INTRODUCTIONChronic lymphocytic leukemia (CLL),

the most prevalent hematologic malig-nancy affecting Caucasian adults, is in-curable (1). The disease is a monoclonalexpansion of a subset of antigen-experienced human B cells expressingsurface membrane CD5 (2,3). A key role

for surface membrane Ig (smIg) is sug-gested by their striking structural simi-larity among unrelated patients (3–5).Furthermore, the presence of somaticmutations in IGHV genes coding thesmIg V-regions segregates patients intosubgroups (6) with dramatically differentclinical outcomes (7,8). Patients with un-

mutated IGHV (U-CLL) have more aggres-sive disease (median survival < 8 years),while patients with mutated IGHV(M-CLL) have a milder course (mediansurvival ≥ 24 years). Such observationsled to the paradigm that developmentand evolution of CLL is influenced byantigen selection and drive (3). There-fore, defining the antigens bound byCLL cells could provide insights into thepathogenesis of the disease.

Clonal selection can be driven by for-eign and self-antigens (9). Apoptosis is amajor source of self-antigens, resulting indisplay of intracellular molecules on cell

M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8 | C A T E R A E T A L . | 6 6 5

Chronic Lymphocytic Leukemia Cells Recognize ConservedEpitopes Associated with Apoptosis and Oxidation

Rosa Catera,1 Gregg J Silverman,2 Katerina Hatzi,1 Till Seiler,1 Sebastien Didier,1 Lu Zhang,1

Maxime Hervé,3 Eric Meffre,3 David G Oscier,4 Helen Vlassara,5 R Hal Scofield,6 Yifang Chen,2

Steven L Allen,1,7 Jonathan Kolitz,1 Kanti R Rai,1,8 Charles C Chu,1,9 and Nicholas Chiorazzi 1,7

1Feinstein Institute for Medical Research, North Shore-LIJ Health System, Manhasset, New York, United States of America;2Rheumatic Diseases Core Center and Laboratory of B-Cell Immunobiology, University of California, San Diego, California, UnitedStates of America; 3Laboratory of Biochemistry and Molecular Immunology, Hospital for Special Surgery, New York, New York,United States of America; 4Molecular Biology and Cytogenetics, Royal Bournemouth Hospital, Bournemouth, United Kingdom; 5Division of Diabetes and Aging Research, Brookdale Department of Geriatrics, Mount Sinai School of Medicine, New York, NewYork, United States of America; 6Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma, United States of America; 7Departments of Medicine, North Shore University Hospital and Albert Einstein College of Medicine, Manhasset, New York, UnitedStates of America; 8Departments of Medicine, Long Island Jewish Medical Center and Albert Einstein College of Medicine, New HydePark, New York, United States of America; and 9Departments of Medicine, North Shore University Hospital and NYU School ofMedicine, Manhasset, New York, United States of America

Chronic lymphocytic leukemia (CLL) represents the outgrowth of a CD5+ B cell. Its etiology is unknown. The structure of mem-brane Ig on CLL cells of unrelated patients can be remarkably similar. Therefore, antigen binding and stimulation could contributeto clonal selection and expansion as well as disease promotion. Initial studies suggest that CLL mAbs bind autoantigens. Sinceapoptosis can make autoantigens accessible for recognition by antibodies, and also create neo-epitopes by chemical modifi-cations occurring naturally during this process, we sought to determine if CLL mAbs recognize autoantigens associated with apo-ptosis. In general, ~60% of CLL mAbs bound the surfaces of apoptotic cells, were polyreactive, and expressed unmutated IGHV.mAbs recognized two types of antigens: native molecules located within healthy cells, which relocated to the external cell sur-face during apoptosis; and/or neoantigens, generated by oxidation during the apoptotic process. Some of the latter epitopesare similar to those on bacteria and other microbes. Although most of the reactive mAbs were not mutated, the use of unmu-tated IGHV did not bestow autoreactivity automatically, since several such mAbs were not reactive. Particular IGHV andIGHV/D/J rearrangements contributed to autoantigen binding, although the presence and degree of reactivity varied based onspecific structural elements. Thus, clonal expansion in CLL may be stimulated by autoantigens occurring naturally during apo-ptosis. These data suggest that CLL may derive from normal B cells whose function is to remove cellular debris, and also to pro-vide a first line of defense against pathogens.Online address: http://www.molmed.orgdoi: 10.2119/2008-00102.Catera

Address correspondence and reprint requests to Nicholas Chiorazzi, The Feinstein Institute

for Medical Research, 350 Community Drive, Manhasset, NY 11030, Phone: 516-562-1090;

Fax: 516-562-1011; Email: NChizzi@nshs.edu.

Submitted September 17, 2008; Accepted for publication September 19, 2008; Epub

(www.molmed.org) ahead of print September 25, 2008.

surfaces (10,11) and generation of neo-antigens by accompanying mechanismssuch as oxidation (12,13). B lymphocytestargeting such epitopes frequently arefound in the pre-immune repertoire,often in the B-1 cell compartment (14).

Because CLL cells likely derive fromautoreactive B cells (15–18), we exploredif apoptosis-associated autoantigenswere relevant to the selection and expan-sion of leukemic cells in this disease.Our data indicate that smIgs, particu-larly from patients with poor outcomeU-CLL, recognize autoantigens madeavailable during apoptosis and/or cre-ated by this catabolic process. Thesefindings suggest that CLL is selectedfrom a B-cell subset that normally helpsclear cellular debris and metabolicbyproducts by recognition of ubiquitous,conserved autoantigens. Response tothis recognition may drive the clonal ex-pansion of leukemic cells, thereby con-tributing to clinical outcome.

MATERIALS AND METHODS

Cloning, Expression, and Purificationof CLL mAbs

Studies were approved by the Institu-tional Review Board of North Shore–LIJJewish Health System in Manhasset, NY,USA, and performed in accordance withthe Helsinki agreement. RNA from bloodmononuclear cells was converted intocDNA, and expressed IG V regions weresequenced as described (6). GenBank ac-cession numbers for these rearrange-ments are provided in Table 1. Cloning,expression, and purification of mAbswere performed as reported (19).

Intracellular Immunostaining of HEp-2Cells

HEp-2 cell-coated slides (INOVA Diag-nostics Inc., San Diego, CA, USA) were in-cubated for 1 h at 4° C with CLL mAbs(2–200 μg/mL) followed by FITC-conjugated goat anti-human IgG, 1 h atroom temperature. Slides were mountedand visualized with an Axiovert 200M in-verted microscope (Zeiss, Thornwood,NY, USA) and analyzed with AxioVision

version 4.5 software (Zeiss), or with anOlympus FluoView 300 confocal micro-scope (Olympus America Inc., Center Val-ley, PA, USA).

Binding of CLL mAbs to Healthy andApoptotic Cell Surfaces

Flow cytometry. Fifteen h after inductionof apoptosis (γ-irradiation, 4000–5000R),2.5 × 105 human T (Jurkat) or B (RAMOS)cells were incubated with CLL mAbs(50 μg/mL) for 1 h at 4° C, followed byeither FITC-conjugated F(ab’)2 goat anti-human IgG (Southern Biotech, Birming-ham, AL, USA) or FITC-conjugatedmouse anti-human IgG (BD Pharmingen,San Jose, CA, USA). Samples then wereexposed to Annexin V-PE and 7-AAD pervendor protocol (BD Pharmingen). Incompetition assays, CLL mAbs wereincubated overnight at 4° C with MDA-BSA (25–100 μg/mL) before processingcells as above. Data were acquired usinga FACS Calibur flow cytometer (BectonDickinson Immunocytometry Systems,San Jose, CA, USA) and analyzed usingFlowJo software version 7.2.4 (Tree StarInc, Ashland, OR, USA).

Confocal microscopy. Jurkat T cellswere incubated with 2 μM camptothecin(Sigma-Aldrich, St. Louis, MO, USA) for3 h and processed as per Radic et al. (11). Inbrief, 2.5 × 105 cells were exposed to CLLmAbs (50 μg/mL) for 1 h at 4° C and re-suspended in a mixture of SYTOX Orange(Molecular Probes Inc., Eugene, OR, USA),Alexa Fluor 647-conjugated Annexin V(Molecular Probes), and FITC-conjugatedF(ab’)2 goat anti-human IgG (INOVA) for30 min at 4° C. In other experiments, Jurkatcells (not camptothecin-treated) were fixed(PBS, 6% paraformaldehyde), permeabi-lized (PBS, 0.1% Triton X-100), and stainedas described. Cells were dried on Poly-D-Lysine Cellware coverslips (BD BioCoat,Bedford, MA, USA) and mounted with flu-orescent mounting solution (DakoCytoma-tion, Carpinteria, CA, USA). Slides wereimaged and analyzed as described.

Enzyme ImmunoassaysTarget antigens were purchased as fol-

lows: Ox-LDL (INTRACEL, Frederick,

MD, USA), malondialdehyde (MDA)-BSA (Cell Biolabs Inc., San Diego, CA,USA), 4-hydroxynonenal (HNE)-BSA(α Diagnostics International, San Anto-nio, TX, USA), PC-BSA (BiosearchTechnologies Inc., Novato, CA, USA).1-palmitoyl-2-(5-oxovaleroyl)-sn-glycero-3-phosphorylcholine (POVPC)-BSA wasprepared from POVPC (Avanti PolarLipids Inc., Alabaster, AL, USA) as de-scribed (20). Polystyrene 96-well plates(NUNC, Rochester, NY, USA) were incu-bated overnight at 4° C with 5 μg/mL ofeach antigen. After blocking with 3%milk, plates were incubated with CLLmAbs (5 and 25 μg/mL) overnight at 4° C, followed by alkaline phosphatase-conjugated goat anti-human IgG (South-ern Biotech). Assays were developedusing p-nitrophenyl phosphatase (Sigma-Aldrich) as substrate.

Proteomic Autoantigen MicroarraysSurveys were performed with a panel

of 65 self-antigens selected for relevanceto autoimmune disease or because of re-activity with natural antibodies as de-scribed recently (21). Briefly, microarrayswere printed on nitrocellulose-coatedFAST slides (Whatman, Florham Park, NJ,USA) with a QSoft QArray Mini, usingQSoft microarray software (Genetix USA,Boston, MA, USA), with six replicates ofeach antigen. After blocking with What-man blocking buffer, slides were incu-bated with CLL mAbs (2 and 10 μg/mL)or control IgG in PBS, pH 7.4. Slides wereprocessed and data determined as de-scribed (21). Independent analyses withduplicate arrays also were performed toconfirm the reproducibility of results.Antigens were removed from the finaldata set if ≤ 3 CLL mAbs were reactive.Significance was assigned for P < 0.05.

RESULTS

CLL mAbs React with IntracellularStructures of Live Human Cells

We expressed recombinant mAbs from19 U-CLL and 9 M-CLL that utilizedIGHV most commonly observed in CLL,1–69, 3–21, 4–34, and 4–39, as well as oth-

6 6 6 | C A T E R A E T A L . | M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8

C L L M A B S B I N D A P O P T O T I C A N D C H E M I C A L O X I D A T I O N E P I T O P E S

R E S E A R C H A R T I C L E

M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8 | C A T E R A E T A L . | 6 6 7

Tab

le 1

.Mo

lec

ula

r ch

ara

cte

ristic

s o

f Ig

H a

nd

IgL

rea

rran

ge

me

nts

in C

LL m

Ab

s u

sed

in t

he

se s

tud

ies

CLL

M

ut.

c%

dIG

Hf

%d

LCD

R3

am

ino

IGLf

No

.aSu

bse

tbIG

HV

sta

tus

Mu

t.IG

HD

IGH

JH

CD

R3

am

ino

ac

id s

eq

ue

nc

ee

Ge

nBa

nk

IGLV

IGLJ

Mu

t.a

cid

se

qu

en

ce

eG

en

Ban

k

068

61-

69U

-CLL

0.0

3-16

3ARGGDYDYVWGSYRSNDAFDI

AY

5536

40K

3-20

KJ4

0.0

QQYGSSP_T

AY

5749

3525

86

1-69

U-C

LL0.

03-

163

ARGGIYDYVWGSYRPNDAFDI

AY

0554

85K

3-20

KJ1

0.0

QQYGSSPGT

AY

5749

38

014

91-

69U

-CLL

0.0

3-03

6ATKNDFWSGYYEGYYYYYYMDV

AF0

2195

1L3

-01

LJ1

0.0

QAWDSSTCYV

AY

0431

0624

69

1-69

U-C

LL0.

33-

036

ARSDQNY_DFWSGYFRYYGMDV

FJ03

9781

K3-

20K

J10.

0QQYGSSPET

FJ03

9795

355

91-

69U

-CLL

0.0

3-03

6ARADLPYYDFWSGMY_YYGMDV

FJ03

9784

K1-

05K

J10.

0QQYNSY_QT

AJ6

9790

4

358

NA

1-69

M-C

LL6.

81-

265

AVLPSPLVGATQIWGDY

FJ03

9785

K3-

20K

J14.

5QQYGSSPPT

AJ6

9790

6

154

11-

18M

-CLL

2.4

6-19

4AREQWLVLSHFDY

AF0

2200

9K

1D-3

9K

J10.

0QQSYSTPPWT

AY

0431

6434

01

1-02

U-C

LL0.

36-

194

AREQWLVLKNFDY

AY

5536

45K

1D-3

9K

J20.

3QQSYSTPPYT

AY

5749

4336

01

1-03

U-C

LL0.

36-

194

AREQWLVLNYFDY

AY

5536

47K

1D-3

9K

J20.

0QQSYSPPPYT

AY

5749

4527

01

1-02

U-C

LL0.

05-

124

ARVQWLGLRHFDY

AY

0554

87K

1D-3

9K

J20.

0QQSYSTPPYT

AY

5749

40

DO

1328

1-02

U-C

LL0.

01-

266

ARQFSGSPTRYYYYYGMDV

FJ03

9793

K4-

01K

J40.

0QQYYSTPQT

FJ03

9803

376

NA

1-24

U-C

LL0.

04-

171

ATSAFTVTHAEYFQH

FJ03

9787

K3-

11K

J10.

3QQRSNWPWT

FJ03

9798

415

NA

1-03

U-C

LL0.

03-

104

ARRPESGYSFVTPFDY

FJ03

9788

K1D

-39

KJ2

0.0

QQSYSTPPHT

FJ03

9799

366

93-

21U

-CLL

0.0

3-03

6ARGVLNY_DFWSVYY_YYGMDV

FJ03

9786

K4-

01K

J40.

0QQYYSTPLT

FJ03

9797

282

23-

21M

-CLL

2.4

1-26

6ARDANGMDV

AY

5536

43L3

-21

LJ3

1.0

QVWDSSSDHPWV

AY

5749

4141

22

3-21

M-C

LL2.

0N

D6

ARDQNGMDV

AY

5536

48L3

-21

LJ3

0.7

QVWDSSSDHPWV

AY

5749

46

562

NA

3-21

U-C

LL1.

76-

195

VRDEITVAATRPCP

FJ03

9789

L3-2

1LJ

31.

4QVWDSSDDHPWV

FJ03

9805

169

NA

3-33

M-C

LL8.

83-

104

AREJJVTGQGGFDY

AY

0554

80K

4-01

KJ4

0.0

QQYYSTPLT

FJ03

9794

141

NA

4-34

U-C

LL0.

02-

025

ARGDWRIVVVPAAVDTAMAANWFDP

AF0

2200

5K

1-27

KJ2

0.3

QKYNSAPRMYT

AY

0431

60

DO

8N

A4-

34U

-CLL

0.0

6-13

4ARGGRAAAGKGLLDY

FJ03

9792

K3-

20K

J10.

0QQYGSSPPRT

FJ03

9802

183

44-

34M

-CLL

3.1

5-18

6ARGYGDTPTIRRYYYYGMDV

AF0

2194

8K

2-30

KJ2

1.7

MQGTHWPPYT

AY

0430

9724

04

4-34

M-C

LL3.

15-

186

ARGYADTPVFRRYYYYGMDV

AY

5536

41K

2-30

KJ2

2.3

MQGTHWPPYT

AY

5749

3734

24

4-34

M-C

LL2.

75-

186

ARGWGDTPMLKRYYYYGLDV

AY

5536

46K

2-30

KJ2

2.3

MQGTHWPPYT

AY

5749

44

114

84-

39U

-CLL

0.7

6-13

5ARRFGYSSSWY_GLDWFDP

AY

2683

72K

1D-3

9K

J10.

7QQSYSTPRT

AY

0430

9465

78

4-39

U-C

LL0.

36-

135

ASKTGYSSSWY_GRDWFDP

FJ03

9790

K1D

-39

KJ1

0.0

QQSYSTPRT

FJ03

9800

845

84-

39U

-CLL

0.0

6-13

5ASSTGYSSSWYSPTNWFDP

FJ03

9791

K1D

-39

KJ4

0.0

QQSYSTPRT

FJ03

9801

260

NA

4-b

U-C

LL0.

02-

026

ARAEIVVVPAAYYYYYGMDV

FJ03

9783

K1-

27JK

30.

0QKYNSAPQVT

FJ03

9796

255

NA

4-59

M-C

LL4.

23-

224

ARHRGYESSGYYSSYFDY

FJ03

9782

L3-0

1LJ

12.

2QAWDSSTVV

FJ03

9804

aC

LL n

um

be

r (N

o.)

.bN

A: N

ot

att

ribu

tab

le t

o a

cu

rre

ntly

de

fine

d s

tere

oty

pic

su

bse

t (2

5,26

).cM

uta

tion

(M

ut.

) st

atu

s o

f IG

HV

: un

mu

tate

d (

U-C

LL)

or m

uta

ted

(M

-CLL

).dPe

rce

nt

mu

tatio

n o

f IG

HV

or I

GLV

as

co

mp

are

d w

ith g

erm

line

ac

co

rdin

g t

o IM

GT

(54)

.eH

CD

R3

or L

CD

R3

am

ino

ac

id s

eq

ue

nc

es

ac

co

rdin

g t

o IM

GT

with

bo

ld c

ha

rac

ters

ind

ica

ting

no

n-id

en

tica

l re

sidu

es

am

on

g m

em

be

rs o

f th

e s

am

e s

ub

set

(exc

lud

ing

CLL

014

).f G

en

Ban

k a

cc

ess

ion

nu

mb

ers

for I

GH

an

d IG

L n

uc

leo

tide

se

qu

en

ce

s.

ers (Table 1). Several mAbs were exam-ples of stereotypic rearrangements thatappear to be CLL-specific (22–26).

Many CLL mAbs react with autoanti-gens in solid phase assays (15–18). There-fore, we determined reactivity by stan-dard immunofluorescence with viableHEp-2 cells that subsequently were fixedand permeabilized to permit mAb accessto intracellular targets; such cells are uti-lized to identify autoantigens in autoim-mune disorders such as systemic lupuserythematosus (SLE). All 28 CLL mAbsbound HEp-2 with varying efficiencies;some showed unequivocal binding at2 μg/mL (for example, No. 114), whileothers required 100–200 μg/mL to dis-criminate binding above background (forexample, No. 412). Unlike SLE Abs thatcommonly bind nuclear structures, 27 of28 CLL mAbs bound predominantly cy-toplasmic and perinuclear structures(18; Figure 1). Only one mAb, CLL255(IGHV4-59, mutated), reacted with nucleiin a homogeneous pattern that sparednucleoli (see Figure 1). A few mAbs alsoreacted with nucleoli (data not shown). Inseveral instances, the pattern of bindingof members of the same stereotypic sub-set was similar (for example, cytoskeletalbinding pattern of CLL mAbs 114, 657,and 845 of subset 8 (18; see Figure 1).

CLL mAbs BIND THE SURFACES OFAPOPTOTIC HUMAN LYMPHOCYTES

Since apoptotic cells are a source of au-toantigens in autoimmune conditions(27), we tested by indirect immunofluo-rescence reactivity of CLL mAbs withB and T cells after in vitro induction ofapoptosis, comparing results with thesame cells in the viable state (Table 2;Figure 2). After induction of apoptosis(Table 2; see Figure 2), 18 of 28 mAbsbound surfaces of human B (RAMOS)and T (Jurkat) cell lines. The vast major-ity were unmutated (16 of 18), althoughseveral U-CLL mAbs failed to bind thesetargets (Nos. 258, 376, and 562). Most unmutated mAbs exhibited antigenicpolyreactivity, binding several unrelated(auto)antigens (18; unpublished data).Moreover, of the 2 M-CLL mAbs that

bound apoptotic cells (Nos. 154 and 342),154 had six IGHV mutations, althoughfour were silent and two replacementsfell within framework regions. The othercell-reactive mutated mAb (No. 342)showed preferential binding to liveB cells (data not shown).

CLL mAbs reacted similarly with apo-ptotic T and B cells from healthy subjectsand from CLL patients as well as withapoptotic murine thymocytes (data notshown). This indicated that the autoanti-gens recognized were conserved acrossspecies.

CLL mAbs Bind ApoptoticAutoantigens in Membrane Blebs andApoptotic Bodies

During apoptosis, autoantigens cantranslocate from intracellular compart-ments to surface membrane blebs and bereleased as apoptotic bodies (10,11). We

set out to determine if CLL mAbs reactedwith such structures. Since this questionwas more conveniently addressed usingcells in solution (as opposed to adherentHEp-2 cells), we first determined thatCLL mAbs bound intracellular antigensof viable RAMOS B and Jurkat T cells.All CLL mAbs that reacted with apopto-tic cell surfaces also bound intracellulartargets of these same lymphocytes whenviable (Figure 3A–C).

Next, Jurkat T cells were incubatedwith camptothecin to induce apoptosis, asconfirmed by surface membrane bindingof Annexin V (Figure 3E,I,M). CLL mAbsreacted with discrete Annexin V+ (AnnV+)surface membrane blebs on apoptotic cells(Figure 3F,G). In addition, mAbs reactedwith apoptotic bodies (Figure 3H–O) thatcontained DNA fragments (Figure 3L,O)or not (Figure 3H,K), indicated by pres-ence or absence of SYTOX orange fluores-cence (Figure 3D,H,L).

CLL mAbs React with AutoantigensAssociated with SystemicAutoimmunity

Based on the above results, we searchedfor reactivity of CLL mAbs with clinically-

6 6 8 | C A T E R A E T A L . | M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8

C L L M A B S B I N D A P O P T O T I C A N D C H E M I C A L O X I D A T I O N E P I T O P E S

Figure 1. Immunoreactivity patterns of CLLmAbs with intracellular structures of livehuman cells. Healthy HEp-2 cells were fixed,permeabilized, and exposed to 50 μg/mL ofmAbs 114 (A,B), 657 (C), 845 (D), 366 (E),and 255 (F). Binding was detected withFITC-labeled goat anti-human IgG. Stan-dard immunofluorescence microscopy im-ages (A,C–F). Confocal microscopy show-ing mAb 114 binding to fiberlike structuressuggestive of cytoskeletal components (B).Studies by Chu et al. indicate that vimentinis the target of this mAb (49).

Figure 2. CLL mAbs bind apoptotic B andT cells. γ-irradiated RAMOS (upper) andJurkat (lower) B and T cells were exposedunder non-permeabilizing conditions toAnnexin V-PE and mAbs 014 (A,C) orDO13 (B,D) (50 μg/mL) plus FITC-labeledgoat anti-human IgG.

relevant autoantigens using a solid phasearray (21) containing 65 autoantigens andadditional control ligands (Figure 4). No-tably, 47 autoantigens displayed signifi-cant binding by three or more mAbs. Themost commonly recognized self antigenwas tubulin, a highly represented speci-ficity of human natural antibodies (28). Inaddition, significant reactivity was de-tected with several autoantigens involvedin systemic autoimmunity, for example,

Sm, Ku, snRNP A, BB’, and C, CENP-B,and others (see Figure 4). Notably, severalof these autoantigens localize to apoptoticblebs during apoptosis (27).

CLL mAbs React with Products ofOxidation Generated DuringApoptosis

In the proteomic arrays, CLL mAbsalso displayed high mean reactivity lev-els with LDL modified by MDA (see

Figure 4), a provocative result becauseapoptotic cells display oxidation-induced neo-epitopes on lipoproteins,lipids, and proteins (12,13), and miceimmunized with syngeneic apoptoticcells produce autoantibodies that bindoxidation-induced neo-epitopes (13).Such oxidation-derived neoantigensresult from conjugation of metabolitesof lipid peroxidation, such as MDA,POVPC, and HNE, to membrane mole-

R E S E A R C H A R T I C L E

M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8 | C A T E R A E T A L . | 6 6 9

Table 2. Antigenic reactivities of CLL mAbs

Reactivity with

Healthy Apoptotic ApoptoticCLL IGHV Mutation HEp-2 RAMOS Jurkat POVPC- OxidizedmAb Subseta gene statusa cellsb B cellsc T cellsc MDA-BSAd BSAd HNE-BSAd PC-BSAd LDLd

068 6 1-69 U-CLL + + + + – – + + –258 6 1-69 U-CLL + – – + – – + –

014 9 1-69 U-CLL + + + + + + + + + + + + – + –246 9 1-69 U-CLL + + + + + + + – – + –355 9 1-69 U-CLL + + + + + + + + + + – – – –

358 NA 1-69 M-CLL + – – – – – – –

154 1 1-18 M-CLL + + + + – – – –340 1 1-2 U-CLL + + + + + – – + –360 1 1-3 U-CLL + + + + + + + – – + –270 1 1-2 U-CLL + + + + + + + – – + +

DO13 28 1-2 U-CLL + + + + + + + + + + + + + + + +

376 NA 1-24 U-CLL + – – – – – – –

415 NA 1-3 U-CLL + + + + – – + –

366 9 3-21 U-CLL + + + + + + + – + +

282 2 3-21 M-CLL + – – – – – – –412 2 3-21 M-CLL + – – – – – – –

562 NA 3-21 U-CLL + – – – – – – –

169 NA 3-33 M-CLL + – – – – – – –

141 NA 4-34 U-CLL + + + – – – – –

DO8 NA 4-34 U-CLL + + + + + – + –

183 4 4-34 M-CLL + – – – – – – –240 4 4-34 M-CLL + – – – – – – –342 4 4-34 M-CLL + + + + – – – +

114 8 4-39 U-CLL + + + + + + + + + + – + +657 8 4-39 U-CLL + + + + + + + + + + + + + – + + +845 8 4-39 U-CLL + + + + + + + + + + – + + +

260 NA 4-b U-CLL + + + + + – – + –

255 NA 4-59 M-CLL + + – – + + – – –

aStereotypic subset and mutation status as in Table 1.bBinding to live HEp-2 cells (subsequently fixed and permeabilized) in arbitrary units based on immunofluorescence intensity.cBinding to apoptotic RAMOS and Jurkat cells by flow cytometry as determined by percent positive events: – = < 2%; + = 2%–5%; + + =5%–10%; + + + = 10%–15%; + + + + = 15%–20%.dBinding to antigens in ELISA based on signal-to-background ratio: – = < 2; + = 2-5; + + = 5-10; + + + = 10-15; + + + + = > 15.

cules (29). To determine the extent ofsuch reactivities, we tested if CLL mAbsreacted with MDA-, POVPC-, and HNE-modified BSA.

Whereas the 28 CLL mAbs exhibitedinsignificant binding to native BSA, sev-eral mAbs reacted with BSA modifiedwith MDA, HNE (Figure 5; Table 2), andPOVPC (Table 2). Reactivity was de-tected more commonly with MDA-BSA(19 of 28) than with POVPC-BSA (8 of 28)and HNE-BSA (1 of 28); all mAbs bind-ing POVPC and HNE reacted withMDA. As was found for apoptotic cellbinding, the most reactive mAbs ex-pressed unmutated IGHV, although

some mAbs with mutated IGHV (Nos.154, 342, and 255) did bind MDA.

The polar head group of phosphoryl-choline (PC), hidden in the membrane ofhealthy cells but exposed during apopto-sis (12), is an immunologically accessible

component of POVPC (30). Therefore, asimilar analysis was carried out with PC-BSA (Table 2). Fifteen CLL mAbs boundPC. Comparing mAb binding to PC withthat to POVPC, 8 other mAbs bound onlyPC, whereas 15 others bound MDA. Four

6 7 0 | C A T E R A E T A L . | M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8

C L L M A B S B I N D A P O P T O T I C A N D C H E M I C A L O X I D A T I O N E P I T O P E S

Figure 4. Hierarchical cluster analysis of autoantigen microarray reactivity of CLL mAbs.Results represent a two-dimensional cluster analysis in which CLL mAbs with similar bindingpatterns were assigned by computational algorithm to adjacent positions. In the seconddimension, a similar computational clustering of antigenic reactivities is shown. Heat map:bright red represents highest relative antibody activity level; bright blue the absence ofrelative activity level. Results denote reactivity with antigens bound by ≥ 3 mAbs.

Figure 5. CLL mAbs bind neoantigens created by oxidation. Reactivity of CLL mAbs (25 μg/mL) with oxidation-induced molecular adducts (MDA and HNE) of BSA wastested by ELISA.

Figure 3. Differential binding of CLL mAbsto healthy versus apoptotic cells. (A–C) Vi-able Jurkat T cells were fixed, permeabi-lized, and exposed to SYTOX Orange andmAb 114 (50 μg/mL) plus FITC-labeled(Fab’)2 goat anti-human IgG. (D–G) Reac-tivity of mAb 114 with surfaces of camp-tothecin-treated Jurkat T cells was ana-lyzed using cells that were notpermeabilized and co-stained with SYTOXOrange and Annexin V-Alexa Fluor 647;reactivity with membranes of apoptoticblebs without (H–K) and with (L-O) DNAwas determined similarly. Blue: nucleicacid binding by SYTOX Orange; red:membrane Annexin V binding; green: CLLmAb binding; yellow: co-localization ofAnnexin V and mAb binding.

MDA-reactive mAbs did not bind PC.Since POVPC can be a component ofmodified LDL, oxidized LDL also wastested as a target (Table 2). Seven CLLmAbs reacted with oxidized LDL; 5 of the8 POVPC-reactive mAbs bound oxidizedLDL. mAbs 014, DO8, and 225 reactedsolely with POVPC; mAbs 270 and 342 re-acted with oxidized LDL and not POVPC.

Correlation between Apoptotic Celland Neo-epitope Binding

To check for a link between CLL mAbbinding to apoptotic lymphocytes and tochemically-induced neo-epitopes, we per-formed an inhibition assay, blocking mAbbinding to apoptotic cells with MDA-BSA,the most reactive of the chemically-modified autoantigens. CLL mAb 114bound 23.5% and DO13 bound 6.8% ofapoptotic Jurkat T cells (Figure 6A,G). Bothreacted more with cells at later (AnnV+/7–AAD+, 55.7% and 18.5% respectively;Figure 6C,I) than earlier (AnnV+/7–AAD–,9.5% and 4.6% respectively; Figure 6B,H)stages of apoptosis.

Notably, with four-fold excess ofMDA-BSA, reactivity with each dyingsubpopulation was inhibited ~30%–50%(Figure 6B,C,H,I). Inhibition was dosedependent (not shown) and specific,since native, unmodified BSA had negli-gible effects on mAb binding at compa-rable doses (Figure 6D–F, J–L).

Correlation between binding of neo-determinants and apoptotic cells wassupported by further comparisons. Of 19MDA-BSA reactive mAbs, 17 recognizedapoptotic lymphocytes; moreover, of 9that did not react with MDA-BSA, only 1bound apoptotic cells. Thus a significantcorrelation (χ2 = 16.33, P = 0.0001; Pear-son’s R = 0.764, P < 0.001) existed be-tween MDA-binding and recognition ofmembrane-associated determinants oncells undergoing apoptosis. Of note, thestrongest neoantigen binders displayedthe broadest reactivity with targets on theautoantigen microarray (see Figure 4).

Structure/Function CorrelatesNext, we looked for correlations be-

tween binding domain structure and

specific antigenic reactivity. Dendro-grams from hierarchical clustering of theproteomic array data showed five majormAb groupings based on antigenic reac-tivity patterns (see Figure 4). The clus-tered set with the greatest autoantigenreactivity included CLL mAbs 355, 360,270, and 246, which all expressed unmu-tated IGHV1 family genes; these werefrom two stereotypic sets, subsets 1 and 9(25,26). These mAbs also displayed sig-nificant MDA-BSA reactivity by enzyme-

linked immunosorbent assay (ELISA).The second most strongly reactive clus-tered set included mAbs 154, 340, 366,412, and 114. All of these but mAb 412displayed reactivity with apoptotic cellsand MDA-BSA. Furthermore, all thesemAbs except for 412 and 154 were codedby unmutated IGHV. Hence, mAbs withthe broadest range of autoantigen bind-ing in the proteomic microarray werethe strongest neoantigen binders (seeFigure 4; Table 2).

R E S E A R C H A R T I C L E

M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8 | C A T E R A E T A L . | 6 7 1

Figure 6. Inhibition of CLL mAb binding to apoptotic cells. γ-Irradiated Jurkat T cells wereexposed to CLL mAbs 114 or DO13 (25 μg/mL) plus FITC-labeled (Fab’)2 goat anti-IgG.(A–C and G–I) mAb binding without (red) or with (green) MDA-BSA (100 μg/mL). (D–Fand J–L) mAb binding without (red) or with (green) native BSA (100 μg/mL).

The strongest reactivity with apoptoticcells was found for CLL mAbs express-ing IGHV1-69 (Table 2). Among these,mAbs belonging to subset 9 (Nos. 014,246, 355), expressing IGHV1-69/IGHD3-3/IGHJ6 rearrangements, bound more ef-fectively than other IGHV1-69-expressingcases (Table 2). Intriguingly, mAb 366,which also falls into subset 9 based on aIGHD3-3/IGHJ6 rearrangement eventhough it expresses IGHV3-21 (25), re-acted well, but not to the same degree asthose subset 9 members using IGHV1-69.This suggests that 1-69 provides an addi-tional binding advantage when co-expressed with this HCDR3. On theother hand, CLL mAb 358 that expressesa mutated IGHV1-69 did not react withapoptotic cells.

Furthermore, CLL mAbs 068 and 258,which express IGHV1-69 but belong tosubset 6, had diminished apoptotic cellbinding compared with subset 9 mem-bers (Table 2). Although CLL mAb 068 re-acted with apoptotic cells, mAb 258 didnot, presumably owing to the few differ-ences in HCDR3 (D→I, position 5 andS→P, position 15) and LCDR3 (1 aminoacid longer – G in 258, position 9) struc-ture (Table 1).

Most CLL mAbs (excepting Nos. 258,141, and 255) that bound apoptotic cellsalso reacted with MDA neoantigens(Table 2), although the degree of reactiv-ity did differ for some mAbs. For in-stance, IGHV1-69-expressing subset 9members that reacted well with apoptoticcells were considerably less reactive withMDA, whereas those CLL mAbs with thebest MDA reactivity (for example, Nos.340, 270, DO13, and 360 expressingIGHV1-02 and IGHV1-03, and Nos.114,657, and 845, expressing IGHV4-39)bound apoptotic cells much less effec-tively. Furthermore, among MDA-reactiveIGHV1-expressing cases, specific IGHVgenes, 1-02 and 1-03, appeared more im-portant for binding than the completeIGHV-D-J rearrangements. The 1-02-expressing mAb DO13 reacted more ef-fectively with products of oxidation, inparticular MDA, than Nos. 154, 340, and360, even though these have remarkably

similar HCDR3 and LCDR3 amino acidsequences and express identical, un-mutated IGKV1D-39 segments (Tables 1and 2). Interestingly, mAb 270 had thesame high reactivity with MDA as DO13.Since mAb 270 is virtually identical inIGKV-J amino acid sequence to Nos. 154,340, and 360, the difference in binding toMDA is likely due to changes in HCDR3sequence, in particular the presence of anextra positively charged residue at theIGHD-J junction in mAb 270 and the lossof a negatively charged residue at theIGHV-D junction in Nos. 154, 340, and360 (Table 1). This is reminiscent of en-hanced autoreactivity in murine SLEmAbs with additional positively chargedamino acids in the HCDR3 (31).

Finally, the importance of specific IGrearrangements for binding was high-lighted by CLL mAbs 114, 657, and 845,which express IGHV4-39/IGHD6-13/IGHJ5 + IGKV1D-39/IGKJ1 rearrange-ments. These mAbs have virtually identi-cal IGHV (one change at the IGHV-Djunction) and similar, albeit not identicalHCDR3 amino acid sequences, pairedwith identical LCDR3s (Table 1). ThesemAbs reacted similarly with all autoanti-gens tested (Table 2).

DISCUSSIONThe goal of this study was to better

understand the role that antigen-smIg in-teractions might play in the develop-ment, progression, and evolution of CLL.Since CLL likely results from transforma-tion of an autoreactive B cell (15–18), wefocused on defining both autoantigensrelevant to the disease and also a processthat yields autoantigens. Our data sug-gest that CLL mAbs react with autoanti-gens associated with cellular apoptosisand the associated oxidation of proteinsand lipids that occurs naturally duringthis process.

Specifically, we found that most, if notall, CLL cells produce mAbs reactivewith intracellular components of severallineages of human cells. These intracellu-lar autoantigens are located predomi-nantly within the cytoplasm (Table 2; seeFigure 1; 18), in contrast to disorders

such as SLE where autoantigens aremainly nuclear (27). Since apoptotic cellsoften are targets of autoantibodies in sys-temic autoimmunity (10,11), we evalu-ated the ability of CLL mAbs to bind thesurfaces of cells undergoing apoptosis.Approximately 60% of CLL mAbs re-acted with apoptotic lymphocytes, butnot with the same cells in the viable state(see Figure 2; Table 2). These principlesare consistent with those recently re-ported by Lanemo et al. (32). In addition,we tracked relevant apoptosis-associatedautoantigens from intracellular locationsto surface membrane blebs and apoptoticbodies (see Figure 3). Furthermore, pro-teomic studies (see Figure 4) indicatedthat CLL autoantigenic targets can be thesame as those in systemic autoimmunity.Both findings suggest a link between thelymphocytes involved in autoimmunediseases and CLL.

Autoantigens can also be created denovo during apoptotic death. Consistentwith this possibility, we (Table 2; see Fig-ure 5) found significant reactivity of CLLmAbs with metabolites of lipid peroxi-dation, as did others (32),. Therefore, acentral role for apoptosis in selectingand expanding CLL cells and their pre-cursors is likely, since oxidation is linkedto apoptosis (12,13).

Binding of CLL mAbs to apoptoticcells could be inhibited partially (~30 to50%) by pre-incubation with MDA-BSA(see Figure 6), suggesting that epitopeson apoptotic cells and chemically-modified molecules could be similarstructurally. Alternatively, a singleparatope conformation could interactwith both native and modified autoanti-gens (33). However, since the inhibitionwith MDA-BSA was not complete, eitherthe mAbs react with unmodified self-antigens or with distinct neoantigens cre-ated by other chemical modifications oc-curring during apoptosis. The latter isreinforced by the findings that somemAbs, especially those with unmutatedIGHV1-69, are more reactive with oxida-tion adducts than apoptotic cells, andthat other mAbs of the IGHV1 family(1-02 and 1-03) react oppositely (Table 2).

6 7 2 | C A T E R A E T A L . | M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8

C L L M A B S B I N D A P O P T O T I C A N D C H E M I C A L O X I D A T I O N E P I T O P E S

In general, CLL mAbs with the greatestautoreactivity were polyreactive and notmutated somatically (Table 2). However,some unmutated mAbs failed to bind thetargets (Nos. 258, 376, and 562), indicatingthat lack of somatic mutation alone didnot necessarily confer autoantigen bind-ing activity. Conversely, a few M-CLLmAbs did react (Nos. 154 and 342). Com-paring autoantigenic reactivities withmAb structure, it is clear that specificIGLV family and gene use, rearrange-ments, and particular IGH-IGL chain pair-ings influence binding, as has been noted(34). Furthermore, unique residues withinHCDR3 appear to augment autoreactivity,as in experimental systems (31). Despitethe dramatic skewing of IgH and IgL Vgene use and association, and the result-ing remarkably similar in H and L CDR3sand overall mAb structure, in some in-stances even limited amino acid differ-ences altered patterns of antigen binding(Tables 1 and 2).

How do these observations relate tothe development and evolution of CLL?The normal pre-immune repertoire isenriched in antibodies reactive with thetypes of epitopes tested. These naturalantibodies serve housekeeping func-tions such as clearance of cellular debrisand products of normal catabolism.They also represent the humoral arm ofinnate-like B-cell immune responses,providing a first-line of defense againstblood-borne pathogens (35). This princi-ple is embodied in the murine TEPC15(T15) mAb, which expresses a canonicalV gene rearrangement that recognizesPC, a component of bacterial cell walls,and protects against lethal sepsis in ani-mal models (35). PC is also a componentof mammalian cell membranes that isexposed during apoptotic death (30).Since T15 and its IgM equivalent, EO6,can bind oxidation-induced epitopesand apoptotic cells (30), reactivity withphospholipid-derived epitopes, cata-bolic byproducts, and microbes arelinked. In this regard, we have shownthat several of the CLL mAbs that bindapoptotic cells, neo-determinants, andPC react with bacteria (36). Collectively,

these findings may relate to the devel-opment of CLL shortly after bacterialpulmonary infections (37).

Our findings also may have implica-tions for understanding the normal cellu-lar precursors of CLL cells. T15-relatedmAbs arise primarily from the murineB-1 cell (CD5+) compartment (14). Simi-larly, CLL cells, or at least those with un-mutated IGHV, could arise from ahuman B-1 equivalent. This subset hasnot been identified in man, althoughdata suggest there is a human analog(38). However, most circulating CD5+

human B cells are not representative of ahuman B-1 equivalent since they lack thepoly/autoreactivity that is the hallmarkof CLL (18). Alternatively, CLL cellscould emanate from marginal zone (MZ)B cells, a subset with features and prop-erties linked to B-1 cells (39). MZ B cellsin mice display autoreactive smIgs (40),some of which lack allelic exclusion (41),as do some CLL cells (42). Derivationfrom an MZ B cell is also consistent withgene expression studies, suggesting thatU-CLL and M-CLL cells arise from thesame precursor (43,44). In this regard, itis intriguing that some CD5+ B cells re-side in the subepithelial spaces of tonsil-lar crypts, a human MZ-equivalent (45).

Finally, our findings may help clarifymechanisms of clonal expansion in CLL.B cells can recognize and respond toapoptosis-associated antigens not onlythrough smIgs but also via Toll-like re-ceptors (TLRs), which interact with mi-crobial pathogen-associated molecularpatterns (46). Since certain TLRs in B cellsare located within endosomal compart-ments, stimulatory ligands must gain en-trance to endosomes via cell surface re-ceptors; smIg can provide this entrance(47). Indeed, in systemic autoimmunity,smIg capturing autoantigens that containDNA and RNA, the stimulatory ligandsfor TLRs 9 and 7, respectively, and deliv-ering these to intracellular TLRs canpropagate autoimmune responses (48).

Several polyreactive CLL mAbs reactwith DNA (15–18) and RNA-proteincomplexes (see Figure 4) that translocateto apoptotic membranes (10,11,27). In ad-

dition, non-muscle myosin heavy chainIIA (49) and vimentin (32,49), proteinsthat also translocate to the surface of ap-optotic cells (50–52) often in associationwith DNA (53), are targets of CLL mAbs.Therefore, since CLL smIgs bind DNAand RNA directly or complexed with na-tive or modified molecules, they couldcapture such nucleic acids from apopto-tic membranes and blebs (see Figure 2)and deliver them to endosomes to enablestimulatory interactions with TLRs 9and 7. Thus, smIg and endosomal TLRco-signaling could enhance CLL cell sur-vival, expansion, and evolution, andthereby influence a patient’s clinicalcourse and outcome.

ACKNOWLEDGMENTSWe appreciate the statistical help of

Peter A Clement, Bull Information Sys-tems, Chelmsford, Massachusetts, UnitedStates of America. Studies were fundedin part by CA87956 (National Cancer In-stitute) and RR018535 (National Centerfor Research Resources). The KarchesFoundation, Peter Jay Sharp Foundation,Prince Foundation, Marks Foundation,Jean Walton Fund for Lymphoma andMyeloma Research, and Joseph ElettoLeukemia Research Fund also providedsupport.

REFERENCES1. Rai KR. (1997) Chronic lymphocytic leukemia in

the elderly population. Clin. Geriatr. Med. 13:245–9.2. Damle RN, et al. (2002) B-cell chronic lympho-

cytic leukemia cells express a surface membranephenotype of activated, antigen-experiencedB lymphocytes. Blood 99:4087–93.

3. Chiorazzi N, Ferrarini M. (2003) B cell chroniclymphocytic leukemia: lessons learned fromstudies of the B cell antigen receptor. Ann. Rev.Immunol. 21:841–94.

4. Stevenson F, Caligaris-Cappio F. (2004) Chroniclymphocytic leukemia: revelations from theB-cell receptor. Blood 103:4389–95.

5. Chiorazzi N, Rai KR, Ferrarini M. (2005) Chroniclymphocytic leukemia. N. Engl. J. Med. 352:804–15.

6. Fais F, et al. (1998) Chronic lymphocytic leuke-mia B cells express restricted sets of mutatedand unmutated antigen receptors. J. Clin. Invest.102:1515–25.

7. Damle RN, et al. (1999) Ig V gene mutation statusand CD38 expression as novel prognostic indica-tors in chronic lymphocytic leukemia. Blood94:1840–7.

R E S E A R C H A R T I C L E

M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8 | C A T E R A E T A L . | 6 7 3

8. Hamblin TJ, Davis Z, Gardiner A, Oscier DG,Stevenson FK. (1999) Unmutated Ig VH genesare associated with a more aggressive form ofchronic lymphocytic leukemia. Blood 94:1848–54.

9. Shlomchik MJ, Marshak-Rothstein A, WolfowiczCB, Rothstein TL, Weigert MG. (1987) The role ofclonal selection and somatic mutation in autoim-munity. Nature 328:805–11.

10. Casciola-Rosen LA, Anhalt G, Rosen A. (1994)Autoantigens targeted in systemic lupus erythe-matosus are clustered in two populations of sur-face structures on apoptotic keratinocytes. J. Exp.Med. 179:1317–30.

11. Radic M, Marion T, Monestier M. (2004) Nucleo-somes are exposed at the cell surface in apopto-sis. J. Immunol. 172:6692–700.

12. Shaw PX, Goodyear CS, Chang MK, Witztum JL,Silverman GJ. (2003) The autoreactivity of anti-phosphorylcholine antibodies for atherosclerosis-associated neo-antigens and apoptotic cells. J. Im-munol. 170:6151–7.

13. Chang MK, et al. (2004) Apoptotic cells withoxidation-specific epitopes are immunogenic andproinflammatory. J. Exp. Med. 200:1359–70.

14. Masmoudi H, Mota-Santos T, Huetz F, CoutinhoA, Cazenave PA. (1990) All T15 Id-positive anti-bodies (but not the majority of VHT15+ antibod-ies) are produced by peritoneal CD5+ B lympho-cytes. Int. Immunol. 2:515–20.

15. Broker BM, et al. (1988) Chronic lymphocyticleukemic cells secrete multispecific autoantibod-ies. J. Autoimmun. 1:469–81.

16. Sthoeger ZM, et al. (1989) Production of autoanti-bodies by CD5-expressing B lymphocytes frompatients with chronic lymphocytic leukemia.J. Exp. Med. 169:255–68.

17. Borche L, Lim A, Binet JL, Dighiero G. (1990) Ev-idence that chronic lymphocytic leukemia B lym-phocytes are frequently committed to productionof natural autoantibodies. Blood 76:562–9.

18. Herve M, et al. (2005) Unmutated and mutatedchronic lymphocytic leukemia derive from com-mon self-reactive B cell precursors despite ex-pressing different antibody reactivity. J. Clin. In-vest. 115:1636–43.

19. Wardemann H, et al. (2003) Predominant autoan-tibody production by early human B cell precur-sors. Science 301:1374–7.

20. Bird DA, et al. (1999) Receptors for oxidizedlow-density lipoprotein on elicited mouse peri-toneal macrophages can recognize both themodified lipid moieties and the modified pro-tein moieties: implications with respect tomacrophage recognition of apoptotic cells. Proc.Natl. Acad. Sci. U. S. A. 96:6347–52.

21. Silverman GJ, et al. (2008) Genetic imprinting ofautoantibody repertoires in systemic lupus erythe-matosus patients. Clin. Exp. Immunol. 153:102–16.

22. Tobin G, et al. (2004) Subsets with restricted im-munoglobulin gene rearrangement features indi-cate a role for antigen selection in the develop-ment of chronic lymphocytic leukemia. Blood104:2879–85.

23. Messmer BT, et al. (2004) Multiple distinct sets of

stereotyped antigen receptors indicate a key rolefor antigen in promoting chronic lymphocyticleukemia. J. Exp. Med. 200:519–25.

24. Widhopf GF 2nd, et al. (2004) Chronic lympho-cytic leukemia B cells of more than 1% of pa-tients express virtually identical immunoglobu-lins. Blood 104:2499–504.

25. Stamatopoulos K, et al. (2007) Over 20% of pa-tients with chronic lymphocytic leukemia carrystereotyped receptors: pathogenetic implicationsand clinical correlations. Blood 109:259–70.

26. Murray F, et al. (2008) Stereotyped patterns of so-matic hypermutation in subsets of patients withchronic lymphocytic leukemia: implications forthe role of antigen selection in leukemogenesis.Blood 111:1524–33.

27. Tan EM. (1994) Autoimmunity and apoptosis.J. Exp. Med. 179:1083–6.

28. Guilbert B, Dighiero G, Avrameas S. (1982) Natu-rally occurring antibodies against nine commonantigens in human sera. I. Detection, isolationand characterization. J. Immunol. 128:2779–87.

29. Esterbauer H, Schaur RJ, Zollner H. (1991)Chemistry and biochemistry of 4-hydroxynone-nal, malonaldehyde and related aldehydes. FreeRadic. Biol. Med. 11:81–128.

30. Chou MY, et al. (2008) Oxidation-specific epitopesare important targets of innate immunity. J. Intern.Med. 263:479–88.

31. Radic MZ, et al. (1993) Residues that mediateDNA binding of autoimmune antibodies. J. Im-munol. 150:4966–77.

32. Lanemo Myhrinder A, et al. (2008) A new perspec-tive: molecular motifs on oxidized-LDL, apoptoticcells, and bacteria are targets for chronic lympho-cytic leukemia antibodies. Blood 111:3838–48.

33. Sethi DK, Agarwal A, Manivel V, Rao KV,Salunke DM. (2006) Differential epitope position-ing within the germline antibody paratope en-hances promiscuity in the primary immune re-sponse. Immunity 24:429–38.

34. Martin T, Crouzier R, Weber JC, Kipps TJ,Pasquali JL. (1994) Structure-function studieson a polyreactive (natural) autoantibody.Polyreactivity is dependent on somatically gen-erated sequences in the third complementarity-determining region of the antibody heavychain. J. Immunol. 152:5988–96.

35. Briles DE, et al. (1981) Antiphosphocholine antibod-ies found in normal mouse serum are protectiveagainst intravenous infection with type 3 strepto-coccus pneumoniae. J. Exp. Med. 153:694–705.

36. Hatzi K, et al. (2006) B-cell chronic lymphocyticleukemia (B-CLL) cells express antibodies reac-tive with antigenic epitopes expressed on thesurface of common bacteria. Blood 108:12a.

37. Landgren O, et al. (2007) Respiratory tract infec-tions and subsequent risk of chronic lymphocyticleukemia. Blood 109:2198–201.

38. Solvason N, Kearney JF. (1992) The human fetalomentum: a site of B cell generation. J. Exp. Med.175:397–404.

39. Martin F, Kearney J. (2000) B-cell subsets and themature preimmune repertoire. Marginal zone

and B1 B cells as part of a “natural immunememory.” Immunol. Rev. 175:70–9.

40. Chen X, Martin F, Forbush KA, Perlmutter RM,Kearney JF. (1997) Evidence for selection of apopulation of multi-reactive B cells into thesplenic marginal zone. Int. Immunol. 9:27–41.

41. Li Y, Li H, Weigert M. (2002) Autoreactive B cellsin the marginal zone that express dual receptors.J. Exp. Med. 195:181–8.

42. Rassenti LZ, Kipps TJ. (1997) Lack of allelic ex-clusion in B cell chronic lymphocytic leukemia.J. Exp. Med. 185:1435–45.

43. Klein U, et al. (2001) Gene expression profiling ofB cell chronic lymphocytic leukemia reveals a ho-mogeneous phenotype related to memory B cells.J. Exp. Med. 194:1625–38.

44. Rosenwald A, et al. (2001) Relation of gene ex-pression phenotype to immunoglobulin mutationgenotype in B cell chronic lymphocytic leukemia.J. Exp. Med. 194:1639–47.

45. Dono M, et al. (2007) CD5(+) B cells with the fea-tures of subepithelial B cells found in humantonsils. Eur. J. Immunol. 37:2138–47.

46. Akira S, Takeda K, Kaisho T. (2001) Toll-like re-ceptors: critical proteins linking innate and ac-quired immunity. Nat. Immunol. 2:675–80.

47. Tian J, et al. (2007) Toll-like receptor 9-dependentactivation by DNA-containing immune com-plexes is mediated by HMGB1 and RAGE. Nat.Immunol. 8:487.

48. Lafyatis R, Marshak-Rothstein A. (2007) Toll-likereceptors and innate immune responses in sys-temic lupus erythematosus. Arthritis Res. Ther.9:222.

49. Chu CC, et al. (2008) Chronic lymphocyticleukemia antibodies with a common stereotypicrearrangement recognize non-muscle myosinheavy chain IIA. Blood 2008, Sep 23 [Epub aheadof print].

50. Kato M, Fukuda H, Nonaka T, Imajoh-Ohmi S.(2005) Cleavage of nonmuscle myosin heavychain-A during apoptosis in human Jurkat T cells.J. Biochem. 137:157–66.

51. Muller K, Dulku S, Hardwick SJ, Skepper JN,Mitchinson MJ. (2001) Changes in vimentin inhuman macrophages during apoptosis inducedby oxidised low density lipoprotein. Atherosclero-sis 156:133–44.

52. Moisan E, Girard D. (2006) Cell surface expres-sion of intermediate filament proteins vimentinand lamin B1 in human neutrophil spontaneousapoptosis. J. Leukoc. Biol. 79:489–98.

53. Shoeman RL, Hartig R, Traub P. (1999) Character-ization of the nucleic acid binding region of theintermediate filament protein vimentin by fluo-rescence polarization. Biochemistry 38:16802–9.

54. Lefranc MP. (2004) IMGT-ONTOLOGY andIMGT databases, tools and Web resources for im-munogenetics and immunoinformatics. Mol. Im-munol. 40:647–60.

6 7 4 | C A T E R A E T A L . | M O L M E D 1 4 ( 1 1 - 1 2 ) 6 6 5 - 6 7 4 , N O V E M B E R - D E C E M B E R 2 0 0 8

C L L M A B S B I N D A P O P T O T I C A N D C H E M I C A L O X I D A T I O N E P I T O P E S