RAPID COMMUNICATION

Autoantibody Activity of Immunoglobulins Isolated From B-Cell Follicular Lymphomas

By G. Dighiero, S. Hart, A. Lim, L. Borche, R. Levy, and R.A. Miller

Previous work with monoclonal Igs (Mlgs) has demonstrated that a high proportion of paraproteins bind to self-antigens such as the Fc fragment of lgG, li blood group antigens, cytoskeleton proteins, DNA, and myelin-associated glycopro- tein (MAG). Recent work in CLL indicates that CD5' B lymphocytes are frequently committed to production of autoantibodies. We have examined the antibody specificity of Mlgs derived from the tumor cells of 31 different patients with CD5- B-cell lymphomas. Our results indicate that the tumor cells from 8 of these 31 patients (25.8%) express Igs

URING RECENT years, considerable evidence has D accumulated indicating that: (1) natural multispecific autoantibodies constitute an important part of normal circulating Ig; (2) precursors of cells producing these natural autoantibodies exist with high frequency in the normal B-cell repertoire, and (3) natural autoantibodies are frequently the product of clones expressing germline Ig variable region genes.'"

Previous work on monoclonal Igs (MIgs) has shown that a high proportion of paraproteins bind to self determinants like the Fc fragment of Ig; Ii blood group antigen^,^ cytoskeleton proteins and DNA,627 and myelin-associated glycoproteins.' More recently, evidence has been obtained indicating that CD5' CLL B lymphocytes are frequently committed to production of natural autoantibodies?." These results support the hypothesis that CD5' B lymphocytes are involved in production of natural autoantibodie~.'~~''

In the present study we have examined the antibody specificity of Igs derived from tumor cells of 31 patients with C D S follicular lymphoma^.'^.'^ Our results indicate that CD5- B-cell follicular lymphoma cells also are frequently committed to production of natural autoantibodies.

MATERIALS AND METHODS

Heterohybridomas were prepared from tumor samples obtained from 31 patients with follicular lymphomas, as previously described.I6 Hybrids secreting Igs of the same heavy-chain and light-chain type as that of the patient's tumor were identified by an enzyme-linked immunosor- bent assay (ELISA). The hybridomas were expanded and Ig was purified from the cell culture supernatant using immunoaffinity chromatography. Monoclonal anti-idiotype antibodies have been developed for each of the tumor Igs as previously described." These antibodies were used in ELISA and immunoperoxidase staining procedures to establish that the Igs secreted from the heterohybrids were derived from the tumor cell^.'^

Tumor Igs were examined by both ELISA and indirect immunofluorescence (IF) techniques for antibody activity against a panel of antigens. Polystyrene flat- bottomed microtiter plates (Nunc, Roskilde, Denmark) were coated with various antigens. These included actin, tubulin, myo- sin, trinitrophenyl-bovine serum albumin (TNP-BSA), and BSA at 5 pdmL each, in 0.1 molL carbonate buffer, pH 9.6. Double- stranded DNA (dsDNA) and single-stranded DNA (ssDNA) were coated at 100 KgimL in 0.1 molL NaCl 0.1 molL sodium citrate buffer, pH 6.0. All antigens were prepared as previously described.'

Isolation of Mlg from B-cell lymphomas.

Antigen reactivity of tumor Igs.

with autoantibody activity. In two cases antibody activity was multispecific. In four cases, antibody activity was exclu- sively directed against the Fc fragment of IgG, whereas the two other cases bound to both Fc fragment of IgG and nuclear antigens. Most nomHodgkin's lymphomas (NHL) are derived from CD5- B cells. These results indicate that like CLL, NHL also express Igs that frequently have autoantibody activity. o 1991 by The American Society of Hematology.

Rabbit IgG (RIgG; Sigma, St Louis, MO) was used for assay of human rheumatoid factor as described by Bampton et a]." The antigen-coated wells were washed five times with phosphate- buffered saline (PBS) pH 7.2 containing 0.05% tween 20 (PBS-T) and incubated overnight at 4"C, with each purified tumor MIg at 5 and 1 &mL in PBS-T containing 0.5% gelatin (PBS-TG). Positive controls included either monoclonal Ig known as IgM A Mar or IgMk Mat derived by fusion of CLL B lymphocytes." IgM A Mar was used as a positive control for actin, myosin, and TNP-BSA, IgMk Mat was used for tubulin and RIgG. As a positive control for anti-DNA activity we used serum obtained from a patient with systemic lupus erythematous (SLE) who displayed high levels of anti-dsDNA and anti-ssDNA antibodies. As negative controls, irrelevant MIgs from CLL B lymphocytes were used. After wash- ing, wells were incubated for 1 hour at 37°C with goat antihuman Ig coupled to P-galactosidase (Biosys, Compiegne, France) at a concentration of 1 pg/mL in PBS-TG. For the anti-RIgG assay, the second antibody was an F(ab'), rabbit antihuman p chain specific antibody conjugated to peroxidase in the case of hybrids expressing IgM or an F(ab'), rabbit anti-y chain conjugated to peroxidase in the case of hybrids expressing IgG (Dakopats, Dako, Copenhagen, Denmark). Enzyme substrate was added after washing and optical densities were measured by Titertek Multiskan (Flow Laborato- ries, Puteaux, France). Optical densities of at least 10 times the mean background level were considered to be positive.

All supernatants were tested by IF, for the presence of antinu- clear (ANA), antismooth muscle (SMA), antimitochondrial (AMA), and stomach (ASA) antibodies by using 6-km frozen sections of rat kidney, liver, and stomach composite blocks. Purified tumor MIgs at 100 and 50 pg/mL concentrations were incubated with frozen sections and the reaction was developed by incubation with sheep antihuman Ig coupled to fluorescein isothyocyanate (FITC; Bio- SYS).

From the Institut Pasteur, Paris, France; IDEC Pharmaceutical Corp, Mountain Kew; and the Stanford University Medical Center, Stanford, CA.

Submitted April 8, 1991; accepted May 17,1991. Address reprint requests to Guillaume Dighiero, MD, PhD, Unit6

d'1mmunoHh"aologie et d 'ImmunoPathologk, Institut Pasteur, 28, rue du Dr Roux, 75724 Paris Cedex 15, France.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. section 1734 solely to indicate this fact.

8 I991 by The American Society of Hematology. 0006-4971 /9I17803-O044$3. OOlO

Blood, Vol78, No 3 (August I), 1991: pp 581-585 581

For personal use only.on September 24, 2017. by guest www.bloodjournal.orgFrom

DlGHlERO ET AL 582

Table 1. Antibody Activity of Hybrids as Detected by ELISA and Immunofluorescence

ELISA (OD x 103) Immunofluorescence Hybrid 19

Patient Histology lsotype Concentration Actin Tubulin Myosin ssDNA dsDNA TNP-BSA RlgG Ig Sto ANA SMA AMA

Sim FSC lgMA - - - 1 P9 3 4 6 10 1 1 30 634 5 Pg Cro FM IgMA - - - 1 P9 7 2 7 5 2 10 30 967 5 P9 11 4 4 5 4 7 36 918 Str FSC lgMA 1 PS 12 7 6 8 7 5 P9 4 4 13 8 4 5 690 1,020 - Wit FSC lgMK - - - 1 P9 4 5 11 12 5 4 102 999

8 6 38 1,023 - 5 P9 7 7 4 6 Dum FSC lgMK - - - 1 P9 5 3 12 3 3 12 48 1,030 5 P9 7 5 1 8 5 5 29 238 - Wil FSC IgGA - - - 1 P9 8 3 7 8 3 16 19 259

Tra FM IgMA - - 1 PS 7 11 10 9 11 6 50 614 (speckled)

Che FSC lgMA - - - 1 Pg 9 3 3 2 3 14 26 1,036

Ell FSC l g M ~ - - - 1 P9 4 2 7 9 2 2 23 986

2 11 53 896 - Del FM lgMh - - - 1 P9 10 4 6 0 5 12 56 932 5 Pg 9 8 4 3 9 11 23 434 - Wils FM lgGA - - - 1 P9 6 6 10 5 3 6 23 358 5 P9 52 8 24 12 14 36 32 612 - Smi FSC lgGA - - - 1 P9 3 7 5 8 2 6 29 579

Sof FSC IgMA - - - 1 P9 6 7 6 4 1 3 50 938 5 P9 5 8 10 30 4 10 146 988 - Bil FM IgMA - - -

9 41 1,000 1 P9 4 6 10 26 16

Bis FSC IgGA - - - 1 P9 4 3 12 6 4 11 3 741

Cal FSC l g M ~ - - - 9 4 5 29 5 5 23 931

5 Pg 11 12 10 10 8 14 67 931 - Vin FSC lgMK - - - 1 P9 1 6 3 10 7 16 24 671 5 P9 3% 981 252 1,207 1,121 1,001 866 1,137 - Daw FSC lgMA 1 Pg 45 293 25 1,067 901 160 194 1,059

905 862 1,004 - 5 P9 131 89 35 149 51 Bir FM l g M ~ - - -

Ala FSC lgMK

5 Pg' 3 2 4 4 3 7 81 945 -

4 5 3 4 11 4 24 1,006 -

ND ND 14 35 897 ND ND

5 P9 3 7 5 2 7 6 289 937 - ++

5 3 2 8 36 1,034 - 5 P9 9 2

5 P9 11 8 4 6 8 9 20 1,009 -

5 P9 7 2 3 3

5 P9 3 8 9 12 12 5 54 981 -

5 P9 6 3 12 17 3 23 2 685 -

5 w9 6 10 3 12 1 6 9 912 -

1 P9

- +++ -

1 P9 26 16 15 10 4 221 188 1,021 5 P9 13 17 16 36 9 31 554 1,033 - +++ - - 1 Pg 3 8 6 14 19 13 157 930 (speckled) 5 P9 41 33 21 37 16 30 689 1,013 - Bin FM IgMK - - - 1 P9 5 5 7 9 4 7 123 974 5 P9 17 28 13 7 8 8 44 1,060 - Bou FSC lgMK - - - 1 Pg 3 10 13 7 5 12 20 1,016

7 14 22 4 6 16 6 947 Cok FSC lgGK

1 P9 5 PS 16 27 25 10 6 8 70 970 - 1 P9 11 6 16 8 6 15 22 890

Mau FSC IgMA

16 201 970 - Mil FM lgMr

5 P9 44 40 51 48 14 25 405 1,124 - 13 17 16 20 1 5 90 1,085

Chr FSC lgMK

9 4 7 10 5 7 204 1,031 - 1 3 4 9 3 4 112 770

Boe FSC l g M ~

5 c9 44 40 51 46 14 25 405 1,124 - 13 17 16 20 1 5 90 1,085

Chr FSC lgMK

5 w9 9 4 7 10 5 7 204 1,031 - 1 3 4 9 3 4 112 770

Boe FSC IgMK

1 3 4 3 2 2 124 285 Lee FM IgMA

2 4 2 3 2 2 190 449 Croc FM lgMA

4 1 1 4 2 2 190 391 Spe FSC lgMA

3 6 4 4 2 1 85 696 FAN FSC IgMA

Positive values are represented by boldface values. Abbreviations: ND, not done; FSC, follicular small cleaved; FM. follicular mixed. *Heterohybrids were diluted at 5 and 1 &g/mLfor ELlSA assays and at 100 Wg/mL for fluorescence tests.

5 PS 21 18 27 14 46 7 8 849 - - - -

- - -

5 PQ 45 40 31 8 20 - - - 1 P9 11 6 26 4 10 1 43 998

1 P9 5 P9 - - 1 P.9

1 w9

1 w9 5 P9 4 6 4 3 3 3 212 355 -

1 P9 5 P9 5 6 7 4 5 4 86 533 -

1 P9 5 P-9 9 5 6 4 2 5 120 435 -

1 t 9 5 PS 10 9 8 5 3 3 110 801 -

1 P9

- - -

-

- - -

- - -

- - -

- - -

- - -

- - -

For personal use only.on September 24, 2017. by guest www.bloodjournal.orgFrom

AUTOANTIBODY ACTIVITY OF CD5 B-CELL LYMPHOMAS

Fig 1. Unfixed yostat ..cHom wtained bv IF with holated Mlgs from prtknts D m and Ala. (A) A typical stalnlng pattern of smooth murck fiben on rat stomach Is observed for Mlg Daw (oflglnal magnlfkotlon x 100). (9) A typkol ANA speckled pattem Is observed on rat Ihrer cells for patlent Ala (origlnal magnHicatlon x 250).

RESULTS

Hybrids wcrc derived from 31 paticnts (20 mcn and 11 womcn) with a diagnosis of follicular lymphomas. Tablc 1 summarizes thc histopathol- ogy and immunophcnotypc of thc malignant B cclls. The isotypcs wcrc IgMk (12 cascs), IgM A (14 cascs), lgGk (onc casc) and IgG A (four cases). In each casc. frozcn tissue scctions wcrc tcstcd using an immunopcroxidasc staining tcchniquc which confirmcd that thc isotypc of thc tumor Ig corrcspondcd to thc isotypcs sccrctcd by the hybridomas. All caws wcrc ncgativc for rcactivity with anti-CDS. In addition, anti-idiotypc antibodies wcrc uwd to establish that thc hctcrohybridoma-dcrivcd Ig also was cxprcsscd by thc tumor spccimcn.

Tablc 1 summarizes the results obscrvcd in ELlSA and IF for the 31 purificd Igs. Daw Ig displaycd a multispccific rcactivity of binding to actin, tubulin, myosin, ssDNA, dsDNA, TNP-BSA, and RIgG in ELlSA and to smooth musclc in IF. Dcspite the prcscncc of high lcvcls of anti-DNA antibodies, ANA wcrc not obscrvcd by IF. Bir Ig bound to actin. &.DNA, TNP-

Characteristic of the tumors.

Anrigen tractiviry by tumor M l p .

BSA, and RlgG. Ig from patients, Fro, Wit, Mil, and Chr bound to RlgG, whereas Tra Ig and Ala Ig bound to RIgG and also displaycd an ANA speckled pattcm. Figurc 1 shows the fluorcsccncc pattcm of Draw Ig and Ala Ig, which dcmonstratcd typical SMA and ANA patterns. All eight Igs that showed reactivity were of the IgM isotypc.

DISCUSSION

We have examined the antibody spccificity of purified Mlgs derived from 31 CDY Bccll lymphomas. Our prcscnt results indicatc there is a high frcqucncy of autoantibody activity among thcsc Mlg. Anti-idiotypc antibodics wcrc used to establish that thc hybrids wcrc dcrivcd from the tumor cells. Thcsc rcsults cxtcnd and confirm the idca that the autorcactivc Bccll rcpcrtoirc may undcrgo malignant transformation.

In thc casc of human monoclonal gammapathics, it has been dcmonstratcd that about 30% of IgM MIgs bind to self dctcrminants.‘.’ Cryoglobulins and cold agglutinins

For personal use only.on September 24, 2017. by guest www.bloodjournal.orgFrom

584 DlGHlERO ET AL

have been shown to express very restricted sets of Ig variable region genes.’s-zl Cryoglobulins sharing the major Wa idiotype, which includes 60% of all cryoglobulins, express a unique VK IIIb gene (Hum Kv 325) and a VH 1 gene. Cold agglutinins also express VK I11 genes and, in some cases, the same Hum Kv 325 gene used by cryoglobu- lins associated with the Ig heavy chain variable genes of the VH 4 family.22 In the case of MIgs with multispecific antibody activity (cytoskeleton proteins, DNA, etc) and of MIgs and anti-MAG activity, there is no structural informa- tion, although the presence of recurrent idiotypes within these MIgs favors their germ line origin.

Recent work on CD5’ B-CLL also has demonstrated a high frequency of natural autoantibody activity among Igs derived from this subset of B Kipps et found that a high proportion of IgMk-bearing B-CLL cells reacted with a murine anti-idiotypic antibody produced against a monoclonal IgM rheumatoid factor expressing the Wa idiotype. These cells express the Hum KV 325 K light chain variable region gene, which is ~nmutated.~’ Similar restric- tion was found when VH genes were analyzed.24 The antibody reactivities observed in our work with CD5- B-cell lymphomas are similar to those observed for CD5’ B-cell CLL. However, in B-CLL there is a high frequency of expression of some idiotypes like Wa idiotype, which was only infrequently seen in CD5- B-cell lymphoma^.'^ In addition, there is an active somatic mutational process in B-cell follicular lymphomas that is rarely observed in

All of these results indicate that there is a high propor- B-CLL.”

tion of B-cell malignancies that express Igs with reactivity against self determinants. Interestingly, antibody specifici- ties are directed against antigens that are highly conserved during evolution. Autoreactive B cells constitute a substan- tial part of the normal B-cell repertoire= and have been claimed to be in an “activated” state through continuous challenge by autoantigens. This “activated” state could create propicious conditions for mutations and chromo- somal translocations that may contribute to malignant transformation. However, the possibility that certain Ig genes could be overexpressed by B-cell tumors should also be considered.

Another important point raised by our results concerns the role of normal CD5’ B cells in the production of natural autoantibodies. Previous work in the mouse with Lyl+ B cells, the murine counterpart of CD5’ B cells, has shown that these cells are involved in the production of autoanti- bodies.’*,’’ The role that CD5- lymphocytes play in natural autoantibody production is not clear. Recent work from other groups indicate that CD5- B cells could also be implicated in the production of natural a u t o a n t i b o d i e ~ . ~ ~ ~ ~ ~ Our results obtained with Igs derived from CD5- B-cell tumors strongly support the possibility that CD5- as well as CD5’ B cells may be involved in autoantibody production. Interestingly, most cases of Waldenstrom’s macroglobulin- emia do not express CD5 antigen in circulating B lympho- cytes and are frequently associated with the presence of MIg, displaying natural autoantibody specificities close to that reported

REFERENCES

1. Guilbert B, Dighiero G, Avrameas S: Naturally occurring antibodies against nine common antigens in normal humans. I. Detection, isolation and characterization. J Immunol 128:2779, 1982

2. Dighiero G, Lymberi P, MaziB JC, Rouyre S, Buttler-Browne S, Whalen RG, Avrameas S: Murine hybridomas secreting natural antibodies reacting with self antigens. J Immunol 131:2267,1983

3. Lymberi P, Dighiero G, Ternynck T, Avrameas S: A high incidence of cross-reactive idiotypes among murine natural autoanti- bodies. Eur J Immunol5:702,1985

4. Crowley JJ, Goldfien RD, Schrohenlober RE, Spiegelberg HI, Silverman GJ, Mageed RA, Jefferis R, Koopman WJ, Carson DA, Fong S: Incidence of three cross-reactive idiotypes on human rheumatoid factor paraproteins. J Immunoll403411,1988

5. Pruzanski W, Shumak KH: Biologic activity of cold reacting autoantibodies. N Engl J Med 297:538,1977

6. Dighiero G, Guilbert B, Fermant JP, Lymberi P, Danon F, Avrameas S: Thirty-six human monoclonal immunoglobulins (MIg) with antibody activity against cytoskeleton proteins, thyroglobulin and native DNA, immunological studies and clinical correlations. Blood 62:264,1983

7. Shoenfeld Y, Amital Teplizki H, Mendlovic S, Blank M, Mozes E, Isenberg D A The role of the human anti-DNA idiotype 1616 in autoimmunity. Clin Immunol Immunopathol51:313,1989

8. Saito T, Sherman W, Latov N: Specificity and idiotype of M-proteins that react with MAG in patients with neuropathy. J Immunol1302496,1983

9. Broker BM, Klajman A, Youinou P, Jouquan J, Worman CP, Murphy J, Mackenzie I, Quartey-Papafio R, Blaschek M, Collins P,

La1 S, Lydyard PM: Chronic lymphocytic leukemic (CLL) cells secrete multispecific autoantibodies. J Autoimmun 1:469,1988

10. Sthoeger ZM, Wakai M, Tse DB, Vinciguerra VP, Allen SL, Budman DR, Lichtman SM, Schulman P, Weiselberg LR, Chiorazzi N: Production of autoantibodies by CD5-expressing B lymphocytes from patients with chronic lymphocytic leukemia. J Exp Med 169:255,1989

11. Borche L, Lim A, Binet JL, Dighiero G: Evidence that chronic lymphocytic leukemia B lymphocytes are frequently com- mited to productions of natural autoantibodies. Blood 76562,1990

12. Hayakawa K, Hardy RR, Parks DR, Herzenber LA: The Ly-1B cell subpopulation in normal immunodefective and autoim- mune mice. J Exp Med 157:202,1983

13. Hardy RR, Hayakawa K Development and physiology of Ly-1 B and its human homolog Leu-1 B. Immunol Rev 93:53,1986

14. Levy R, Dilley J: Rescue of immunoglobulin secretion from human neoplastic lymphoid cells by somatic cell hybridization. Proc Natl Acad Sci USA 75:2411,1978

15. Cleary ML, Meeker TC, Levy S, Lee E, Trela M, Sklar J, Levy R: Clustering of extensive somatic mutations in the variable region of an immunoglobulin heavy chain gene from a human B cell lymphoma. Cell 44:97,1986

16. Carrol WL, Thielemans K, Dilley J, Levy R Mouse x human heterohybridomas as fusion partners with human B cell tumors. J Immunol Methods 8961,1986

17. Bampton JLM, Cawston TE, Kyle MV, Hazleman B L Measurement of rheumatoid factors by an enzyme-linked- immunosorbent assay (ELISA) and comparison with other meth- ods. Ann Rheum Dis 44:13,1989

For personal use only.on September 24, 2017. by guest www.bloodjournal.orgFrom

AUTOANTIBODY ACTIVITY OF CD5- B-CELL LYMPHOMAS 585

18. Kritzman J, Kunkel HG, McCarthy J, Mellors R C Studies of a Waldenstrom-type macroglobulin with rheumatoid factor proper- ties. J Lab Clin Med 57:905, 1961

19. Newkirk MM, Mageed RA, Jefferis R, Chen PP, Capra JD: Complete amino acid sequences of variable regions of two human IgM rheumatoid factors. BOR and KAS of the Wa idiotypic family reveal restricted use of heavy and light chain variable and joining region gene segments. J Exp Med 166:550,1987

20. Chen PP, Robbins DL, Jirik FR, Kipps TJ, Carson D A Isolation and characterization of a light chain variable region gene for human rheumatoid factors. J Exp Med 166:1900,1987

21. Capra JD, Williams RC, Feizi T, Kunkel HG: Light chain sequences of human IgM cold agglutinin. Proc Natl Acad Sci USA 69:40,1972

22. Silverman GJ, Chen PP, Carson D A Cold agglutinins: Specificity, idiotypy and structural analysis, in Carson DA, Chen PP, Kipps TJ (eds): Idiotypes in Biology and Medicine. Basel, Switzerland, Karger, 1990, p 109

23. Kipps TJ, Tomhave E, Chen PP, Carson DA: Autoantibody associated K light chain variable region gene expressed in chronic lymphocytic leukemia with little or no somatic mutation, implica- tions for etiology and immunotherapy. J Exp Med 167:840,1988

24. Kipps TJ, Tomhave E, Pratt LF, DufQ S, Chen PP, Carson D A Developmentally restricted immunoglobulin heavy chain vari- able region gene expressed at high frequency in chronic lympocytic leukemia. Proc Natl Acad Sci USA 86:5913,1989

25. Dighiero G, Lymberi P, Guilbert B, Ternynck T, Avrameas S: Natural auto-antibodies constitute a substantial part of normal circulating immunoglobulins. Ann NY Acad Sci 475:135,1986

26. Deane M, Norton JD: Immunoglobulin heavy chain variable region family usage is independent of tumor cell phenotype in human B lineage leukemias. Eur J Immunol20:2209,1990

27. Conger JD, Pike BL, Nossal GJV: Clonal analysis of the anti-DNA repeptoire of murine B lymphocytes. Proc Natl Acad Sci USA 84:2931,1987

28. Kaushik A, Lim A, Poncet P, Ge XR, Dighiero G: Compara- tive analysis of natural antibody specificities among hybridomas originating from spleen and peritoneal cavity of adult NZB and BALB/c mice. Scand J Immunol27:461,1988

29. Kaushik A, Mayer R, Fidanza V, Zaghouani H, Lim A, Bona C, Dighiero G: Lyl and V-gene expression among hybridomas secreting natural autoantibody. J Autoimmun 3:687,1991

For personal use only.on September 24, 2017. by guest www.bloodjournal.orgFrom

1991 78: 581-585  

G Dighiero, S Hart, A Lim, L Borche, R Levy and RA Miller follicular lymphomasAutoantibody activity of immunoglobulins isolated from B-cell 

http://www.bloodjournal.org/content/78/3/581.full.htmlUpdated information and services can be found at:

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American

For personal use only.on September 24, 2017. by guest www.bloodjournal.orgFrom