocytic leukemia: cau- tionary notes and additional considerations and possibilities. Blood. 117:1781–91. 51. Griffin DO, Holodick NE, Rothstein TL. (2011) Human B1 cells in umbilical cord and adult peripheral blood express the novel phenotype CD20+ CD27+ CD43+ CD70. J. Exp. Med. 208:67–80. 52. Fagarasan S, et al. (2000) Mechanism of B1 cell differentiation and migration in GALT. Curr. Top. Microbiol. Immunol. 252:221–9. 53. Dick JE. (2008) Stem cell concepts renew cancer research. Blood. 112:4793–807. 54. Hayakawa K, Hardy RR, Herzenberg LA, Herzenberg LA. (1985) Progenitors for Ly-1 B cells are distinct from progenitors for other B cells. J. Exp. Med. 161:1554–68. 55. Chiorazzi N. (2007) Cell proliferation and death: forgotten features of chronic lymphocytic leukemia B cells. Best Pract. Res. Clin. Haematol. 20:399–413. 56. Burger JA, Ghia P, Rosenwald A, Caligaris-Cap- pio F. (2009) The microenvironment in mature B-cell malignancies: a target for new treatment strategies. Blood. 114:3367–75. 57. Alfonso-Perez M, et al. (2006) Anti-CCR7 monoclonal antibodies as a novel tool for the treatment of chronic lymphocyte leukemia. J. Leukoc. Biol. 79:1157–65.

1382 | CALISSANO ET AL. | MOL MED 17(11-12)1374-1382, NOVEMBER-DECEMBER 2011