Relationship of the CD5 B cell to human tonsillar lymphocytes that express autoantibody associated cross reactive idiotypes

Relationship of the CD5 B cell to human

tonsillar lymphocytes that express

autoantibody-associated cross-reactive

idiotypes.

T J Kipps, S F Duffy

J Clin Invest.

1991;

87(6)

:2087-2096.

https://doi.org/10.1172/JCI115239

.

We examined human tonsillar B cells for expression of autoantibody heavy-chain or kappa

light-chain cross-reactive idiotypes (CRIs), respectively defined by murine MAbs G6 or

17.109. We find 17.109 or G6 each specifically binds a subpopulation of B cells,

respectively reacting with 3.8 +/- 3% (mean +/- SD) or 2.0 +/- 1.2% of all tonsillar

lymphocytes. Cells reactive with both 17.109 and G6 comprise only 0.4 +/- 0.3% of tonsillar

lymphocytes. Although each tested specimen had 17.109-positive cells, 2 of 19 tonsils

(11%) did not have any G6-reactive cells. We find that CRI-positive cells and CD5 B cells

both co-express slgD but fail to bind peanut agglutinin or MAbs specific for CD10, indicating

that both cell types reside in the mantle zones of secondary B cell follicles. However, less

than half of the B cells bearing one or both of these CRIs express detectable levels of CD5.

Nevertheless, we find that G6-reactive lymphocytes constitute a multiclonal population of

cells that express homologous heavy chain variable region genes, each rearranged to one

of several distinct and apparently nonmutated D and JH gene segments. Collectively, these

studies indicate that expression of nondiversified autoantibody-encoding variable region

genes may not be an exclusive property of B cells that bear detectable levels of the CD5

surface antigen.

Research Article

Find the latest version:

Relationship

of

the

CD5 B

Cell

to

Human

Tonsillar

Lymphocytes

That Express

Autoantibody-associated

Cross-reactive

Idiotypes

Thomas J.Kippsand Stuart F.Duffy

Department ofMedicine, University of California, San Diego, La Jolla, California 92093-0945

Abstract

Weexamined humantonsillarB cellsforexpressionof autoan-tibody

heavy-chain

or kappa light-chain cross-reactive idio-types

(CRIs),

respectively defined by murine MAbs G6 or

17.109. Wefind 17.109orG6 each specifically binds a sub-population of B cells, respectively reacting with 3.8±3% (mean±SD) or 2.0±1.2% of all tonsillar lymphocytes. Cells reactive with both 17.109andG6 comprise only 0.4±0.3% of tonsillar lymphocytes. Although each tested specimen had 17.109-positive cells, 2of 19 tonsils(11%) did nothave any

G6-reactive

cells. Wefindthat

CRI-positive

cells and CD5 B cells both co-expressslgDbutfail to bind peanutagglutininor

MAbsspecificforCD10, indicatingthat both cell typesreside inthe mantle zonesofsecondary B cellfollicles.However, less than half ofthe B cellsbearingone or both of these CRIs

ex-press detectable levelsof CD5. Nevertheless,wefindthat G6-reactive lymphocytes constitute a multiclonal population of cells that express homologous heavy chain variable region genes, eachrearrangedtooneof several distinctandapparently nonmutated D and JH gene segments.Collectively,these stud-iesindicatethatexpression of nondiversified autoantibody-en-codingvariableregiongenes may not be anexclusiveproperty ofBcells that beardetectablelevelsofthe CD5 surface

antigen.

(J. Clin. Invest. 1991.

87:2087-2096.) Key

words: autoanti-body * CD5 B cell *cross-reactiveidiotype-V geneexpression

Introduction

MalignantBcellsfrompatients with chronic lymphocytic leu-kemia

(CLL)'

or small lymphocytic lymphoma (SLL) fre-quently express antibodies bearing cross-reactive idiotypes (CRIs) associated withIgMautoantibodies.Inastudyofover 30 CLLpatients, one-fourth ofthepatients withkappa light chain-expressing leukemia cells had malignant lymphocytes thatexpressedaCRIdefined by reactivity withaMAb, desig-nated17.109(1). Thisantibody,preparedagainstanIgM

rheu-Addressreprintrequests to Dr. Kipps, Department of Medicine,

Uni-versityof California,SanDiego,La Jolla, CA 92093-0945.

Receivedfor publication5November 1990 and in revised form 4

February1991.

1.Abbreviations used in this paper: APC, allophycocyanin; CD, cluster of differentiation antigen; CDR, complementarity determiningregion;

CLL, chronic lymphocytic leukemia; CRI, cross-reactive idiotype;D,

diversity;FR, antibody variable region framework; JH, heavy-chain

joining region; PCR, polymerase chain reaction;PE, phycoerythrin;

PNA, peanutagglutinin;RF, rheumatoidfactor;VHgene, heavy chain

variable regiongene; VK gene, kappa light chain variable region gene.

matoid factor (RF) recognizesakappa

light

chain-associated CRIpresent on manyIgM

paraproteins

withRF

activity (2, 3).

Furthermore, nearly

one-fifth

of all surface

Ig-expressing

CLL werefoundtoreactwith G6 (4). This MAb is

specific

foran

Ig

heavy

chain-associated

CRI present on several RF parapro-teins (5).

Interestingly,

nearly half of the CLLcaseswith leuke-miacellsreactive with 17.109 also reacted with

G6, reflecting

a

biased

co-expression

of these CRIs (4,

6).

Evaluation of the molecular

basis

for these CRIs in CLL revealed that each is a serologic marker for

expression

ofa

conserved immunoglobulinV

gene(s)

with littleor nosomatic mutation. The 17.109-reactive neoplastic B cellsfrom unre-lated patientsexpressnearly identical kappa

light

chain vari-ableregiongenes(VK genes)thatbelongtotheVKIIIb sub-sub-group(7, 8). Finally, thesequencesof the VKgenes

expressed

by 17.109-reactive leukemia cells share>99%nucleic acid

se-quence homology with a

germline

VKIII geneisolated from placentalDNA,designated

Humkv325

(9).

Similarly,

G6-reac-tive leukemia cells from unrelated patientsexpress homolo-gousheavy chain variable

region

genes(VH

genes)

of the VH1 subgroup (10).Moreover,

comparisons

of the deduced amino acidsequences of G6-reactive CLL and that of

G6-negative

antibody heavy chains encoded by VH1 genes suggestthat the G6-CRI in CLL is relatively resilientto substitutions within complementarity determining

region

3(CDR3), butaffected by

permutations

withinCDR1and CDR2.Together these data arguethattheG6-CRIinCLL isaserologic marker fora con-served VH1 gene alsoexpressed with littleor nosomatic muta-tion.

Conceivably, the

frequent expression

of these

autoanti-body-associated CRIsinCLLorSLL mayreflecttherepertoire ofthe presumed normal cellcounterpart tothesemalignancies, namely the clusterof differentiation antigen 5 (CD5) B cell. CD5Bcellsconstituteasmallsubpopulation of humanB lym-phocytes in lymphoid organs andperipheral blood ofnormal adults( 1-15). Thesecellscomprisethemajority ofBcellsin theprimary follicles of human fetalspleen(16, 17),wherehigh proportions ofcells express these CRIs (18). Finally, CD5 B lymphocyteshave been reported to include nearly all the cells thatspontaneously produce IgM autoantibodies (19, 20), in-cludingRF. Asnoted,IgMautoantibodies frequentlybear the 17.109 CRI or the G6 CRI (3, 5, 21). Also, IgM molecules co-expressing both CRIs almost invariablyhave RFactivity (22).However, therelationshipbetween CD5 B cells and non-malignant B cells that co-express autoantibody-associated CRIshas not been evaluated. Accordingly,weexamined hu-man tonsillar lymphocytes to determine the prevalence and surface antigen phenotype ofthe Bcellsthatbear the 17.109 and/ortheG6 CRI.

Methods

Patientmaterials. Weobtained freshresidual lymphoid material from

patients (ages 7-69, mean 26yr) undergoing surgical tonsillectomy

J.Clin. Invest.

©TheAmericanSocietyfor ClinicalInvestigation,Inc.

0021-9738/91/06/2087/10 $2.00

(TableI). Sections of each tonsil were frozen at -700C for immunohis- Chromaprobe Immunological) as described (1). Biotinylated antibod-tochemical analysesasdescribed (4, 23). A single-cell suspension was ies used to stain cells for three- or four-color immunofluorescence anal-madefrom the remainder of each specimen. Mononuclear cells were yses were developed with Texas Red-avidin (Molecular Probes, Inc.). isolated on density gradients of Ficoll-Hypaque, washed in Hanks' bal- Fluorescein-conjugated peanutagglutinin(FITC-PNA) was obtained ancedsaltsolution and then suspended in FCS containing 10% dimeth- from Vector Laboratories, Inc., Burlingame, CA. With the exception of ylsufoxide (Sigma Chemical Co., St. Louis, MO), for storage in liquid PE or APC-labeled MAbs, all reagents were spun at 100,000gfor 15 nitrogen. Viability of recovered cells exceeded 90% before analyses. min before use.

Antibodies andfluorescenceprobes. G6, a murine

IgG,

MAb reac- Flow cytometric analyses. Three- and four-color flow cytometric

tive with aautoantibodyheavychain-associated CRI (5) was provided analyses were performed using a FACScan (Becton, Dickinson & Co.,

byDrs.RizgarMageed and RoyJefferis(University of Birmingham, San Jose, CA) equipped with Consort30 (Hewlett-Packard Co., Palo Birmingham, England). We obtained an IgG1 MAb specific for a Alto, CA) or a dual laser FACStarP` (Becton Dickinson & Co.)

VKIIIbsubgroup determinant(s) (24), from Dr. George Abraham (Uni- equipped with a MicroVAX computer (Digital Equipment Corp., versityofRochester, Rochester, NY). MAb 17.109 is as described (2). Marlboro, MA), asdescribed (30). The latter allowed for use ofElectric

Fluorochrome-conjugated MAbs specific for CDIO (CALLA), and Desksoftware (Dr. R. Hardy, Fox Chase Cancer Center, Philadelphia, CDl9(Leu 12)werepurchasedfromBecton-Dickinson Immunocy- PA). Unless otherwise indicated, cells were labeled with

fluorochrome-tometrySystems, San Jose, CA. Hybridomas obtained from the Ameri- conjugated MAbs and reagents in staining medium consisting of bio-canType Culture Collection, Rockville,MDwereOKT3, an IgG2. tin-deficientRPMI-1640 (Irvine Scientific, Santa Ana, CA), 3% FCS

anti-CD3 (25), DA4-4, anIgG, anti-human mu heavychain(26), (Gibco Laboratories, Grand Island, NY), and 50 mM Hepes (Irvine dTA4- 1,anIgG3 anti-human delta heavy chain (27), and THB-5, an Scientific),pH7.6. Although viability of the analyzed cell populations

IgG2,anti-CD21 (C3dR) (28).SC-l anIgG2,MAbreactive withhu- generally exceeded 90%, cells were stained also with propidium iodide

manCD5 (29), was provided by Dr. Robert Fox (Scripps Clinic and (1

;&g/ml)

toallow us to discriminate between living and dead cells ResearchFoundation, La Jolla, CA). Anti-human kappa or anti-hu- duringFACScan analysis (30).

manlambdaMAb-producinghybridomas were as described (1, 7).An- IsolationofG6-reactive cells. Tonsillar lymphocytes, 108, from ton-tibodieswerepurified from ascitesviaprecipitationwith ammonium silno.7(Table I),wereincubatedwithFITC-conjugated G6 for 30 min

sulfateandeitherabsorptionwith quartenaryaminoethyl(Pharmacia at4°C, washed,and thenincubated with 1.6 x 107 anti-mouse IgG FineChemicals,Uppsala, Sweden) or Protein ASepharose column magnetic microspheres(Dynabeads, Dynal, Inc., Oslo, Norway) for 1 h

chromatography (BioRad Laboratories, Richmond, CA).Purified anti- at4°Cwithgentleagitation.FACscan analysesof the G6-labeled cell

bodieswereconjugatedwitheitherfluorescein isothiocyanate (FITC; population beforeaddition of themagneticbeads demonstrated 4.1% Molecular Probes, Inc., Eugene, OR), normal human serum-biotin of the cellsdisplayed brightgreenfluorescencewhenexcited at 488 nm

(PierceChemical Co.,Rockford, IL), R-phycoerythrin(PE; Chroma- (TableI).Magneticbead-bound cellswereisolated by twice decanting probeImmunological,RedwoodCity,CA),orallophycocyanin (APC; the cellsuspensionin the presence ofastrongmagneticfield. FACScan

TableI.LymphocyteSubpopulationsin Human Tonsil

Sampleno. Sex Dx Patient age CD3 CD5 CD19 CD5 B IgM IgD K L 17.109 VK3b G6 G6/17.109

yr %

I M CT 16 4 27 97 21 76 74 49 41 4.2 6.1 3.4 0.2

2 F CT 20 6 25 93 16 75 69 51 38 6.3 9.1 4.5 0.7

3 M AT 22 26 31 73 2 58 54 39 30 4.4 5.5 2.0 1.0

4 F AT 20 15 29 78 13 45 29 50 33 1.3 2.2 0.8 0.1

5 F AT 7 39 44 64 3 59 42 40 30 4.0 4.3 1.5 0.5

6 M CT 41 22 29 77 6 55 45 44 31 5.1 5.5 2.5 0.2

7 M CT 32 19 50 74 31 54 50 44 31 5.1 7.8 4.1 0.8

8 M OSA 69 34 40 64 5 33 30 34 28 2.9 3.0 1.6 0.2

9 F AT 19 49 55 45 10 24 19 20 19 3.3 ND 1.2 0.2

10 M CT 41 18 32 62 12 59 54 40 23 2.4 2.6 0.0 0.0

11 M CT 27 32 35 63 2 36 23 26 28 3.1 ND 2.2 0.2

12 F CT 17 52 55 44 6 22 19 21 17 2.9 3.1 2.2 0.3

13 M CT 24 14 26 87 11 48 36 42 38 3.7 ND 2.0 0.5

14 F AT 4 13 36 77 26 62 59 40 39 4.6 5.8 1.8 0.3

15 M CT 30 19 28 77 12 64 59 44 35 3.4 5.9 3.4 0.3

16 M OSA 42 11 29 91 18 69 71 51 37 6.0 ND 0.0 0.0

17 F CT 21 27 47 71 17 48 45 40 30 2.6 5.6 2.3 0.2

18 M AT 14 4 40 97 36 83 83 62 31 4.3 ND 2.7 0.3

19 M CT 21 38 44 55 8 31 28 30 22 3.9 5.0 3.3 0.7

Average 26 23 37 73 13 52 47 40 30 3.9 5.1 '2.2 0.4

SD 15 14 10 16 10 18 19 11 7 1.2 2.0 1.2 0.3

Inthe column labeled "Dx"arethediagnoses underlyingthe medicalindicationfortonsillectomy:AT,acutetonsillitis(history<6mo);CT,

chronicorrecurrenttonsillitis;andOSA,obstructivesleepapnea. In the column labeled"patient age"arethe ages of thepatientsfrom whom

biopsyspecimenswereobtained. Values inremainingcolumns indicate the percentagesofcellsthat have the surfaceantigen phenotypelisted

atthe top of each column. Below each columnarethemeanpercentages(average)andstandard deviations of themean(SD)of the values in each column.ND,notdone.

analysisdemonstrated thedecanted cell populationdid notcontain

anyresidual labeled cells.Cellsadheringtothe tubewallinthe presence of the magnetwererecovered, washed inRPMI-1640,and then used as the source ofG6-positivecells. DNAwasisolated fromG6-positive

cellsasdescribed(8,10),except thatsubsequenttodigestionwith Pro-teinase-K,themagneticbeadswereremovedfromthepreparation by

1-min centrifugationat13,000 g.

Polymerase chainreaction(PCR)amplification ofgenomic DNA.

PCR wasperformedon 1 Agof DNA isolated fromG6-positivecells

using Taqpolymerase(31) (Cetus Corp.,Emoryville, CA)and primers

thatcorrespondtothesensestrandofasequencecommontothe leader

sequence of allVHl genes

[dGTTCHTCACCATGGACTGGACTG,

H=A/C/TJ and to the anti-sense strand of a JH consensus

[dCCTGAGGAGACMGTGACC, M=C/T],asdescribed

(10).

Atthe

5'endof eachprimer, restriction sitesfor HindIllorEcoRI, respec-tively,wereaddedtofacilitatecloningof theamplifiedgenefragments.

After 30 cycles ofPCR,theamplifiedgenefragmentwasdigestedwith

EcoRIandHindIII,purifiedvia agarosegelelectrophoresis,andthen

cloned intoaHindIII/EcoRI-digested pUC19vector

(Bluescribe,

Stra-tagene, LaJolla,CA)for transformation into competent XL-1 Blue (Stratagene). Coloniescontaining plasmidswithrecombinantinserts

wereidentifiedasdescribed(7, 10).PlasmidDNAcontaining amplified

recombinant insert DNA was isolated for double-stranded DNA

se-quencingasdescribed (32). Inaddition, 1/20thof the PCR reaction mixture was analyzed by the method of Southern, as described (8).

Oligonucleotideprobescorrespondingtotheanti-sense of a sequence in the second complementarity determining region (G6-CDR2)

[dGCGTAGTTTGCTGTACC]

or to the anti-sense of a sequence in the second framework region (VHI-FR2)

[dCCATCCACT-CAAGCCCJof the VHI gene expressed by G6-reactive CLL(10)were usedtoprobe thefilter-immobilizedDNA.Oligonucleotide synthesis, labeling, hybridization, and washing conditions were as described (1), with afinal 1-minwashing temperature of

490C.

For DNA sequence analyses, weusedtheGenetics Computer Group Sequence Analysis Software Packageon aVAX750mainframe computer.

Results

Multiparameter

flowcytometricanalyses of tonsillar lympho-cytes. We examined the lymphocytes from 19 tonsils using

multiparameter

flow cytometric analyses. Althoughmost speci-mens werefrom youngadultswithacute orchronic tonsillitis, two surgical specimens(nos. 8 and 16in Table

I)

were from patients with obstructive sleep apnea that required tonsillec-tomy to

ameliorate

partialupperairway obstruction.Wecould notdiscernanyconsistent phenotypic difference between these twoandthe other lymphocyte populations (Table

I).

The pro-portion ofCDI9-positiveBcells (73±16%;mean±standard de-viation) generally exceededtheproportion of CD3-positiveT cells(23±14%) by twofold (Table I).

Wereadily detected CD5 B cellsin each tonsil specimen (Table I). The proportion of CD5B cells wascalculatedas the percentage ofBcells labeled witha

fluorochrome-labeled

anti-CD5 MAb minus thebackground percentage (< 0.5%) of B cells stained with a

fluorochrome-labeled

anti-CD3 MAb of the samemouseIgisotype. Similarto thevariationnoted in the proportions ofB cellsexpressingCD5 in theperipheral blood of young healthyadults

(15),

theproportionsoftonsillar lym-phocytesthatco-expressedCD5 and pan-Blymphocyte surface

antigens

ranged from2% to36% ofthe totallymphocytesor from 3%to 42% ofthe total B cells (Table I).

We examined tonsillar lymphocytes for reactivity with 17.109 orG6. With few exceptions, each tonsil contained a subpopulation of CRI-positive B cells that expressed the 17.109

and/or

theG6 CRI

(Table

I).

Cells

bearing

either CRI

uniformly

expressed

B celldifferentiation

antigens

CDl9 and CD21

_(data

not

shown). Lymphocytes

thatbound17.109

com-prised

a

subpopulation

ofB cells that

expressed

kappa

light

chainof the

VKIIIb subgroup.

The

proportion

of

17.109-posi-tive cells

ranged

from 1.3%to6.3%

(m

=

3.9±1.2%)

ofthe total

lymphocytes,

from 3%to16%

(m

=

10±3%)

ofthe

kappa

light

chain-bearing cells,

andfrom 46%to97%

(m

=

77±16%)

ofthe cells

expressing kappa light

chains of the

VKIIIb subgroup

(Ta-ble

_{I). Except}

in tonsil

specimens

fromtwo

individuals,

G6-pos-itivecells

expressed

either

kappa

orlambda

light

chains and

comprised

0. 1-4.5%

_(m

=

2.1±1.2%)

ofthe total

lymphocytes.

Cells

expressing

both the17.109andG6CRI

comprised

avery small

subpopulation

constituting

<0.1-1%

(m

=

0.4±0.3%)

of thetotal

lymphocyte population (Table I).

Tonsillar

lympho-cytes fromtwoindividuals

(nos.

10 and 16 in Table

I)

didnot

possess any cells that

expressed

detectable levelsoftheG6-CRI

(Table

I).

Immunohistochemical

analysis

offreshfrozen tissue sections of these two tonsil

_specimens

confirmed that these tonsilslackedanyG6-reactive

lymphocytes (data

not

shown).

Four-color immunofluorescence

analyses

demonstrated that

CRI-positive

tonsillar

lymphocytes

share with tonsillar CD5Bcells the

phenotype

of mantlezonelymphocytes. Ton-sillarBcellscanbedelineated into

subpopulations

basedupon the differential

expression

of

sIgD,

CD10 and receptors for PNA

(Fig. 1).

Cells

residing

withinthe mantlezonebear

sIgD

and fail to bind PNA orMAbs

specific

forCD1O

(CALLA),

whereasBcellswithin the germinalcentergenerallylack sub-stantial

sIgD,

express low levels ofCDIOandavidlybindPNA

(33-35).

Wefind that> 95% of tonsillarBcellsreactive with 17.109

and/or

G6expresssIgD andfailtobindPNAorMAbs

specific

forCDI0

(Fig. 1, top).

CD2

1-positive

B cells that bind anti-CD5 MAbs also have this phenotype (Fig. 1,

middle).

However,wenotethat thegatedCD21-positiveB cell popula-tionsfailtobindanti-CD3 MAbsofthesame mouseIg

isotype

(Fig.

1,

bottom).

These data indicatethat tonsillarCD5 Bcells reside in the sameanatomic sites asdo CRI-positive B

cells,

namely within the mantle zones encapsulating thegerminal

centersof

secondary

B cell follicles.

Despite

sharingthephenotypeofmantle zone B cells,

CRI-bearing lymphocytes,

andCD5 B cells each appearto

comprise

independent

but partially overlappingBcell subsets.First,the proportions of CD5Bcellswithinthespecimens apparentlydo not correlate with the proportions of cells found to express eitherone orbothof thetwoCRIs.For example,inspecimen

no. 3, MAbs 17.109 or G6 respectively reacted with 4.4% or 2.0% of the lymphocytes, whereas only 2% of thisspecimen's cells expressed bothpan-B cellsurface antigens and detectable levels ofCD5 (Table I). In contrast, specimen no. 18 had a much greater proportion ofCD5 B cells (36%) than tonsil no. 3, but stillhad proportions ofCRI-bearing cells that were compa-rable tothat ofspecimen no. 3 (Table I). Secondly, over one half of the tonsillar B cellsexpressing 17.109 and/or G6 do not expressdetectablelevelsofthe CD5 surface antigen. For exam-ple,cellsstaining with PE- 17.109 displaybrightergreen

cc

LO

a 0

CIO

IgD PNA CD10

andG6 CRIsalso do not expressdetectablelevels of CD5(Fig. 3,

right panel).

Analyses

of

VH genes

rearranged

in G6-reactive tonsillar

lymphocytes.

To examine the molecular basis forCRI expres-sion in normaltonsil,weevaluatedtheVHgenesrearrangedin

G6-reactive lymphocytes using the PCR. Approximately 4

X 106 G6-reactive cells were isolated from tonsillar lympho-cytes of tonsil no. 7 (Table I). Because all G6-reactive IgM proteins of knownsequence haveheavy chain variable regions of theVHI subgroup (10, 36),weperformedPCRongenomic DNA extracted from

G6-positive

cells

using

oligonucleotide

0)

0 0)0

CD3

Figure 1.Probability distribution of tonsil-lar Blymphocytes electronically gated to depict only those cells staining positively with APC-conjugated anti-CD21. Contour plots describe the logarithmic red (ordi-nate) or green(abscissa)fluorescence in-tensity of cells stained with PE anti-CRI

MAb G6(top row), PE anti-CD5 (middle row), or PE anti-CD3 (bottom row) and

FITCanti-IgD (left column),FITC-PNA

(middle column), or FITC anti-CDIO (right column).

primers designed to amplify any VHl that hadundergone Ig gene rearrangement with DJH gene segments (10). Southern analyses of

l/20th

of the reactionproduct indicated that such PCR generatedfragmentsof - 600bp that hybridized

specifi-cally with an oligonucleotide complementary to the second framework region of all known VHl genes (VH1-FR2) (10) (data not shown). Furthermore, PCR of 1

gg

of DNA from G6-reactivetonsillar lymphocytesgeneratedamountsofVH1 geneproductcomparabletothatof PCRusing 1 uig ofgenomic DNA fromG6-reactiveCLL B cells(datanotshown).

Nucleic acidsequenceanalysesofthe clonedPCRproducts

... Figure2. FACScan immunofluorescence

;--. . analyses of tonsillar lymphocytes labeled with PE-conjugated 17.109 and

fluores-cein-conjugatedanti-CD3(left) or

fluores-2kit cein-conjugated anti-CD5 (right). Num-bers correspond to the relative red

(ordi-nate)

orgreen

(abscissa)

fluorescence intensityof the cells when excited at 488

CD5

nm.

U)

0

E

z

0

a)

ci)

a:

17.109+

Cells

Fluorescence

revealthatG6-reactive tonsillar cellsexpresshomologous VH1 genes.Independentcolonies of bacteriatransformed with plas-midscontaining inserts derived from PCR of G6-reactive ton-sillarlymphocyte DNAwere screened byhybridization with oligonucleotide

"VHl-FR2"

oran

oligonucleotide specific

for the second CDR ofVHl genes found to encode theG6-CRI (G6-CDR2) (10). The DNA from all such colonies annealed with either oligonucleotide probe under conditions of high

stringency

(datanot shown). 15colonieswerechosenat

ran-dom for nucleic acidsequenceanalyses. Allclones have VH1 genesthatarehighlyhomologoustothatofL7, whichin turn is identical with thatofaVH l geneexpressed inG6-reactive CLL (10). Theheterogeneity noted in the fifth base ofeachsequence reflects the sequence of theVH 1 primerused toamplify the rearranged VH genes(Fig.4).Thisprimer hasanA/C/T wobble atthisposition. Except for clone L8, allclonescontain func-tionally rearrangedVHl genes.The VHgeneof L8 isan appar-entpseudogene.ThisVH gene has asingle point deletion inthe second baseof codon 71 (Fig. 4)that produces downstream termination codons inthethirdframework and

complemen-tarity determining

region (Figs.4and5).Allother cloneshave functional VH1 geneswithout stop codonsor gene rearrange-ments that shift the reading frame ofdownstream gene seg-ments.

Despite each clone having nearly identical VH regions, thereisgreatdiversity amongthevariousclonesinthethird CDR(CDR3)andfourthframework regions.Theseregionsare encodedbythe D and JH gene segments, that undergo recombi-nationandamino-terminalnucleic acid base insertion immedi-ately before VHgene rearrangement(42, 43).As such, the se-quenceoftheCDR3generallyisidiosyncraticto each Ig V gene rearrangement. Only two clones,

L17

andL37, share identity in the deducedCDR3region,indicatingthatthese two variable region genes most likely were derived from the same B cell clone

(Fig.

5). All other clones have highly distinctive se-quences in the deduced CDR3region,indicating that each was derived fromanindependentIg Vgene rearrangement (Fig. 5). Most clones have a segment within CDR3 that has a high de-greeof sequence homologywith aknown germlineDminigene (37-41)(Fig.5). Ofthe 14 distinct clones analyzed, at least 11 different germlineDsegments may be represented. In the re-gion presumably encoded by a rearranged JH gene segment, nearly half (6/13) ofthe 13distinctandapparently functional clones use

JH4 (L17/37,

L25,L34, L36, L39, and L41), three

17.109/G6+

Cells

Figure3. Relative greenfluorescence intensityof tonsillarlymphocyteselectronically gatedto

in-cludeonlythose cellsstaining positivefor

PE-conjugated

17.109(left)or_{PE-conjugated} 17.109 andAPC-conjugated G6(right). Histograms 10o 1100

correspond

tothe

logarithmic

greenfluorescence

intensityof suchcellswhen co-stained with

FITC-conjugatedanti-CD3(solid line)or FITC-conju-gated anti-CD5(dashedline).

useJH5 (L22, L26, L42), four use JH6 (L7, L26, L30, and L33) (Fig. 5). The VHl pseudogene, L8, apparently isjuxtaposed with JHl. In addition to the variety noted in the D and JH gene segments, thelengthsoftheregions presumablyencodedbythe DJH gene segments varies greatly among the various clones, ranging from 18 codonsfor the shortest(L26)to38 codons for L30(Fig. 5).

Discussion

Thisstudyexamines therelationship betweenCD5 B cells and tonsillar lymphocytes that express major autoantibody-asso-ciated CRIs respectively defined by MAbs 17.109 or G6. The former constitute a significant subpopulation of B cells that readily isidentified in each tonsil population (Table I). The latter definesignificantly smaller proportionsof Bcells,and for G6-CRI-expressingcells, are notrepresented in all tonsils ex-amined. Both populations have surface antigen phenotypesof mantle zonelymphocytes. Although originally claimedto re-side in thegerminalcentersof human tonsil(1 1), tonsillar B cellsexpressing CD5 subsequentlywere notedtoexpress low levelsofLeu 16(CD20)using multiparameter flow cytometric analyses (34). Inaddition, we notethatboth tonsillar CD5 B cells andCRI-positiveBcells expresssurface IgDand lack de-tectablebinding affinityforanti-CD 10 MAbs and PNA. These phenotypic characteristics distinguish B cells located in the mantle zone from lymphocytes residing within thegerminal

center(34, 35).As

such,

thesestudies indicatethatCD5Bcells reside inthe same anatomic location asdocellsbearingthe 17.109orG6 CRI.

However,not allCRI-bearingBcellsare"CD5-positive"B cells. Despite sensitive immunofluorescence techniques, we findthatlessthanhalf ofthe B cellsbearingthe17.109-CRI or theG6-CRIexpressdetectable levels of CD5. Thisalsoistrue for B lymphocytesco-expressing both CRIs. Asindicated by studiesonisolatedIgMparaproteins(22), such 17.109 and G6 reactive cells have a high probability of bearingsIgwith RF

activity.

This suggeststhatRF-bearingBcells may not express detectablelevelsof CD5.

.

@

* * * * * @ * * * * * @ * * * * * @

******

*****@

* * * * * @

*****@

* * * * * .

******

* . . . . . * * * * * @ * * * * * @ * * * * * @ * * * * * * * * * * * * * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @

*****@

*****@

*****@

* * * * * * * * * * * * * * * * * e

* * * * * *

*****@

* * * * * @

*****s

* * * * * @

*****@

*****@

*****@

* * * * * @ * * * * * @

*****@

* * * * * * * * * * * @ * * * * * @ * * * * * @ * * * * * § * * * * * @ * * * * * @

*****@

*.

***@

* * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * * * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @

*****@

* * * * * @

*****@

*

*****@

* * * * @ * * * * * @ * * * * * @

*****@

*****@

*****@

* * * * * @ * * * * * @ * * * * * * * . . * * *

*****@

* * * * * @ * * * * * @

*****@

* * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @

*****@

* * . * * @ * * * * * * * * * * * e

* * * * * e

******

*****@

*****e

* * * * * e * * * * * *

Eq

*****@

*****@

*****@

* . . . .

* * * . * @

*****@

* * . * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @

*****@

*****@

* * * * * @ * * * * * @

*****@

* * * * * * * * * * * @

*****Z

*****@

* * * * * @

*****@

* * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ . * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @ * * * * * @

*****@

*****@

* * * * * @ * * * * * @ * * * * * @ * * * * * e

* * * * * e * * * * * @

* * * * * @

*****@

* * * * * @

*****@

C) U . U U .

* * * * * § * * * * * @ *****@

*****@

* * . @ * * * @ * * * @ * * * Z

* * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * * * * * * * * * * * * * @ * * * @ * * * * * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * . @ * * * @ * . . @ * * * | * * . @ * * * @ * * . @ * * * @ * * . @ * * * * * * * @ * * * @ * * * * * * * * * * . * * * * * * * . * * @ @ @ * * . @ * * * e * . . e

* * * @ * * * * * * * * * * * * * * * @ * * . @ * * * * * * * * * * * @ * * . @ * * * e * * . @

* * * @

* * . @ * * * s * * . s * * * @ * * * @ * * * * * * . @ * * * s * * * @ * * * @ * * * @ * * @ @ * * * @ * * * @ * * * @ * * * @ * * * Z * * * * * * * @

. . * Eq

* * * q * * * q * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * e * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @

* * * e * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @ * * * @

EqEq U U

* * * @

* * * @ * * * @ * * * e

* * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * e * * @ * * @

* * e

* * e

* * @ * * @ * * @ * * @ * * @ * * * * * * * * * * * @ * * @ * * * * * 9 * * @ * * @ * * 9 * * 9

* * @ * * @

* * e

* * @ * * @ * * @

* * *

* * @ * * Z * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * * * * @ * * @ * * @ * * Z * * | * * @ * * @ * * @ * * * * * e * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * * * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * @ @

* * 9 * * 9 * * 9 * * 9 * * 9 * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * @ * * § * * v * * @ * * @ * * @

* * @ * * e * * @ * * @ 4J H E-4

LI

E90-4 2t 14H to (>1 00 tooU :>U VO 'OE C2) + UH 0Iu

_Utn

U

U)E-;

EH :ON coins 4)U U) coc i4 tCD Ln i4u to O >1 Co CD

CoH<

r t7 t7 r- OCrun w o m v w r- m N rON v

uO

w o m v w r m_-4cs

r- w iN N N N m m m M m m v -w r Go r- C1 CN c N04 (I m m m m m qw

a- G- 0 G- 1. 6. 0. 1. 0 Fa 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 0 1- 0 0

* ... *

rCcN v LA D 0o v w r- _aM -4 N

r

(4

( N N (N (') ( (f ( ( C)

0 0 :> H E-4 u E-4 H 0 0 H

9

tJ

H 0

K t

0 0: uE ZE-4 u E-4U E-4 u

these cells acquired phenotypic characteristics ofgerminal centerlymphocytes; i.e., they lost sIgD expression and became positive for CD 10. However, as noted, CRI-bearing B cells that do not express CD5 have thephenotypeof mantle zone lym-phocytes. Thus,CD5-negative CRI-bearingBcells at leastdo not representsuchdifferentiatedCD5 B cells.

Alternatively, CRI-bearing B lymphocytes may constitute a B cell population with characteristics ofthe CD5 B cell. Al-thoughused asamarkerforasubpopulationofBlymphocytes, theexpression of CD5 hasnoknownrelationshipto the appar-entphysiology of the CD5Bcell,and may in fact be an epiphe-nomenon(reviewedinreference45). Furthermore, B cells can-not bedelineated readily into discrete subpopulations basedon thedetectable levelof CD5 surface antigen. Enumeration of CD5Bcellsrequires subtraction methods,inwhich thepercent numberof isotype-control-stained B cells that have fluores-cence above acertain threshold is subtracted frompercent num-berofanti-CD5-MAb-stained cells thataredetected above that samethreshold. Cells with extremelylowlevels of CD5 may not fluoresce brightly enough to be scored as positive, even whenstained withthebrightest of fluorochrome-labeled anti-CD5MAbs. Assuch, itmay beargued thatall

CRI-bearing

B cellsactuallymay expressCD5, butat levelsbelowdetection using flow cytometric analyses.

Finally, CRI-bearing"CD5-negative"Bcells may be simi-lar to theCD5-negative "sister" population ofBcells detected in the mouse (46-48). Inexperimentsaimedatcharacterizing the CD5 B cell populationandits progenitors in the mouse peritoneum, peritonealBcells werefoundthatdidnot express surface CD5 but apparentlyhadotherphenotypic traits pecu-liartoCD5 Bcells (46), such as thesurface expression of MAC-1

(CDl

lb) (49),extendedlongevity in vitro,orabilityto gener-atespontaneousoligoclonalBcellexpansions in vivothat re-semble human B cell CLL(50). Intransfer experiments,such

CD5-negative

sister cells initially were found toreconstitute miceonly whenever theadoptivehosts were reconstituted with CD5 Bcells (46). More recentstudies, however, suggest that such cells mayconstitute a relatedbut possibly independent and

self-renewing

Bcellsubpopulationthat mayarise latter in ontogenythanBcells thatexpressdetectable levelsof the CD5 surface

antigen

(46, 48). By analogy,CD5-negative CRI-bear-ingBcellsmay begeneratedat alaterstageofBcell develop-mentthanCRI-bearing CD5Bcells,butstillexpress manyof thetraitsthat areassociated with the CD5Bcell.

Onepeculiar phenotypic characteristicproposedforCD5 B cells isstable IgV gene expression with little orno somatic mutation. Independent clones of CD5 B cells may express identicalVgeneswithout evidence for somaticmutation (51, 52). The antibodiesencoded by such V genes frequently are reactive withavariety of selfantigens, i.e., proteolytically pro-cessederythrocytemembranes, denatured DNA, and/or the Fc portion of self IgG (53-57). Stable expression ofsuch V genes

with little or no somatic mutation may account for the finding that CD5 B cells are enriched for cells that spontaneously pro-duceIgM autoantibodies (19, 20).

Inthis light, itwasimportanttoexaminethe V genes ex-pressed by CRI-bearing cells. Previous studies indicated that the G6 CRI was found onIg heavy chains oftheVH1subgroup (58). In CLL, G6 reactive leukemia cells were found to express identical VH 1 genes, indicating that the G6 CRI is a marker for oneVH1 gene(s) expressed with little or no somatic mutation (10). However, itwas not known whether thisrelationshipis valid for nonmalignant G6-reactive B cells. Conceivably, B cells free to permute theirexpressedV genesthroughthe pro-cessof somatic mutationmay express G6-reactiveIgencoded by other VHl genes or VHl genes that have diversified from those present in the germline DNA. Indeed, the VHl heavy chains of a few IgM autoantibodies that share only modest homologytotheheavy chainexpressed byG6-reactive CLLare

knownto reactwith the G6 MAb(58). Particularlyinlightof the finding that a large fraction of G6-reactive cells do not

expressdetectablelevelsofCD5,itwasimportanttoexamine theVH1 genes usedbycellsexpressingthe G6-CRI.

Nucleic acid sequence analyses of the Ig VH1 genes rearranged and presumablyexpressedin G6-reactive cells re-vealedlittle evidencefor somatic mutation. Otherthan clones LI7,L30,andL34,all otherclones haveVHl genes that were identicalto L7(Fig. 4). The VH genesofclones LI 7and L34, thatapparentlyarederived from thesame Bcell clone(LI 7/ L34), differ from L7 atonly twosites

(Fig.

4). LI 7 andL34, moreover, share with an independent clone, L30, the same nonconservative G-> Asubstitution in codon74.Conceivably theL17/L34 and the L30VHl genes arederived from thesame allele.Inthiscase,thecommon basesubstitution notedin these clones may represent an allelic polymorphism in the VH1 gene(s)

encoding

G6-reactive heavy chains, rather than a so-matic mutation event. In any case,inover5,600 bases evalu-ated forthe VHgenesofthe 13 independentand functional clones, we find only three base substitutions, representing a

fidelity

in the DNA sequencesof>99.94%. Theabsence of even nonconservative base changes indicates that the con-served VHgene sequences are not the mere consequence of selection forcellshaving reactivity fortheG6MAb. Rather, these data indicate that G6-reactive tonsillar B cells express homologous VH1 geneswith littleor nosomatic mutations.

Evenin the CDR3 region,muchofthediversityapparently isnot the result of somatic mutation subsequent to Ig gene rearrangement. Except atsites flankingthe Dregion that may be generated by amino-terminal sequence insertions, three-fourthsof the independent clones (1 1/15)have short stretches of complete homology with known germlineDsegments (L7, L8, L22, L24, L26, L30, L33, L34, L36, L39, L42). Also, ex-ceptforcodonsimmediately contiguoustothe D-JHjunction, all clonesthat use the same JH gene segment have sequences

Figure4. Nucleic acidsequencecomparison ofrearrangedIgVH I genes amplified from G6-reactive tonsillar lymphocytes. The name of each

sequenceislistedontheleft.All sequences are compared withL7.Adot (-)indicates homology between the compared sequences and a letter

indicatesabasedifference.In L8, thesite of the single base deletion is framed with brackets. Listed above the sequence ofL7are the amino acid

positionnumbersaccordingtoKabat et al. (37), and descriptions indicating the location of the leader sequence, theCDR1 and CDR2. The 5' endof the leadersequence that corresponds to the PCR primer isindicated by a double underline. The sequences recognized by the

oligonucle-otide probes"FR2"(aapositions 44-49)and "CDR2"(aapositions55-60) are underlined. These sequence data are available from EMBL/

D

SEGMENT

GlyGlyLysGlyIleThrMetThrSer

L7 GGGGGCAAAGGAATTACTATGACTTCT

Dxp'1 T...GT-CGGGGAGTTATTATAAC LeuGlyGlyValEndGlnLeuValArgV

L8 GAGGGGGAGTATAGCAGCTCGTCCGCG DN4 ...

GlySerThrValThrThrGlyAsp

L17 GGGAGTACGGTGACTACGGGAGAT

DAl T--C.

A--A---GluGlyTyrCysSerSerThrSerCysSer

L22 GAGGGATATTGTAGTAGTACCAGCTGCTCG

DLR4 A...ATGCCC GluValValGlyCysSerSerThrSerCys

L24 GAGGTGGTTGGTTGTAGTAGTACCAGCTGC

DLR2 A-GATA...TATGCCC SerLeuArgArgTyrSerSerGlyTrpTyrGluGly

L25 TCCCTCCGGCGGTATAGCAGTGGCTGGTACGAGGGT

DN1 G. CA.

GlyAsp

L26 GGGGAC

Dq52 CTAACT...

L30 Dxpl

JH

SEGMENT

TyrCysTyrTyrTyrGlyMetAspValTrpGlyGlnGlyThrThrValThrValSerSer TACTGCTACTACTACGGTATGGACGTCTGGGGCCAAGGGACCACGGTCACCGTCTCCTCA

A... JH6

alLeuProAlaLeuGlyProGlyHisProGlyHisArgLeuLeu

TACTTCCAGCACTGGGGCCAGGGCACCCTGGTCACCGTCTCCTCA

GCTGAA... JH1 PheAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer TTTGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA TAC...A... JH4 AsnTrpPheAspProTrpGlyGlnGlyThrLeuValThrValSerSer AACTGGTTCGACCCCTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

. _T.A.

JH5

TyrTyrTyrTyrGlyMetAspValTrpGlyGlnGlyThrThrValThrValSerSer TACTACTACTACGGTATGGACGTCTGGGGCCAAGGGACCACGGTCACCGTCTCCTCA

TAC...G... JH6 PheAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer

TmTGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

TAC...A... JH4 AsnTrpPheAspProTrpGlyGlnGlyThrLeuValThrValSerSer AACTGGTTCGACCCCTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

. _T.A.

JH5

GlyArgThrArgValSerValSerThrLeuTyrAspSerSerGlyTyrTyrAspPheSerGly TyrTyrGlyMetAspValTrpGlyGlnGlyThrThrValThrValSerSer

GGCCGGACCCGGGTATCGGTTTCGACCCTCTATGATAGTAGTGGTTATTACGACTTCTCCGGC TACTACGGTATGGACGTCTGGGGCCAAGGGACCACGGTCACCGTCTCCTCA

*TATTA-G--AT-TTG-C...TA--Gly

L33 GGA

DLR1 ...TATTGTACTGGTGGTGTATGCTATACC

GlySerSerTrpThr

L34 GGCAGCAGCTGGACC DN1

A...TA-MetTyrAsnTrpAsnPhe

L36 ATGTATAACTGGAACTTT DM1 G...AC

GlySerThrValThrThrGlyAsp

L37 GGGAGTACGGTGACTACGGGAGAT

DAl T--C

A--A----AspThrSerSerGlyTyrTyrArg

L39 GATACTAGTAGTGGTTATTATCGC

Dxp4 GTATTAC..

TT-T-G...AC-GlyGlyValAlaGlyArgProHis

L41 GGCGGAGTTGCTGGTAGGCCCCAT DLR4

AGGATATT-TA-T-CCA----C--T,--GlyGlyMetGlyTyrCysSerGlyGlySerCys

L42 GGGGGGATGGGATATTGTAGTGGTGGTAGCTGC DLR2 A...TACTCCC

TACTACTAC... G... JH6

TyrTyrTyrTyrTyrGlyMetAspValTrpGlyGlnGlyThrThrValThrValSerSer

TACTACTACTACTACGGTATGGACGTCTGGGGCCAAGGGACCACGGTCACCGTCTCCTCA

G.

-G---JH6

PheAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer

TTTGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

TAC...A... JH4 PheAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer

TLTGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

TAC...A... JH4

PheAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer TTTGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

TAC...A... JH4 LeuAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer

CTTGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

TACT...A... JH4

PheAspTyrTrpGlyGlnGlyThrLeuValThrValSerSer

TTCGACTACTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA TAC--T...A... JH4

TyrTrpPheAspProTrpGlyGlnGlyThrLeuValThrValSerSer

TACTGGTTCGACCCCTGGGGCCAGGGAACCCTGGTCACCGTCTCCTCA

A.T .A. JH5

Figure5.DeducedDand JHsegmentsof theIgheavy chain variable

region

sequences.Eachsequenceis

compared

totheknown D and

JH

seg-mentsthat have thegreatesthomology(37-41).Sequence

comparisons

areindicatedasin

Fig.

2. Thesesequencedataareavailable from

EMBL/GenBank/DDBJ under accession numbers M65090-M65 104.

thatdiffer fromknown

germline JH

genesegmentsatthesame

siteand with thesamebase substitution

(Fig. 5).

For

_example,

all clones withJH4 differ from

published

germline JH4

mini-genes

(59)

withaA-- G substitutioninthe sixth codon of the

JHsegment(Fig. 5).In addition,anA- Gsubstitution at this

samelocationwasnotedinthededuced

_JH4

gene

segment

used

by G6-reactive leukemiccells

(10)

and

4B4,

ahuman-humanB

cell

hybridoma

expressing

a

VH4

gene with littleor nosomatic

mutation (56). Itis

unlikely

thatsuchsubstitutions would oc-cur

repeatedly

indifferent clones thathad

_undergone

indepen-dent gene rearrangements. As

such,

exceptatsites that maybe

permuted

during

the process of

Ig

generearrangement,the

dif-ferencesfrompublished JHsequence maybesecondaryto ge-netic

polymorphisms,

rather thantosomaticmutations in the

rearrangedJH gene segment.

Itis

unlikely

that the methods used in this

study

skewedour analysestofavor isolation of

rearranged

V_genesthatnot un-dergonesomaticmutation. We used

primers

corresponding

to consensussequencescommon toall

_JH

genesegmentsandtoa

sequenceatthe 5'

region

of all

VH

1 leader sequences.

ably, chancemutationsin arearranged VH1 gene at thesites recognizedbythese primersmaydisrupttheabilityof thatVHl gene to be amplified by our methods. However, these same

primershavesuccessfully amplified

VHl

genesotherthanthose foundinthisstudy, including VHl genes that haveundergone apparent somatic mutationsubsequent to V gene rearrange-ment. (T.J. Kipps, R. Kobayashi, andS.Duffy, unpublished observations). Also, theamount ofVHlgeneproductproduced in PCR of G6-reactive tonsillar lymphocytes is comparableto that produced in PCR of G6-reactive leukemia cell DNA. Thus,itseemsunlikely that the bias introduced byour meth-odscouldaccountfor the results thatweobtainedinthis

study.

Rather, a morelikely

explanation

is that G6-reactive tonsillar Bcells expresshomologousVH 1genes with little or nosomatic mutations.

Consistent with this interpretation isourfindingthatsome

ofthetonsilstesteddidnothave any cellsreactivewithG6. If the G6 MAb could bindtoheavychainsencodedby several different VHIgenes, thenfindingpersons without any

G6-posi-tive cells wouldbeunlikely. Thatover I10%ofthe tonsils

exam-ineddidnot have any cellsreactive with the G6 MAb, indicates that afairly high

proportion

ofthepopulationmaylack expres-sion ofanyVHgenesthatcanencodeIg heavy chainswiththis major CRI. Considering the

conservation

noted in the

VHl

gene sequencesof G6-reactive cells([10] and this study), the frequent expression ofthe VH 1 gene(s) encoding the G6-CRI the fetal immune repertoire (18, 60), the association of the G6-CRI with IgM autoantibodies (5, 61), and relative high-fre-quency expression of this CRI in CLL and SLL(4, 6), the absence of G6expressionmayreflectapolymorphisminVH1 gene use thatpossiblyhas

physiologic significance.

Finally, these studies indicate thatarelativelyhigh propor-tion of mantelzonelymphocytesexpresses autoantibody-asso-ciated CRIs.sIgD-bearingcellsconstituteonlyafraction ofthe total Bcells, ranging from 37% to 87% ofthe

CD19-positive

cells(Table I).Asnoted,the 17.109 CRIand/or the G6 CRI predominantly is expressedby cells thatco-expresssIgD.Since suchsIgD bearingcellspredominantlyrepresentmantlezone lymphocytes, the 17.109-reactive, orG6-reactivecells, respec-tively, mayconstitute 4-17% and0-12%of the mantlezone lymphocytes. Thesefrequenciesapproach those notedforCRI

expression

in CLL. Conceivably transformation of lympho-cytessuchasthosefound inthemantelzone may accountfor thehigh

frequency

expression ofautoantibody-associated CRIs inCLL and related malignancies.However, thereisan appar-entbiasfor co-expression ofboth 17.109 and G6 by leukemia cell

populations

thatisnotnotedin mosttonsillarlymphocyte populations studied. Biased co-expression ofthe 17.109 CRI and the G6 CRIby leukemiacellpopulationsmayreflect selec-tion of CRI-bearingcellsbasedupon thebinding activities of theexpressedIg. Furthercomparativeanalysesofthe immuno-globulinsproduced by malignant and normal B cells may help resolve whetherneoplastic CRI-bearingcells

arise

from a

sub-setofnormalCRI-bearingBlymphocytes.

Acknowledgments

Wethank Drs. Rizgar Mageed and Roy Jefferis (University of Bir-mingham, BirBir-mingham, England) fortheG6 MAbandDr. George Abraham (University of Rochester, Rochester, NY) for the anti-VHIIIb MAb.WethankDr.DennisA.Carson(University of

Califor-nia,SanDiego) for his carefulreadingof thismanuscript andhelpful

advice. In

addition,

we

acknowledge

the finetechnicalassistanceofMr. Grant Meisenholder.

Dr.

Kipps

isaScholar of the Leukemia

Society

ofAmerica,funded

in partbythe Scott

_Helping

Hand Fund. This workwas

supported

in

partbyNational Institutes of HealthgrantsAR38475 andAGO4100.

References

1.Kipps,T.J.,S.Fong,E.Tomhave,P. P.Chen,R. D._Goldfien,and D.A.

Carson. 1987.High-frequency expressionofaconservedkappalight-chain

vari-able-regiongene inchroniclymphocyticleukemia.Proc.Nall. Acad.Sci. USA.

84:2916-2920.

2.Carson,D.A.,and S.Fong.1983.Acommonidiotopeonhuman

rheuma-toid factorsidentifiedbya_hybridomaantibody.Mol.Immunol. 20:1081-1087.

3. Carson,D.A.,P.P. Chen,T. J. Kipps,V.Radoux,F. R. Jirik,R.D.

Goldfien,R.I.Fox, G. J.Silverman,and S.Fong. 1987.Idiotypicandgenetic

studies ofhumanrheumatoidfactors. Arthritis Rheum.30:1321-1325.

4.Kipps,T.J.,B.A.Robbins,P.Kuster,and D.A.Carson. 1988.

Autoanti-body-associatedcross-reactiveidiotypes expressedat_{high frequency}inchronic

lymphocyticleukemia relativetoB-celllymphomasoffollicularcentercell

ori-gin.Blood.72:422-428.

5.Mageed,R.A.,M.Dearlove,D.M.Goodall,and R.Jefferis.1986.

Immun-ogenicandantigenic epitopesofimmunoglobulins.XVII. Monoclonal

antibod-iesreactivewithcommonandrestrictedidiotopestotheheavychainof human

rheumatoidfactors.Rheumatol.Int.6:179-183.

6.Kipps,T.J.,B. A.Robbins,A.Tefferi,G.Meisenholder,P.M.Banks,and D.A.Carson. 1990.CD5-positiveB-cell_malignanciesfrequentlyexpress

cross-reactiveidiotypesassociatedwithIgMautoantibodies.Am.J.Pathol.

136:809-816.

7.Kipps,T.J.,E.Tomhave,P. P.Chen,and D.A.Carson. 1988.

Autoanti-body-associatedkappalightchain variableregiongeneexpressedinchronic lym-phocyticleukemia with littleornosomatic mutation.Implicationsforetiology

andimmunotherapy.J.Exp.Med. 167:840-852.

8.Pratt,L.F.,L.Rassenti,J.Larrick,B.Robbins,P.Banks,and T. J.Kipps.

1989.Immunoglobulingeneexpressioninsmalllymphocyticlymphomawith littleornosomatichypermutation.J.Immunol.143:699-705.

9.Radoux, V.,P. P.Chen,J.A._Sorge,and D.A.Carson. 1986.Aconserved humangermlineV_kappagenedirectlyencodesrheumatoidfactorlightchains. J.

Exp.Med. 164:2119-2124.

10.Kipps,T.J.,E.Tomhave,L.F.Pratt,S.Duffy,P. P.Chen,and D.A.

Carson. 1989.DevelopmentallyrestrictedVHgeneexpressedathigh frequency

inchronic

lymphocytic

leukemia.Proc.NaWl. Acad.Sci. USA. 86:5913-5917.

11.Caligaris-Cappio,F.,M.Gobbi,M.Bofill,and G.Janossy. 1982.

Infre-quentnormalBlymphocytesexpress features of B-chroniclymphocytic

leuke-mia. J.Exp.Med.155:623-628.

12.Gobbi, M.,F.Caligaris-Cappio,and G.Janossy.1983. Normalequivalent

cellsofBcellmalignancies:analysiswith monoclonalantibodies.Br.J.

Haema-tol. 54:393-403.

13.Gadol, N.,and K.A.Ault. 1986.Phenotypicandfunctional

characteriza-tion of human Leu I(CD5)cells.Immunol.Rev. 93:23-34.

14.Hardy,R.R.,and K.Hayakawa. 1986.Developmentandphysiologyof

Ly-lBand its humanhomolog,Leu-l B.Immunol.Rev.93:53-79.

15._Kipps,T.J.,and J. H.Vaughan.1987.Geneticinfluenceonthe levelsof

circulatingCD5B

lymphocytes.

J.Immunol. 139:1060-1064.

16.Bofill,M.,G. Janossy,M. Janossa,G. D. Burford,G. J.Seymour,P. Wernet,and E. Kelemen. 1985. Human B celldevelopment.II.Subpopulations

in the human fetus. J.Immunol. 134:1531-1538.

17.Antin,J.H.,S.G.Emerson,P.Martin,N.Gadol,andK A. Ault.1986. Leu-I+_(CD5+)Bcells:a_major

lymphoid

subpopulationin fetalspleen.

Pheno-typicand functional studies. J.Immunol.136:505-5 10.

18.Kipps, T.J.,B. A. Robbins,and D. A.Carson. 1990. Uniformhigh frequencyexpressionofautoantibody-associatedcross-reactiveidiotypesin the

primaryBcellfollicles of human fetalspleen.J.Exp.Med. 171:189-196. 19.Casali, P.,S. E.Burastero,M.Nakamura,G.Inghirami,andA.L. Not-kins. 1987. Human

lymphocytes

makingrheumatoid factor andantibodyto

belongtoLeu-l+B-cell subset. Science_{(Wash. DC).}236:77-81.

20.Hardy,R.R.,K._Hayakawa,M.Shimizu,K.Yamasaki,andT.

Kishi-moto.1987. Rheumatoid factor secretion from human Leu-l+Bcells. Science

(Wash.DC).236:81-83.

21.Carson,D.A.,P.P.Chen,T. J._Kipps,V.Radoux,F.Jirik,R.D.Goldfien,

R.I.Fox,G. J.Silverman,and S.Fong.1987. Molecular basis forthe

cross-reac-tive idiotypeson human _anti-IgGautoantibodies(rheumatoid factors).Ciba

Found.Symp. 129:123-134.

22.Silverman,G.

_J.,

R.D.Goldfien,P.Chen,R. A.Mageed,R.Jefferis,F. Goni,B._Frangione,S.Fong,and D. A.Carson.1988.Idiotypicandsubgroup

analysisof human monoclonal rheumatoidfactors:implicationsfor structural and_geneticbasisof autoantibodiesin humans. J.Clin.Invest.82:469-475.

and relateddisorders: practical aspects of frozen section technique and interpreta-tion. J.Clin. Lab.Anal. 1:104-115.

24.Greenstein, J. L., A. Solomon, and G. N. Abraham. 1984. Monoclonal antibodies reactive with idiotypic and variable-region specific determinants on humanimmunoglobulins.Immunology. 51:17-25.

25.Reinherz, E. L., R. E. Hussey, and S. F. Schlossman. 1980. A monoclonal antibody blocking human T cell function. Eur. J. Immunol. 10:758-762.

26.Maruyama, S., H. Kubagawa, and M. D. Cooper. 1985. Activation of human B cells andinhibition of their terminal differentiation by monoclonal anti-mu antibodies. J.Immunol. 135:192-199.

27. Kubagawa, H., W. E.Gathings, D. Levitt, J. F. Kearney, and M. D. Cooper. 1982.Immunoglobulinisotype expression of normal pre-B cells as deter-mined byimmunofluorescence. J. Clin. Immunol. 2:264-269.

28. Tedder, T. F., L. T.Clement, and M. D. Cooper. 1984. Expression ofC3d receptors during human B celldifferentiation:Immunofluorescenceanalysis with theHR-5monoclonalantibody. J. Immunol. 133:678-683.

29. Fox, R. I., D. Harlow,I. Royston, and J. Elder. 1982. Structural character-ization of the human T cell surface antigen (p67) isolated from normal and neoplastic lymphocytes. J.Immunol. 129:401-405.

30.Parks, D. R.,L. L.Lanier, and L. A. Herzenberg. 1986. Flow cytometry

andfluorescenceactivated cellsorting (FACS). In Handbook of Experimental

Immunology. D. M.Weir,L. A.Herzenberg, and C. Blackwell,editors.Blackwell ScientificPublications, Oxford. 29.1-29.21.

31.Saiki, R. K., D. H.Gelfand,S.Stoffel,S. J.Scharf,R.Higuchi, G. T. Horn, K. B.Mullis, and H. A. Erlich. 1988. Primer-directed enzymatic amplification of DNAwith a thermostable DNA polymerase.Science(Wash.DC). 239:487-491.

32. Hattori, M., and Y.Sakaki. 1986. Dideoxy sequencing method using denaturedplasmid templates.Anal. Biochem. 152:232-238.

33. Hsu, S.-M., and E. S.Jaffe. 1984. Phenotypic expression of B-lympho-cytes. 2. Immunoglobulin expression ofgermialcentercells. Am. J. Pathol.

114:396-402.

34.Gadol, N.,M. A.Peacock, andK. A.Ault.1988.Antigenicphenotype and functional characterization of human tonsil B cells.Blood.71:1048-1055.

35. Weinberg, D. S., K. A. Ault, M.Gurley, and G. S. Pinkus. 1986. The humanlymph nodegerminalcentercell:characterization and isolation byusing

two-colorflowcytometry. J.Immunol. 137:1486-1494.

36.Newkirk, M. M., H. Gram, G. F. Heinrich, L. Ostber g, J. D. Capra, and R. L.Wasserman. 1988. Complete protein sequences of the variableregionsof the clonedheavy andlight chainsofahumananticytomegalovirus antibody reveal astriking similarity to human monoclonal rheumatoid factors of the Wa idiotypic family. J.Clin.Invest.81:1511-1518.

37. Kabat, E.,T.T.Wu, M.Reid-Miller, H. M. Perry, and K. S. Gottesmann. 1987. Sequences of Proteins of Immunological Interest. Publication No. 1987:165-462. U.S. Government PrintingOffice, Bethesda, MD.

38.Siebenlist,U., J. V.Ravetch,S.Korsmeyer,T.Waldmann,and P. Leder.

1981.HumanimmunoglobulinDsegments encoded in tandemmultigenic fami-lies. Nature(Lond.).294:631-635.

39.Ichihara, Y., M. Abe,H.Yasui,H.Matsuoka,andY.Kurosawa.1988.At

leastfive DH genes of humanimmunoglobulin heavychainsareencoded in 9-kilobaseDNAfragments.Eur. J.Immunol. 18:649-652.

40.Ichihara, Y.,H.Matsuoka,and Y. Kurosawa. 1988.Organizationof

hu-manimmunoglobulinheavy chaindiversitygene loci. EMBO(Eur.Mol. Biol.

Organ.)J.7:4141-4150.

41.Matsuda, F., K. H. Lee, S.Nakai,T.Sato,M.Kodaira,S.Q.Zong,H.

Ohno,S.Fukuhara,andT.Honjo.1989.Dispersedlocalization ofD segments in

thehumanimmunoglobulin heavy-chainlocus.EMBO(Eur.Mol. Biol.Organ.)

J. 7:1047-1051.

42.Yancopoulos, G. D., andF. W.Alt. 1986.Regulationof theassemblyand

expressionofvariable-regiongenes.Annu.Rev. Immunol. 4:339-368.

43.Desiderio,S.V., G.D.Yancopoulos,M.Paskind,E.Thomas,M.A.Boss,

N.Landau,F. W.Alt,and D.Baltimore. 1984. Insertion ofNregionsinto

heavy-chain genes is correlated withexpressionof terminaldeoxytransferasein B cells. Nature(Lond.).311:752-755.

44.Caligaris-Cappio,F.,M.Riva,L.Tesio,M.Schena,G.Gaidano,and L.

Bergui.1989.Human normalCD5+Blymphocytescanbeinducedto differen-tiatetoCD5- Blymphocyteswithgerminalcentercell features. Blood.

73:1259-1263.

45.Kipps,T. J. 1989. TheCD5 B cell. Adv. Immunol. 47:117-185.

46.Herzenberg,L. A., A. M.Stall,P. A.Lalor,C.Sidman,W. A.Moore,and

D.R.Parks. 1986. TheLy-lBcelllineage. Immunol.Rev.93:81-102.

47.Herzenberg,L. A., andA.M.Stall.1989.ConventionalandLy-I B-cell

lineagesin normal andmutransgenicmice.Cold Spring Harbor Symp. Quant.

Biol.54:219-225.

48.Herzenberg,L. A., andL.A.Herzenberg.1989. Toward a layered immune

system.Cell. 59:953-954.

49.Herzenberg,L. A., A.M. Stall, J. Braun, D. Weaver, D. Baltimore, and R.

Grosschedl. 1987.Depletionofthe predominant B-cell population in mu

heavy-chaintransgenic mice.Nature(Lond.). 329:71-73.

50.Stall,A.M.,M.C. Farinas, D. M. Tarlinton, P. A. Lalor, L. A.

Herzen-berg,and S.Strober.1988. Ly-I B-cell clonessimilar to human chronic

lympho-cytic leukemia routinelydevelop in older normal mice and youngautoimmune

Zealand Black-relatedanimals.Proc.Natl.Acad.Sci. USA. 85:7312-7316. 51.Forster, I., H. Gu, and K. Rajewsky. 1988. Germline antibody V regions

asdeterminants ofclonalpersistenceandmalignantgrowth in the B cell

compart-ment.EMBO(Eur. Mol. Biol. Organ.)J.7:3693-3703.

52.Tarlinton,D.,A.M.Stall, and L. A. Herzenberg. 1988. Repetitive usage of

immunoglobulinVHand D gene segments inCD5+ Ly-l B clonesof (NZB

xNZW)F1mice. EMBO (Eur. Mol. Biol.Organ.)J. 7:3705-10.

53.Reininger,L., P.Ollier,P.Poncet,A.Kaushik, and J. C. Jaton. 1987.

NovelVgenes encodevirtually identicalvariableregions of six murine

monoclo-nalanti-bromelain-treatedred blood cellautoantibodies.J.Immunol.

138:316-323.

54.Sanz, I.,P.Casali,J. W.Thomas,A.L.Notkins,and J. D.Capra. 1989.

Nucleotide sequencesof eighthumannaturalautoantibodyVHregions reveals apparentrestricteduseofVHfamilies.J.Immunol.142:4054-4061.

55.Nakamura, M.,S. E.Burastero,Y.Ueki,J. W.Larrick,A.L.Notkins, and

P.Casali. 1988. Probingthe normalandautoimmuneBcell repertoire with

Epstein-Barr virus: frequencyofBcellsproducingmonoreactivehigh affinity

autoantibodies inpatientswith Hashimoto'sdiseaseandsystemiclupus

erythe-matosus.J.Immunol. 141:4165-4172.

56.Sanz, I.,H.Dang,M.Takei,N.Talal,andJ. D.Capra.1989.VHsequence

ofahumananti-Smautoantibody.Evidence thatautoantibodiescanbe unmu-tatedcopiesofgermlinegenes. J.Immunol. 142:883-887.

57.Baccala, R.,T. V.Quang,M.Gilbert,T.Ternynck,andS. Avrameas.

1989. Twomurine naturalpolyreactive autoantibodiesareencodedby

nonmu-tatedgerm-linegenes. Proc. Natl.Acad.Sci. USA. 86:4624-4628.

58.Newkirk,M.M.,R.A.Mageed,R.Jefferis,P. P.Chen,and J.D.Capra.

1987.Completeamino acid sequences of variableregionsoftwohumanIgM rheumatoidfactors,BOR and KASofthe Waidiotypic family,revealrestricted

useofheavyandlightchainvariableandjoining regiongene segments. J.Exp.

Med. 166:550-564.

59.Ravetch,J.V.,U.Siebenlist,S.Korsmeyer,T.Waldmann,and P. Leder.

1981. Structureof the humanimmunoglobulinmulocus:characterization of

embryonicandrearrangedJ andDgenes.Cell.27:583-591.

60.Schroeder,H.W., Jr.,J. L.Hillson,andR.M. Perlmutter. 1987.Early

restrictionofthe humanantibody repertoire.Science(Wash.DC).238:791-793. 61.Shokri,F.,R.A.Mageed,E.Tunn,P.A.Bacon,and R.Jefferis. 1990.

Qualitativeandquantitative expressionofVHIassociatedcrossreactiveidiotopes

withinIgMrheumatoidfactorfrompatientswithearlysynovitis.Ann.Rheum.

Dis.49:150-154.